From drewpc at colorado.edu  Thu Feb  1 22:36:11 2001
From: drewpc at colorado.edu (drewpc at colorado.edu)
Date: Thu, 01 Feb 2001 21:36:11 GMT
Subject: Python and Industry
References: <20010111172342.10148.00000934@ng-ca1.aol.com>
Message-ID: <95cksa$e2d$1@nnrp1.deja.com>

In article <20010111172342.10148.00000934 at ng-ca1.aol.com>,
  crutoy at aol.com (Crutoy) wrote:
> Hi , does anyone know how widly in an instries does python being used
? I
> looked at the syntax of it and read some stuff on it,it's interested i
just
> want to know if it's worth learning it. Thanks
>

I work at a large networking company, and it's used all over the place.
Anywhere from maintaining over a hundred DNS server configs to
configuring routers to pulling accounting information off of various
telco gear.  I'm constantly amazed at how widely it's used.


Sent via Deja.com
http://www.deja.com/


From qrczak at knm.org.pl  Sat Feb  3 14:44:45 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 3 Feb 2001 13:44:45 GMT
Subject: python compiled to native in less than a year?
References: <kcF86.6006$J%.594457@news.flash.net> <Pine.GSO.4.21.0101151106200.7702-100000@z.glue.umd.edu> <8pF86.6015$J%.595435@news.flash.net> <0luc6toovfccbqspkvc0dkqkuc5dck7u51@4ax.com> <3A67341D.D5022902@businesscollaborator.com> <mailman.979927472.470.python-list@python.org>
Message-ID: <slrn97o2qd.60f.qrczak@qrnik.zagroda>

Fri, 19 Jan 2001 12:44:59 -0500, D-Man <dsh8290 at rit.edu> pisze:

> BTW,  I've heard several references to C# on this list, with a high
> concentration on this thread.  Where can I find some sort of (short)
> docs that describe what C# is?  (an intro or overview type of thing)

http://msdn.microsoft.com/vstudio/nextgen/technology/csharpintro.asp

C# is very similar to Java.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From rerla at channing.harvard.edu  Thu Feb  1 19:24:43 2001
From: rerla at channing.harvard.edu (Ross Lazarus)
Date: Thu, 01 Feb 2001 13:24:43 -0500
Subject: Random and whrandom
References: <mailman.980371172.18811.python-list@python.org> <94p31e02135@news2.newsguy.com>
Message-ID: <3A79A9EB.701D1E96@channing.harvard.edu>

Nice - and I found the documentation at
http://www.stecf.org/~npirzkal/python/rv/rv-ref.html 

Also of potential interest to fellow prng consumers who need real speed
- http://www.sbc.su.se/~per/crng/ - gives about 6 million/sec from the
mersenne twister cf 7600/sec from rv twister (rv running on a slower
pentium). 

Alex Martelli wrote:

> As partial atonement for the sin and heresy of "Defending
> Unnecessary Complexity In The Name Of Unwarranted Generality",
> I've taken advantage of Frohne's very-generous copyright
> provisions to put an rv.zip containing his rv.py 1.1 (from
> November 1998) up for download at http://www.fenx.com/alex/rv.zip
> (it's not a particularly fast or well-connected machine, but
> the only one I can place something for anonymous dowload
> right now -- I hope somebody will pick it up and place it
> somewhere better accessible!).
>


From frohne at gci.net  Fri Feb  2 01:35:33 2001
From: frohne at gci.net (Ivan Frohne)
Date: Thu, 1 Feb 2001 15:35:33 -0900
Subject: Random and whrandom
References: <mailman.980371172.18811.python-list@python.org> <94p31e02135@news2.newsguy.com> <3A79A9EB.701D1E96@channing.harvard.edu>
Message-ID: <t7k0knr158smcc@corp.supernews.com>

"Ross Lazarus" <rerla at channing.harvard.edu> wrote in message
news:3A79A9EB.701D1E96 at channing.harvard.edu...
> Nice - and I found the documentation at
> http://www.stecf.org/~npirzkal/python/rv/rv-ref.html
>
> Also of potential interest to fellow prng consumers who need real speed
> - http://www.sbc.su.se/~per/crng/ - gives about 6 million/sec from the
> mersenne twister cf 7600/sec from rv twister (rv running on a slower
> pentium).

You can also find the documentation for the random number module
rv 1.1 at

http://www.frohne.westhost.com/rv11reference.htm

and a ZIP archive containing the documentation and Python
module rv.py at

http://www.frohne.westhost.com/rv11.zip

If you're still hungry, download

http://www.frohne.westhost.com/mr_random.py

which is a whrandom clone, 60% faster and with a
period of 10**28 (vs 10**12 for whrandom).

Ivan Frohne




From aleaxit at yahoo.com  Fri Feb  2 10:27:08 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 2 Feb 2001 10:27:08 +0100
Subject: Random and whrandom
References: <mailman.980371172.18811.python-list@python.org> <94p31e02135@news2.newsguy.com> <3A79A9EB.701D1E96@channing.harvard.edu>
Message-ID: <95duhc030kb@news1.newsguy.com>

"Ross Lazarus" <rerla at channing.harvard.edu> wrote in message
news:3A79A9EB.701D1E96 at channing.harvard.edu...
> Nice - and I found the documentation at
> http://www.stecf.org/~npirzkal/python/rv/rv-ref.html
>
> Also of potential interest to fellow prng consumers who need real speed
> - http://www.sbc.su.se/~per/crng/ - gives about 6 million/sec from the
> mersenne twister cf 7600/sec from rv twister (rv running on a slower
> pentium).

Mersenne Twister looks like just the ticket when what you're
doing is "shuffling a sequence" (or "stirring" it, in crng
terms -- or, also, using 'choose' or 'sample' to get subsets
or subsequences, with or without repetition/permutation),
the point being that its huge period gives you a _chance_ of
selecting subsets/permutations/samples in a way that makes
all possible ones roughly equally-likely (but I don't think
this has ever been *proved* -- it's just that MT has some
*chance*, when RNG's with shorter periods clearly don't).

Speed can also become quite relevant here.

Say you're shuffling a deck of 52 cards -- there are about
8.0e+67 permutations (52!), and getting one permutation
requires about 50 random numbers (with the normal Knuth
approach).  A typical task in bridge studies is to deal
out a huge variety of deals and throw away all except those
meeting certain specified criteria -- and it's not unusual
for those criteria to have an a priori chance of, say, one
in a million, so one needs about 50 million random numbers
per each significant deal produced for study (at least
when such safe brute-force Montecarlo approaches are
used; clever accelerations are possible but run serious
risks of double-counting or subtle distortions if one is
not incredibly careful in setting conditions up).

This translates to about 8 seconds per significant deal
with crng, versus about 2 hours per deal with rv, with
the numbers you quote (actually, it WILL take, even in
the best case, a couple of milliseconds at least to study
and discard each deal produced, so the ratio is not all
that dramatic: 2000+ seconds per produced deal with
crng, vs almost 9000 with rv -- which is why one stays
motivated to produce the 'clever accelerations' above
mentioned, but that's another issue -- still, a 4:1
ratio *is* achievable for a brute force approach here).


Alex





From Pekka.Pessi at nokia.com  Mon Feb  5 01:14:29 2001
From: Pekka.Pessi at nokia.com (Pekka Pessi)
Date: 05 Feb 2001 02:14:29 +0200
Subject: How do I know all thrown exceptions of a function?
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <Qeya6.7932$AH6.1260668@newsc.telia.net> <3A6B24F0.EA164698@gte.net> <94icr2$6kv$1@animus.fel.iae.nl> <3A6CDBDA.C3120F98@gte.net> <94ipci$afo$1@animus.fel.iae.nl> <3A6CFD8E.BD473DFD@gte.net> <iEab6.8765$Qb7.1233360@newsb.telia.net> <3A6D54CB.2261D535@gte.net> <94jo9u$h45$1@nnrp1.deja.com> <94jt7e01fil@news1.newsguy.com> <Zvjb6.8419$AH6.1356839@newsc.telia.net>
Message-ID: <pv8znm7zq2.fsf@agni.research.nokia.com>

In message <Zvjb6.8419$AH6.1356839 at newsc.telia.net> "Fredrik Lundh" <fredrik at effbot.org> writes:
>might so be, but I still cannot make any sense of his post.

        He says that according to the docs, "from tkinter include *" should
        be used instead of "include tkinter".  The latter adds lot of
        unnecessary text, ie. cruft, into your program.  Cruft makes your
        code harder to read.

        Are we feeling a little bit obtuse on purpose today?

                                        Pekka


From fredrik at effbot.org  Mon Feb  5 02:03:35 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 05 Feb 2001 01:03:35 GMT
Subject: How do I know all thrown exceptions of a function?
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <Qeya6.7932$AH6.1260668@newsc.telia.net> <3A6B24F0.EA164698@gte.net> <94icr2$6kv$1@animus.fel.iae.nl> <3A6CDBDA.C3120F98@gte.net> <94ipci$afo$1@animus.fel.iae.nl> <3A6CFD8E.BD473DFD@gte.net> <iEab6.8765$Qb7.1233360@newsb.telia.net> <3A6D54CB.2261D535@gte.net> <94jo9u$h45$1@nnrp1.deja.com> <94jt7e01fil@news1.newsguy.com> <Zvjb6.8419$AH6.1356839@newsc.telia.net> <pv8znm7zq2.fsf@agni.research.nokia.com>
Message-ID: <HZmf6.12204$AH6.1891634@newsc.telia.net>

Pekka Pessi wrote:
> >might so be, but I still cannot make any sense of his post.
>
>         He says that according to the docs, "from tkinter include *" should
>         be used instead of "include tkinter".  The latter adds lot of
>         unnecessary text, ie. cruft, into your program.  Cruft makes your
>         code harder to read.

umm.  *I* wrote the part on Tkinter, and I'm pretty sure I
know what I meant with the original text.  still don't know
what Steve meant, and frankly, I don't really care.

Cheers /F




From sholden at holdenweb.com  Mon Feb  5 00:14:05 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 4 Feb 2001 18:14:05 -0500
Subject: How do I know all thrown exceptions of a function?
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <Qeya6.7932$AH6.1260668@newsc.telia.net> <3A6B24F0.EA164698@gte.net> <94icr2$6kv$1@animus.fel.iae.nl> <3A6CDBDA.C3120F98@gte.net> <94ipci$afo$1@animus.fel.iae.nl> <3A6CFD8E.BD473DFD@gte.net> <iEab6.8765$Qb7.1233360@newsb.telia.net> <3A6D54CB.2261D535@gte.net> <94jo9u$h45$1@nnrp1.deja.com>
Message-ID: <OvBf6.32$x4.160@e420r-atl2.usenetserver.com>

<fredrik at pythonware.com> wrote in message
news:94jo9u$h45$1 at nnrp1.deja.com...
> "sandj.williams at gte.net" wrote:
> > I parse that to mean 'cruft'.
>
> Sorry, I cannot find 'cruft' in my dictionary.  Did you mean
> that Python's import statement doesn't work as described, that
> Tkinter hasn't been designed to support from-import, or what?
>
> </F>
>
>
> Sent via Deja.com
> http://www.deja.com/
>

It's defined in the jargon file. See http://www.tuxedo.org/~esr/jargon.html

regards
 Steve




From sholden at holdenweb.com  Fri Feb  2 22:07:37 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 2 Feb 2001 16:07:37 -0500
Subject: How do I know all thrown exceptions of a function?
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <slrn96rk66.ugd.scarblac@pino.selwerd.nl> <mailman.980287769.21816.python-list@python.org>
Message-ID: <loFe6.6912$o5.45360@e420r-atl1.usenetserver.com>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.980287769.21816.python-list at python.org...
> On Tue, Jan 23, 2001 at 06:48:45PM +0000, Remco Gerlich wrote:
[ ... ]
>
> | - Obviously trying to change exceptions that aren't caused by your code
> |   (like KeyboardInterrupt) is of no use.
>
> How about ComputerExplodingError :-).
>
Surely the only code which would cause this would be something which
naiively executed the infamous HCF opcode?  Is there an HCF equivalent in
the Python bytecodes?

regards
 Steve



From sandj.williams at gte.net  Sat Feb  3 05:05:41 2001
From: sandj.williams at gte.net (Steve Williams)
Date: Sat, 03 Feb 2001 04:05:41 GMT
Subject: How do I know all thrown exceptions of a function?
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <slrn96rk66.ugd.scarblac@pino.selwerd.nl> <mailman.980287769.21816.python-list@python.org> <loFe6.6912$o5.45360@e420r-atl1.usenetserver.com>
Message-ID: <3A7B850F.E7188161@gte.net>

Steve Holden wrote:

> "D-Man" <dsh8290 at rit.edu> wrote in message
> news:mailman.980287769.21816.python-list at python.org...
> > On Tue, Jan 23, 2001 at 06:48:45PM +0000, Remco Gerlich wrote:
> [ ... ]
> >
> > | - Obviously trying to change exceptions that aren't caused by your code
> > |   (like KeyboardInterrupt) is of no use.
> >
> > How about ComputerExplodingError :-).
> >
> Surely the only code which would cause this would be something which
> naiively executed the infamous HCF opcode?  Is there an HCF equivalent in
> the Python bytecodes?
>
> regards
>  Steve

Or even the IBM 7010 opcode 'Branch on Sleepy Operator.'



From jafo at tummy.com  Sat Feb  3 17:46:08 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Sat, 3 Feb 2001 09:46:08 -0700
Subject: How do I know all thrown exceptions of a function?
In-Reply-To: <loFe6.6912$o5.45360@e420r-atl1.usenetserver.com>; from sholden@holdenweb.com on Fri, Feb 02, 2001 at 04:07:37PM -0500
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <slrn96rk66.ugd.scarblac@pino.selwerd.nl> <mailman.980287769.21816.python-list@python.org> <loFe6.6912$o5.45360@e420r-atl1.usenetserver.com>
Message-ID: <20010203094608.D4938@tummy.com>

On Fri, Feb 02, 2001 at 04:07:37PM -0500, Steve Holden wrote:
>naiively executed the infamous HCF opcode?  Is there an HCF equivalent in
>the Python bytecodes?

Rockwell named a modem after it.  For those who don't recall, HCF is
the mythical Halt and Catch Fire opcode.  As ISP tech support folks
will attest, the Rockwell modems might as well have...

On the exceptions thrown by a function, this is one of the things I liked
about Java.  It required that your functions either catch all exceptions
generated within it, or that they declare that they will throw certain
exceptions.  Unfortunately, that's kind of hard to do in Python.

Sean
-- 
 It often shows a fine command of a language to say nothing.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From dnew at san.rr.com  Sun Feb  4 01:22:47 2001
From: dnew at san.rr.com (Darren New)
Date: Sun, 04 Feb 2001 00:22:47 GMT
Subject: How do I know all thrown exceptions of a function?
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <slrn96rk66.ugd.scarblac@pino.selwerd.nl> <mailman.980287769.21816.python-list@python.org> <loFe6.6912$o5.45360@e420r-atl1.usenetserver.com> <mailman.981218824.27465.python-list@python.org>
Message-ID: <3A7CA0D7.D8A40228@san.rr.com>

Sean Reifschneider wrote:
> For those who don't recall, HCF is
> the mythical Halt and Catch Fire opcode. 

It's not mythical. It's one of the undocumented opcodes in a 6502,
presumedly for testing during manufacturing. When you execute the HCF
opcode, the chip starts cycling all the address lines, just counting thru
all 65536 patterns, until you remove power. The person who discovered it
assumed it was used before the chip was bonded into the carrier.  I think
all this was described in Byte back when mainstream programmers cared about
6502s.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
                 Ignorance can be cured. Naivety cures itself.


From jafo at tummy.com  Sun Feb  4 05:47:38 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Sat, 3 Feb 2001 21:47:38 -0700
Subject: How do I know all thrown exceptions of a function?
In-Reply-To: <3A7CA0D7.D8A40228@san.rr.com>; from dnew@san.rr.com on Sun, Feb 04, 2001 at 12:22:47AM +0000
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <slrn96rk66.ugd.scarblac@pino.selwerd.nl> <mailman.980287769.21816.python-list@python.org> <loFe6.6912$o5.45360@e420r-atl1.usenetserver.com> <mailman.981218824.27465.python-list@python.org> <3A7CA0D7.D8A40228@san.rr.com>
Message-ID: <20010203214737.A29866@tummy.com>

On Sun, Feb 04, 2001 at 12:22:47AM +0000, Darren New wrote:
>It's not mythical. It's one of the undocumented opcodes in a 6502,

Though Rockwell was probably the closest, I don't believe anyone has
correctly implemented that opcode.  ;-)

Sean
-- 
 I took a Meyers-Briggs personality test and came out an SMTP.
                 -- Sean Reifschneider, 2000
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From ullrich at math.okstate.edu  Fri Feb  2 15:58:08 2001
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Fri, 02 Feb 2001 14:58:08 GMT
Subject: pyXML support for XSL tranformations??
References: <rtga6.18098$bb.708443@news1.rdc1.tx.home.com> <94d1tj$6sa$1@nnrp1.deja.com> <3a6aef10.3738521@nntp.sprynet.com> <3A7232DB.C576B547@ogbuji.net> <94vg3k$ed9$1@nnrp1.deja.com> <3A7442FF.E1CFDF6D@ogbuji.net>
Message-ID: <3a7ac7f1.3651537@nntp.sprynet.com>

On Sun, 28 Jan 2001 16:04:18 GMT, Uche Ogbuji <uche at ogbuji.net> wrote:

>"David C. Ullrich" wrote:
>
>> Once I got it working it went right onto my I-don't-see-
>> how-you-guys-can-give-this-stuff-away list, btw - hope
>> you don't have the idea I've been wanting my money
>> back or anything.
>
>No, I just wanted to have things clear on the thread because there *was*
>a time when installing 4Suite was a huge pain, and I'd like all to know
>we've worked very hard to make it a breeze (with some help from Greg
>Ward and distutils).

The install (probably a year or so ago) _was_ very easy. I may
as well say what was so stupid about me: There was a spot
in the instructions about unzipping to a "suitable" directory,
with no definition of "suitable". Didn't take me long to figure
out what was unsuitable about the directory I used. That
led to copying some things over some things. I didn't replace
a certain subdirectory that already existed cuz I wasn't
sure whether anything was using that stuff already. The
XSLT didn't work. Then I replaced python\xml\dom with
the one that appeared when I unzipped your package
and everything worked fine.

The "uncleanliness" I was alluding to was having to
modify part of an existing installation. But if I were
not an idiot I would have chosen a suitable directory
to begin with and I wouldn't have noticed the change.
Or I would have noticed that the stuff in the directory
I was reluctant to replace was _also_ an xml\dom
from 4Thought - I thought for no good reason that
it was a module from an entirely independent
source so that replacing it would likely break something
else.

>> > There is no
>> > reason why you can't download 4Suite-0.10.1.tar.gz, untar it, run
>> > "python setup.py install", and be completely ready to go.
>> 
>> You only think that because you're not stupid enough to find
>> ways to screw it up...
>
>Well, we'd actually like 4Suite to be easy to install even for
>non-Pythoneers. 

I think I just slipped through the cracks there - knew just
enough to do it wrong. Not your fault.

> That's because some of the features are completely
>unique in XML space, and we think we can actually draw some XML users to
>Python if they don't have to work too hard to get started.  That's why
>we put together the HOWTOs which even walk you though installing Python.
>
>> > But just in case, there are a 4Suite Installation HOWTOs for UNIX and
>> > Windows.
>> >
>> > See
>> >
>> > http://services.4Suite.org/topics/rdf/Top/Documentation/HOWTOs
>
>
>-- 
>Uche Ogbuji
>Personal:   uche at ogbuji.net		http://uche.ogbuji.net
>Work:       uche.ogbuji at fourthought.com	http://Fourthought.com



From dave at dogwood.com  Wed Feb  7 06:51:45 2001
From: dave at dogwood.com (Dave Cornejo)
Date: Wed, 07 Feb 2001 05:51:45 GMT
Subject: 2nd CFV: comp.soft-sys.wxwindows
References: <980126365.16946@isc.org>
Message-ID: <981525104.6624@isc.org>

                      LAST CALL FOR VOTES (of 2)
              unmoderated group comp.soft-sys.wxwindows

Newsgroups line:
comp.soft-sys.wxwindows	Discussions related to wxWindows framework.

Votes must be received by 23:59:59 UTC, 12 Feb 2001.

This vote is being conducted by a neutral third party.  Questions
about the proposed group should be directed to the proponent.

Proponent: Vadim Zeitlin <vadim at wxwindows.org>
Votetaker: Dave Cornejo <dave at dogwood.com>

RATIONALE: comp.soft-sys.wxwindows

This group should host the discussions about wxWindows (portable,
free framework for writing GUI and console applications in C++,
Python and Perl for a variety of platforms). wxWindows currently
has a number of very busy mailing lists, creating a newsgroup
should help people to participate in discussions about it without
being overwhelmed by the number of messages.

CHARTER: comp.soft-sys.wxwindows

The comp.soft-sys.wxwindows newsgroup should be used for any
discussions related to wxWindows framework without limitation to any
language (although only C++, Python and Perl are currently supported)
or platform (porting to a currently unsupported platform may be
discussed here).

Examples of correct postings include (but are not limited to) the
following subjects:

- questions related to using wxWindows
- discussion of current or future wxWindows developments
- bug reports and feature requests

Examples of unwelcome postings:

- questions about C or C++ language except when directly related to
  support for C++ feature or lack of thereof in wxWindows (e.g.
  "Should destructor of the base class be virtual?" is inappropriate
  while "How can I use exceptions with wxWindows?" is not).
- platform-specific programming questions unrelated to wxWindows
  (specifically questions about Win32 API are unwelcome)
- questions about using wxWindows answered in the FAQ
- long (more than a few lines) patches - they should be submitted
  instead to the patch manager at the project Web site

The newsgroup is not moderated. Posting of binaries and
advertisement is not allowed. As an exception, advertisement
is allowed for the announcement of products directly related to
wxWindows (either written for wxWindows or using wxWindows).

END CHARTER.

HOW TO VOTE:

Follow these instructions *exactly*!  Votes are counted by computer.
You should send E-MAIL (posts to a newsgroup are invalid) to:

    vote at dogwood.com

Please do not assume that just replying to this message will work.
Check the address before you mail your vote.  Your mail message
should contain one and only one of the following vote statements:

    I vote YES on comp.soft-sys.wxwindows
    I vote NO on comp.soft-sys.wxwindows

    Voter name:

If your mail software does not indicate your real name (for example,
AOL does not), include _exactly_ the statement above on a _separate_
line and add your name after the colon.  Having your name in your
signature line is NOT enough!  Do NOT join the lines together or
remove the words "Voter name"!

You may also vote ABSTAIN (which records an empty vote) or CANCEL
(which removes any earlier votes).  ABSTAIN does not affect the final
vote count in any way but is listed, whereas CANCEL is not.

IMPORTANT VOTING PROCEDURE NOTES:

Standard Guidelines for voting apply.  One vote per person, one
account per voter.  Votes must be mailed directly from the voter to
the votetaker.  Anonymous, forwarded or proxy votes are not valid.
Votes mailed by WWW/HTML/CGI forms are considered to be anonymous
votes.  Votes from non-existent email addresses are not valid.

Vote counting is automated.  Failure to follow these directions may
mean that your vote does not get counted.  If you do not receive an
acknowledgment of your vote within three days contact the votetaker
about the problem.  It's your responsibility to make sure your vote is
registered correctly.  Duplicate votes are resolved in favor of the
most recent valid vote.  Addresses, names and votes of all voters will
be published in the final voting results post.

The purpose of a Usenet vote is to determine the genuine interest of
persons who would read a proposed newsgroup.  Soliciting votes from
disinterested parties defeats this purpose.  Please do not distribute
this CFV.  If you must, direct people to the official CFV as posted to
news.announce.newgroups.  Distributing pre-marked or otherwise edited
copies of this CFV is generally considered to be vote fraud.  When in
doubt, ask the votetaker.

DISTRIBUTION:

In addition to the groups named in the Newsgroups: header, the CFV
and the eventual RESULT posts will be mailed to these mailing lists:

Mailing list name: wxwindows-devel
Submission address: wxwindows-devel at lists.sourceforge.org
Request address (optional):

Mailing list name: wxwindows-users
Submission address: wxwindows-users at lists.sourceforge.org
Request address (optional):

Mailing list name: wxpython-users
Submission address: wxpython-users at lists.sourceforge.org
Request address (optional):

-- 
Voting question & problems: Dave Cornejo <dave at dogwood.com>
Voting address: vote at dogwood.com


From dave at dogwood.com  Fri Feb 23 08:31:14 2001
From: dave at dogwood.com (Dave Cornejo)
Date: Fri, 23 Feb 2001 07:31:14 GMT
Subject: RESULT: comp.soft-sys.wxwindows passes 164:13
References: <980126365.16946@isc.org> <981525104.6624@isc.org>
Message-ID: <982913473.1281@isc.org>

                                RESULT
       unmoderated group comp.soft-sys.wxwindows passes 164:13

There were 164 YES votes and 13 NO votes, for a total of 177 valid
votes.  There were 3 abstentions and 3 invalid ballots.

For a group to pass, YES votes must be at least 2/3 of all valid
(YES and NO) votes.  There must also be at least 100 more YES votes
than NO votes.

A five day discussion period follows this announcement.  If no
serious allegations of voting irregularities are raised, the
moderator of news.announce.newgroups will create the group shortly
thereafter.


Newsgroups line:
comp.soft-sys.wxwindows	Discussions related to wxWindows framework.

Voting closed at 23:59:59 UTC, 12 Feb 2001.

This vote was conducted by a neutral third party.  Questions
about the proposed group should be directed to the proponent.

Proponent: Vadim Zeitlin <vadim at wxwindows.org>
Votetaker: Dave Cornejo <dave at dogwood.com>

RATIONALE: comp.soft-sys.wxwindows

This group should host the discussions about wxWindows (portable,
free framework for writing GUI and console applications in C++,
Python and Perl for a variety of platforms). wxWindows currently
has a number of very busy mailing lists, creating a newsgroup
should help people to participate in discussions about it without
being overwhelmed by the number of messages.

CHARTER: comp.soft-sys.wxwindows

The comp.soft-sys.wxwindows newsgroup should be used for any
discussions related to wxWindows framework without limitation to any
language (although only C++, Python and Perl are currently supported)
or platform (porting to a currently unsupported platform may be
discussed here).

Examples of correct postings include (but are not limited to) the
following subjects:

- questions related to using wxWindows
- discussion of current or future wxWindows developments
- bug reports and feature requests

Examples of unwelcome postings:

- questions about C or C++ language except when directly related to
  support for C++ feature or lack of thereof in wxWindows (e.g.
  "Should destructor of the base class be virtual?" is inappropriate
  while "How can I use exceptions with wxWindows?" is not).
- platform-specific programming questions unrelated to wxWindows
  (specifically questions about Win32 API are unwelcome)
- questions about using wxWindows answered in the FAQ
- long (more than a few lines) patches - they should be submitted
  instead to the patch manager at the project Web site

The newsgroup is not moderated. Posting of binaries and
advertisement is not allowed. As an exception, advertisement
is allowed for the announcement of products directly related to
wxWindows (either written for wxWindows or using wxWindows).

END CHARTER.

DISTRIBUTION:

In addition to the groups named in the Newsgroups: header, the CFV
and the eventual RESULT posts will be mailed to these mailing lists:

Mailing list name: wxwindows-devel
Submission address: wxwindows-devel at lists.sourceforge.org
Request address (optional):

Mailing list name: wxwindows-users
Submission address: wxwindows-users at lists.sourceforge.org
Request address (optional):

Mailing list name: wxpython-users
Submission address: wxpython-users at lists.sourceforge.org
Request address (optional):


comp.soft-sys.wxwindows Final Voter list

Voted YES
-------------------------------------------------------------------------------
peter#scully.com.au                                                Peter Scully
geldridg#progsoc.uts.edu.au                                      Geoff Eldridge
toivo#ucs.uwa.edu.au                                              Toivo Pedaste
Chris.Fama#uq.net.au                                                 Chris Fama
praxhon#crl.be                                                  Philippe Raxhon
Humphrey.Clerx#eurocontrol.be                                    Humphrey Clerx
ranma#mail.gargoyles.ch                                         Thomas Haeberli
robin#alldunn.com                                                    Robin Dunn
teague#apicom.com                                               Teague Sheridan
Achim.Neumann#base-system.com                                     Achim Neumann
Alain.Culos#bigfoot.com                                             Alain CULOS
michaelteo#bigfoot.com                                              Michael Teo
pvlad#bigfoot.com                                                        P Vlad
jumpo#bitex.com                                                     Angel Popov
movits#BLOOMBERG.COM                                            Mordechai Ovits
biswapesh.chattopadhyay#bt.com                          Biswapesh Chattopadhyay
grw#cabernet.com                                                 Greg Whitehead
michael.kennedy#cognotec.com                                    Michael Kennedy
Onorio_Catenacci#compuware.com                                 Onorio Catenacci
dan#sol.control.com                                              Dan L. Pierson
francis.irving#creaturelabs.com                                  Francis Irving
rjh#cyberscience.com                                              Richie Hindle
carl#dgi.com                                                        Carl Godkin
john#dgi.com                                                         John Skiff
erlando#dnspilot.com   =?iso-8859-1?Q?S=F8ren?= Erland =?iso-8859-1?Q?Vest=F8?=
peter#engcorp.com                                                  Peter Hansen
jfn#shell-1.enteract.com                                             J.F.Nelson
gtasker#fastpicsystems.com                                        George Tasker
emile#fcfw.fenx.com                                           Emile van Sebille
db3l#fitlinxx.com                                                   David Bolen
Duane.Kaufman#med.ge.com                                          Duane Kaufman
Klaus-Guenter.Leiss#de.heidelberg.com                     Klaus - Guenter Leiss
donod#home.com                                                    Don O'Donnell
earthscibbs#home.com                                              David Ramalho
gregglsc#home.com                                                      Gregg=20
parz#home.com                                                   Parzival Herzog
steve_frampton#hotmail.com                                       Steve Frampton
mtebeka#iil.intel.com                                               Miki Tebeka
lionel.allorge#lunerouge.com                                     Lionel Allorge
z3p#mailandnews.com                                                         z3p
jbail#mak.com                                                         Jeff Bail
tom#malcolmson.com                                               Tom Malcolmson
perry#modelspace.com                                               Perry Miller
fajuto#my-deja.com                                             Francisco Fajuto
gtalvola#nameconnector.com                                     Geoffrey Talvola
jc+usenet#port25.com                                        Juergen Christoffel
julians#redhat.com                                                 Julian Smart
procyon#sensewave.com                                            Lars Helgeland
dbryson#techass.com                                                Derry Bryson
ry#teleport.com                                                      Ryan Gavin
dima#testinsight.com                                                Dima Angert
kbrogers#thedigirati.com                                        Kerry B. Rogers
clstamper#theglobe.com                                            Chris Stamper
Alex.Martelli#think3.com                                          Alex Martelli
mfletch#tpresence.com                                          Mike C. Fletcher
xanthian#well.com                                               Kent Paul Dolan
Allen.Van.Sickel#usa.xerox.com                                 Allen Van Sickel
Carlo.Sans#usa.xerox.com                                             Carlo Sans
Loreene.Dixon#usa.xerox.com                                       Loreene Dixon
Marvin.Aviles#usa.xerox.com                                       Marvin Aviles
Richard.L.Clark#usa.xerox.com                                     Richard Clark
cpaul#xybion.com                                                   Carl A. Paul
bwilk_97#yahoo.com                                               Bill Wilkinson
clarkcox3#yahoo.com                                           Clark S. Cox, III
jeffduska#yahoo.com                                 Jeff Duska (Jeff#duska.com)
jurus#cs.cas.cz                                                     Pavel Jurus
v.slavik#volny.cz                                                 Vaclav Slavik
axel#aroxis.de                                                   Axel Schlueter
gerhard#bigfoot.de                                              Gerhard Haering
Gerhard.Gruber#coconet.de                                        Gerhard Gruber
Armin#exklusiv.de                                   Armin K=F6nigsdorfer</FONT>
benndorf#pb.izm.fhg.de                                             Kai Benndorf
Wolfgang.Strobl#gmd.de                                          Wolfgang Strobl
amanjit.gill#gmx.de                                                Amanjit Gill
andorxor#gmx.de                                               Stephan Tolksdorf
bob.g#gmx.de                                                        Robert Gurk
carsten.block#gmx.de                                              Carsten Block
Markus.Fieber#gmx.de                                 Markus Fieber</FONT></DIV>
SchaeferFFM#gmx.de                                              Oliver Schaefer
Thomas_Krebs#gmx.de                                                Thomas Krebs
w_keller#gmx.de                                                 Wolfgang Keller
a.larsen#identecsolutions.de                                      Anders Larsen
christof#pastors.de                                            Christof Pastors
svpeters#t-online.de                                                Sven Peters
hsauer#marnie.teuto.de                                            Henning Sauer
manfred.heumann#uni-bielefeld.de                                  Manni Heumann
Wolfram.Gloger#dent.med.uni-muenchen.de                          Wolfram Gloger
hbachmann#vellum.de                                              Horst Bachmann
Ruediger.Maehl#web.de                                            Ruediger Maehl
bouvin#daimi.au.dk                                            Niels Olof Bouvin
dpw#CS.Arizona.EDU                                                 Don Waugaman
fungus#OCF.Berkeley.EDU                                               Hank Fung
bkropf#ONLINE.EMICH.EDU                                             Brian Kropf
kjcole#gri.gallaudet.edu                                             Kevin Cole
steriana#claymore.engineer.gvsu.edu                              Andrew Sterian
wsryu#fas.harvard.edu                                               William Ryu
jrl1#lehigh.edu                                                            John
jmbowman#alum.mit.edu                                             Jeremy Bowman
amcmicha#Princeton.EDU                                         Andrew McMichael
dominicb#cvs.rochester.edu                                  Dominic Barraclough
matt#nephi.tamu.edu                                          Matthew W. Roberts
stabler#ucla.edu                                                      E Stabler
jonathan.gilligan#vanderbilt.edu                           Jonathan M. Gilligan
jeff#rufinus.cs.widener.edu                                           J Rufinus
clee#gnwy100.wuh.wustl.edu                                      christopher lee
paul.magwene#yale.edu                                              Paul Magwene
guille#iies.es                                       Guillermo Rodriguez Garcia
irm#nexo.es                                                        I.R.Maturana
pedraza#datsi.fi.upm.es                                         Jose L. Pedraza
ts#UWasa.Fi                                                          Timo Salmi
zeitlin#dptmaths.ens-cachan.fr                                    Vadim Zeitlin
af#eppra.polytechnique.fr                                          Anton Frolov
Gilles.Depeyrot#wanadoo.fr                                      Gilles Depeyrot
leonard.williams1#jsc.nasa.gov                                 Leonard Williams
newtons#gofree.indigo.ie                                            Greg Newton
dfrost#maths.tcd.ie                                                Dermot Frost
mau#beatles.cselt.it                                           Maurizio Codogno
mtbros#libero.it                                                     Mario Tola
vtm#libero.it                                                        Lello Mele
Nicola.Musatti#ObjectWay.it                                      Nicola Musatti
mbarbon#dsi.unive.it                                              Mattia Barbon
rol9999#attglobal.net                                          Roland Schlenker
greglandrum#earthlink.net                                          Greg Landrum
projectobjects#earthlink.net                                          Earthlink
rhkramer#fast.net                                                  Randy Kramer
illume#gmx.net                                                  Paul C. Kunysch
alexis.manning#gte.net                                           Alexis Manning
cbarker#jps.net                                              Christopher Barker
derek#leewo.net                                              Derek M. A. Lee-Wo
agold1#mediaone.net                                          Arthur  Goldhammer
franke#meso.net                                                   Stefan Franke
kneecap#netnet.net                                                  Neal Cappel
dcfoxmail#netscape.net                                                David Fox
votefraud#pyrophore.ogoense.net                                             UVV
schaffer#optonline.net                                             Les Schaffer
randy#spoke.net                                                 Randall F. Kern
skrenta#textport.net                                               Rich Skrenta
dogusanh#tr.net                                                   Hakki Dogusan
rhet#turnbull.net                                                 Rhet Turnbull
reed#zerohour.net                               Reed Hedges <reed#zerohour.net>
rcamesz#dds.nl                                                     Robert Amesz
scarblac#pino.selwerd.nl                                          Remco Gerlich
joukj#hrem.stm.tudelft.nl                                              J.Jansen
a.t.hofkamp#tue.nl                                               Albert Hofkamp
p.g.m.vandermeulen#student.utwente.nl                     Pieter van der Meulen
A.H.vandenBoogaard#wb.utwente.nl                          A.H. van den Boogaard
faulds#es.co.nz                                            Stuart Brodie Faulds
BgPorter#acm.org                                                 Brett g Porter
mbickel#asc-hq.org                                                Martin Bickel
tim.hochberg#ieee.org                                          Timothy Hochberg
richard#magicality.org                                           Richard Palmer
kedziers#mml.ch.pwr.wroc.pl                                    Pawel Kedzierski
orlov#diasoft.ru                                                     Oleg Orlov
avv#quasar.ipa.nw.ru                                        Alexander V. Voinov
phd#mail2.phd.pp.ru                                              Oleg Broytmann
federico#carmen.se                                           Federico Hernandez
jonas#flintv210.sn.umu.se                                         Jonas Rydberg
zdravko.bas#iskratr.si                                              Zdravko Bas
paul.gammans#brunel.ac.uk                                          Paul Gammans
dag1000#eng.cam.ac.uk                                            Diana Galletly
jpmg#eng.cam.ac.uk                                              Patrick Gosling
biol75#york.ac.uk                                                 chris elliott
jdg#diogenes.sacramento.ca.us                                   John David Galt

Voted NO
-------------------------------------------------------------------------------
klaas#cs.dal.ca                                                   Michael Klaas
LMcmil2586#aol.com                                Jeremy McMillan</FONT></HTML>
kimdv#best.com                                                     Kim DeVaughn
stainles#bga.com                                                   Dwight Brown
JEckart#mail.com                                                    John Eckart
squid#panix.com                                                        Yeoh Yiu
naddy#mips.inka.de                                         Christian Weisgerber
meg#Steam.Stanford.EDU                                               meg worley
rick#bcm.tmc.edu                                               Richard H Miller
dlganger#earthlink.net                                          Devin L. Ganger
mmontcha#OregonVOS.net                                       Matthew Montchalin
chriseb#ukshells.co.uk                                           Chris Ebenezer
jmulder#mighty.co.za                                              Jeritt Mulder

Abstained
-------------------------------------------------------------------------------
mmrosa#bigpond.com                                                  Michal Rosa
barkjr#home.com                                                        BarkerJr
aahz#pobox.com                                                             Aahz

Invalid ballots
-------------------------------------------------------------------------------
pttlapinblanc#chez.com                                         Stephane Junique
    ! Invalid email address
someone#somewhere.com                                              darren smith
    ! Invalid email address
meow#best.com                                                         Meow Meow
    ! No vote statement in message


To restore the email addresses above, pipe the ack list through the
following command:

    sed -e 's/#/@/g'

-- 
Voting question & problems: Dave Cornejo <dave at dogwood.com>
Voting address: vote at dogwood.com


From doug at postsmart.net  Sat Feb  3 07:00:34 2001
From: doug at postsmart.net (Doug Ball)
Date: Sat, 3 Feb 2001 00:00:34 -0600
Subject: Python/C++ for graphics vs. other languages?
References: <mailman.980225485.12027.python-list@python.org> <200101231452.IAA15405@starbase.neosoft.com>
Message-ID: <001301c08da6$9fffc180$8a00000a@darwin.net>

Cameron Laird Wrote:

> Yes.  Everything Mr. Wilson has written is
> correct and applicable.
>
> I want to add a few remarks.  Dylan, Common
> Lisp, and OCaML are great languages.  You
> might have specific requirements that make
> any of these ideal.  Part of our confidence
> in urging Python on you, though, is that even
> for the situations where, say, OCaML is per-
> fect, a well-crafted Python solution is likely
> to lie only a small multiple away.
>
> "... [A] potentially large graphics intensive
> program that will be speed sensitive ..."
> covers a great deal of territory.  We can give
> more precise advice with more precise know-
> ledge, of course.
>
.
[snip]
.
>It's impos-
> sible to know whether you're in the category
> of projects where there's a significant differ-
> ence, in the absence of more details.
> --
>


Thank you to all that have responded to my question pertaining to the
suitability of programming in a combination of  Python and C++ for graphical
applications. Your guidance is insightful and I am very grateful.

It has been pointed out that I have not provided enough information about my
application to allow a proper assessment.  Please forgive me, I have been
unintentionally vague while I trying to be concise = )

I am working on a real-time 3d simulation of a medical procedure, to run on
an Athlon 1.2 GHz machine under windows 2000.  Probably using DirectX as an
interface ( I don't know any OpenGL ).  No force feed back or tactile
simulation will be implemented.

Thank you again to all who have responded, and if this information enables
you to offer any further guidance I would be very grateful thanks!

Doug








From doug at postsmart.net  Sat Feb  3 18:30:14 2001
From: doug at postsmart.net (Doug Ball)
Date: Sat, 3 Feb 2001 11:30:14 -0600
Subject: Python/C++ for graphics vs. other languages?
References: <mailman.980225485.12027.python-list@python.org> <200101231452.IAA15405@starbase.neosoft.com> <mailman.981180244.14710.python-list@python.org> <63bo7t4qrf0oebbj6hnni9tjjdip8hg5u8@4ax.com>
Message-ID: <000501c08e06$f83a6b60$8a00000a@darwin.net>

les schaffer wrote in response to a prior posting:


> >I am working on a real-time 3d simulation of a medical procedure, to run
on
> >an Athlon 1.2 GHz machine under windows 2000.
>
> this sounds interesting. are you at liberty to divulge more details of
> the project?


Thanks for your interest!

We will be working on an interactive simulation of
electromyographically-guided Botulinum toxin injection.  In this procedure,
the practitioner places a needle into a muscle, generally guided by surface
landmarks, then more specifically guided by the electrical signals produced
when the muscle contracts (summated compound action potentials).  Once the
needle is placed, Botulinum toxin is injected, subsequently blocking the
chemical signals than are responsible for controlling muscle contraction (it
blocks the presynaptic vesicle from binding to the presynaptic membrane).
This procedure is typically used for patients with severe muscle spasticity.

The simulation user will have visual and auditory feedback from a 3D model
of the patient and from the EMG-machine monitor.  The user will indicate the
initial location of needle placement, then will request that the patient
model activates the target muscle. This action will result in a simulated
signal on the EMG-monitor (with both visual and auditory components) that
aids the user in further localizing the muscle, while adjusting  the needle
for optimal placement within the muscle.

Our goal is to develop this application for physician education.

Thanks again.

Doug





From nas at arctrix.com  Thu Feb  1 11:59:55 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Thu, 1 Feb 2001 02:59:55 -0800
Subject: Python vs Ruby
In-Reply-To: <959mv0$qke$1@nnrp1.deja.com>; from hansboehm@my-deja.com on Wed, Jan 31, 2001 at 06:53:24PM +0000
References: <20010129111820.C15924@harmony.cs.rit.edu> <mailman.980835459.24363.python-list@python.org> <959mv0$qke$1@nnrp1.deja.com>
Message-ID: <20010201025954.B24620@glacier.fnational.com>

On Wed, Jan 31, 2001 at 06:53:24PM +0000, hansboehm at my-deja.com wrote:
> [Neil Schemenauer]
> > RC also works quite well with data caches and VM managers.  Using
> > the Boehm-Demers GC with Python and disabling reference counting
> > slows the interpreter down by a significant amount.
>
> Are you sure the collector is working correctly, and not retaining
> garbage for some reason?  I looked at the source tree quickly, and it
> appeared to me that the interpreter maintained a doubly linked list of
> all Python objects?  That would effectively keep the collector from
> reclaiming anything.  Is that list essential?

That linked list is only used if the cyclic garbage collector is
enabled.  It has been quite a while since I played with your GC,
before the linked list stuff was added to Python. 

> When I last looked at this 4 or 5 years ago, there were some other
> issues that prevented a direct plug-in of a tracing collector.  I think
> some subsystems did their own free-list management.  This tends to be a
> pessimization with a tracing GC, especially if the objects on the free
> list still point to other long-dead objects, etc.  I couldn't find such
> issues now, but I didn't look very hard.

Those freelists are still there.  I don't doubt that I could have
made your GC perform better.  My tests were very un-scientific
and should not be given much weight.  Mark and sweep GC will
probably be considered for Python 3000 (at least I'm going to
look at it again).

> Reference counting tends to look considerably worse if you fully support
> concurrency, so that you have to update reference counts with atomic
> operations.

Right, that is one of the big problems with "free threading" the
current interpreter.  Currently the interpreter is protected by
one big lock.  Again, this is probably something that will be
considered in the Python 3000 design.

Thanks for your comments.

  Neil



From rtrocca at libero.it  Thu Feb  1 20:12:05 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Thu, 01 Feb 2001 19:12:05 GMT
Subject: what GUI, please advise
References: <3A6F0B27.A88D11C9@crepido.com> <756399CF731B46D9.57530C0D0C9C2171.472E737DF76189F7@lp.airnews.net> <955na0$m4r2@news.qualitynet.net>
Message-ID: <1eo5bx2.1jcwqo1dvno74N%rtrocca@libero.it>

bashar A. Asad <baasad at qualitynet.net> wrote:

>
> >>There seems to be several good options to choose from when
> >>it comes to GUI developent with Python.
> >>
> >>Can someone say somthing, without killing each other with flames, what
> >>GUI one should choose?
> >>
> >>Stupid and impossible question by the way as everything depends as
> >>usual... but anyway..
 Personally I've tried tkinter and pygtk.
I really like pygtk, also because I can create the GUI with glade, save
it in a XML file and load it on the fly. It is possible to modify the
look and feel without rewriting your code.
Anyway I've just started using it.

Riccardo


From pablo.prieto at dulcesol.es  Tue Feb 13 11:42:05 2001
From: pablo.prieto at dulcesol.es (Pablo Prieto)
Date: Tue, 13 Feb 2001 10:42:05 -0000
Subject: Python and postgreSQL ??
Message-ID: <009401c095a9$a54e7d10$150310ac@pablo>

> You may also wish to consider the other postgresql/python interface
> package, PoPy:  http://popy.sourceforge.net
>
> I find it quite useful.

Hi!
I'd like to connect with PostgreSQL via Python. The program is in Windows
(now), but will be executed, in few months, in Linux. I know I can use ODBC,
but is there a Library or so that I can use in both OS', by meant to make
the code 100% portable?

I've downloaded popy, but looks like only for Linux env.

Thanks in advance

Yours sincerely,

Pablo Prieto




From stephen_purcell at yahoo.com  Tue Feb 13 10:54:55 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 13 Feb 2001 10:54:55 +0100
Subject: Python and postgreSQL ??
In-Reply-To: <009401c095a9$a54e7d10$150310ac@pablo>; from pablo.prieto@dulcesol.es on Tue, Feb 13, 2001 at 10:42:05AM -0000
References: <009401c095a9$a54e7d10$150310ac@pablo>
Message-ID: <20010213105455.A7121@freedom.puma-ag.com>

Pablo Prieto wrote:
> I'd like to connect with PostgreSQL via Python. The program is in Windows
> (now), but will be executed, in few months, in Linux. I know I can use ODBC,
> but is there a Library or so that I can use in both OS', by meant to make
> the code 100% portable?

There's a note about PyGreSQL on Windows at FAQTS.com:-

  http://www.faqts.com/knowledge_base/view.phtml/aid/1420/fid/278

-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From gerhard.nospam at bigfoot.de  Wed Feb 14 11:00:34 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Wed, 14 Feb 2001 11:00:34 +0100
Subject: Python and postgreSQL ??
References: <009401c095a9$a54e7d10$150310ac@pablo> <mailman.982058113.12720.python-list@python.org>
Message-ID: <slrn98kmll.1bu.gerhard.nospam@lilith.hqd-internal>

On Tue, 13 Feb 2001 10:54:55 +0100, Steve Purcell wrote:
>Pablo Prieto wrote:
>> I'd like to connect with PostgreSQL via Python. The program is in Windows
>> (now), but will be executed, in few months, in Linux. I know I can use ODBC,
>> but is there a Library or so that I can use in both OS', by meant to make
>> the code 100% portable?
>
>There's a note about PyGreSQL on Windows at FAQTS.com:-
>
>  http://www.faqts.com/knowledge_base/view.phtml/aid/1420/fid/278

PyGreSQL is the only one from the 3 Postgres Python interfaces available
that I could build on Windows. The other two were too unixoid. And it's also
the original, which is included in the Postgres distribution. The PyGreSQL
web site at druid.net is currently only showing the default Apache docs. I
hope they will be back soon.

In the meantime you can still use the Windows version from my homepage (it's
a superset of the source distribution with a few patches from my part).

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de
web:    http://highqualdev.com


From darcy at vex.net  Thu Feb 15 22:40:54 2001
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: Thu, 15 Feb 2001 21:40:54 +0000 (UTC)
Subject: Python and postgreSQL ??
References: <009401c095a9$a54e7d10$150310ac@pablo> <mailman.982058113.12720.python-list@python.org> <slrn98kmll.1bu.gerhard.nospam@lilith.hqd-internal>
Message-ID: <96hid6$2v4r$1@news.tht.net>

Gerhard H?ring <gerhard.nospam at bigfoot.de> wrote:
> On Tue, 13 Feb 2001 10:54:55 +0100, Steve Purcell wrote:
>>Pablo Prieto wrote:
>>> I'd like to connect with PostgreSQL via Python. The program is in Windows
>>> (now), but will be executed, in few months, in Linux. I know I can use ODBC,
>>> but is there a Library or so that I can use in both OS', by meant to make
>>> the code 100% portable?
>>
>>There's a note about PyGreSQL on Windows at FAQTS.com:-
>>
>>  http://www.faqts.com/knowledge_base/view.phtml/aid/1420/fid/278

That page acts funny.  I eventually found the page that the link there
is supposed to send you to.  http://home.t-online.de/home/err666/.

> PyGreSQL is the only one from the 3 Postgres Python interfaces available
> that I could build on Windows. The other two were too unixoid. And it's also
> the original, which is included in the Postgres distribution. The PyGreSQL
> web site at druid.net is currently only showing the default Apache docs. I
> hope they will be back soon.

Back now.  I just moved druid.net to its new home and you caught me in
transition.

> In the meantime you can still use the Windows version from my homepage (it's
> a superset of the source distribution with a few patches from my part).

Do I have all the changes?  I'm sure that I could fold them in with #ifdefs
at least.  Didn't I already make some changes for Windows?

BTW I will soon be making the source available by anonymous CVS.

-- 
D'Arcy J.M. Cain <darcy at vex.net>           |  Democracy is three wolves
http://www.vex.net/                        |  and a sheep voting on         
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.


From greg at cosc.canterbury.ac.nz  Thu Feb  1 01:32:13 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 01 Feb 2001 13:32:13 +1300
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org>
Message-ID: <3A78AE8D.C9D25694@cosc.canterbury.ac.nz>

Sean Reifschneider wrote:
> 
> It didn't really
> seem like they were missing so much as it seemed like they would act
> more like everything else now...

Everything? What about lists, dicts, tuples, file
objects...?

What's so special about functions that we need to
be able to plonk arbitrary attributes on them, but
not any other builtin types?

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From mlh at idi.ntnu.no  Thu Feb  1 02:02:59 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 1 Feb 2001 02:02:59 +0100
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz>
Message-ID: <95acpj$sbe$1@tyfon.itea.ntnu.no>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
news:3A78AE8D.C9D25694 at cosc.canterbury.ac.nz...
> Sean Reifschneider wrote:
> >
> > It didn't really
> > seem like they were missing so much as it seemed like they would act
> > more like everything else now...
>
> Everything? What about lists, dicts, tuples, file
> objects...?
>
> What's so special about functions that we need to
> be able to plonk arbitrary attributes on them, but
> not any other builtin types?

Now *that* would be nice (if we could do that, I mean).
Everything as objects... SmallTalk lurking in the
background. Sounds good to me...

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From katz at Glue.umd.edu  Thu Feb  1 04:46:41 2001
From: katz at Glue.umd.edu (Roy Katz)
Date: Wed, 31 Jan 2001 22:46:41 -0500
Subject: function attributes are like function objects
In-Reply-To: <95acpj$sbe$1@tyfon.itea.ntnu.no>
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz> <95acpj$sbe$1@tyfon.itea.ntnu.no>
Message-ID: <Pine.GSO.4.21.0101312232120.6444-100000@z.glue.umd.edu>

Oh, good, perhaps a calll for a Py3k suggestion?

idea:

Perhaps replace def/class with a general 'def' which would
allow the definition of 
 
   - a standard class
   - a function
   - an instance method
   - a class method
   - a singleton class

all through a single, consistent syntax. 
lists, strings, dicts, ints and other types would be subclass-able. 

This would solve the following current problems:

   - Python lacks class methods 
   - Current Python has no support for subclassing from primitive
     types. 
    
I really would like that -- provided it is consistent and not
occam's-razor syntax extension. 

I had an idea where one could write

  def f(x):   pass
  def C:
    def f(self):  pass
    def g():      pass  # where some new syntax indicates that g is static

    print 'hi'          # C is like a function and a class in one. 

and you could call f(3), C.g(), C() to call C as a function.. 
that'd fill those requirements... sort of. 


On the other hand, maybe there needs to be a distinct separation between
class and function;  I think right now that function attributes are a
waste when the functions can use (1) global values,
(2) 'call-by-reference' (er, mutable, right?) object parameters in which
to hold values, or (3) be implemented as function objects, with all being
viable alternatives.  


(1)  

    state = 123

    def f(): 
      return state

(2) 

     class stateHolder:
       state = 0
       def __init__(self, state): self.state = state

     state = stateHolder( 123 )

     def f( s )    # s is a stateHolder
        return s.state 


(3) 

     class func:
        state = 123
        def f(self):
            return self.state

-------


there.  I prefer (3), (I'm CS-oriented :))


Roey
 



On Thu, 1 Feb 2001, Magnus Lie Hetland wrote:

> "Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
> news:3A78AE8D.C9D25694 at cosc.canterbury.ac.nz...
> > Sean Reifschneider wrote:
> > >
> > > It didn't really
> > > seem like they were missing so much as it seemed like they would act
> > > more like everything else now...
> >
> > Everything? What about lists, dicts, tuples, file
> > objects...?
> >
> > What's so special about functions that we need to
> > be able to plonk arbitrary attributes on them, but
> > not any other builtin types?
> 
> Now *that* would be nice (if we could do that, I mean).
> Everything as objects... SmallTalk lurking in the
> background. Sounds good to me...
> 
> --
> 
>   Magnus Lie Hetland      (magnus at hetland dot org)
> 
>  "Reality is what refuses to disappear when you stop
>   believing in it"                 -- Philip K. Dick
> 
> 
> 
> 



From sholden at holdenweb.com  Fri Feb  2 22:13:00 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 2 Feb 2001 16:13:00 -0500
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz> <95acpj$sbe$1@tyfon.itea.ntnu.no>
Message-ID: <stFe6.7032$o5.45454@e420r-atl1.usenetserver.com>

"Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
news:95acpj$sbe$1 at tyfon.itea.ntnu.no...
> "Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote in message
> news:3A78AE8D.C9D25694 at cosc.canterbury.ac.nz...
> > Sean Reifschneider wrote:
> > >
> > > It didn't really
> > > seem like they were missing so much as it seemed like they would act
> > > more like everything else now...
> >
> > Everything? What about lists, dicts, tuples, file
> > objects...?
> >
> > What's so special about functions that we need to
> > be able to plonk arbitrary attributes on them, but
> > not any other builtin types?
>
> Now *that* would be nice (if we could do that, I mean).
> Everything as objects... SmallTalk lurking in the
> background. Sounds good to me...
>
Sounds more like Javascript lurking in the background to me.

regards
 Steve



From mwh21 at cam.ac.uk  Thu Feb  1 13:26:50 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Feb 2001 12:26:50 +0000
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz>
Message-ID: <m3g0hy7fn9.fsf@atrus.jesus.cam.ac.uk>

Greg Ewing <greg at cosc.canterbury.ac.nz> writes:

> Sean Reifschneider wrote:
> > 
> > It didn't really
> > seem like they were missing so much as it seemed like they would act
> > more like everything else now...
> 
> Everything? What about lists, dicts, tuples, file
> objects...?
> 
> What's so special about functions that we need to
> be able to plonk arbitrary attributes on them, but
> not any other builtin types?

The fact that people were using the one attribute they could get at
(__doc__) for things far from its original purpose?

Cheers,
M.

-- 
  The use of COBOL cripples the mind; its teaching should, therefore,
  be regarded as a criminal offence.
           -- Edsger W. Dijkstra, SIGPLAN Notices, Volume 17, Number 5


From donn at u.washington.edu  Thu Feb  1 18:05:01 2001
From: donn at u.washington.edu (Donn Cave)
Date: 1 Feb 2001 17:05:01 GMT
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz> <m3g0hy7fn9.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <95c4vt$g0i$1@nntp6.u.washington.edu>

Quoth Michael Hudson <mwh21 at cam.ac.uk>:
| Greg Ewing <greg at cosc.canterbury.ac.nz> writes:
|> Everything? What about lists, dicts, tuples, file
|> objects...?
|> 
|> What's so special about functions that we need to
|> be able to plonk arbitrary attributes on them, but
|> not any other builtin types?
|
| The fact that people were using the one attribute they could get at
| (__doc__) for things far from its original purpose?

No, that can't be it.  Programmers will abuse all kinds of
language features;  sometimes it is indeed in the service
of good, but just as often it's utterly misguided.

My curiosity is killing me.  How does function attributes
make Python better?

	Donn Cave, donn at u.washington.edu


From fredrik at effbot.org  Thu Feb  1 18:22:45 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 01 Feb 2001 17:22:45 GMT
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz> <m3g0hy7fn9.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <FXge6.10628$AH6.1747678@newsc.telia.net>

Michael Hudson wrote:
> > What's so special about functions that we need to
> > be able to plonk arbitrary attributes on them, but
> > not any other builtin types?
>
> The fact that people were using the one attribute they could get at
> (__doc__) for things far from its original purpose?

well, if people are too lazy to write

    def f(...):
        ...
    a[f] = "something"

do you really think they'll find it much easier to write:

    def f(...):
        ...
    f.a = "something"

(fwiw, I think function attributes is about the dumbest
thing we've added to python lately -- face it, "because
we can" isn't a very good argument when deciding what
to add to the language core...)

Cheers /F




From bobhicks at adelphia.net  Thu Feb  1 19:27:14 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Thu, 01 Feb 2001 18:27:14 GMT
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz> <m3g0hy7fn9.fsf@atrus.jesus.cam.ac.uk> <FXge6.10628$AH6.1747678@newsc.telia.net>
Message-ID: <B69F14A9.9B8D%bobhicks@adelphia.net>


> From: "Fredrik Lundh" <fredrik at effbot.org>
> Organization: Telia Internet
> Newsgroups: comp.lang.python
> Date: Thu, 01 Feb 2001 17:22:45 GMT
> Subject: Re: function attributes are like function objects
> 
> Michael Hudson wrote:
>>> What's so special about functions that we need to
>>> be able to plonk arbitrary attributes on them, but
>>> not any other builtin types?
>> 
>> The fact that people were using the one attribute they could get at
>> (__doc__) for things far from its original purpose?
> 
> well, if people are too lazy to write
> 
> def f(...):
> ...
> a[f] = "something"
> 
> do you really think they'll find it much easier to write:
> 
> def f(...):
> ...
> f.a = "something"
> 
> (fwiw, I think function attributes is about the dumbest
> thing we've added to python lately -- face it, "because
> we can" isn't a very good argument when deciding what
> to add to the language core...)

Was that the rational? Let's add it because we can? 



From tim.one at home.com  Fri Feb  2 06:02:06 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 2 Feb 2001 00:02:06 -0500
Subject: function attributes are like function objects
In-Reply-To: <FXge6.10628$AH6.1747678@newsc.telia.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEELGIMAA.tim.one@home.com>

[/F]
> well, if people are too lazy to write
>
>     def f(...):
>        ...
>     a[f] = "something"
>
> do you really think they'll find it much easier to write:
>
>     def f(...):
>         ...
>     f.a = "something"

Believe it or not, I do.  Bets?  Don't forget methods when deciding how easy
the first alternative is, and that decorating methods (not top-level
functions) is the primary cause of docstring abuse.  But I bet people would
find the latter substantially easier even without methods complicating the
first gimmick.

> (fwiw, I think function attributes is about the dumbest
> thing we've added to python lately

I happen to like sre better myself <wink>.

> -- face it, "because we can" isn't a very good argument
> when deciding what to add to the language core...)

Hmm.  I suppose the PEP would be that much stronger had Barry thought to
mention this argument <ahem>:

    http://python.sourceforge.net/peps/pep-0232.html




From dieter at handshake.de  Fri Feb  2 21:48:22 2001
From: dieter at handshake.de (Dieter Maurer)
Date: 02 Feb 2001 21:48:22 +0100
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz> <m3g0hy7fn9.fsf@atrus.jesus.cam.ac.uk> <FXge6.10628$AH6.1747678@newsc.telia.net>
Message-ID: <x7bsskvmjt.fsf@lindm.dm>

"Fredrik Lundh" <fredrik at effbot.org> writes on Thu, 01 Feb 2001 17:22:45 GMT:
> (fwiw, I think function attributes is about the dumbest
> thing we've added to python lately -- face it, "because
> we can" isn't a very good argument when deciding what
> to add to the language core...)
Zope makes quite good use of function attributes:

  It uses them to associate permissions, for example.

  It is better to have them directly attached to the
  function rather than maintained in a separate object, such
  as a dictionary. For, if the function is passed around
  the program, its attributes are automatically passed
  with it. Otherwise, the auxiliary object would need
  to be passed explicitly.


Dieter


From fredrik at effbot.org  Fri Feb  2 10:30:32 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 09:30:32 GMT
Subject: function attributes are like function objects
References: <mailman.981098831.20002.python-list@python.org>
Message-ID: <Y6ve6.11078$AH6.1774108@newsc.telia.net>

Tim Peters wrote:
> [/F]
> > well, if people are too lazy to write
> >
> >     def f(...):
> >        ...
> >     a[f] = "something"
> >
> > do you really think they'll find it much easier to write:
> >
> >     def f(...):
> >         ...
> >     f.a = "something"
>
> Believe it or not, I do.  Bets?  Don't forget methods when deciding how easy
> the first alternative is, and that decorating methods (not top-level
> functions) is the primary cause of docstring abuse.

Works for methods too:

    class Spam:
        a = {}
        def f(...):
            ...
        a[f] = "something"

You can dig it out with a one-liner (or 2-3 lines, if you want a
more general version -- put it in the "code" module, where it
belongs...)

> > -- face it, "because we can" isn't a very good argument
> > when deciding what to add to the language core...)
>
> Hmm.  I suppose the PEP would be that much stronger had Barry thought to
> mention this argument <ahem>:
>
>     http://python.sourceforge.net/peps/pep-0232.html

Sure looks like "because we can" to me...

Or maybe "because it's more fun to hack the core than to hack
the library".  (For some reason, the python-devers seem to prefer
writing C/Java over Python...)

On the other hand, this isn't such a big deal, especially not com-
pared to some of the stuff I've seen on python-dev lately.  I'm
pretty sure a third-party python implementer can ignore function
attributes, and nobody will ever notice...

Cheers /F




From tim.one at home.com  Sat Feb  3 08:38:55 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 02:38:55 -0500
Subject: function attributes are like function objects
In-Reply-To: <Y6ve6.11078$AH6.1774108@newsc.telia.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDHINAA.tim.one@home.com>

[/F]
> well, if people are too lazy to write
>
>     def f(...):
>        ...
>     a[f] = "something"
>
> do you really think they'll find it much easier to write:
>
>     def f(...):
>         ...
>     f.a = "something"

[Tim]
> Believe it or not, I do.  Bets?  Don't forget methods when
> deciding how easy the first alternative is, and that decorating
> methods (not top-level functions) is the primary cause of
> docstring abuse.

[/F]
> Works for methods too:
>
>     class Spam:
>         a = {}
>         def f(...):
>             ...
>         a[f] = "something"

Now, now, let's be serious:

> You can dig it out with a one-liner

Yes, and if you *showed* that one-liner, the "easier" would be shot all to
pieces.  These aren't meant to be write-only attributes <wink>.

> (or 2-3 lines, if you want a more general version

Or, under Barry's patch, simply

    C.f.a

> -- put it in the "code" module, where it belongs...)

Na, C.f.a is too simple to hide in "code" <wink>.

> ...
> Sure looks like "because we can" to me...

Well, my employer is famous for using the presence or absence of a docstring
to determine whether a method "is publishable".  It's handier and cleaner to
have an attribute for that instead.  You want to use a dict, and then there
are multiple dicts, and then multiple "conceptual" attributes get
distributed in an unhelpful way.  Like the trick of hiding a function in a
class, it's an alternative that's been available for years, and people
simply won't do it.

Heading on 50 years ago, early LISP systems had the radical idea that you
could "attach" any number of key+value pairs to any symbol, via setprop,
getprop, remprop functions.  Python has objects instead of symbols, and
dicts instead of association lists, and attr notation instead of functional,
but other than those it's the same ancient idea (except still less general
in Python!).  Nothing bad happened to LISP as a result, and it was handy for
all sorts of things.

> Or maybe "because it's more fun to hack the core than to hack
> the library".  (For some reason, the python-devers seem to prefer
> writing C/Java over Python...)

And writing code rather than docs; and anything rather than volunteer to
investigate a bug report.  They're developers <0.6 wink>.

> On the other hand, this isn't such a big deal, especially not com-
> pared to some of the stuff I've seen on python-dev lately.  I'm
> pretty sure a third-party python implementer can ignore function
> attributes, and nobody will ever notice...

We would certainly notice the absence of function.__doc__, and any 3rd-party
implementer realizing that first would likely take a much more uniform
approach to implementing object attributes than did Guido at the start.
Remember that func.__doc__ didn't exist at the start, so got backstitched in
as an afterthought.  One hopes that later implementers might even learn from
that <wink>.

hope-springs-eternal-ly y'rs  - tim




From geoff at homegain.com  Fri Feb  2 20:31:00 2001
From: geoff at homegain.com (Geoffrey Gerrietts)
Date: Fri, 2 Feb 2001 11:31:00 -0800 
Subject: function attributes are like function objects
Message-ID: <393D961EA4FCD311B03F00D0B746D65802625F21@HQSXCH01>

I'm pretty new to the discussion, and definitely underinformed,
but I have a question or two.

Has the addition of function attributes included the addition of an
introspection mechanism to allow the function access to its own
attributes? In other words, can a function use its attributes to
provide the moral equivalent of a static variable, or a closure?

Or are function attributes pretty much accessible only to those
who know the function by name?

Thanks,
--G.

---
Geoff Gerrietts <geoff at homegain.com>
Software Engineer, HomeGain.com
510-655-0800 x4320



From barry at digicool.com  Fri Feb  2 23:41:04 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Fri, 2 Feb 2001 17:41:04 -0500
Subject: function attributes are like function objects
References: <393D961EA4FCD311B03F00D0B746D65802625F21@HQSXCH01>
Message-ID: <14971.14208.294333.479174@anthem.wooz.org>

>>>>> "GG" == Geoffrey Gerrietts <geoff at homegain.com> writes:

    GG> Or are function attributes pretty much accessible only to
    GG> those who know the function by name?

Well, if you have the function passed to you, then you don't need to
know what it's called.  You just access the attribute.

def publish(func):
    try:
	if func.publish:
	    print func()
    except AttributeError:
        log('Non-conforming function given to publish(): %s' % func)

-Barry



From fredrik at effbot.org  Fri Feb  2 21:54:57 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 20:54:57 GMT
Subject: function attributes are like function objects
References: <mailman.981142385.2222.python-list@python.org>
Message-ID: <B8Fe6.11158$Qb7.1631992@newsb.telia.net>

Geoffrey Gerrietts wrote:
> Has the addition of function attributes included the addition of an
> introspection mechanism to allow the function access to its own
> attributes? In other words, can a function use its attributes to
> provide the moral equivalent of a static variable, or a closure?

nope.

> Or are function attributes pretty much accessible only to those
> who know the function by name?

exactly.

on the other hand, the programmer defining a function
often knows what it's called, so things like the following
will work (if spam is defined in the global namespace,
and as long as nobody else overloads spam with some-
thing else):

    def spam(x):
        # fetch attribute via global namespace
        # lookup and attribute lookup
        print spam.x
    spam.x = "hello"

but if that's what you want, you can of course use
a global instead:

    def spam(x):
        # fetch attribute via global namespace lookup
        print spam_x
    spam_x = "hello"

Cheers /F




From geoff at homegain.com  Sat Feb  3 00:31:22 2001
From: geoff at homegain.com (Geoffrey Gerrietts)
Date: Fri, 2 Feb 2001 15:31:22 -0800 
Subject: function attributes are like function objects
Message-ID: <393D961EA4FCD311B03F00D0B746D65802625F24@HQSXCH01>

Barry writes:
> >>>>> "GG" == Geoffrey Gerrietts <geoff at homegain.com> writes:
> 
>     GG> Or are function attributes pretty much accessible only to
>     GG> those who know the function by name?
> 
> Well, if you have the function passed to you, then you don't need to
> know what it's called.  You just access the attribute.

Yeah, well, right, so you don't know it by its birth name, but you
still know it by the name you gave it, just like you call me GG.

Not to pick nits, just that the question is about introspection, 
not about whether regular namespace semantics work, or whether 
our functions are really first-class objects.

Thanks,
--G.

---
Geoff Gerrietts <geoff at homegain.com>
Software Engineer, HomeGain.com
510-655-0800 x4320



From barry at digicool.com  Sat Feb  3 04:32:37 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Fri, 2 Feb 2001 22:32:37 -0500
Subject: function attributes are like function objects
References: <393D961EA4FCD311B03F00D0B746D65802625F24@HQSXCH01>
Message-ID: <14971.31701.714377.397629@anthem.wooz.org>

>>>>> "GG" == Geoffrey Gerrietts <geoff at homegain.com> writes:

    GG> Yeah, well, right, so you don't know it by its birth name, but
    GG> you still know it by the name you gave it, just like you call
    GG> me GG.

In other words, you're accessing the object though the variable it's
locally bound to (nothing special there).

    GG> Not to pick nits, just that the question is about
    GG> introspection, not about whether regular namespace semantics
    GG> work, or whether our functions are really first-class objects.

Okay, so introspection works no differently for functions as any other
object with writable attributes (e.g. classes, instances, modules).
You can use getattr() to test whether a function has a particular
attribute, and setattr() to set that attribute.

But maybe I'm missing the point of the question?  It's been a long
couple of days.
-Barry



From aahz at panix.com  Sat Feb  3 07:58:14 2001
From: aahz at panix.com (Aahz Maruch)
Date: 2 Feb 2001 22:58:14 -0800
Subject: function attributes are like function objects
References: <393D961EA4FCD311B03F00D0B746D65802625F24@HQSXCH01> <mailman.981171622.11894.python-list@python.org>
Message-ID: <95ga66$nbj$1@panix2.panix.com>

In article <mailman.981171622.11894.python-list at python.org>,
Barry A. Warsaw <barry at digicool.com> wrote:
>>>>>> "GG" == Geoffrey Gerrietts <geoff at homegain.com> writes:
>
>    GG> Not to pick nits, just that the question is about
>    GG> introspection, not about whether regular namespace semantics
>    GG> work, or whether our functions are really first-class objects.
>
>Okay, so introspection works no differently for functions as any other
>object with writable attributes (e.g. classes, instances, modules).
>You can use getattr() to test whether a function has a particular
>attribute, and setattr() to set that attribute.

I think the issue is that Geoffrey suddenly realized that there's no
"self" equivalent for functions.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From geoff at homegain.com  Sat Feb  3 00:57:29 2001
From: geoff at homegain.com (Geoffrey Gerrietts)
Date: Fri, 2 Feb 2001 15:57:29 -0800 
Subject: function attributes are like function objects
Message-ID: <393D961EA4FCD311B03F00D0B746D65802625F26@HQSXCH01>

After I sent, I realized how smarmy that last message sounds.

I don't mean to imply that Barry's information was inappropriate. I can 
totally see how what I wrote could be read to indicate I didn't realize
that function objects can be aliased and passed around and yada yada.

I just want to avoid losing the point; is the function going to have 
access to this bit of state its carrying around, or is that reserved
for outside parties?

Sorry, Barry, I appreciate you trying to help. I should be more careful
about my tone.

Thanks,
--G.

---
Geoff Gerrietts <geoff at homegain.com>
Software Engineer, HomeGain.com
510-655-0800 x4320
 



From tim.one at home.com  Sat Feb  3 09:17:21 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 03:17:21 -0500
Subject: function attributes are like function objects
In-Reply-To: <393D961EA4FCD311B03F00D0B746D65802625F26@HQSXCH01>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEDJINAA.tim.one@home.com>

[Geoffrey Gerrietts]
> ...
> I just want to avoid losing the point; is the function going to have
> access to this bit of state its carrying around,

No.

> or is that reserved for outside parties?

Also no.  I know you don't want to hear about namespace issues, but those
*are* the issues here.  As Aahz said, there is no form of "self" for
function objects (or class objects, or file objects, or list objects, or
...).  The only way to get to the attr is via the object, for functions as
for everything else.  Whether you're "in" or "out" of the function makes no
difference to that.  Visibility of names is the entire story in both cases.

>>> def f():
...     print "f.a is", f.a
>>> f.a = 42
>>> f()
f.a is 42
>>

But if-- for whatever reason --the name of "f" is not visible inside the
body of "f", that won't work, and there's no special syntax then that can
make it work.  For example, run this under 2.1a2:

def f():
    def f():
        print "inner f.a is", f.a
    f.a = 666
    f()
    print "outer f.a is", f.a
f.a = 42
f()

Both "print" stmts see the inner redefinition of f then, so both references
to "f.a" yield 666.  But that's mondo perverse code!

don't-punch-yourself-in-the-throat-and-you-won't-gag-ly y'rs  - tim




From geoff at homegain.com  Sat Feb  3 09:39:21 2001
From: geoff at homegain.com (Geoffrey Gerrietts)
Date: Sat, 3 Feb 2001 00:39:21 -0800 
Subject: function attributes are like function objects
Message-ID: <393D961EA4FCD311B03F00D0B746D65802625F2A@HQSXCH01>

Aahz writes:
> I think the issue is that Geoffrey suddenly realized that there's no
> "self" equivalent for functions.

Not so suddenly as maybe it seems. I wasn't aware of any mechanism that
would help me out there; I just wondered whether one might be part of the
package. It seemed like an obvious corollary to me; if the function is going
to be carrying around this bit of state, it should probably have access to
it, sorta in the way that globals() lets you at a module's __dict__.

Fredrik has answered my question, though: there's nothing on the drawing
board that will let me at the attributes without first having an outside
handle on the object, and nothing that will give me an outside handle on it
from the inside. Maybe nested scopes will help that? Maybe not.

If function attributes make it easier for some folks, that's great. I 
thought maybe I saw a niche they might fill for me, but now I'm back to
wait-and-see; I'll wait for a final release before I adopt, anyway. :)

Thanks,
--G.




From fredrik at effbot.org  Sat Feb  3 13:40:58 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 12:40:58 GMT
Subject: function attributes are like function objects
References: <mailman.981185944.16489.python-list@python.org>
Message-ID: <u%Se6.11240$Qb7.1652423@newsb.telia.net>

Tim Peters wrote:
> Yes, and if you *showed* that one-liner, the "easier" would be shot all to
> pieces.  These aren't meant to be write-only attributes <wink>.

here's how I would solve your publishing example:

ALWAYS = "ALWAYS"
SOMETIMES = "SOMETIMES"

#
# user code

class ClassicSpam:

    publish = {}

    def egg(self, x):
        pass
    publish[egg] = ALWAYS
    # 2.1: egg.publish = ALWAYS

class NewSpam(ClassicSpam):

    publish = {}

    def bacon(self, x):
        pass
    publish[bacon] = SOMETIMES
    # 2.1: bacon.publish = SOMETIMES

#
# code for the publisher framework

def getpublish(klass, name):
    # figure out if/when we can publish this method
    try:
        method = getattr(klass, name)
        return method.im_class.publish[method.im_func]
        # 2.1: return method.publish
    except (AttributeError, KeyError):
        return None # nope

#
# try it out

for function in ["bacon", "egg", "spam"]:
    print function, getpublish(NewSpam, function)

advantages:
+ works with pre-2.1 python
+ requires one dictionary per class instead of one
  per method
+ no memory penalty for classes that doesn't belong
  to the publishing framework

disadvantages:
+ the 1-3 lines you need to dig out the attribute
  varies slightly, depending on the application

> > On the other hand, this isn't such a big deal, especially not com-
> > pared to some of the stuff I've seen on python-dev lately.  I'm
> > pretty sure a third-party python implementer can ignore function
> > attributes, and nobody will ever notice...
>
> We would certainly notice the absence of function.__doc__,
> and any 3rd-party implementer realizing that first would likely
> take a much more uniform approach to implementing object
>> attributes than did Guido at the start.

not so sure about that -- __doc__ strings are very common,
and are better stored in a separate slot to save memory (just
like CPython does, of course).

and I doubt many people would notice if __doc__ were made
readonly...

Cheers /F




From tanzer at swing.co.at  Sat Feb  3 14:27:07 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Sat, 03 Feb 2001 14:27:07 +0100
Subject: function attributes are like function objects 
In-Reply-To: Your message of "Sat, 03 Feb 2001 12:40:58 GMT."
             <u%Se6.11240$Qb7.1652423@newsb.telia.net> 
Message-ID: <m14P2iZ-000wcEC@swing.co.at>

"Fredrik Lundh" <fredrik at effbot.org> wrote:

> and I doubt many people would notice if __doc__ were made
> readonly...

Well, I'm not many but I certainly would notice (and complain, too).

> advantages:
> + works with pre-2.1 python
> + requires one dictionary per class instead of one
>   per method

- requires one dictionary per type of method attribute.

Worse, it is a FORTRANique [*] instead of a pythonic solution. Instead
of encapsulating all information about an object in one place,
corresponding attributes of different objects are bundled up together.
For a single method attribute, it doesn't matter much but if several
attributes are used per method it sucks.

[*] (read FORTRAN IV or FORTRAN 77) 

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From pieter at equinox.co.za  Mon Feb  5 12:15:52 2001
From: pieter at equinox.co.za (Pieter Nagel)
Date: Mon, 5 Feb 2001 13:15:52 +0200
Subject: function attributes are like function objects
In-Reply-To: <mailman.981185944.16489.python-list@python.org>
References: <mailman.981185944.16489.python-list@python.org>
Message-ID: <Pine.LNX.4.21.0102051314260.8491-100000@bast.jhb.equinox.co.za>

On Sat, 3 Feb 2001, Tim Peters wrote:

> Well, my employer is famous for using the presence or absence of a docstring
> to determine whether a method "is publishable".

We were sincerely hoping that the Python core team would teach their
employers how to code Python, instead of the other way around... 

}:-)

-- 
     ,_
     /_)              /| /
    /   i e t e r    / |/ a g e l



From dbrueck at edgix.com  Mon Feb  5 18:04:54 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Mon, 5 Feb 2001 10:04:54 -0700
Subject: function attributes are like function objects
In-Reply-To: <Pine.LNX.4.21.0102051314260.8491-100000@bast.jhb.equinox.co.za>
Message-ID: <BLEOJGEMBNBNEAKMIFDHOEIBPIAA.dbrueck@edgix.com>

> [mailto:python-list-admin at python.org]On Behalf Of Pieter Nagel
> > Well, my employer is famous for using the presence or absence 
> of a docstring
> > to determine whether a method "is publishable".
> 
> We were sincerely hoping that the Python core team would teach their
> employers how to code Python, instead of the other way around... 

Er, I read 'my employer' to mean Tim's boss, the BDFL himself.

-Dave



From fredrik at effbot.org  Sat Feb  3 15:36:07 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 14:36:07 GMT
Subject: function attributes are like function objects
References: <mailman.981207245.23192.python-list@python.org>
Message-ID: <rHUe6.11254$Qb7.1656310@newsb.telia.net>

Christian Tanzer wrote:
> > and I doubt many people would notice if __doc__ were made
> > readonly...
>
> Well, I'm not many but I certainly would notice (and complain, too).

Did you read the language specification before you started
abusing the __doc__ attribute?  If you look under function
objects, you'll notice that it says that the __doc__ attribute
MAY be writable.  Not that it always will be.

(the real reason to have writable func_doc/func_code attributes
is to allow a development environment to update functions in a
running program, not so much to support doc-string hacks in user
code).

And if you're relying on implementation-dependent features, you
only have yourself to blame...

> Worse, it is a FORTRANique [*] instead of a
> pythonic solution.

Really?  I didn't know Fortran had builtin dictionaries and hashable
function objects, or that it builds class objects by executing code
in a special namespace...

> For a single method attribute, it doesn't matter much but if several
> attributes are used per method it sucks.

Yeah, right -- we all know that writing two lines instead of
two lines suck...

</F>




From me at nospam.net  Wed Feb  7 06:23:29 2001
From: me at nospam.net (Scottie)
Date: Tue, 6 Feb 2001 21:23:29 -0800
Subject: function attributes are like function objects
References: <mailman.981207245.23192.python-list@python.org>
Message-ID: <t81mvi63vnl0bc@corp.supernews.com>

The dictionary (a in your example) will force a reference to the function to
exist, thus preventing it from being collected.  An attribute evaporates
with
the object it is connected to.  The difference provides another reason to
use
an attribute to perform the legerdemain rather than some (probably global)
dictionary.

On a separate note, can't a function acces itself through an exception's
 traceback frame?  Something like:
    try:
        raise MyExcept
    except MyExcept:
        obtain info thru via sys.exc_info()

"Christian Tanzer" <tanzer at swing.co.at> wrote:
> "Fredrik Lundh" <fredrik at effbot.org> wrote:
> > advantages:
> > + works with pre-2.1 python
> > + requires one dictionary per class instead of one
> >  per method
> - requires one dictionary per type of method attribute.
- prevents funtions referenced from being garbage collected.

--
Scott David Daniels          Scott.Daniels at Acm.Org






From fredrik at effbot.org  Wed Feb  7 08:19:23 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 07 Feb 2001 07:19:23 GMT
Subject: function attributes are like function objects
References: <mailman.981207245.23192.python-list@python.org> <t81mvi63vnl0bc@corp.supernews.com>
Message-ID: <%F6g6.13180$AH6.1981166@newsc.telia.net>

Scottie wrote:
> The dictionary (a in your example) will force a reference to the function to
> exist, thus preventing it from being collected.

That dictionary sits inside a *class* object.   That class
also holds references to the function objects, as well as
to the dictionary.  No cycles here.

> > - requires one dictionary per type of method attribute.

Which is a problem only if you have more attributes than
methods in your classes, of course.  Not very likely.

(and easy to solve with a global attribute dictionary)

> - prevents funtions referenced from being garbage collected.

Which is a problem only if you spend a lot of time adding and
removing methods from classes.  Not very likely.

(and easy to solve with a weak dictionary, or by changing
to a better design)

Cheers /F




From pieter at equinox.co.za  Tue Feb  6 14:45:50 2001
From: pieter at equinox.co.za (Pieter Nagel)
Date: Tue, 6 Feb 2001 15:45:50 +0200
Subject: function attributes are like function objects
In-Reply-To: <mailman.981392847.30947.python-list@python.org>
References: <mailman.981392847.30947.python-list@python.org>
Message-ID: <Pine.LNX.4.21.0102061544270.4343-100000@bast.jhb.equinox.co.za>

On Mon, 5 Feb 2001, Dave Brueck wrote:

> Pieter Nagel wrote:
> > Tim Peters wrote:

> > > Well, my employer is famous for using the presence or absence 
> > of a docstring
> > > to determine whether a method "is publishable".
> > 
> > We were sincerely hoping that the Python core team would teach their
> > employers how to code Python, instead of the other way around... 
> 
> Er, I read 'my employer' to mean Tim's boss, the BDFL himself.

I thought that referred to the perpetrators of Zope.

-- 
     ,_
     /_)              /| /
    /   i e t e r    / |/ a g e l



From tim.one at home.com  Thu Feb  1 07:55:46 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 01:55:46 -0500
Subject: invalid-token syntax hook (was Re: Hack request: rational numbers)
In-Reply-To: <20010129102836.B1279@inetnebr.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEIGIMAA.tim.one@home.com>

[Jeff Epler]
> ...
> I have no illusion that our BDFL would accept something like Mobius
> into core Python,

Hey, in the end he accepted xreadlines(), and you may remember his initial
patch comment for that <wink>.

Guido confessed to *trying* to dream up a Pythonic way to allow syntax
extensions when he was designing the language, but failed.  And despite what
anyone may think, that's really not an invitation to discuss hygienic macros
for the 50th time <0.7 wink>.

life's-short-and-brutal-but-if-you're-lucky-you-die-ly y'rs  - tim




From tim.one at home.com  Thu Feb  1 08:38:25 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 02:38:25 -0500
Subject: simple threads kill  my IDLE
In-Reply-To: <mpac6.1$yH6.175@news.world-online.no>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEIHIMAA.tim.one@home.com>

[posted & mailed]

[Henning VON ROSEN]
> While playing around width threads executing either interactively
> or from the editmode by "f5", in many cases IDLE froze and forced
> me to restart it.
>
> Have anyone experienced such problems?

Just everyone who has tried it <wink>.  Guido believes it would take a major
rearchitecture to get IDLE playing nice with threads (Tk isn't really happy
with threads).  An entry from the "feature request" PEP at

    http://python.sourceforge.net/peps/pep-0042.html:

"""
- IDLE has deep problems running threaded programs.  Re-architect.

http://sourceforge.net/bugs/?func=detailbug&bug_id=121963&group_id=5470
"""

> ...
> 3) Just execute in DOS Windows (If there is a bug, these ugly
>    windows dies off, leaving me without feddback error messages!
>    How do I solve that?)

Are you starting your program by double-clicking on its icon?  Then don't do
that:  bring up a DOS box *first*, and type

    python path-to-your-program

at the command prompt.  When mucking with threads, I often keep an IDLE
window open for editing, and a DOS box open for running the program.  If you
enter

    doskey

upon opening your DOS box, you'll get command history (among other stuff --
do "doskey /?") so that reentering your last "python path-..." line is just
a matter of hitting the up-arrow key.

luckily-programmers-can-get-used-to-anything-ly y'rs  - tim




From chaluba at yahoo.com  Thu Feb 22 18:11:57 2001
From: chaluba at yahoo.com (chaluba at yahoo.com)
Date: Thu, 22 Feb 2001 17:11:57 -0000
Subject: win32com.client - missing methods?
In-Reply-To: <94set0$ecf$1@news5.svr.pol.co.uk>
Message-ID: <973h8t+prq2@eGroups.com>

I ran into the same problem below, the first day I tried using
win32com on NT. I eventually fixed it and the following is my account
hoping it helps you or others.

Setup (work): WinNT4.0SP6a + python1.52 + win32all-136 + MS Office 97

BuiltInDocumentProperties is a DocumentProperties object defined in
MS Office type library. I had generated type libraries for MS Office,
and DocumentProperties did not seem to have any methods. However, when
I browsed with COM Browser, I was able to see its methods.

Setup1 (home): Win2K + python1.52 + win32all-136 + MS Word 2000
Setup2 (home): Win2K + ActivePython 2.02 + MS Word 2000

At home I tried to access methods of BuiltInDocumentProperties and it
worked under both setups. I had not generated any type libraries
manually assuming they'd be auto-generated. They were not generated.

Back to work, I deleted the type libraries and voila - it worked. Type
libaries were not generated. I'm not sure if it's supposed to work
like that but it sure confused me quite a bit (the fact that it did
not work with generated type libraries). 

Maybe someone can comment.

Ayhan

--- In python-list at y..., "Kirby James" <kirby at j...> wrote:
> Hi, I've written a Python program which uses the win32com interface
to
> obtain information from a number of Word-97 files.
> The program runs fine on my Win-98 system but fails on two NT
systems.
> The problem line is :-
> 
>     sAuthor =
myDoc.BuiltInDocumentProperties(constants.wdPropertyAuthor)
> 
> which produces the error
> 
>     AttributeError: no __call__ method defined
> 
> When I remove this line the other COM functions that I call work
fine. On
> all three systems I've run the COM Makepy utility - but I notice
that
> the generated files are of different sizes on the different systems
(is it
> safe to copy - and rename the win-98 file to NT).
> 
> I'd appreciate any pointers
> 
> ta, Kirby
> 
> 
> Traceback (most recent call last):
>   File "D:\Python20\Pythonwin\pywin\framework\scriptutils.py", line
301, in
> RunScript
>     exec codeObject in __main__.__dict__
>   File "D:\python\utilities\word2.py", line 178, in ?
>     os.path.walk('d:\\relocation', walker, 10)
>   File "D:\Python20\lib\ntpath.py", line 270, in walk
>     walk(name, func, arg)
>   File "D:\Python20\lib\ntpath.py", line 264, in walk
>     func(arg, top, names)
>   File "D:\python\utilities\word2.py", line 153, in walker
>     readWord(pathfilename)
>   File "D:\python\utilities\word2.py", line 64, in readWord
>     sAuthor =
myDoc.BuiltInDocumentProperties(constants.wdPropertyAuthor)
> AttributeError: no __call__ method defined
> 
> 
>     WORD  = 'Word.Application.8'
> 
>     myWord = Dispatch(WORD)
>     myWord.Visible = 1
>     myDoc = myWord.Documents.Open(fn)
>     sAuthor =
myDoc.BuiltInDocumentProperties(constants.wdPropertyAuthor)
> ####  line 64
>     sAuthor = str(sAuthor)
>     print '  Author = ' + sAuthor
> 
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From NOstarfighterSPAM at freeuk.com  Sat Feb 24 11:56:47 2001
From: NOstarfighterSPAM at freeuk.com (Richard Townsend)
Date: Sat, 24 Feb 2001 10:56:47 -0000
Subject: Python2.0 and Pmw 0.8.4 counter problem
References: <3A71D728.EE4B9F03@icestormfx.com> <PVmc6.8815$W94.473070@nnrp4.clara.net> <3A75937B.14C8D022@icestormfx.com>
Message-ID: <PtMl6.6183$vI.930231@nnrp3.clara.net>

There is a new version (0.8.5) of Pmw available at
http://pmw.sourceforge.net/ it includes the fix for PmwCounter.py and a few
other changes.

regards,
Richard


"Sylvie Bruneau" <sylvieb at icestormfx.com> wrote in message
news:3A75937B.14C8D022 at icestormfx.com...
>
> Thank you !!  Now it works well.  But why the Pmw team doesn't make the
> modification in their last version ?  Anyway.  Do you know if there are
other
> modifications like this one that as to be done to make sure that all the
Pmw widgets
> works fine ?
>
> Thanks again...
>
>
>
> Richard Townsend wrote:
>
> > >     We used Python1.5 with Pmw 0.8.3 to develop an in-house
application.
> > For some
> > > reasons, we changed to Python2.0 and we decided to use the new Pmw
version
> > (
> > > 0.8.4 ).  In our application, we have a Pmw.Counter of integers and it
> > doesn't work
> > > anymore.  The behaviour is strange :  if we type a new integer in the
> > entry field,
> > > everything is fine.  But using the arrows ( left or right, up or
down )
> > just clean
> > > the entry field, without updating its content.  I try the demos in the
Pmw
> > > distributions and they don't work either.  We installed Python1.6 with
Pmw
> > 0.8.3, and
> > > it still doesn't work !!!  I tried all the other types of counters and
> > they all work
> > > fine, but the integer.
> > >
> > >     Anybody has a suggestion ?
> > >
> >
> > I posted the same question a few months ago and got the following
solution
> > from Roberto Amorim
> >
> > > Hi, to correct it is simple. You'll have to edit PmwCounter.py. Here
it
> > > is:
> > >
> > > - Find the function called _changeNumber() (line 273)
> > > - On line 279 there's the return statement for this function. Change
it
> > > to return str(value) (and not return str(value)[:-1])
> > >
> > > That ocurred because str(long) on Python 2.0 doesn't return a number
> > > with L on the end like Python 1.5.2 did.
> > >
> >
> > regards,
> > Richard Townsend
>
> --
> /=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/=-/
> Sylvie Bruneau ( sylvieb at icestormfx.com ) Tel: (514) 527-3963
> Software Developer, IceStorm Digital.     Fax: (514) 527-5165
> 2595 Place Chasse, Montreal, Quebec, H1Y 2C3.
>
> Where would you have gone today if Window's hadn't crashed?
>
>
>




From tim.one at home.com  Thu Feb  1 19:46:27 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 13:46:27 -0500
Subject: c(++)python?
In-Reply-To: <20010131124910.B26971@harmony.cs.rit.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEJJIMAA.tim.one@home.com>

[D-Man]
> What is the difference between C89 and C99?  ( a URL or a quick
> overview is sufficient )

http://www.google.com/search?q=C89+C99

That will lead you to about 2,000 URLs and quick overviews <wink>.




From fredrik at effbot.org  Thu Feb  1 17:32:03 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 01 Feb 2001 16:32:03 GMT
Subject: c(++)python?
References: <3A71EB9C.D258C1C0@comm.mot.com> <LNBBLJKPBEHFEDALKOLCEEJNILAA.tim.one@home.com> <mailman.981044329.3296.python-list@python.org>
Message-ID: <7cge6.10615$AH6.1745820@newsc.telia.net>

"D-Man" wrote:
> What is the difference between C89 and C99?  ( a URL or a quick
> overview is sufficient )

http://anubis.dkuug.dk/JTC1/SC22/WG14/www/newinc9x.htm

the final committee draft can be found here:
http://www.lysator.liu.se/c/

Cheers /F




From dsh8290 at rit.edu  Fri Feb  2 01:24:29 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 1 Feb 2001 19:24:29 -0500
Subject: c(++)python?
In-Reply-To: <7cge6.10615$AH6.1745820@newsc.telia.net>; from fredrik@effbot.org on Thu, Feb 01, 2001 at 04:32:03PM +0000
References: <3A71EB9C.D258C1C0@comm.mot.com> <LNBBLJKPBEHFEDALKOLCEEJNILAA.tim.one@home.com> <mailman.981044329.3296.python-list@python.org> <7cge6.10615$AH6.1745820@newsc.telia.net>
Message-ID: <20010201192429.A6542@harmony.cs.rit.edu>

On Thu, Feb 01, 2001 at 04:32:03PM +0000, Fredrik Lundh wrote:
| "D-Man" wrote:
| > What is the difference between C89 and C99?  ( a URL or a quick
| > overview is sufficient )
| 
| http://anubis.dkuug.dk/JTC1/SC22/WG14/www/newinc9x.htm

Thanks, exactly what I was looking for.  I think I like a lot of the
features in C99.  (Basically my reasons for using C++ even if I wasn't
going to use (hardly) any C++ features)

| 
| the final committee draft can be found here:
| http://www.lysator.liu.se/c/
| 
| Cheers /F
| 

-D




From Pekka.Pessi at nokia.com  Mon Feb  5 02:32:34 2001
From: Pekka.Pessi at nokia.com (Pekka Pessi)
Date: 05 Feb 2001 03:32:34 +0200
Subject: . Python 2.1 function attributes
References: <Pine.GSO.4.21.0101271158190.3501-100000@y.glue.umd.edu>, <Pine.GSO.4.21.0101270135190.7677-100000@y.glue.umd.edu> <mailman.980593892.725.python-list@python.org> <$5A+ZwAGjtc6EwE0@jessikat.fsnet.co.uk> <14962.62953.304511.523826@anthem.wooz.org> <mailman.980613639.23439.python-list@python.org> <mailman.980617472.28020.python-list@python.org> <Pine.GSO.4.21.0101271303540.3501-100000@y.glue.umd.edu>
Message-ID: <pv3ddt9aod.fsf@agni.research.nokia.com>

In message <Pine.GSO.4.21.0101271303540.3501-100000 at y.glue.umd.edu> Roy Katz <katz at Glue.umd.edu> writes:
>Let's add bless().  I want to change the class of an object at
>runtime. How 'bout it?  

        How about this:

class sm:
    "State machine accepting ab*"
    def __init__(self):
	self.__class__ = a

    valid = 0

class a(sm):

    def a(self):
	self.__class__ = b

class b(sm):

    def b(self):
	pass

    valid = 1

        BTW, have you never looked in the source code of pickle?
       
                                        Pekka


From n8gray at caltech.edu.is.my.email.address  Thu Feb  1 09:43:07 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Thu, 1 Feb 2001 00:43:07 -0800
Subject: "in" for dicts (was: Python 2.1 function attributes)
References: <9535qg$8mu@gap.cco.caltech.edu> <LNBBLJKPBEHFEDALKOLCEEOPILAA.tim.one@home.com> <mailman.980765676.8978.python-list@python.org>
Message-ID: <95b7tj$ib3@gap.cco.caltech.edu>

Thomas Wouters wrote:
> 
> My main beef with the 'x in dict' idea is something Nathaniel didn't
> raise, though: we all seem to think 'x in dict' is naturally 'key in dict'
> -- but why ? Is it because we all know you can't *get* a value from a dict
> without extracting them all, and all values could be the same ? How
> obvious is that to people without a technical idea of dictionaries ? What,
> exactly, is the target group for Python ? :)

I think this may be what truly underlies my discomfort with 'in dict'.  At 
some fundamental level _values_ are what we store in dictionaries, not 
keys.  (Unless we're kludging together a Set :^) The only reason we have 
the keys around is to get to the values.  So when I talk about something 
being "in" a dictionary, it's the value.  I can understand, though, that 
from a practical standpoint the keys may be accessed and iterated over more 
frequently.

> > [Timbot]
> > We still don't have anything else for 2.1; only "a_key in dict" in a
> > *boolean* context has been implemented.  And, yes, we've already got
> > dict.has_key(a_key) for that.  And, yes, Guido just *approved* of adding
> > "for a_key in dict" too, although implementing that is a puzzle given
> > the current iteration protocol's use of __getitem__ (it doesn't jibe
> > with what dict.__getitem__ does), and "what to do" about programs that
> > go mutating the dict during iteration.

Sorry if I munged up the for/if distinction.  The idea's the same:  To me, 
'in dict' doesn't logically imply 'in dict.keys()'.

I'm reassured that I'm right by the fact that, until recently Guido agreed 
with me.  Has he ever used the time machine to go _forward_ in time?

In-a-battle-between-old-Guido-and-new-Guido-who-would-win?-ly-yrs,

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From n8gray at caltech.edu.is.my.e-mail.address  Fri Feb  2 11:43:13 2001
From: n8gray at caltech.edu.is.my.e-mail.address (Nathaniel Gray)
Date: Fri, 02 Feb 2001 02:43:13 -0800
Subject: "in" for dicts (was: Python 2.1 function attributes)
References: <mailman.980760334.2776.python-list@python.org>
Message-ID: <3A7A8F41.4B59D327@caltech.edu.is.my.e-mail.address>

Tim Peters wrote:
> > (I sure hope we're not doing it just to save a few keystrokes every
> > now and then!)
> 
> No.  I don't even believe that was mentioned.  Most people seem to think
> that these uses of "in" for dicts are *obvious*, and that has been
> mentioned -- often.  Some have also mentioned that "x in dict" will be
> faster than "dict.has_key(x)", and ditto for iteration (in which latter case
> it may also save an unbounded amount of additional storage).  The latter has
> been a FAQ for years:
> 
>     http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.012.htp

Hmmm...  It's always disturbing to hear a construct described as
"obvious" in a feature debate.  It often means "obvious in the context
of the program I wrote yesterday," or "obvious if you follow my coding
style."  A minor shift in perspective can quickly render an "obvious"
feature misleading, as T. Wouters' anecdotal evidence suggests.

Interesting.  I can see how "for key in dict" is faster than "for key in
dict.keys()", but why is "if key in dict" faster than "if
dict.has_key(key)"?  

> > "It's not a big dog.  Perhaps if we clipped it's ears and taught it to
> > meow it could pass for a cat!"
> 
> Oh, now I'm getting confused!  Is a dict the dog or the cat?  Is a list the
> other one?  And what are the ears?  Is meowing iteration?  In that case
> we're taking the cat out of the dungeon and removing the cruel muzzle from
> its cute little kitty mouth, while giving the doghouse a fresh coat of
> paint.

Well why didn't you just say so in the first place?  In that case I
withdraw my complaint and I'll submit a foolproof "for x in dict" patch
to SourceForge forthwith!  ;^)

> > And yes, I dislike list.pop() too.  ;^)
> 
> Really?  Or did you intend to say you dislike dict.pop() (which would fit
> better with the animal analogies <wink>).

Nope.  I've got no beef with dict.pop() (isn't it popitem or some such?)
-- it's obvious <wink> to me why that's useful.  list.pop(), on the
other hand, troubles me because it encourages the programmer to use a
list in place of a stack.  Trouble is, the list can do things the stack
can't, and this opens up a mess of potential problems when the guy who
maintains the code isn't the guy who wrote it.  Java makes the same mess
by subclassing Stack from Vector!  (A Stack "is a" Vector?!)

I guess I just don't understand why Python needs to open up this can of
worms with "x = list.pop()" when it already has "x = list[-1]; del
list[-1];".  Is there some subtlety I'm missing?  Performance
considerations?  Thread safety?  Atomic Pop?  (Great name for a band
:^)  Educate me.

> > Lest I come off as too much of a grumpy Gus, I should add that I really
> > like most of the new stuff in Python 2.1a1, particularly xreadlines.
> 
> Figures -- that's the one Guido likes least.  He dislikes xrange too.  He
> was forced into xreadlines by timing considerations (we couldn't match its
> performance cleanly, not with finite effort).  The only reason to like
> xreadlines is speed (which is reason enough!) -- but in that case you should
> like both flavors of "x in dict" too, and for the same reason.

I like xreadlines because it makes the most simple idiom the same as the
most efficient idiom.  True, the name is unfortunate, mostly because it
begs an explanation when you introduce it to a newbie.  In a language
like Python, though, the easy way should be the best way for
ultra-common tasks, and vice-versa.  With the frequency that Python
programmers want to iterate over the lines of text files we ought to
expect that plain old readline should be doing buffering cartwheels and
breathing magical platform-dependent fire to get us those lines
post-haste!  :^)

Yeah, I know.  Submit a patch.  The bloody noosphere won't homestead
itself...

> > I'll  be first in line to upgrade when 2.1 stabilizes, even if "in dict"
> > survives.
> 
> Hey, CVS is open for business!  The only way a new release stablilizes is
> via people *using* it.  

No can do.  The only reason I'm able to play with Python at all right
now is because 1.52 comes preinstalled on RH.  I'm doing a one-term
rotation with a research group where I'm the only one who's even _heard_
of Python.  It's a bit of a "don't ask, don't tell" kind of situation.

Big bummer:  when the end of the term comes I'll have to port all my
code to C.  I can already feel my productivity slipping away...

Anybody know of a good C implementation of dictionaries?  (Other than
Python's, that is :^)

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._


From quinn at hork.ugcs.caltech.edu  Fri Feb  2 19:53:24 2001
From: quinn at hork.ugcs.caltech.edu (Quinn Dunkan)
Date: 2 Feb 2001 18:53:24 GMT
Subject: "in" for dicts (was: Python 2.1 function attributes)
References: <mailman.980760334.2776.python-list@python.org> <3A7A8F41.4B59D327@caltech.edu.is.my.e-mail.address>
Message-ID: <slrn97m0h3.prn.quinn@hork.ugcs.caltech.edu>

On Fri, 02 Feb 2001 02:43:13 -0800, Nathaniel Gray
<n8gray at caltech.edu.is.my.e-mail.address> wrote:
>No can do.  The only reason I'm able to play with Python at all right
>now is because 1.52 comes preinstalled on RH.  I'm doing a one-term
>rotation with a research group where I'm the only one who's even _heard_
>of Python.  It's a bit of a "don't ask, don't tell" kind of situation.
>
>Big bummer:  when the end of the term comes I'll have to port all my
>code to C.  I can already feel my productivity slipping away...
>
>Anybody know of a good C implementation of dictionaries?  (Other than
>Python's, that is :^)

Well, two ideas:

Say "I use this C library called libpython.a that provides some basic
facilities.  Oh, and I also use a special notation to call libpython.a
procedures called 'python'---you should be able to learn the special notation
quicker than the C library interface.  But if you really want to learn the C
interface, the special notation can be converted with a utility called
'Python2C'".

Or:

Port your code to eiffel.  It's not quite python, but it should be an easier
port than to C.  Then use SmallEiffel to compile your eiffel to C.  Ok, so it
won't be the prettiest C in the world, but hey, it's C and that's what they
want, right?  Roll that baby through 'indent' and you're set!

;)


From n8gray at caltech.edu.is.my.email.address  Sat Feb  3 22:08:17 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Sat, 3 Feb 2001 13:08:17 -0800
Subject: "in" for dicts (was: Python 2.1 function attributes)
References: <mailman.980760334.2776.python-list@python.org> <3A7A8F41.4B59D327@caltech.edu.is.my.e-mail.address> <slrn97m0h3.prn.quinn@hork.ugcs.caltech.edu>
Message-ID: <95hsar$ruv@gap.cco.caltech.edu>

If the resulting code can fit in the 256 KB onboard the Khepera mobile 
robot, I'm there!

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._

Quinn Dunkan wrote:

> On Fri, 02 Feb 2001 02:43:13 -0800, Nathaniel Gray
> <n8gray at caltech.edu.is.my.e-mail.address> wrote:
> >Anybody know of a good C implementation of dictionaries?  (Other than
> >Python's, that is :^)
> 
> Well, two ideas:
> 
> Say "I use this C library called libpython.a that provides some basic
> facilities.  Oh, and I also use a special notation to call libpython.a
> procedures called 'python'---you should be able to learn the special
> notation
> quicker than the C library interface.  But if you really want to learn the
> C interface, the special notation can be converted with a utility called
> 'Python2C'".
> 
> Or:
> 
> Port your code to eiffel.  It's not quite python, but it should be an
> easier
> port than to C.  Then use SmallEiffel to compile your eiffel to C.  Ok, so
> it won't be the prettiest C in the world, but hey, it's C and that's what
> they
> want, right?  Roll that baby through 'indent' and you're set!
> 
> ;)



From quinn at retch.ugcs.caltech.edu  Tue Feb  6 22:32:04 2001
From: quinn at retch.ugcs.caltech.edu (Quinn Dunkan)
Date: 6 Feb 2001 21:32:04 GMT
Subject: "in" for dicts (was: Python 2.1 function attributes)
References: <mailman.980760334.2776.python-list@python.org> <3A7A8F41.4B59D327@caltech.edu.is.my.e-mail.address> <slrn97m0h3.prn.quinn@hork.ugcs.caltech.edu> <95hsar$ruv@gap.cco.caltech.edu>
Message-ID: <slrn980raj.23m.quinn@retch.ugcs.caltech.edu>

On Sat, 3 Feb 2001 13:08:17 -0800, Nathaniel Gray
<n8gray at caltech.edu.is.my.email.address> wrote:
>If the resulting code can fit in the 256 KB onboard the Khepera mobile 
>robot, I'm there!

I've been mentioning lua too many times lately, but this seems appropriate:
lua is about 100K, the std libraries add an additional 60K.  Without the
stdlib, it has very few posix dependencies, so you won't need huge posix
libraries linked in either.  It ought to fit in about 200K, with some left for
the actual bytecode.

lua is very python-like (in fact, I wonder if some aspects, such as the
debugging api, weren't inspired by python).  Porting ought to be a breeze
(unless you used lots of fancy python modules).

And it would be much easier to make the case for it just being a C utility
library w/ special notation :)


From tim_one at email.msn.com  Thu Feb  1 07:36:19 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 1 Feb 2001 01:36:19 -0500
Subject: "in" for dicts (was: Python 2.1 function attributes)
In-Reply-To: <20010130095928.A962@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEIEIMAA.tim_one@email.msn.com>

[Tim]
> we also haven't pulled the platform-dependent tricks Perl does
> to optimize char-at-a-time reading the way vendors *should*
> optimize fgets() but almost never do (Tru64 Unix appears to be
> the sole probable exception to date).

[Thomas Wouters]
> FreeBSD, actually, had insane performance in that area as well.
> (getc() being just as fast as getc_unlocked(), in processes with
> only a single thread active.)

This one doesn't have anything to do with locking or not, it has to do with
whether the inner loop is optimized to use register shadows of _iobuf
members, and to use the platform's moral equivalent of the _cnt member to
avoid needing to check for EOF in the inner loop.  It's not "legal" to do
that in user-level code, because it requires breaking into the FILE*
abstraction.  Vendors can do it under the covers, although apparently almost
never do.  The unique thing about Tru64 Unix was the report that using the
fgets() method was substantially faster than using the getc_unlocked()
method; that's darned hard to account for unless Tru64 optimizes the inner
loop of fgets() (as Perl does).

IOW, just using getc_unlocked() still leaves a pile of *potential*
improvement on the floor, even on a single-thread run, but improvement you
can't get at without cheating (unless the vendor libc authors were savvy
enough to do it for you).

If Perl while(<>) is still substantially faster on FreeBSD than Python 2.1
using the fgets() method in single-thread runs, platform failure to optimize
fgets() in this way is a top contender for "why".




From greg at cosc.canterbury.ac.nz  Thu Feb  1 05:26:46 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 01 Feb 2001 17:26:46 +1300
Subject: "in" for dicts (was: Python 2.1 function attributes)
References: <mailman.980770173.14038.python-list@python.org>
Message-ID: <3A78E586.4638E05B@cosc.canterbury.ac.nz>

Tim Peters wrote:
> 
> Nobody searches phone books or Webster's
> "backwards" in real life.

Indeed! If I were to ask my colleague "Is the word 'bamboozle'
used anywhere in a definition in Webster's dictionary?" the
answer would probably be "Buggered if I know."

Which leads us to what the semantics of
"x in dict" should obviously be:

>>> if x in dict:
...   print "boo:
<stdin>, line 1: BuggeredIfIKnowError

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From tim.one at home.com  Thu Feb  1 10:34:32 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 04:34:32 -0500
Subject: . Python 2.1 function attributes
In-Reply-To: <95432j014on@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEIIIMAA.tim.one@home.com>

[Tim]
>> ...
>> The testimony of one person who tried a thing and found it
>> lacking is more compelling than 100 virgins speculating about
>> why they're certain they would despise it if only they could
>> bear stooping low enough to try it <wink>.

[Alex Martelli]
> Actual experience with using some tool may be a _part_ of what
> gives 'impressiveness' to reports that said tool is inferior,
> but it's generally not a major component thereof -- e.g., the
> tool may have been misunderstood, misused, etc.

That's important too, though.  That is, if people misunderstand or misuse,
that may point to serious flaws in the docs or the design.  There's just no
substitute for experience.  Not that that's *all* there is, but it's
important.  In this case, we learned that *nobody* misunderstood what it
did.  I suspect some people still take it way too seriously, though (like
reduce, map and lambda, it's meant to be a minor convenience, not a new
foundation on which to build all your file output!  maybe that's a lesson we
failed to learn from reduce & friends).

>> telling fact that Guido rejected Barry's print>> proposal
>> before giving it a test drive in his own code -- pure argument
>> never had much to do with this one.

> Interesting -- I don't recall hearing about this detail previously.

There's a bit of it in the SourceForge patch history:

http://sourceforge.net/patch/index.php?func=detailpatch&patch_id=
    100970&group_id=5470

although there are many gaps in the discussion.  The Python-Dev archives may
or may not have more, but I've really got no interest in reliving this one.

> It _does_ interestingly contradict Guido's posts at the time,
> requesting _arguments_ (NOT "experiences") -- which, to me and I
> guess to other respondents, implied that the requested arguments
> *would* have a bit to do with something; I'm less surprised, now
> that I hear this, that said arguments mattered not a whit.

We need some context here.  Stephen Figgins wrote a great (IMO) "outsider"
summary of the brouhaha at the time:

  http://www.oreillynet.com/pub/a/python/2000/08/30/pythonnews.html

Against the background of "a flurry of posts that accused the Python
development team of creeping featurism, selling out the language to
corporate interests, moving too fast, and turning a deaf ear to the Python
community", Guido was trying to insist on some sanity.  Don't take it
personally, I don't think it was ever directed at you.  But the overall
debate had gone nuts.  As Moshe reminded already, what Guido asked NOT to
get was "votes", not "experience".  A killer pure argument would have been
great too.  However, in his estimation he didn't get one, much as in your
estimation he didn't make one.  So it goes.

> ...
> Oh, I do realize this -- Python *will* have this wart forevermore.
> That doesn't mean I intend to suffer it *silently*.

No indeed, so long as it's understood that the right to speak is not also
the right to be listened to -- don't you think there are more *interesting*
things to do in the Python world than beat this dead horse?  I do.

> ...
> I thought I had posted (must have been 4-5 months ago) about my
> "print>>" experiments?

I could well have missed that -- sorry.  In the crunch to get Python 2.0 out
the door, we were all pulling 20-hour days for a while, and facing the
prospect of being out of work anyway (indeed, as things turned out, we
finished the release on our own time).  It was darned hectic.

> Briefly -- I took one module that was using about 20 print
> statements in two functions (and used the sys.stdout setting for
> redirect when needed, from a separate wrapping function) and
> recoded it using '>>blotch' on every print (with 'blotch' a
> function parameter defaulting to None on the operating functions).

I don't think I would have liked ">>blotch" in that context either:  if I
find myself using print>> in more than one or two lines in a function, I
don't like it.  It's simply too wordy for me then.  It's the bee's knees,
though, for an infrequent print to a rarely-used file, or for echo'ing the
same msg to multiple files (like a log and stdout).  Then it's pleasant.

BTW, I see you also picked up on my dislike for sys.stderr.  In the doctest
context, yes, that's 100% my baby and I wrote it to please me.  I don't have
that luxury when working on other peoples' code, though, and if someone is
already printing to sys.stderr, I'll play along for the sake of keeping
their code true to *their* vision.  And then

   print >> sys.stderr, "Hey! I got", `got`, "but expected", `want`

1 line in 250 is pure pleasure.

> The small wrapper was not bad:
>
> from [before 'print>>']:
>
> def foobarRedirect(fileob, *args):
>     save = sys.stdout
>     try: return foobar(*args)
>     finally: sys.stdout = save
>
> to the simpler:
>
> def foobarRedirect(fileob, *args):
>     return foobar(*args, blotch=fileob)
>
> However, function foobar itself was cluttered by a more
> substantial amount than the wrapper was simplified, what with
> each 'print x' becoming 'print>>blotch,x' (plus the extra
> blotch=None argument to both foobar and the other function,
> also exposed from the module, that foobar calls).

As above, I don't believe I would use print>> here either.

> This has as little 'scientific' value as any other similar
> experiment, of course (which is part of why I find your
> 'more-impressed' somewhat unlikely) -- it's not a controlled
> ergonomics experiment:

We have no funding for that, so anecdotal evidence-- in the absence of a
killer argument --is the best we're going to get.

> in the end, the evaluation of "how readable and maintainable is
> the resulting code" is about as subjective and aesthetic as one's
> original appreciation of the construct.  But then, actual human-
> factors controlled experiments are just as scarce on the side
> _supporting_ 'print>>', aren't they?

There are none in either direction that I know of, and I doubt there ever
will be.  The realm of the subjective is where a BDFL is *most* needed!

> I don't think I posted on the similar experience I did with
> code previously using flob.write, but it was even worse from
> my point of view.  Changing the calls, typically originally
> all in the form:
>     flob.write(format % (a,b,c))
> to print-with-redirect:
>     print >> flob, format % (a,b,c),
> not only had no observable benefit, but actually gave a
> slight problem in a case where an unwanted blank space got
> inserted by two back-to-back 'print>>' (while the .write
> calls, of course, did not insert it).

The space-inserting gimmicks of plain "print" are best for output intended
to be read by humans, and I think that's just as true of "print>>".  I
*suspect* (but don't know) that one reason Guido found print>> attractive is
the number of msgs (over the years) directed to sys.stderr that are
*missing* crucial spaces when done in the sys.stderr.write(msg) form -- and
since they're "error msgs", sometimes years go by before we see that one of
those is damaged.

Now a space-eager printing gimmick could have been done via a function too,
but Python already *had* a stmt for that purpose.  From that POV, it was
natural to extend it.

> But then, I had even lower expectations this time -- it seems
> that 'print>>' is studied to substitute for print-plus-
> sys.stdout-changes, not for .write (which is what I most often
> use for output).

I don't think print>> is ever a good substitute for .write(), except in
those cases where .write() is being used for human-readable output and the
author didn't want to do the stdout redirection dance around plain "print".
Else automagical space-insertion gets in the way more than it helps.  Don't
know whether Guido would agree, but that's my take on it.

> The "lower expectations" may be important -- but I can't
> brainwash myself (and wouldn't if I could) to expect the
> sky and stars from a construct that strikes me so badly:
> giving it "a fair chance" may thus be impossible for me.

Try thinking of it as a minor convenience?  I do!  I could live without it;
I'd rather not.

> Still, we're all humans (or bots), and thus, this is no
> doubt a widespread issue.  I've spent an hour or so on it,
> confirmed the preconceived ideas I had before I did, and
> my opinions about 'print>>' are, as could be expected,
>
> Be that as it may, I'm still waiting to see real-life
> examples, in my code *or others'*, where the use of
> 'print>>' will strike me as "ah yes, clearly the one
> right, obvious way to do it".

A quick regexp search turns up about 135 uses in the current CVS tree.  Some
I don't like.  Some I do.  I'll mention two of the latter because I expect
to hear about all the others from you <wink>.

1. The last line of Lib/test/test_largefile.py:

    print >>sys.stderr, name, "exists:", os.path.exists(name)

A human-readable msg and the only use of sys.stderr in the module.  There is
no (IMO) alternative as obvious or convenient as this way of writing it.

2. Lib/warnings.py:  3 instances in 280 lines of code, again all trying to
write something human-readable to sys.stderr.

And I'll toss in one marginal one:  Lib/smtpd.py has 12 instances of

    print >> DEBUGSTREAM, some, human, readable, msg

That's in over 531 lines of code, and it never appears often in a single
function, so it's not being overused by my lights.  But I prefer using a
debug *function* for this kind of thing:  it's more flexible.  For example,
the module's default value for DEBUGSTREAM is

DEBUGSTREAM = Devnull()

where the author wrote a Devnull class with a do-nothing .write() method:

class Devnull:
    def write(self, msg): pass
    def flush(self): pass

While just a few lines of code, it's on the edge of being strained.  I'd
also rather see these lines as

    DEBUG(some, human, readable, msg)

because it's the DEBUG part that's important, not the "print" part (and,
indeed, the *normal* use of these lines is *not* to print!).

Ack -- it's after 4AM again, and we've got a release to get out tomorrow
(today?  who can tell anymore <wink/sigh>).  One quickie:

> ...
> I guess, when he comes around to it, since he has no C background
> to understand the rules for the format string; I'd better write up
> something for him -- or can somebody help with an URL about %-
> formatting already written up for somebody with zero C background?).

I don't, but there's a very *good* writeup at:

    http://www-ccs.ucsd.edu/c/lib_prin.html

Lots of examples, explanation, and even a railroad diagram for the syntax of
a format code (which I'm afraid is more confusing than helpful!).

At least the docs for 2.1 contain *a* description of each format code.  The
contributed patch Fred started from can be found here:

http://sourceforge.net/patch/index.php?func=detailpatch&
    patch_id=103412&group_id=5470

good-topic-for-a-"howto"!-dense-little-languages-aren't-obvious-ly
    y'rs  - tim




From barry at digicool.com  Fri Feb  2 09:24:26 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Fri, 2 Feb 2001 03:24:26 -0500
Subject: . Python 2.1 function attributes
References: <95432j014on@news2.newsguy.com>
	<LNBBLJKPBEHFEDALKOLCKEIIIMAA.tim.one@home.com>
Message-ID: <14970.28346.972083.948158@anthem.wooz.org>

>>>>> "TP" == Tim Peters <tim.one at home.com> writes:

    TP> And I'll toss in one marginal one: Lib/smtpd.py has 12
    TP> instances of

And I agree, it's marginal.  But I wrote that code a looong time
ago[*], I think just after print>> was added, and mostly to experiment
with it.  If I did it again today, I probably /would/ use a debug
function (much like I still use a syslog() debug function in
Mailman).  Tim's absolutely right, there's still a place for debug
functions, but when print's behavior is what you want, I see no reason
to go through the headaches and ugliness of sys.stdout redirection, or
to re-implement all that behavior in a homegrown function.

I do use print>> quite a bit in my mimelib package (Generator.py)
because it's just so damn convenient and (to me, and I wrote it so I
get to choose :) so much cleaner than either a function or wrapping in
sys.stdout redirection.  I really do believe either of those latter
solutions would be uglier, if done correctly, compatibly, and
robustly.

and-email-is-(arguably)-human-readable-ly y'rs,
-Barry

[*] even though it was just recently checked in.



From donn at u.washington.edu  Thu Feb  1 20:05:42 2001
From: donn at u.washington.edu (Donn Cave)
Date: 1 Feb 2001 19:05:42 GMT
Subject: . Python 2.1 function attributes
References: <mailman.981046930.4615.python-list@python.org>
Message-ID: <95cc26$k0k$1@nntp6.u.washington.edu>

Quoth "Tim Peters" <tim.one at home.com>:
[... MASSIVE WEIGHT LOSS in ONE DAY! ...]

| A quick regexp search turns up about 135 uses in the current CVS tree.  Some
...

To cruelly prolong the misery of this thread - I personally find
"print >>" kind of distasteful in terms of the notation, but as
long as we have it, we may as well enjoy it.

Now that the build procedure is partly implemented in Python, I wish
that a bunch of "print" commands in distutils could be "print >> sys.stderr"
instead.  The right way to run a build is with output redirected to a
disk file, so you can review it at your leisure, and that makes stdout
block buffered.  The output doesn't get to the file for a long time,
or at all if you have some reason to kill the process, comes out in
blocks irrespective of line structure, and meanwhile all the incidental
diagnostics are arriving that were issued to stderr/unit 2.  Since
stderr is line buffered, it's really much better to go there with
diagnostic outputs, in fact that's what it's for and why it works the
way it does.  Likewise, the one line in test_*.py should be about
every instance of print in any such file.

	Donn Cave, donn at u.washington.edu


From tim.one at home.com  Fri Feb  2 05:43:16 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 23:43:16 -0500
Subject: . Python 2.1 function attributes
In-Reply-To: <95cc26$k0k$1@nntp6.u.washington.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCIELEIMAA.tim.one@home.com>

[Donn Cave, stderr's best buddy]
> ...
> Since stderr is line buffered, it's really much better to go
> there with diagnostic outputs, in fact that's what it's for and
> why it works the way it does.

Maybe on your OS this is helpful.  On mine output to stderr scrolls off the
shell window irretrievably, and can't be redirected from the command line.
Not much fun to see it interleaved in seemingly random order with stdout
either.

> Likewise, the one line in test_*.py should be about every
> instance of print in any such file.

Sorry, this one's wrong independent of OS.  Python's std regression tests
(test_*.py) work by capturing stdout and comparing it to canned "expected
output" files.  stderr *may* have a marginal role there, but only for
exceedingly rare "what the f**k?!  this isn't just wrong, it's an utterly
unexpected disaster!" cases.  But even those would be better directed to
stdout, because the stdout comparison mechanism would point them out if they
happen, and won't forget to look for them.  People do forget, and, as is,
when one of these stderr cases crops up, regrtest.py doesn't realize the
test failed either.

and-no-i-had-no-part-in-writing-regrtest.py<wink>-ly y'rs  - tim




From aleaxit at yahoo.com  Fri Feb  2 15:24:29 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 2 Feb 2001 15:24:29 +0100
Subject: . Python 2.1 function attributes
References: <mailman.981046930.4615.python-list@python.org>
Message-ID: <95efv80hih@news1.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.981046930.4615.python-list at python.org...
    [snip]
> We need some context here.  Stephen Figgins wrote a great (IMO) "outsider"
> summary of the brouhaha at the time:
>
>   http://www.oreillynet.com/pub/a/python/2000/08/30/pythonnews.html
    [snip]
> > Oh, I do realize this -- Python *will* have this wart forevermore.
> > That doesn't mean I intend to suffer it *silently*.
>
> No indeed, so long as it's understood that the right to speak is not also
> the right to be listened to -- don't you think there are more
*interesting*
> things to do in the Python world than beat this dead horse?  I do.

There is, in my opinion, a subtle and important distortion in what you
call "a great outsider summary" above -- specifically in its summing
things up (as of August 2000) with "This has quieted some of the complaints
down. I expect them to taper off this week" and "am looking forward to
seeing what controversy will rise up next as this one finally dies down".

This implies that the controversy was a tempest in a teapot, just one
of many, that everybody was by now resigned to what Mr. Figgins calls
"a done deal".  This is not the way I see the "print>>" issue.  I
consider it _by far_ the worst thing to ever have happened to Python
(in the limited timespan in which I've been following Python; but,
judging from what I've read in old mailing list archives and what I
can see now in the language as it stands, then 'ever' does apply).

"Those who cannot remember the past are condemned to repeat it" (this
Santayana quote is often subject to many variations -- I apologize,
but I don't know what the exact wording should be).  If there is
anything at all that I can do, which may decrease by one part in a
million the probability that Python may suffer a repetition of the
"print>>" disaster, it may be worth it; and, if this quote contains
any germ of truth, then ensuring the remembrance is not buried may
perhaps be some part of that one-in-a-million decrease in the risk
of such a repetition.

No, I do not believe there is any Python related endeavour, that
I might be capable of, that is more IMPORTANT ("interesting" is a
different issue) than whatever fraction of a part in a million I may
contribute to avoiding such a repetition (if I thought there was ANY
chance, no matter how tiny, to have the horrid wart REMOVED from the
language, then that would be something else again -- but I don't; so,
the _most_ I can hope for is to perhaps contribute to lowering the
probability of a repetition).

Of course I have no right to be listened to -- killfiles have not
been abrogated, have they?  Everybody, as usual, is _welcome_ to
killfile me if they're not interested in what I have to say -- have
I ever said or implied otherwise?


> their code true to *their* vision.  And then
>
>    print >> sys.stderr, "Hey! I got", `got`, "but expected", `want`
>
> 1 line in 250 is pure pleasure.

How is this "better" than

    sys.stderr.write("Hey! I got %r but expected %r\n" % (got, want))

...?

> *suspect* (but don't know) that one reason Guido found print>> attractive
is
> the number of msgs (over the years) directed to sys.stderr that are
> *missing* crucial spaces when done in the sys.stderr.write(msg) form --
and
> since they're "error msgs", sometimes years go by before we see that one
of
> those is damaged.

When the msg is put together with a %-format, which seems to be
the way that most comes natural to me, it's hard to think that
missing spaces can be an important issue.

> Now a space-eager printing gimmick could have been done via a function
too,
> but Python already *had* a stmt for that purpose.  From that POV, it was
> natural to extend it.

I have no bones to pick against the print statement for its most
elementary, daily use -- rough & ready debugging; no matter how
refined the debuggers I have at hand, I'm enough of a dinosaur
that much of my debugging is inserting and removing print's.  But
for that use case there is no added value in redirecting each of
them separately -- if I find a sudden need to have all of the
print's I currently have use OutputDebugString, or whatever, then
a temporary setting of sys.stdout around the interesting parts
remains preferable (which I think is part of what you were agreeing
on, in the parts of your post that I snipped, above).

_For a specific need to output something onto a given file_, I
find absolutely nothing natural or worthwhile in "extending" a
statement whose main use case is so different (impromptu tracing).

An output function would have required no syntax blotch (I have
just reread Guido's response and see that he does mention the
analogies with sh's use of >> -- he does fail to mention that
those analogies are _misleading_ ones, exactly those that seemed
to trip up my son), and the objection "but for a function I have
no way to say ``and don't use a linebreak''", also part of Guido's
response, is unbelievable -- why not just
    printfun(x,y,z,newline=None)
...?!

*Fortunately*, it's NOT natural in Python for a statement that
already exists for a purpose to be extended, with kludgy syntax,
to other similar purposes a function would serve better.  For
parallel iteration over lists, we didn't get some syntactically
horrid kludge based on the existing for statement -- we got a
usable and elegant zip function instead, good for that one AND
other secondary uses too.  Let's hope the "well, there IS one
statement that does something a BIT like that, let's think of
some C++ish abuse of a binary operator into a unary not-quite-
an-operator-whatever-the-heck-IS-it to have it do that too"
remains a once-a-decade occurrence in Python, rather than have
it become a way of life.


> I don't think print>> is ever a good substitute for .write(), except in

So far, we agree, up to the "except" excluded:-).

> those cases where .write() is being used for human-readable output and the
> author didn't want to do the stdout redirection dance around plain
"print".

The redirection minuet is one thing, but I claim that write
with a %-format doesn't leave me wanting for decent ways to
do output.  A printlike function would be better, but I
don't hunger for it often, because write-plus-% does it well;
if I _did_ hunger for it, it would not be terrible to write
it, either (though having it as a builtin would encourage
its use in small pieces of code meant to be portable).


> Else automagical space-insertion gets in the way more than it helps.
Don't
> know whether Guido would agree, but that's my take on it.

FWIW (little, of course), I agree with your take, from the 'Else' (excluded)
onwards.


> > The "lower expectations" may be important -- but I can't
> > brainwash myself (and wouldn't if I could) to expect the
> > sky and stars from a construct that strikes me so badly:
> > giving it "a fair chance" may thus be impossible for me.
>
> Try thinking of it as a minor convenience?  I do!  I could live without
it;
> I'd rather not.

"convenience" is the most-used euphemism, in today's programming,
for "horrid blotch that anybody should be ashamed of", to justify
complication, irregularity, rampant ad-hocitis and kludges.

Thus, I have no trouble at all thinking of 'print>>' as a
"convenience" (visualize Dr. Evil doing the ``quotes''
finger-gesture...).

It's not "minor", IMHO, mostly because its syntax stands out
like a sore thumb (and BY DESIGN, no less!) from Python's
generally pretty nice picture.  It DOES, as intended, recall
C++, or sh -- but, in THOSE punctuation-obsessed contexts,
what's a >> more or less.  Anything that takes Python's
syntax closer, _by design_, to that of such collections of
syntactic warts as C++ or sh is a MAJOR "convenience" indeed!


> A quick regexp search turns up about 135 uses in the current CVS tree.
Some
> I don't like.  Some I do.  I'll mention two of the latter because I expect
> to hear about all the others from you <wink>.

No need... the "likeable" ones are bad enough!-)  Taking just one:

> 1. The last line of Lib/test/test_largefile.py:
>
>     print >>sys.stderr, name, "exists:", os.path.exists(name)
>
> A human-readable msg and the only use of sys.stderr in the module.  There
is
> no (IMO) alternative as obvious or convenient as this way of writing it.

What about:

    sys.stderr.write("%s exists: %s\n" % (name, os.path.exists(name))

...?  What's so horribly inconvenient or too-subtle about %-formatting?


> While just a few lines of code, it's on the edge of being strained.  I'd
> also rather see these lines as
>
>     DEBUG(some, human, readable, msg)
>
> because it's the DEBUG part that's important, not the "print" part (and,
> indeed, the *normal* use of these lines is *not* to print!).

I fully agree with this assessment.  If a function printfun and/or a
method printmethod existed to do printlike things, I think this would
encourage by example the use of a function here and in similar
contexts -- one might just have DEBUG=fileobject.printmethod to
turn it on, and DEBUG=functionthatdoesnothing to turn it off.


> > I guess, when he comes around to it, since he has no C background
> > to understand the rules for the format string; I'd better write up
> > something for him -- or can somebody help with an URL about %-
> > formatting already written up for somebody with zero C background?).
>
> I don't, but there's a very *good* writeup at:
>
>     http://www-ccs.ucsd.edu/c/lib_prin.html
>
> Lots of examples, explanation, and even a railroad diagram for the syntax
of
> a format code (which I'm afraid is more confusing than helpful!).

I'm not a diagram-oriented person myself, but the rest does
seem useful -- thanks!


> At least the docs for 2.1 contain *a* description of each format code.
The
> contributed patch Fred started from can be found here:
>
> http://sourceforge.net/patch/index.php?func=detailpatch&
>     patch_id=103412&group_id=5470

Thanks again, I'll have a look.


> good-topic-for-a-"howto"!-dense-little-languages-aren't-obvious-ly

Good point...


Alex





From mbel44 at dial.pipex.net  Mon Feb  5 15:15:48 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Mon, 05 Feb 2001 14:15:48 +0000
Subject: . Python 2.1 function attributes
References: <mailman.981046930.4615.python-list@python.org> <95efv80hih@news1.newsguy.com>
Message-ID: <1jct7tg5u6j3fir09edclfm5jmf5n59eto@4ax.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote:

>> I don't think print>> is ever a good substitute for .write(), except in
>
>So far, we agree, up to the "except" excluded:-).
>
>> those cases where .write() is being used for human-readable output and the
>> author didn't want to do the stdout redirection dance around plain
>"print".
>
>The redirection minuet is one thing, but I claim that write
>with a %-format doesn't leave me wanting for decent ways to
>do output.  A printlike function would be better, but I
>don't hunger for it often, because write-plus-% does it well;

Not well enough, apparently, judging by the number of packages that
grew out of the need for exactly this type of output. A need that
wasnt satisfied by file.write("formatting"%(data))

The two that I am familiar with are the templating parts of quixote
http://www.mems-exchange.org/software/python/quixote/
and DC's DTML. DTML is part of Zope - It currently currently does alot
more than just formatting, but I understand that was its original (and
only) goal.

The real problem (IMO) with file.write("formatting"%(data)) is that
the expressions that build up the data are out of line with the static
content. Its harder for the programmer to visualize the bigger picture
of his output, because he has to mentally interleave the dynamic and
static parts. The problem is greater in functions whose only job is
output formatting.

The disadvantage is comparable to the difference between using an
infix notation, or function notation for arithmetic.


ps. Ive used both dtml and quixote templates, but largely abandonned
them both in favor of print >>



Toby Dickenson
tdickenson at geminidataloggers.com


From aleaxit at yahoo.com  Mon Feb  5 17:20:45 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 5 Feb 2001 17:20:45 +0100
Subject: . Python 2.1 function attributes
References: <mailman.981046930.4615.python-list@python.org> <95efv80hih@news1.newsguy.com> <1jct7tg5u6j3fir09edclfm5jmf5n59eto@4ax.com>
Message-ID: <95mjsv01t2t@news1.newsguy.com>

"Toby Dickenson" <mbel44 at dial.pipex.net> wrote in message
news:1jct7tg5u6j3fir09edclfm5jmf5n59eto at 4ax.com...
    [snip]
> >The redirection minuet is one thing, but I claim that write
> >with a %-format doesn't leave me wanting for decent ways to
> >do output.  A printlike function would be better, but I
> >don't hunger for it often, because write-plus-% does it well;
>
> Not well enough, apparently, judging by the number of packages that
> grew out of the need for exactly this type of output. A need that
> wasnt satisfied by file.write("formatting"%(data))

Oh, of course, there ARE plenty of human-readable-output needs
that neither file.write, nor print in any form, support decently.

You mention HTML-oriented templating output, which has no special
need to be human-readable: there, file.write with format is _much_
superior, since more easily localized (print is a disaster in this
sense -- no problem for quick&dirty debug-oriented output, but of
course NOBODY is writing deliberately non-localizable applications,
*ARE THEY*...?-); but 'full templating' is often more powerful, and
it's simple to implement well in Python, too -- so, like everybody
else, I built my own "Yet Another Python Template Utility" (YAPTU)
and use that one for those (not-infrequent) text-output needs that
are well satisfied via templating.

_Some_ output needs require more logic than such templating can
supply while remaining simple and well-usable.  Hierarchical stuff
that nests to arbitrary depth, for example; I know of no template
solution that fully supports the unlimited-depth recursion that is
typical (e.g.) of some ways to present a "bill of materials".  For
THAT, the recursive logic is better handled in Python (IMHO).  But
when the time comes to output a given 'block' of info (typically a
line or small block of lines, with the code-logic determining, e.g.,
to what 'depth' it is to be indented), the specifics are still best
handled with a formatting-line (with a dictionary on the RHS of the
%, no doubt, and named fields -- we ARE talking about somewhat
complicated needs after all!), and that formatting line had better
be retrieved by reasonably flexible logic, else, one can wave
good-bye to localization and customization (fortunately, a DB tends
to be around when one is doing stuff at THIS level of complexity,
so, a way is not lacking to flexibly store and recoved the needed
formatting-lines!-)


> The real problem (IMO) with file.write("formatting"%(data)) is that
> the expressions that build up the data are out of line with the static
> content. Its harder for the programmer to visualize the bigger picture
> of his output, because he has to mentally interleave the dynamic and
> static parts. The problem is greater in functions whose only job is
> output formatting.

If the ONLY job is output formatting, then there are going to be
no 'expressions' to speak of -- 'data' will be a dictionary that
associates names with values, and the computation will have been
done elsewhere -- a very sensible way to split computation from
presentation tasks, of course.  But then, where is the difference
between, e.g.,

flob.write("An a %(one)s an a %(two)s an a %(three)s!\n" % data)

and

print >> flob, "An a",one,"an a",two,"an a",str(three)+"!"

except the sundry ways in which the second form is inferior?

E.g., it's trouble to get the final '!' adjacent to the last
item being output (you need to switch to str and concatenation
rather than using the same approach as for all other items),
localization is a bear (how do you get the pieces, how many
do you get, how do you order the items...?), so is formatting
to align columns to the left or right (trivial with %), the
function itself needs to know the names of the items and get
them into its namespace appropriately, AND flob itself may
get several write calls and softspace settings and queries
rather than dealing with just one nice solid write period.


> The disadvantage is comparable to the difference between using an
> infix notation, or function notation for arithmetic.

It's true that a few more parentheses are needed for a write
call, since Python syntax requires them (both around the
whole %-expression, and inside the format-string when the
RHS is a dictionary).  But print>> requires a similar amount
of commas, so it balances out; the _order_ in which the
names are used is the same, and there are no "function names"
vs "operators" involved, so the analogy is strained.


> ps. Ive used both dtml and quixote templates, but largely abandonned
> them both in favor of print >>

If templates were not suitable for your needs, you no doubt
did well to abandon their use.  But using fileob.write rather
than print>>fileob would have ensured further benefits -- see
above.  So, your choice need not have been the optimal one.

Betraying the "one obviously correct way" philosophy is just
one more way in which 'print>>' is unPythonic.  There never
really IS just one way, of course, nor is the right one, often,
"obviously" right; but providing one more way to do yet the
same thing -- particularly one whose disadvantages (difficulty
in localization, column alignment, etc, etc) may not be obvious
to inexperienced users -- was a step in the wrong direction.

(Reflecting on how all needs might have been best served -- a
futile exercise, some would say, but I'm a Santayana follower --
leads me to believe that a Print function [also supplied as a
method on built-in objects -- much like xreadlines is going to
be both a method of built-in objects AND a utility function to
wrap user-written objects, if I understand that correctly] would
have been Just Right.  It would take an optional 'format='
argument, which, if present and not None, overrides the implicit
``formatting'' of "paste stuff together with spaces and a \n
and the end"; one could easily prototype with a call to Print
almost identical to today's print statement, then add to it a
suitable format= for localization, just as one might add a
suitable file= etc.).


Alex





From mbel44 at dial.pipex.net  Tue Feb  6 15:37:13 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Tue, 06 Feb 2001 14:37:13 +0000
Subject: . Python 2.1 function attributes
References: <mailman.981046930.4615.python-list@python.org> <95efv80hih@news1.newsguy.com> <1jct7tg5u6j3fir09edclfm5jmf5n59eto@4ax.com> <95mjsv01t2t@news1.newsguy.com>
Message-ID: <fv108t0rmqqs86g815skrm9be2s0a967kl@4ax.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote:

>> The real problem (IMO) with file.write("formatting"%(data)) is that
>> the expressions that build up the data are out of line with the static
>> content. Its harder for the programmer to visualize the bigger picture
>> of his output, because he has to mentally interleave the dynamic and
>> static parts. The problem is greater in functions whose only job is
>> output formatting.
>
>If the ONLY job is output formatting, then there are going to be
>no 'expressions' to speak of -- 'data' will be a dictionary that
>associates names with values, and the computation will have been
>done elsewhere -- a very sensible way to split computation from
>presentation tasks, of course.  But then, where is the difference
>between, e.g.,
>
>flob.write("An a %(one)s an a %(two)s an a %(three)s!\n" % data)
>
>and
>
>print >> flob, "An a",one,"an a",two,"an a",str(three)+"!"

If your elements comes neatly packed in a dict, then yes thats true.
The % operator is adept at picking named elements out of your dict.

More commonly you need to pick the elements out of some other, equally
simple data structure:

print >> flob, "An a",things[1],"an a",things[2],"an a",things[3]+"!"

print >> flob, "An a",things.pop(),"an a",things.pop(),"an
a",things.pop()+"!"


>It's true that a few more parentheses are needed for a write
>call, since Python syntax requires them (both around the
>whole %-expression, and inside the format-string when the
>RHS is a dictionary).  But print>> requires a similar amount
>of commas, so it balances out; the _order_ in which the
>names are used is the same, and there are no "function names"
>vs "operators" involved, so the analogy is strained.

No, the analogy is about visual locality, not names. If your data
comes packaged in something other than a dict:

flob.write("An a %s an a %s an a %s!\n"%(one,two,three))



Toby Dickenson
tdickenson at geminidataloggers.com


From greg at cosc.canterbury.ac.nz  Wed Feb  7 23:08:08 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 08 Feb 2001 11:08:08 +1300
Subject: . Python 2.1 function attributes
References: <mailman.981046930.4615.python-list@python.org> <95efv80hih@news1.newsguy.com> <1jct7tg5u6j3fir09edclfm5jmf5n59eto@4ax.com>
Message-ID: <3A81C748.527272C@cosc.canterbury.ac.nz>

Toby Dickenson wrote:
> 
> The real problem (IMO) with file.write("formatting"%(data)) is that
> the expressions that build up the data are out of line with the static
> content. Its harder for the programmer to visualize the bigger picture
> of his output,

It seems that different people may see this differently,
since I often find it easier to visualise the output
when using a format string. When the output items are
interleaved with string literals, I find the result
is often a confusing clutter of commas and quotes,
especially when there are commas in the string
literals as well.

Besides, with the %(name) formatting item,
you can get the best of both worlds.

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From joconnor at cybermesa.com  Thu Feb  8 00:03:55 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Wed, 07 Feb 2001 23:03:55 +0000
Subject: . Python 2.1 function attributes
References: <mailman.981046930.4615.python-list@python.org> <95efv80hih@news1.newsguy.com> <1jct7tg5u6j3fir09edclfm5jmf5n59eto@4ax.com> <3A81C748.527272C@cosc.canterbury.ac.nz>
Message-ID: <3A81D45B.B90C8F45@cybermesa.com>

Greg Ewing wrote:
> 
> Toby Dickenson wrote:
> >
> > The real problem (IMO) with file.write("formatting"%(data)) is that
> > the expressions that build up the data are out of line with the static
> > content. Its harder for the programmer to visualize the bigger picture
> > of his output,
> 
> It seems that different people may see this differently,
> since I often find it easier to visualise the output
> when using a format string. When the output items are
> interleaved with string literals, I find the result
> is often a confusing clutter of commas and quotes,
> especially when there are commas in the string
> literals as well.

FWIW - I find 'printf()' style formatting string easier to write but
harder to reaf

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython


From donn at u.washington.edu  Fri Feb  2 18:36:04 2001
From: donn at u.washington.edu (Donn Cave)
Date: 2 Feb 2001 17:36:04 GMT
Subject: . Python 2.1 function attributes
References: <mailman.981098775.19961.python-list@python.org>
Message-ID: <95er64$q36$1@nntp6.u.washington.edu>

Quoth "Tim Peters" <tim.one at home.com>:
| [Donn Cave, stderr's best buddy]
| > ...
| > Since stderr is line buffered, it's really much better to go
| > there with diagnostic outputs, in fact that's what it's for and
| > why it works the way it does.
|
| Maybe on your OS this is helpful.  On mine output to stderr scrolls off the
| shell window irretrievably, and can't be redirected from the command line.
| Not much fun to see it interleaved in seemingly random order with stdout
| either.

That is so sad.

|> Likewise, the one line in test_*.py should be about every
|> instance of print in any such file.
|
| Sorry, this one's wrong independent of OS.  Python's std regression tests
| (test_*.py) work by capturing stdout and comparing it to canned "expected
| output" files.  stderr *may* have a marginal role there, but only for
| exceedingly rare "what the f**k?!  this isn't just wrong, it's an utterly
| unexpected disaster!" cases.  But even those would be better directed to
| stdout, because the stdout comparison mechanism would point them out if they
| happen, and won't forget to look for them.  People do forget, and, as is,
| when one of these stderr cases crops up, regrtest.py doesn't realize the
| test failed either.

You're right, I should have looked at the stuff before saying that.
The "prints" I was really thinking of turn out to be in regrtest.py
itself.

	Donn Cave, donn at u.washington.edu


From juergen.erhard at gmx.net  Wed Feb  7 01:11:07 2001
From: juergen.erhard at gmx.net (Jürgen A. Erhard)
Date: Wed, 07 Feb 2001 01:11:07 +0100
Subject: Interested in a Crypto-SIG?
In-Reply-To: <slrn978o29.gn.amk@mira.erols.com>
References: <3A72C8A3.C4B4D69A@stroeder.com> <slrn97650f.i1.amk@mira.erols.com> <3A731F26.15461A27@stroeder.com> <slrn978o29.gn.amk@mira.erols.com>
Message-ID: <07022001.1@wanderer.local.jae.ddns.org>

>>>>> "A" == A M Kuchling <amk at mira.erols.com> writes:

    A> On Sat, 27 Jan 2001 20:19:02 +0100, 
    A> 	Michael Str?der <michael at stroeder.com> wrote:
    >> <P>The list is hosted inside the US, as are its archives, so
    >> to avoid falling afoul of the US export restrictions, it's
    >> recommended that postings not contain complete programs or modules. 
    >> 
    >> This does not attract me very much.

    A> Given the relaxing of US export laws last year, this text is
    A> really no longer necessary, and I should edit the list
    A> description accordingly.  (Shows how much time I spend on
    A> crypto these days...)

It's not export *laws* it's just the *regulations* of the *current*
Administration.

And, as we all know, the Administration just changed...

I'm not expecting too much good from *that* guy.  And I'd *certainly*
wait with moving in crypto-related matters until those regulations had
turned into real, Senate/HR-fortified, law.

Sorry if I spoiled you day,

Bye, J

-- 
J?rgen A. Erhard    juergen.erhard at gmx.net   phone: (GERMANY) 0721 27326
     MARS: http://members.tripod.com/Juergen_Erhard/mars_index.html
          George Herrimann's Krazy Kat (http://www.krazy.com)
  "Windows NT" is an acronym for "Windows? No thanks." -- Russ McManus



From michael at stroeder.com  Wed Feb  7 15:44:47 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Wed, 07 Feb 2001 15:44:47 +0100
Subject: Interested in a Crypto-SIG?
References: <3A72C8A3.C4B4D69A@stroeder.com> <slrn97650f.i1.amk@mira.erols.com> <3A731F26.15461A27@stroeder.com> <slrn978o29.gn.amk@mira.erols.com> <mailman.981511010.7121.python-list@python.org>
Message-ID: <3A815F5F.C37B76E2@stroeder.com>

"J?rgen A. Erhard" wrote:
> 
>     A> On Sat, 27 Jan 2001 20:19:02 +0100,
>     A>  Michael Str?der <michael at stroeder.com> wrote:
>     >> <P>The list is hosted inside the US, as are its archives, so
>     >> to avoid falling afoul of the US export restrictions, it's
>     >> recommended that postings not contain complete programs or modules.
>     >>
>     >> This does not attract me very much.
> 
>     A> Given the relaxing of US export laws last year, this text is
>     A> really no longer necessary,
> 
> And I'd *certainly*
> wait with moving in crypto-related matters until those regulations had
> turned into real, Senate/HR-fortified, law.

The solution for a python-crypto list is:

  http://listserv.surfnet.nl/archives/python-crypto.html

Ciao, Michael.


From drewpc at colorado.edu  Thu Feb  1 22:44:59 2001
From: drewpc at colorado.edu (PC Drew)
Date: Thu, 01 Feb 2001 21:44:59 GMT
Subject: Interested in a Crypto-SIG?
References: <3A72C8A3.C4B4D69A@stroeder.com>
Message-ID: <95clco$ecq$1@nnrp1.deja.com>

In article <3A72C8A3.C4B4D69A at stroeder.com>,
  Michael =?iso-8859-1?Q?Str=F6der?= <michael at stroeder.com> wrote:
> HI!
>
> (I know that SIG proposals should go to the META-SIG list but I
> guess not many developers are subscribed to this list. Therefore I
> start the discussion here.)
>
> I would like to propose a Crypto-SIG. The goal is to define Pythonic
> class APIs for various cryptographic and related standards to enable
> developers to use a unified API for implementing crypto-enabled
> applications in Python independent of the underlying implementations
> (basically inspired by Sun's Java crypto-related APIs).
>

I'm the one who proposed the idea to the meat-sig list the other day.
Since then, I've joined the python-crypto list and have been having
interesting discussions regarding this idea.

There seems to be a genuine interest in this idea and I think that we
should start drafting up an API.  I'm starting to look into amkCrypto
and m2crypto heavily so that I can see what things I haven't thought
about.  I hope to have some sort of draft in the next week. I'll
coordinate this with you.

Also, I think once we get started on this, it'll be easier to try and
get a SIG.  I too think that it'd be better to get a SIG!!


Sent via Deja.com
http://www.deja.com/


From schlatterbeck at tttech.com  Mon Feb  5 13:28:33 2001
From: schlatterbeck at tttech.com (Ralf Schlatterbeck)
Date: 5 Feb 2001 13:28:33 +0100
Subject: Python for embedded systems (was Re: What does Python do)
References: <94gmeu$4fc$1@sooner.brightok.net> <t1Sa6.4569$Yv.544288@juliett.dax.net> <94gvv5$7mo$1@sooner.brightok.net> <LPUa6.4598$Yv.546853@juliett.dax.net> <2F4DC6312B7FE9DA.74AB95EBEF3C09D4.1D6BDBE5660A1914@lp.airnews.net> <eqZa6.1433$eI2.490960@ruti.visi.com> <3A6D1E04.AABCC7DE@engcorp.com> <riib6.1969$eI2.601187@ruti.visi.com> <3A6E66EE.38FEB4F0@engcorp.com> <E14NBls-0000PP-00@ttt14.tttech.com> <3A731E86.9FB8930D@fxtech.com>
Message-ID: <95m69h$768$1@ttt14.tttech.com>

Peter Hansen <peter at engcorp.com> writes:
> We have Python running on a 1MB Flash/1MB RAM PC/104 module.
> With the compiler.  Quite cool.  I know from posts here that
> at least one other company has achieved something similar.  
> I'm still pleased to find how well many of Python's advantages
> translate through into the embedded world.  This is, of course,
> an extremely long way from a 16-bit chip with a 64K address
> space, but for that there's always FORTH. :-)

I replied to Peter by email and asked him:
> How did you do it? Will you make this publicly available?

He asked me to make his answer available in this group, here it is:

Peter Hansen <peter at engcorp.com> writes:

Perhaps you would be good enough to post my reply to the forum for me,
as the  system on which I'm typing this has no Usenet access.  Thanks.

Making Python run on a PC104 system with 1MB-flash was not at
all difficult.  Among the few things we had to do was to strip
out a large number of "unnecessary" (to us) modules, including
floating point, files, long integers, all platform-specific   
stuff, and Unicode.  We had to provide a customized thread.h  
for our third-party real-time kernel.  Maybe the only thing   
that was at all difficult to figure out was how to modify     
marshal.c to avoid allocating a buffer in RAM and copying     
the code object there when a module is imported (since we     
are running code directly from flash, unlike many PC104       
systems where the flash is only used as a file system and     
all code is loaded into RAM before executing).  The final     
item to produce a useful environment was to provide a
serial port implementation of sys.stdout and sys.stdin
for the interactive interpreter.  It is very gratifying
to see the ">>> " prompt show up in Hyperterminal on a 
PC after all the hard work. :)

Our intention was to make this publicly available, but it
is not yet in any shape to do so.  Furthermore, because of
a change in strategy with our product, we are likely to   
upgrade to a more powerful processor board which will     
require migrating to a more typical configuration with    
flash file system and code running from RAM.  Rather      
than release the coarse results of our work so far,       
I would be happy to respond to individuals with
non-frivolous requests for specific assistance,
by helping point out the areas where we've made changes
and providing 'patches' or portions of our work
rather than the whole thing.  (Packaging up an 
entire work product for release, even as a free
'no warranty' type of thing, takes a lot of doing
from a corporate point of view, and I'm sorry to have
to say I can't spare the time do that right now.)

(Once we have our final configuration polished up and
at least our first product based on it released, I   
definitely intend to educate my upper management     
about why supporting the open-source community by    
contributing back to it is a Good Thing.  It really  
shouldn't be a hard sell, but it has to be done
in just the right way, as you may know. :-)

-- 
Dr. Ralf Schlatterbeck                  mailto:schlatterbeck at tttech.com
TTTech Computertechnik AG, Sch?nbrunnerstra?e 7, A-1040 Vienna, Austria
Voice: +43/1/5853434-0,  Fax: +43/1/5853434-90,  http://www.tttech.com/


From ngps at madcap.dyndns.org  Thu Feb  1 18:05:00 2001
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 1 Feb 2001 17:05:00 GMT
Subject: SSL EOF
References: <mailman.980662149.14963.python-list@python.org> <980790776.835347129@news.silcom.com> <3A75DEB9.308FE746@san.rr.com> <980877001.384793985@news.silcom.com>
Message-ID: <95c4vs$kss$1@violet.singnet.com.sg>

According to Clarence Gardner  <clarence at netlojix.com>:
> Here is the relevant code.  I don't know what error==6 signifies, but even
> if the other side closed improperly, I don't think that's so significant as to
> warrant making these sockets behave differently from non-ssl ones.
> 
>     count = SSL_read(self->ssl, PyString_AsString(buf), len);
>     res = SSL_get_error(self->ssl, count);
> 
>     switch (res) {
>     case 0:         /* Good return value! */
>         break;
>     case 6:
>         PyErr_SetString(SSLErrorObject, "EOF");
>         Py_DECREF(buf);
>         return NULL;
>         break;
>     case 5:
>     default:
>         return PyErr_SetFromErrno(SSLErrorObject);
>         break;
> 

For comparison, here's what M2Crypto's ssl_read() looks like:

PyObject *ssl_read(SSL *ssl, int num) {
    PyObject *obj;
    void *buf;
    int r, err;

    if (!(buf = PyMem_Malloc(num))) {
        PyErr_SetString(PyExc_MemoryError, "ssl_read");
        return NULL;
    }
    Py_BEGIN_ALLOW_THREADS
    r = SSL_read(ssl, buf, num);
    Py_END_ALLOW_THREADS
    switch (SSL_get_error(ssl, r)) {
        case SSL_ERROR_NONE:
        case SSL_ERROR_ZERO_RETURN:
            buf = PyMem_Realloc(buf, r);
            obj = PyString_FromStringAndSize(buf, r);
            break;
        case SSL_ERROR_WANT_WRITE:
        case SSL_ERROR_WANT_READ:
        case SSL_ERROR_WANT_X509_LOOKUP:
            Py_INCREF(Py_None);
            obj = Py_None;
            break;
        case SSL_ERROR_SSL:
            PyErr_SetString(_ssl_err, ERR_reason_error_string(ERR_get_error()));
            obj = NULL;
            break;
        case SSL_ERROR_SYSCALL:
            err = ERR_get_error();
            if (err)
                PyErr_SetString(_ssl_err, ERR_reason_error_string(err));
            else if (r == 0)
                PyErr_SetString(_ssl_err, "unexpected eof");
            else if (r == -1)
                PyErr_SetFromErrno(_ssl_err);
            obj = NULL;
            break;
    }
    PyMem_Free(buf);
    return obj;
}


And Python code using that, from M2Crypto's https extension to Medusa:

    def recv(self, buffer_size):
        try:
            result = self.socket._read_nbio(buffer_size)
            if result is None:
                return ''
            elif result == '':
                self.socket.set_shutdown(SSL.SSL_SENT_SHUTDOWN|SSL.SSL_RECEIVED_SHUTDOWN)
                return ''
            else:
                self.server.bytes_in.increment(len(result))
                return result
        except SSL.SSLError, why:
            self.socket.set_shutdown(SSL.SSL_SENT_SHUTDOWN|SSL.SSL_RECEIVED_SHUTDOWN)
            self.close()
            self.log_info('recv: closing channel %s %s' % (repr(self), why))
            return ''


So, in M2Crypto, ssl_read() does return a '' on eof; if the other side
failed to close properly, an SSL.SSLError exception is raised with the
error "unexpected eof".

M2Crypto works with both Python 1.5.2 and Python 2.0. Get it here:

    http://www.post1.com/home/ngps/m2

Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From tim.one at home.com  Fri Feb  2 00:28:53 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 18:28:53 -0500
Subject: SSL EOF
In-Reply-To: <95c4vs$kss$1@violet.singnet.com.sg>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEKMIMAA.tim.one@home.com>

[Ng Pheng Siong]
> ...
> So, in M2Crypto, ssl_read() does return a '' on eof;
> ...

The good news is that it will in Python 2.1a2 also (to be released
tomorrow).  However, the more we (Jeremy, Guido, me) looked at Python's SLL
support, the more questionable code we found.  Alas, not being experts in
this protocol, we're unlikely to fix it.  So we would be grateful for a
patch against CVS Python for doing this stuff more intelligently, including
corresponding repairs to the httplib.py code that uses it.

"case 6:"-brrrrrrrr-ingly y'rs  - tim




From clarence at netlojix.com  Fri Feb  2 20:16:05 2001
From: clarence at netlojix.com (Clarence Gardner)
Date: Fri, 2 Feb 2001 11:16:05 -0800
Subject: SSL EOF
References: <mailman.981087116.16733.python-list@python.org>
Message-ID: <981141471.767582214@news.silcom.com>

On Thu, 01 Feb 2001, Tim Peters wrote:
>[Ng Pheng Siong]
>> ...
>> So, in M2Crypto, ssl_read() does return a '' on eof;
>> ...
>
>The good news is that it will in Python 2.1a2 also (to be released
>tomorrow).  However, the more we (Jeremy, Guido, me) looked at Python's SLL
>support, the more questionable code we found.  Alas, not being experts in
>this protocol, we're unlikely to fix it.  So we would be grateful for a
>patch against CVS Python for doing this stuff more intelligently, including
>corresponding repairs to the httplib.py code that uses it.
>
>"case 6:"-brrrrrrrr-ingly y'rs  - tim

Glad to hear about the EOF behavior change.

To: The hypothetical person who knows about this stuff
    and will do it for Python
From: A person who doesn't, but did anyway

Yes, I implemented an SSL module for Python 1.5, and always felt
guilty about not contributing it.  I didn't because I implemented it
without being able to find any documentation on SSL, and though it
worked for my purposes, I have no idea how many holes it has.

Here are some things to keep in mind when thinking of a re-do for
Python:
  - The 2.0 stuff doesn't include a shutdown method.  I had one in
    my implementation, but since I didn't know what I was doing, I
    just had it do a shutdown on the socket.  It worked, but is
    clearly wrong.
  - As far as I can tell, the 2.0 stuff only allows the Python program
    to act as a client.  Remembering my trial-and-error method of
    implementing server side, I can see why.  :)

                                  ###

I noticed that the M2Crypto code that was posted still, at the Python
level, doesn't treat encrypted sockets the same as the others (e.g.,
returning a None result in some circumstances).  I may be out of touch
with the rest of the community, (or just objectively wrong), in thinking
that they should...  It seems to me that the application probably just
cares about whether it can communicate with the other end, not which of
the zillion different ways an ssl channel can go wrong.
----
Clarence Gardner
Software Engineer
NetLojix Communications
clarence at netlojix.com



From ngps at madcap.dyndns.org  Sat Feb  3 04:55:05 2001
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 3 Feb 2001 03:55:05 GMT
Subject: SSL EOF
References: <mailman.981087116.16733.python-list@python.org> <981141471.767582214@news.silcom.com>
Message-ID: <95fvep$tqb$1@clematis.singnet.com.sg>

According to Clarence Gardner  <clarence at netlojix.com>:
>   - The 2.0 stuff doesn't include a shutdown method.  I had one in
>     my implementation, but since I didn't know what I was doing, I
>     just had it do a shutdown on the socket.  It worked, but is
>     clearly wrong.

M2Crypto provides a shutdown() method for an SSL socket which calls
SSL_shutdown() which DTRT.

>   - As far as I can tell, the 2.0 stuff only allows the Python program
>     to act as a client.  Remembering my trial-and-error method of
>     implementing server side, I can see why.  :)

M2Crypto supports forking, threading, and async-io servers. 

(Sorry for the commercial. ;-)


> I noticed that the M2Crypto code that was posted still, at the Python
> level, doesn't treat encrypted sockets the same as the others (e.g.,
> returning a None result in some circumstances).  

None is overloaded to mean "try again" in non-blocking mode only.
In blocking mode OpenSSL promises (heh) to not reach that code path.

I could've invented an arbitrary object type, or return a tuple 
(return status, data), but they seem like too much work. So I overload
None. ;-)


> It seems to me that the application probably just
> cares about whether it can communicate with the other end, not which of
> the zillion different ways an ssl channel can go wrong.

I disagree. Since you're using SSL, presumably you care about security.
An unexpected shutdown on the other side is a security event, and you'll
want to know about it.

Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From ngps at madcap.dyndns.org  Sat Feb  3 04:43:07 2001
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 3 Feb 2001 03:43:07 GMT
Subject: SSL EOF
References: <mailman.981087116.16733.python-list@python.org>
Message-ID: <95fuob$t3q$1@clematis.singnet.com.sg>

According to Tim Peters <tim.one at home.com>:
> The good news is that it will in Python 2.1a2 also (to be released
> tomorrow).  However, the more we (Jeremy, Guido, me) looked at Python's SLL
> support, the more questionable code we found.  Alas, not being experts in
> this protocol, we're unlikely to fix it.  So we would be grateful for a
> patch against CVS Python for doing this stuff more intelligently, including
> corresponding repairs to the httplib.py code that uses it.

I'll see what I can do. (I recall saying that when 2.0 came out... ;-)

Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From rvprasad at cis.ksu.edu  Thu Feb  1 23:42:43 2001
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 01 Feb 2001 16:42:43 -0600
Subject: PyXML
References: <m37l3f76i7.fsf@boss.dreamsoft.com> <3A776B59.FDEC3975@ogbuji.net>
Message-ID: <m3snlyxbx8.fsf@boss.dreamsoft.com>

Hi,

The example is 

from xml.dom.ext.reader.Sax2 import Reader
from xml.dom.ext import PrettyPrint
import sys

i = open(sys.argv[1])
j = Reader(1).fromStream(i)
i.close()
PrettyPrint(j)

print "++++", j.doctype.__systemId, "----"

Nothing is printed between ++++ and ----

Also, shouldn't Line 61 in _xmlplus/dom/NamedNodeMap.py be 
	return self.get(('', name))
and Line 65 in _xmlplus/xpath/XPathParserBase.py be
	if XPath.cvar.g_errorOccured == 1:

-- 
Venkatesh Prasad Ranganath


From greg at cosc.canterbury.ac.nz  Thu Feb  1 05:19:10 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 01 Feb 2001 17:19:10 +1300
Subject: curly-brace-aphobic?
References: <3A74EBD6.3C87FD15@seebelow.org> <slrn97a9fu.7sf.clgonsal@keeshah.penguinpowered.com> <3A757FBE.E54C5ECD@seebelow.org> <mailman.980789494.4239.python-list@python.org> <slrn97c6bs.bjo.scarblac@pino.selwerd.nl> <mailman.980870616.378.python-list@python.org> <95718r018s8@drn.newsguy.com>
Message-ID: <3A78E3BE.AED8CABF@cosc.canterbury.ac.nz>

Grant Griffin wrote:
> 
> However, the rule that I seem to want to apply is "All indexers are indexed
> using the style of brace used to bind them."

But tuples and strings aren't constructed using any
kind of brackets at all:

   t = 1,2,3
   s = "parrot"

How do you propose to index those, using your rule?

What's more, there are many more indexable things than
the built-in ones. Extension modules can implement new
indexable types, and any class can be made indexable
by giving it a __getitem__ method.

If you want to index every indexable thing using a
unique syntax, you're going to run out of characters
very quickly!

It could be argued, I suppose, that sequence-like
objects should be indexed with [] and mapping-like
objects with {}. But is a class which has __getitem__
a sequence or a mapping? 

You could say that there should be two methods, such as 
__getitem__ for [] and __mapitem__ for {}. But that 
seems like needlessly multiplying entities. Why go to
all that trouble to provide two different indexing
notations when one is quite sufficient?

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From cjc26 at nospam.cornell.edu  Thu Feb  1 16:00:24 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Thu, 01 Feb 2001 15:00:24 GMT
Subject: curly-brace-aphobic?
References: <3A74EBD6.3C87FD15@seebelow.org> <slrn97a9fu.7sf.clgonsal@keeshah.penguinpowered.com> <3A757FBE.E54C5ECD@seebelow.org> <mailman.980789494.4239.python-list@python.org> <slrn97c6bs.bjo.scarblac@pino.selwerd.nl> <mailman.980870616.378.python-list@python.org> <95718r018s8@drn.newsguy.com> <3A78E3BE.AED8CABF@cosc.canterbury.ac.nz>
Message-ID: <cSee6.1237$o91.133183@typhoon.nyroc.rr.com>

* Greg Ewing <greg at cosc.canterbury.ac.nz> menulis:
| Grant Griffin wrote:
| > 
| > However, the rule that I seem to want to apply is "All indexers are indexed
| > using the style of brace used to bind them."
| 
| But tuples and strings aren't constructed using any
| kind of brackets at all:
| 
|    t = 1,2,3
|    s = "parrot"
| 
| How do you propose to index those, using your rule?

Well, >obviously<, you would do the following:

>>> t 2
2
>>> s"4"
r

Of course, you have to be careful with tuples, because sometimes
parentheses are used when creating them:

>>> t2 = (4, 5, 6)
>>> t2(3)
6


-- 
Cliff Crawford               http://www.people.cornell.edu/pages/cjc26/
                             print "Just another Python hacker"


From quinn at cruzeiro.ugcs.caltech.edu  Thu Feb  1 00:40:50 2001
From: quinn at cruzeiro.ugcs.caltech.edu (Quinn Dunkan)
Date: 31 Jan 2001 23:40:50 GMT
Subject: curly-brace-aphobic?
References: <3A74EBD6.3C87FD15@seebelow.org> <slrn97a9fu.7sf.clgonsal@keeshah.penguinpowered.com> <3A757FBE.E54C5ECD@seebelow.org> <mailman.980789494.4239.python-list@python.org> <slrn97c6bs.bjo.scarblac@pino.selwerd.nl> <mailman.980870616.378.python-list@python.org> <slrn97eob8.gjm.scarblac@pino.selwerd.nl>
Message-ID: <slrn97h8k2.aod.quinn@cruzeiro.ugcs.caltech.edu>

On 31 Jan 2001 00:57:48 GMT, Remco Gerlich <scarblac at pino.selwerd.nl> wrote:
>Well, writing "string.split" is inconsistent, shouldn't it be string[split]
>or string["split"]?
>
>It's like indexing the module. It is actually a dict lookup internally.
>They're not the same thing really since an identifier suddenly becomes a
>string and so on, but it's just a connection my mind made...

You'd like lua, then.  Everything is a table.  Modules (which are not built
into the langauge, naturally) are tables.  Arrays are tables indexed with
numbers.  foo.bar is sugar for foo['bar'].  It's theoretically appealing, but
sometimes causes additional complexity, when, for example, you want to know
how long an array is (the convention is to have an 'n' key with the length).

And tables are created with {} but indexed with [] :)


From mwh21 at cam.ac.uk  Thu Feb  1 13:28:50 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Feb 2001 12:28:50 +0000
Subject: Unicode pb in Python 2.0 solved - no thanks to WinPython
References: <95b65r$l27$1@wanadoo.fr>
Message-ID: <m3d7d27fjx.fsf@atrus.jesus.cam.ac.uk>

"Jurie Horneman" <jhorSPAMBGONneman at wanadoo.fr> writes:

> I would very much like to "RTFM", and believe me I tried, but there isn't
> much M to FR. How could I have avoided this problem with less agony? Should
> I report the bug to the authors of WinPython?

Do you mean PythonWin?  I would, yes.

Cheers,
M.

-- 
31. Simplicity does not precede complexity, but follows it. 
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From fuess at att.net  Sat Feb  3 16:07:23 2001
From: fuess at att.net (David A. Fuess)
Date: Sat, 03 Feb 2001 15:07:23 GMT
Subject: Unicode pb in Python 2.0 solved - no thanks to WinPython
References: <95b65r$l27$1@wanadoo.fr> <m3d7d27fjx.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <GR58Otg8OgaaZPJ2DkrjTxJfjA5Y@4ax.com>

Actually this is not a bug in PythonWin, although it might appear to
be. What's happened is that PythonWin does not reload an already
loaded module until told to do so. You do that with the save and
reload icon on the toolbar (right next to the save and run icon).

Dave

Michael Hudson <mwh21 at cam.ac.uk> wrote:

>"Jurie Horneman" <jhorSPAMBGONneman at wanadoo.fr> writes:
>
>> I would very much like to "RTFM", and believe me I tried, but there isn't
>> much M to FR. How could I have avoided this problem with less agony? Should
>> I report the bug to the authors of WinPython?
>
>Do you mean PythonWin?  I would, yes.
>
>Cheers,
>M.



From mwh21 at cam.ac.uk  Sat Feb  3 17:51:47 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Feb 2001 16:51:47 +0000
Subject: Unicode pb in Python 2.0 solved - no thanks to WinPython
References: <95b65r$l27$1@wanadoo.fr> <m3d7d27fjx.fsf@atrus.jesus.cam.ac.uk> <GR58Otg8OgaaZPJ2DkrjTxJfjA5Y@4ax.com>
Message-ID: <m31ytf676k.fsf@atrus.jesus.cam.ac.uk>

David A. Fuess <fuess at att.net> writes:

> Actually this is not a bug in PythonWin, although it might appear to
> be. What's happened is that PythonWin does not reload an already
> loaded module until told to do so. You do that with the save and
> reload icon on the toolbar (right next to the save and run icon).

As I understood the report, the problem was PythonWin ignoring
SyntaxErrors encountered during startup - which sounds like a bug in
PythonWin to me.

Cheers,
M.

-- 
  If you don't use emacs, you're a pathetic, mewling, masochistic
  weakling and I can't be bothered to convert you.    -- Ron Echeverri


From loewis at informatik.hu-berlin.de  Fri Feb  2 22:20:16 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:20:16 +0100
Subject: Displaying Unicode text in Tkinter Canvas widget
References: <3A75D600.60F7CD0D@mega-nerd.com> <3A769B29.77C3CE84@mega-nerd.com>
Message-ID: <p6q66ispysv.fsf@informatik.hu-berlin.de>

Erik de Castro Lopo <nospam at mega-nerd.com> writes:

> I'm sorry, I should have mentioned that I'm trying to display
> the Hiragana as text in a Tkinter Canvas widget.

Why does it have to be a Canvas? I'd suggest to start with a Label, or
Text widget.

>>> import Tkinter
>>> t=Tkinter.Label(text=u"\N{HIRAGANA LETTER KA}")
>>> t.pack()                                                                    
works fine for me (on Linux).

Regards,
Martin


From mikael at isy.liu.se  Fri Feb  2 11:50:57 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 02 Feb 2001 11:50:57 +0100 (MET)
Subject: PEP-0218 (Was: RE: Sets in Python)
In-Reply-To: <LNBBLJKPBEHFEDALKOLCCEFAIMAA.tim.one@home.com>
Message-ID: <XFMail.010202115057.mikael@isy.liu.se>

Tim, Greg and others,

On 31-Jan-01 Tim Peters wrote:
 >  Note Greg Wilson's PEP for adding a set type to Python:
 >  
 >      http://python.sourceforge.net/peps/pep-0218.html

This looks promising. I am especially glad to see that the PEP proposes
the logical operations '|', '&', and '^' to implement union, intersection 
and symmetric difference. However, I'm not sure I like the proposed use
of '-' for difference. Mathematician use '\' (so called setminus). That 
would have been my favourite, but we use '\' for so many things, and 
especially not as an arithmetic operation. The visually closest thing 
would be '/', which is what I have used myself. That would for instance 
make it possible to let -S mean {-x for x in S}.

Let's just hope that there will be a UserSet class that I can subclass, 
for instance to implement direct sums as A+B and direct difference as
A-B (interpreted as A+(-B) with -B as above), provided that '-' is not
used for difference. I could also implement cartesian products as A*B. 
Someone else could perhaps think of a resonable way to interprete the 
modulo operator (%) in terms of sets...

It sure does look promising.

I've-said-this-before-
programming-is-doing-math-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    02-Feb-01
Time:    11:18:01

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From michael at stroeder.com  Fri Feb  2 18:38:19 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 02 Feb 2001 18:38:19 +0100
Subject: PEP-0218 (Was: RE: Sets in Python)
References: <mailman.981111125.24370.python-list@python.org>
Message-ID: <3A7AF08B.DB88EC56@stroeder.com>

Mikael Olofsson wrote:
> 
> On 31-Jan-01 Tim Peters wrote:
>  >  Note Greg Wilson's PEP for adding a set type to Python:
>  >
>  >      http://python.sourceforge.net/peps/pep-0218.html
> 
> This looks promising. I am especially glad to see that the PEP proposes
> the logical operations '|', '&', and '^' to implement union, intersection
> and symmetric difference. However, I'm not sure I like the proposed use
> of '-' for difference.

I second that. '-' could also be associated to be S^-1 (inverse set)
of set S.

Ciao, Michael.


From mlh at idi.ntnu.no  Thu Feb  1 01:56:59 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 1 Feb 2001 01:56:59 +0100
Subject: Sets in Python
References: <mailman.980905187.10012.python-list@python.org>
Message-ID: <95aceb$s79$1@tyfon.itea.ntnu.no>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.980905187.10012.python-list at python.org...
[...]
> Note Greg Wilson's PEP for adding a set type to Python:

Nice!

Any chance that this may be included some time in the future?

And as regards "... in dict"; It would perhaps be more natural
to see a dict as a set of pairs, so that the dictionary would
act similarly to dict.items()?

I.e. "(key, val) in dict" would be the same as
"(key, val) in dict.items()" except that it would take O(1)
time...?

Oh, well... If we get a set type I guess that isn't necessary.

(And I would *love* to have a set type. :-)))

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From l.szyster at ibm.net  Thu Feb  1 16:29:47 2001
From: l.szyster at ibm.net (Laurent Szyster)
Date: Thu, 01 Feb 2001 16:29:47 +0100
Subject: Sets in Python
References: <956hcd$qef$1@tyfon.itea.ntnu.no>
Message-ID: <3A7980EA.F16596D5@ibm.net>

Magnus Lie Hetland wrote:
> 
> I quite often use dictionaries as sets in Python... But
> it doesn't really seem pretty enough to me... Some
> arbitrariness (as with "while 1: ... break" I still
> don't like that one...) about using what to me is a
> value as a key, and assigning it some value -- like 1,
> for instance... And then checking it with has_key, and
> removing it with del...

if you are looking for set manipulations, get the
kjbuckets module and do:

>>> from kjbuckets import kjSet
>>> A = kjSet ([1,2,3,4,'X','Y'])
>>> A.has_key ('X')
1
>>> B = kjSet ([1,2,5,'Z'])
>>> A-B
kjSet([3, 4, 'X', 'Y'])
>>> A+B
kjSet([1, 2, 3, 4, 'X', 'Y', 5, 'Z'])
>>> A&B
kjSet([1, 2])

The module also provides a graph class (and everything
you need to write your own relational algebra, since
it's been developped by the Gadfly author).


Laurent Szyster


From phd at phd.pp.ru  Fri Feb  2 10:26:07 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 2 Feb 2001 12:26:07 +0300 (MSK)
Subject: Sets in Python
In-Reply-To: <3A7980EA.F16596D5@ibm.net>
Message-ID: <Pine.LNX.4.30.0102021225160.11957-100000@fep132.fep.ru>

On Thu, 1 Feb 2001, Laurent Szyster wrote:
> if you are looking for set manipulations, get the
> kjbuckets module and do:

   Anyone succeed compiling kjbuckets for Python 2.0?

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From ray_drew at yahoo.co.uk  Fri Feb  2 10:48:11 2001
From: ray_drew at yahoo.co.uk (Ray Drew)
Date: Fri, 02 Feb 2001 09:48:11 GMT
Subject: Sets in Python
References: <mailman.981106027.22946.python-list@python.org>
Message-ID: <3a7a8253.1394768@reading.news.pipex.net>

On Fri, 2 Feb 2001 12:26:07 +0300 (MSK), Oleg Broytmann
<phd at phd.pp.ru> wrote:

>On Thu, 1 Feb 2001, Laurent Szyster wrote:
>> if you are looking for set manipulations, get the
>> kjbuckets module and do:
>
>   Anyone succeed compiling kjbuckets for Python 2.0?
>
>Oleg.
>----
>     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
>           Programmers don't die, they just GOSUB without RETURN.
>
>
 www.malcolmson.com/python


From phd at phd.pp.ru  Fri Feb  2 11:20:13 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 2 Feb 2001 13:20:13 +0300 (MSK)
Subject: Sets in Python
In-Reply-To: <3a7a8253.1394768@reading.news.pipex.net>
Message-ID: <Pine.LNX.4.30.0102021318390.11957-100000@fep132.fep.ru>

On Fri, 2 Feb 2001, Ray Drew wrote:
> >   Anyone succeed compiling kjbuckets for Python 2.0?
> >
>  www.malcolmson.com/python

   Thank you very much! It is not what I searched (I had problem on UNIX),
but anyway it's nice to know someone succeed compiling it. If he did it on
Windows I should build it on UNIX, I suppose :) Will try harder.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From phd at phd.pp.ru  Fri Feb  2 12:44:34 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 2 Feb 2001 14:44:34 +0300 (MSK)
Subject: kjbuckets for Python 2.0 (was Re: Sets in Python)
In-Reply-To: <Pine.LNX.4.30.0102021318390.11957-100000@fep132.fep.ru>
Message-ID: <Pine.LNX.4.30.0102021432110.11957-100000@fep132.fep.ru>

Hello!

On Fri, 2 Feb 2001, Oleg Broytmann wrote:
> On Fri, 2 Feb 2001, Ray Drew wrote:
> > >   Anyone succeed compiling kjbuckets for Python 2.0?
> > >
> >  www.malcolmson.com/python
>
>    Thank you very much! It is not what I searched (I had problem on UNIX),
> but anyway it's nice to know someone succeed compiling it. If he did it on
> Windows I should build it on UNIX, I suppose :) Will try it harder.

   Ok, I managed to build kjbuckets for Python 2.0. I needed rename2.h from
Python 1.5.2 :) and I was forced (by gcc, perhaps) to replace all occurence
of OVERFLOW to OVERFLOWx (the replace string was choosed arbitrary - I just
choosed a name to satisfy compiler).
   Aaron, any chance the kjbuckets module will be officially updated? Or
you lost interests and the module dropped on the floor?

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From bhoel at starship.python.net  Fri Feb  2 21:36:27 2001
From: bhoel at starship.python.net (Berthold Höllmann)
Date: 02 Feb 2001 21:36:27 +0100
Subject: kjbuckets for Python 2.0 (was Re: Sets in Python)
References: <mailman.981114302.25279.python-list@python.org>
Message-ID: <m2puh0u8j8.fsf@pchoel.privat.uni-hamburg.de>

Oleg Broytmann <phd at phd.pp.ru> writes:

> Hello!
> 
> On Fri, 2 Feb 2001, Oleg Broytmann wrote:
> > On Fri, 2 Feb 2001, Ray Drew wrote:
> > > >   Anyone succeed compiling kjbuckets for Python 2.0?
> > > >
> > >  www.malcolmson.com/python
> >
> >    Thank you very much! It is not what I searched (I had problem on UNIX),
> > but anyway it's nice to know someone succeed compiling it. If he did it on
> > Windows I should build it on UNIX, I suppose :) Will try it harder.
> 
>    Ok, I managed to build kjbuckets for Python 2.0. I needed rename2.h from
> Python 1.5.2 :) and I was forced (by gcc, perhaps) to replace all occurence
> of OVERFLOW to OVERFLOWx (the replace string was choosed arbitrary - I just
> choosed a name to satisfy compiler).
>    Aaron, any chance the kjbuckets module will be officially updated? Or
> you lost interests and the module dropped on the floor?

I don't know about Aaron, but I got it compiled with python 2.0
without copying rename2.h. The changes were quite straightforward, but
a "diff -c" file is larger than the original (-: The simple diff is
49K, a little large for the newsgroup/mailinglist. I would put it onto
starship, but I have problems with my account (what the heck was my
password). So I will send the file kjbucketsmodule.c to anyone
interested and without any warranty. Just drop me an email.

Greetings

Berthold
-- 
bhoel at starship.python.net / http://starship.python.net/crew/bhoel/
        It is unlawful to use this email address for unsolicited ads
        (USC Title 47 Sec.227). I will assess a US$500 charge for
        reviewing and deleting each unsolicited ad.


From phd at phd.pp.ru  Mon Feb  5 11:03:59 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 5 Feb 2001 13:03:59 +0300 (MSK)
Subject: kjbuckets for Python 2.0
In-Reply-To: <m2puh0u8j8.fsf@pchoel.privat.uni-hamburg.de>
Message-ID: <Pine.LNX.4.30.0102051303230.32462-100000@fep132.fep.ru>

On 2 Feb 2001, Berthold H?llmann wrote:
> I don't know about Aaron, but I got it compiled with python 2.0
> without copying rename2.h. The changes were quite straightforward, but
> a "diff -c" file is larger than the original (-: The simple diff is
> 49K, a little large for the newsgroup/mailinglist. I would put it onto
> starship, but I have problems with my account (what the heck was my
> password). So I will send the file kjbucketsmodule.c to anyone
> interested and without any warranty. Just drop me an email.

   Please, send it to me, I'll test it and send furter to Aaron.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From hinsen at cnrs-orleans.fr  Mon Feb  5 17:51:58 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 05 Feb 2001 17:51:58 +0100
Subject: MMTK and Scientific python for NT 4.0
References: <3A76D096.B1747887@helsinki.fi>
Message-ID: <m3pugx3wep.fsf@chinon.cnrs-orleans.fr>

Kimmo P??kk?nen <Kimmo.Paakkonen at helsinki.fi> writes:

>   If there are no such binaries, are there some instructions how to
> compile them for NT? I have the free Borland 5.5 compiler, and I have

According to the little I know about Windows, you must use the same
compiler for compiling the Python interpreter and any extension
modules. So the first step would be to recompile Python with your
compiler.

> ), but building MMTK extensions seems confusing, as I don't have
> experience of compiling bigger projects. Any help appreciated!!

I can't help with Windows, sorry. MMTK extension modules use no
special tricks, except that the Setup file is preprocessed before
being handed to the standard Python Makefile.pre.in mechanism. You
might have to do this preprocessing manually under Windows. It
involves replacing the string "@SCIENTIFIC@" by the path to the
directory Scientific/Include of your ScientificPython installation,
and replacing the string "@LONG@" by

1) "-D_LONG64_" if sys.maxint > 2147483647
2) an empty string otherwise

With that information, Windows compilation experts should be able
to tell you what you need to do. And please let me know if you succeed;
I get an inquiry about Windows binaries about once per month.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From sholden at holdenweb.com  Thu Feb  1 01:45:19 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 31 Jan 2001 19:45:19 -0500
Subject: mxDateTime for Python 2.0 for Win32 - binary PYD file
References: <qAAd6.8144$Oe.37720@tor-nn1.netcom.ca> <kEAd6.8145$Oe.37320@tor-nn1.netcom.ca>
Message-ID: <jo2e6.773$j4.36414@e420r-atl2.usenetserver.com>

"Warren Postma" <embed at geocities.com> wrote in message
news:kEAd6.8145$Oe.37320 at tor-nn1.netcom.ca...
> Okay, I think I feel stupid now.
>
> Warren
>
Please feel free to borrow my head-thwacker. But don't keep it tool long, as
I'll undoubtedly need it myself the next time I get too smart for my own
good.

regards
 Steve




From suzuki at acm.org  Sat Feb  3 08:10:00 2001
From: suzuki at acm.org (SUZUKI Hisao)
Date: Sat, 03 Feb 2001 16:10:00 +0900
Subject: Timer module?
References: <3A76F442.7BB0B9A9@americasm01.nt.com>
Message-ID: <95gaik$12vo$1@nwall2.odn.ne.jp>

In article <3A76F442.7BB0B9A9 at americasm01.nt.com>, Rick wrote:
> I wonder if there is already any timer module out there that allows me
> to:
> 
> - set a timer without blocking the current thread

I have written such one :^)

> - cancel the timer

Yes.

> - maybe also, able to tell it to recur at a given interval for so many
> times; and able to change that

Yes, by re-scheduling it (= a Task instance) explicitly.

> - create thousands of timers with a performance penalty that grows
> slower than linearly with the number of timers; preferably a performance
> cost that is small and almost constant

Probably yes; O(1) at best and O(n) at worst theoretically.

The module is pypage/lib/timer.py, which is a part of PyPage
1.4, a handy multi-threaded Web server with JSP-like features.
I wrote and posted it to a newsgroup in Japan last year:
| Newsgroups: fj.sources
| From: SUZUKI Hisao <suzuki at acm.org>
| Subject: PyPage 1.4
| Date: Sun, 12 Nov 2000 23:20:41 +0900
| Message-ID: <8um90n$fc9$1 at nwms2.odn.ne.jp>

You can get it also via HTTP (thanks to TAKAHASHI Hideaki):
 http://www.zob.ne.jp/~hide-t/comp/python/pypage/pypage-1.4.tar.bz2

It is a free software WITHOUT ANY WARRANTY.  Anyone may freely
use, alter and redistribute it.  All I ask is a brief mention of
me as the original author in the source or accompanying file(s).
However, the documentation is in Japanese.

If you don't mind it, I introduce the module to you.  First,
copy two files, timer.py and jthread.py, from pypage/lib/ to the
current directory.  Then, execute the script:

----------------------------------------------------------------
#!/usr/bin/env python
import timer

def foo(msg):
    print msg
    task.when += 3
    timer.schedule(task)                # re-schedule the task

# enable the timer
timer.start()
# create and schedule a task to call foo("spam") 3 seconds later
task = timer.after(3, foo, "spam")

raw_input('hit "Enter" to stop\n')
----------------------------------------------------------------

It will repeat "spam" until you hit Enter key.

Hope this helps.

--
SUZUKI Hisao          >>> def fib(n): return reduce(lambda x, y:
suzuki at acm.org        ... (x,x[0][-1]+x[1]), [()]*n, ((0L,),1L))



From andymac at bullseye.apana.org.au  Thu Feb  1 21:11:25 2001
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Fri, 2 Feb 2001 07:11:25 +1100 (EDT)
Subject: Data management products (was: How do you lock your web app
 data?)
In-Reply-To: <95a49u11uao@news1.newsguy.com>
Message-ID: <Pine.OS2.4.30.0102020710080.679-100000@CENTRAL>

On Wed, 31 Jan 2001, Alex Martelli wrote:

> [To be honest, personally, I'm having a hard time
> thinking up anything I'd *want* to code in Access, other VBA
> platforms, Foxpro, or similar languages, rather than Python --
> but, I'm sure there are niches where such tools help a lot].

The only thing I use it for is its report writer.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From claird at starbase.neosoft.com  Thu Feb  1 21:27:47 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 1 Feb 2001 14:27:47 -0600
Subject: Data management products (was: How do you lock your web app data?)
References: <956uqd$6st$1@newsreaderg1.core.theplanet.net> <958j810o2e@news2.newsguy.com> <D62094CDC48C0FCA.44A0CB1CEB4E1F76.B8EA27DB5535EBAD@lp.airnews.net> <95a49u11uao@news1.newsguy.com>
Message-ID: <6457C4C831E9B110.6049FFDB799E041D.AA4A6C480AE5F526@lp.airnews.net>

In article <95a49u11uao at news1.newsguy.com>,
Alex Martelli <aleaxit at yahoo.com> wrote:
>"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
>news:D62094CDC48C0FCA.44A0CB1CEB4E1F76.B8EA27DB5535EBAD at lp.airnews.net...
>    [snip]
>> Actually, all these Alex has mentioned have
>> capabilities that'll knock the socks off vic-
>> tims habituated to Access, for example.
>
>Just one attempt at terminology clarification...:
>
>Access is a possibly-handy GUI IDE front-end for cobbling together
>possibly-intricated DB operations; at least if you don't mind
			.
			.
			.
>What was in a far, distant past misnamed 'the Access database'
>by Microsoft marketroids (the same misbegotten ilk who managed
>for years to throw confusion around COM by misnaming everything
>COM *as if* it had anything to do with various technologies who were
>and are *ON TOP* of COM itself -- at first, OLE, later, Active/X), is
>now fortunately again known only by its true name, 'Jet'.  It's a
>very old-technology, legacy engine, with no pluses I can think of
>when compared to MSDE, Postgres, etc.
			.
		[other salient points]
			.
			.
Do Java people use property files?  I have no sense
about this.  I think the answer is, no.  In any case,
CPython has no particular need to read and write 'em,
right?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From aleaxit at yahoo.com  Fri Feb  2 09:53:25 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 2 Feb 2001 09:53:25 +0100
Subject: property files (was Re: Data management products (was: How do you lock your web app data?))
References: <956uqd$6st$1@newsreaderg1.core.theplanet.net> <958j810o2e@news2.newsguy.com> <D62094CDC48C0FCA.44A0CB1CEB4E1F76.B8EA27DB5535EBAD@lp.airnews.net> <95a49u11uao@news1.newsguy.com> <6457C4C831E9B110.6049FFDB799E041D.AA4A6C480AE5F526@lp.airnews.net>
Message-ID: <95dsid02v03@news1.newsguy.com>

"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
news:6457C4C831E9B110.6049FFDB799E041D.AA4A6C480AE5F526 at lp.airnews.net...
    [snip]
> Do Java people use property files?  I have no sense
> about this.  I think the answer is, no.  In any case,
> CPython has no particular need to read and write 'em,
> right?

Sorry, what do you mean by "property files" in this
context [discussion of RDBMS engines]?  I think I
may be confused...


Alex





From cobrien at Radix.Net  Fri Feb  2 02:26:15 2001
From: cobrien at Radix.Net (Cary O'Brien)
Date: 1 Feb 2001 20:26:15 -0500
Subject: Data management products (was: How do you lock your web app data?)
References: <956uqd$6st$1@newsreaderg1.core.theplanet.net> <958j810o2e@news2.newsguy.com> <D62094CDC48C0FCA.44A0CB1CEB4E1F76.B8EA27DB5535EBAD@lp.airnews.net> <95a49u11uao@news1.newsguy.com>
Message-ID: <95d2bn$s7b$1@saltmine.radix.net>

In article <95a49u11uao at news1.newsguy.com>,
Alex Martelli <aleaxit at yahoo.com> wrote:
>"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
>news:D62094CDC48C0FCA.44A0CB1CEB4E1F76.B8EA27DB5535EBAD at lp.airnews.net...
>    [snip]
>> Actually, all these Alex has mentioned have
>> capabilities that'll knock the socks off vic-
>> tims habituated to Access, for example.
>
>Just one attempt at terminology clarification...:
>
>Access is a possibly-handy GUI IDE front-end for cobbling together
>possibly-intricated DB operations; at least if you don't mind

[snip]
>
>Fortunately, Access of course works perfectly well with MSDE (or
>other editions of SQL Server), Oracle, and no doubt other good
>RDBMS's yet. 

Well...  The PostgreSQL people had one heck of a time dealing with
the sql that access generated.  But that was a year or two ago.

-- cary

>               Even if one IS habituated to Access (and there may
>be perfectly good reasons for choosing it to develop certain DB
>apps, I guess), there is no reason not to switch to a better engine
>underneath the chrome (given that MSDE is freely redistributable

[snip]


From nielsenjf at my-deja.com  Thu Feb  1 14:02:43 2001
From: nielsenjf at my-deja.com (John Nielsen)
Date: Thu, 01 Feb 2001 13:02:43 GMT
Subject: How do you lock your web app data?
References: <956uqd$6st$1@newsreaderg1.core.theplanet.net>
Message-ID: <95bmpg$g8k$1@nnrp1.deja.com>

Since it will not be a heavy load app, for a windows solution, you can
also look at python's win32file extension, it exposes LockFileEx which
will lock files for you.

How to use it is described in the help section of pythonwin:
Help->Other->Win32Extensions and then look for locking.

The lock, however, is process based. This means you'd need to run it
in a separate process from IIS (the reason you don't want it to be a
heavy-load app).

Have a good day,

john



In article <956uqd$6st$1 at newsreaderg1.core.theplanet.net>,
  "Franz GEIGER" <fgeiger at datec.at> wrote:
>
> Hi all,
>
> I began to play around with cgi scripts and of course use Python for that.
> And man, it is a pleasure to do that with Python! It really is a snap.
After
> about a day I had written an Issue Tracker, where People can add software
> feature requests and bug reports. At this time I add functionality to edit
> and delete the entries - until now they only could have been added.
>
> So  an important issue enters the game: How do I lock others out, when one
> edits/writes records? Please get me right: This will not be a heavy
load app
> nor will it be to handle requests and reports from hundreds of people.
> Therefore and because I just settled out to explore cgi I decided to use
> text files and pickle files for maintaining data and state, which is very
> convenient.
>
> Of course I could take a database as backend but I wonder if there are
other
> possibilities to achieve that "serialization"? How do you do it? Do you
> always use a database? Is it worth the effort doing it without a database?
>
> The server is an NT box running IIS4. Python is of version 2.0
(AvtiveState
> distribution).
>
> Any hint is appreciated.
>
> Thanks in advance and best regards
> Franz GEIGER
>
>

--
nielsenjf at my-Deja.com


Sent via Deja.com
http://www.deja.com/


From fgeiger at datec.at  Thu Feb  8 08:26:50 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Thu, 8 Feb 2001 08:26:50 +0100
Subject: How do you lock your web app data?
References: <956uqd$6st$1@newsreaderg1.core.theplanet.net>
Message-ID: <95thnk$vr9$1@newsreaderm1.core.theplanet.net>

Thanks a lot to all who contributed. The posts were numerous, comprehensive
and of great help. It's always a pleasure to be here!

Best regards
Franz GEIGER

"Franz GEIGER" <fgeiger at datec.at> wrote in message
news:956uqd$6st$1 at newsreaderg1.core.theplanet.net...
>
> Hi all,
>
> I began to play around with cgi scripts and of course use Python for that.
> And man, it is a pleasure to do that with Python! It really is a snap.
After
> about a day I had written an Issue Tracker, where People can add software
> feature requests and bug reports. At this time I add functionality to edit
> and delete the entries - until now they only could have been added.
>
> So  an important issue enters the game: How do I lock others out, when one
> edits/writes records? Please get me right: This will not be a heavy load
app
> nor will it be to handle requests and reports from hundreds of people.
> Therefore and because I just settled out to explore cgi I decided to use
> text files and pickle files for maintaining data and state, which is very
> convenient.
>
> Of course I could take a database as backend but I wonder if there are
other
> possibilities to achieve that "serialization"? How do you do it? Do you
> always use a database? Is it worth the effort doing it without a database?
>
> The server is an NT box running IIS4. Python is of version 2.0
(AvtiveState
> distribution).
>
> Any hint is appreciated.
>
> Thanks in advance and best regards
> Franz GEIGER
>
>
>
>




From m.bless at gmx.de  Thu Feb  1 01:07:24 2001
From: m.bless at gmx.de (Martin Bless)
Date: Thu, 01 Feb 2001 00:07:24 GMT
Subject: Roadmap wanted: Python access to SPIRIT.OCX (Win9x)
References: <3a7b39da.39409649@news.muenster.de> <958i3j0mk4@news2.newsguy.com>
Message-ID: <3a78a445.7210135@news.muenster.de>

It's past midnight - but it works!!!!!!!
It was all there, and it's easy once you know what to to.
makepy.py produced spirit.py, and

import win32com, spirit
brick = win32com.client.Dispatch('SPIRIT.SpiritCtrl.1')

instantiates the object. And now I can say brick.On((0,1,3)) to get
three motors going. Absolutely fascinating. IMHO, beats the proposed
visual basic solutions by far. Of course I haven't understood a bit of
the inner workings of makepy - the good news is, I obviously don't
have to at the moment. That's real magic! I'd like to express my
respect. Makepy does a marvellous job.

It took me some time to discover this little hint in the generated
spirit.py:

# This CoClass is known by the name 'SPIRIT.SpiritCtrl.1'

>Not me (another fascinating timesink being the LAST thing I
>need:-) but I do understand why one would!-)

Oh yes oh yes, it seems it has hit ME. Hope I can save myself ...

Still very excited
Martin


From robertcdio at my-deja.com  Fri Feb  2 20:19:20 2001
From: robertcdio at my-deja.com (robertcdio at my-deja.com)
Date: Fri, 02 Feb 2001 19:19:20 GMT
Subject: fork, exec no wait
References: <3A774C91.40835F30@lbl.gov> <958m19$s2j@news.or.intel.com>
Message-ID: <95f17d$fll$1@nnrp1.deja.com>

Heres the deal:
a process vfork/execs a child
child does stuff1 while...
parent runs stuff2  (without waiting for child to finish)

*But* upon  completion of child, i want the parent to switch to stuff3

simple right?

well here is the problem i cant find a way for the rent to know that
there little guy is done...  could call wait or waitpid but both of
those stall the parents process... i looked at some early versions of
system(function)  and thought i was on to something because i cam
across the wait command embedded in a while loop with no body than i
thought well perhaps the wait command doesnt actually wait at all but
is stalled by the while loop so i dropped stuff2 into the body of the
while loop... allas this didn't work. So i had a smoke...
then i thought well if the child is done its pid goes away so i figured
well thats simple i just nest a call to get the childs pid into a while
loop so that the parent routinely checks for the childs expected pid.
hmm the parent can't talk to the child only the child can talk to the
parent... am i right wrong confused in the wrong proffecion?  HELP
I know i should be a writer check out the run on sentences!!! Kudos to
the brilliance that aids me(and bring anti perspirant)!


Sent via Deja.com
http://www.deja.com/


From erno-news at erno.iki.fi  Sun Feb  4 16:33:34 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 04 Feb 2001 17:33:34 +0200
Subject: fork, exec no wait
References: <3A774C91.40835F30@lbl.gov> <958m19$s2j@news.or.intel.com> <95f17d$fll$1@nnrp1.deja.com>
Message-ID: <ku4rya4g4x.fsf@lasipalatsi.fi>

In article <95f17d$fll$1 at nnrp1.deja.com>, robertcdio at my-deja.com writes:

| could call wait or waitpid but both of those stall the parents
| process...

os.waitpid(..., os.WNOHANG)

  -- erno


From reborn at neozone.net  Thu Feb  1 18:02:48 2001
From: reborn at neozone.net (??)
Date: Fri, 2 Feb 2001 02:02:48 +0900
Subject: Q : About Jython & Applet (i'm begginer -_-)
References: <957pjq$mg9$1@news1.kornet.net> <mailman.981044377.3323.python-list@python.org>
Message-ID: <95c4r1$bv9$1@news1.kornet.net>

Thank you for your regard..^ ^

> How are you using the appletviewer or explorer?  To display an applet,
> it must first be in an HTML file.  Example:
>
> <html><body>
> <applet classpath="HelloApplet.class" main="HelloApplet">
> </body></html>

I made HTML file like that

<html><body>
<applet code="HelloApplet.class" width=300 height=250>
</applet>
</body></html>

and also test as your HTML code..
but, it's same result..

I insert some code for test it in jython :

if __name__ == '__main__':
    import pawt
    pawt.test(HelloApplet(), size=(500, 120))


then, I execute it like that :

jython HelloApplet.py

It display the result well.

but, after I complie it like that :
jythonc --core -d HelloApplet.py

It make two java file (HelloApplet.java, HelloApplet.class)

It don't work no more ...

my project is that make web-based tutoring Authoring tool.

If this problem doesn't solve, we have to throw away jython..

Please... give me some help..




From d98aron at dtek.chalmers.se  Fri Feb  2 17:24:11 2001
From: d98aron at dtek.chalmers.se (Fredrik Aronsson)
Date: 2 Feb 2001 16:24:11 GMT
Subject: Q : About Jython & Applet (i'm begginer -_-)
References: <957pjq$mg9$1@news1.kornet.net>
    <mailman.981044377.3323.python-list@python.org>
    <95c4r1$bv9$1@news1.kornet.net>
Message-ID: <95emvb$sai$1@nyheter.chalmers.se>

I used your code but compiled with

jythonc -c -d -j Hello.jar HelloApplet.py

and used HTML:

<html><body>
<applet code="HelloApplet.class"
        archive="Hello.jar"
        width=300
        height=250>
</applet>
</body></html>

Now it works in both netscape and appletviewer for me.


[Posted and mailed]

In article <95c4r1$bv9$1 at news1.kornet.net>,
	"??" <reborn at neozone.net> writes:
[snip]
> 
> I made HTML file like that
> 
> <html><body>
> <applet code="HelloApplet.class" width=300 height=250>
> </applet>
> </body></html>
> 
[snip]
> 
> but, after I complie it like that :
> jythonc --core -d HelloApplet.py
> 
> It make two java file (HelloApplet.java, HelloApplet.class)

didn't it make a HelloApplet$_PyInner.class also?
You also need the a lot of classes from the org.python.core
in the jar file, but that is taken care of by jythonc.

Hope this helps

/Fredrik


From dale at out-think.NOSPAMco.uk  Thu Feb  1 09:08:12 2001
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Thu, 01 Feb 2001 08:08:12 +0000
Subject: Python 2.1 and Unicode
References: <hpse7tc1h31hosvt4ojeu419c531mm76ij@4ax.com> <958gnc0ksj@news2.newsguy.com> <ivig7t0dvpblnqlqa348jop70dbtmd2hjo@4ax.com> <95a49s01uao@news1.newsguy.com>
Message-ID: <pa6i7tkvj786j06j5cudd9ss9na6jn4umc@4ax.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote:

>"Dale Strickland-Clark" <dale at out-think.NOSPAMco.uk> wrote in message
>news:ivig7t0dvpblnqlqa348jop70dbtmd2hjo at 4ax.com...
>    [snip]
>> Thanks, again, Alex.
>
>You're welcome!
>
>> I'm not familiar with the site configuration stuff. Could I also use this
>> for identifying our own module library?
>>
>> Where's it documented?
>
>http://www.python.org/doc/current/lib/module-site.html and environs.
>
>Basically, strategically placing .pth files will help you "identify your
>module libraries" etc.  Then, sitecustomize.py, if it exists, lets you do
>any last-minute site-specific tweaks -- it's imported first.
>
>
>Alex
>
>
>
>

Cheers. I'll take a look.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From loewis at informatik.hu-berlin.de  Fri Feb  2 22:26:18 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:26:18 +0100
Subject: Python 2.1 and Unicode
References: <hpse7tc1h31hosvt4ojeu419c531mm76ij@4ax.com> <958gnc0ksj@news2.newsguy.com> <ivig7t0dvpblnqlqa348jop70dbtmd2hjo@4ax.com> <95a49s01uao@news1.newsguy.com>
Message-ID: <p6q3ddwpyit.fsf@informatik.hu-berlin.de>

"Alex Martelli" <aleaxit at yahoo.com> writes:

> http://www.python.org/doc/current/lib/module-site.html and environs.
> 
> Basically, strategically placing .pth files will help you "identify your
> module libraries" etc.  Then, sitecustomize.py, if it exists, lets you do
> any last-minute site-specific tweaks -- it's imported first.

Furthermore, starting from Python 2.1, you can have "code" in a .pth
file, i.e. lines starting with import. That way, you can easily add to
__builtins__ if you want without editing sitecustomize.py.

Regards,
Martin


From loewis at informatik.hu-berlin.de  Fri Feb  2 22:32:32 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:32:32 +0100
Subject: Python 2.1 and Unicode
References: <hpse7tc1h31hosvt4ojeu419c531mm76ij@4ax.com>
Message-ID: <p6qzog4ojnz.fsf@informatik.hu-berlin.de>

Dale Strickland-Clark <dale at out-think.NOSPAMco.uk> writes:

> However, I was disapointed not to find any improvements to the way
> Unicode is handled.

Did you report the problems you had as bugs? If not, why do you think
it might have "improved"?

> Currently we are having to use constructs such as this all over the place when dealing with values
> from international databases:
> 
> 	if type(val) == types.UnicodeType:
> 		val = val.encode('Latin-1', 'ignore')
> 	else:
> 		val = str(val)

It surely sounds wrong if you do this in many places. However, it
would be equally wrong if Python would set string conversion to
Latin-1 by default - what do you do if your terminal does not support
Latin-1?

> And it gets even more ugly when lists of values are handled.
>
> encode() blows if you give it an integer and str() blows if it
> doesn't like the Unicode.

Well, explicit is better than implicit. This fragment looks like an
application-specific conversion function - so it should be one,
instead of being inlined in many places.

> On top of that, print needs to default to rugged handling of Unicode
> and not raise an exception.

That turns out to be difficult, since print uses str() for each value.
What print *really* should do is to use the maximum capabilities of
the terminals (i.e. loading fonts, whatever is necessary). That is
even more difficult.

> Inserting print statements is normally the quickest way to gather
> debugging info but it becomes a big problem when dealing with
> Unicode, requiring code like that above.

Nah, for debugging, I think

  print `val`

would work fine in most cases.

> 1. A print command that defaults to ignoring Unicode conversion
> errors - or at least has the option to do so.

So how does repr() sound to you?

> 2. A function, such as str() that converts *anything* to an ASCII
> string - ignoring Unicode errors.

Well, this is actually the question to which repr() is the answer...

Regards,
Martin


From joseamartin at venmex.com  Thu Feb  1 15:43:50 2001
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 1 Feb 2001 10:43:50 -0400
Subject: Benchmarking module for Python
References: <2P53OlqNc5Cg7uVAjN=rZZ5t6Jix@4ax.com>
Message-ID: <t7iu7jo3dp3b5a@corp.supernews.com>

hI PAUL

you can use the clock function, in python 2.0 this function was improved.


I am using clock function to crack some passwords and works very well.

good look !.


Paul Moore <paul.moore at uk.origin-it.com> escribi? en el mensaje de noticias
2P53OlqNc5Cg7uVAjN=rZZ5t6Jix at 4ax.com...
> I'm looking for a method of benchmarking snippets of Python code, a
> bit like Perl's Benchmark module. I have looked at the profile module,
> but it doesn't seem to do quite what I want.
>
> What I'm after, given a line of code, is something that will let me
> run that line a specified number of times, say 1000, and see the
> average time that the line took.
>
> I can do something like that with profile, by wrapping my code in a
> function, and profiling 'for i in xrange(1000): fn()', and then
> extracting the time spent per call in fn(). But it's a bit of a long
> way round to get what I want.
>
> Ideally, I'd like to be able to reuse the timing code from profile,
> but if not, then an independent benchmark module would be OK. All I
> found on Parnassus was PyBench, which tests a standard set of tasks.
>
> Is there any such thing, or should I start coding?
>
> Paul.
>
> PS Yes, I know that benchmarks are meaningless....
>




From ngreco at softhome.net  Thu Feb  1 16:29:38 2001
From: ngreco at softhome.net (Nahuel Greco)
Date: Thu, 1 Feb 2001 12:29:38 -0300
Subject: PyKDE / PyQT debian packages
In-Reply-To: <slrn97gioe.j0u.grey@teleute.dmiyu.org>
References: <mailman.980945617.21811.python-list@python.org>
	<slrn97gioe.j0u.grey@teleute.dmiyu.org>
Message-ID: <20010201122938.09ba6419.ngreco@softhome.net>

There is debian packages of PyKDE and PyQT somewhere in the net?

(some off-topic, the khtml widget supports edition of the html?, and can use
it from pyqt?)



Nahuel Greco



From phil at river-bank.demon.co.uk  Sun Feb  4 11:57:46 2001
From: phil at river-bank.demon.co.uk (phil at river-bank.demon.co.uk)
Date: Sun, 04 Feb 2001 10:57:46 -0000
Subject: COMMERCIAL: BlackAdder Windows/Linux Python GUI Development Environment
In-Reply-To: <slrn97gioe.j0u.grey@teleute.dmiyu.org>
Message-ID: <95jcja+75br@eGroups.com>

--- In python-list at y..., grey at d... (Steve Lamb) wrote:
> On Wed, 31 Jan 2001 12:52:34 -0000, phil at r...
> <phil at r...> wrote:
> >BlackAdder is available in two editions, the Personal Edition and
the
> >Business Edition. The Business Edition allows the run-time elements
of
> >BlackAdder (except for mxODBC) to be re-distributed with your
> >applications at no extra cost.
> 
>     This is the part that gets me wondering.  How pervasive are
these run time
> portions?  I'd like to use QT as the graphical toolkit for a project
I have
> and that project will be GPL'd.  What problems would I have in
interfacing GPL
> code and the personal edition upon distribution?  I'm confident I
could
> distribute the code freely but what would that code evetually run
on?

The run-time elements consist of Python, and a special build of PyQt
that includes Qt statically linked. With the Personal Edition you can
distribute Python (because it has a different license anyway) but not
the BlackAdder version of PyQt.

To distribute your application (is doesn't matter if it is free or
not), you have the following choices...

1. Your user has to have their own copy of the Personal Edition.

2. You buy the Business Edition which allows you to re-distribute the
BlackAdder version of PyQt at no extra charge.

3. Your user has to have the free version of PyQt and Qt already
installed. For UNIX this isn't a problem, for Windows they would also
need a commercial Qt license which probably make it too expensive to
be practical.

Phil




From embed at geocities.com  Thu Feb  1 15:07:15 2001
From: embed at geocities.com (Warren Postma)
Date: Thu, 1 Feb 2001 09:07:15 -0500
Subject: COMMERCIAL: BlackAdder Windows/Linux Python GUI Development Environment
References: <mailman.980945617.21811.python-list@python.org>
Message-ID: <M6ee6.8725$Oe.40269@tor-nn1.netcom.ca>

Quoth phil at river-bank.demon.co.uk:
> The cunning plan continues...
>
> theKompany.com is pleased to announce the immediate >availability of  the
first Beta release of BlackAdder, a
>commercial Windows/Linux IDE for Python and the Qt
>GUI toolkit.

Looks interesting.  Besides this one (for QT) and Boa Constructor (wxPython
based) are there any other RAD/IDE tools are in the works that include
integrated forms builders and syntax-highlighting editor, etc?

I'm trying to figure out which one to contribute to the development of. So
far I'm leaning towards Boa because of the more liberal Windows licensing of
wxWindows (free) compared to QT on Windows(generally not free).

Warren




From bobhicks at adelphia.net  Thu Feb  1 15:51:39 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Thu, 01 Feb 2001 14:51:39 GMT
Subject: COMMERCIAL: BlackAdder Windows/Linux Python GUI Development
	Environment
References: <mailman.980945617.21811.python-list@python.org> <M6ee6.8725$Oe.40269@tor-nn1.netcom.ca>
Message-ID: <B69EE223.9ADB%bobhicks@adelphia.net>

Why not make a pure Tkinter IDE...?

> From: "Warren Postma" <embed at geocities.com>
> Organization: AT&T Canada IES
> Newsgroups: comp.lang.python
> Date: Thu, 1 Feb 2001 09:07:15 -0500
> Subject: Re: COMMERCIAL: BlackAdder Windows/Linux Python GUI Development
> Environment
> 
> 
> Quoth phil at river-bank.demon.co.uk:
>> The cunning plan continues...
>> 
>> theKompany.com is pleased to announce the immediate >availability of  the
> first Beta release of BlackAdder, a
>> commercial Windows/Linux IDE for Python and the Qt
>> GUI toolkit.
> 
> Looks interesting.  Besides this one (for QT) and Boa Constructor (wxPython
> based) are there any other RAD/IDE tools are in the works that include
> integrated forms builders and syntax-highlighting editor, etc?
> 
> I'm trying to figure out which one to contribute to the development of. So
> far I'm leaning towards Boa because of the more liberal Windows licensing of
> wxWindows (free) compared to QT on Windows(generally not free).
> 
> Warren
> 
> 



From fgeiger at datec.at  Thu Feb  8 08:28:54 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Thu, 8 Feb 2001 08:28:54 +0100
Subject: COMMERCIAL: BlackAdder Windows/Linux Python GUI Development Environment
References: <mailman.980945617.21811.python-list@python.org> <M6ee6.8725$Oe.40269@tor-nn1.netcom.ca>
Message-ID: <95thrg$vt6$1@newsreaderm1.core.theplanet.net>

wxDesigner is such a tool, but is commercial: www.roebling.de

Regards
Franz GEIGER

"Warren Postma" <embed at geocities.com> wrote in message
news:M6ee6.8725$Oe.40269 at tor-nn1.netcom.ca...
>
> Quoth phil at river-bank.demon.co.uk:
> > The cunning plan continues...
> >
> > theKompany.com is pleased to announce the immediate >availability of
the
> first Beta release of BlackAdder, a
> >commercial Windows/Linux IDE for Python and the Qt
> >GUI toolkit.
>
> Looks interesting.  Besides this one (for QT) and Boa Constructor
(wxPython
> based) are there any other RAD/IDE tools are in the works that include
> integrated forms builders and syntax-highlighting editor, etc?
>
> I'm trying to figure out which one to contribute to the development of. So
> far I'm leaning towards Boa because of the more liberal Windows licensing
of
> wxWindows (free) compared to QT on Windows(generally not free).
>
> Warren
>
>




From pieter at nagel.co.za  Thu Feb  1 03:45:00 2001
From: pieter at nagel.co.za (Pieter Nagel)
Date: Thu, 1 Feb 2001 04:45:00 +0200
Subject: Dynamically Defined Functions in Classes?
In-Reply-To: <mailman.980958702.3721.python-list@python.org>
References: <mailman.980958702.3721.python-list@python.org>
Message-ID: <Pine.LNX.4.21.0102010443250.4948-100000@bast.jhb.equinox.co.za>

On Wed, 31 Jan 2001, Jim Meyer wrote:

> I have an application in which I wish to treat the members of a
> dictionary as attributes;

class Bar:
    def __init__(self, initial_dict):
        self.__dict__ = initial_dict

b = Bar({'joe' : 'cool', 'frank' : 'lee', 'ron' : 'dell'})
print b.joe
b.frank = 'foo'

-- 
     ,_
     /_)              /| /
    /   i e t e r    / |/ a g e l



From wolfgang.grafen at marconi.com  Thu Feb  1 10:42:45 2001
From: wolfgang.grafen at marconi.com (Wolfgang Grafen)
Date: Thu, 01 Feb 2001 10:42:45 +0100
Subject: Dynamically Defined Functions in Classes?
References: <mailman.980958702.3721.python-list@python.org> <Pine.LNX.4.21.0102010443250.4948-100000@bast.jhb.equinox.co.za>
Message-ID: <3A792F95.E9872F6@marconi.com>

Pieter Nagel wrote:
> 
> On Wed, 31 Jan 2001, Jim Meyer wrote:
> 
> > I have an application in which I wish to treat the members of a
> > dictionary as attributes;
> 
> class Bar:
>     def __init__(self, initial_dict):
>         self.__dict__ = initial_dict
> 
> b = Bar({'joe' : 'cool', 'frank' : 'lee', 'ron' : 'dell'})
> print b.joe
> b.frank = 'foo'
> 
> --
>      ,_
>      /_)              /| /
>     /   i e t e r    / |/ a g e l

Good idea!

Let's do a little additional stuff and you get:


class Bar(UserDict):
    def __init__(self,dict={}):
        UserDict.__init__(self,dict)
        self.data.update(self.__dict__)
        self.__dict__=self.data

>>> b=Bar({'a':1,'b':2)
>>> b
{'data': {...}, 'b': 2, 'a': 1}

>>> b.b
2

>>> b.a
1

>>> b['c']=3
>>> b
{'data': {...}, 'b': 2, 'c': 3, 'a': 1}

>>> b.c
3

>>> b['b']
2

Cheers,

Wolfgang


From timr at probo.com  Fri Feb  2 05:41:34 2001
From: timr at probo.com (Tim Roberts)
Date: Thu, 01 Feb 2001 20:41:34 -0800
Subject: from-import on non-module objects?
References: <959jna$nfl$1@nnrp1.deja.com>
Message-ID: <qiek7tssid1k23egr3km39g86g46lm4r8c@4ax.com>

Hamish Lawson <hamish_lawson at yahoo.co.uk> wrote:

>Have there been thoughts on having from-import used on non-module
>objects in order to introduce attributes from a given object into the
>current scope?

Unless I misunderstood the explanation, I believe this exact thing is one
of the features being included in Python 2.1.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From tim.one at home.com  Sun Feb  4 11:31:38 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 4 Feb 2001 05:31:38 -0500
Subject: from-import on non-module objects?
In-Reply-To: <qiek7tssid1k23egr3km39g86g46lm4r8c@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEGFINAA.tim.one@home.com>

[Hamish Lawson]
> Have there been thoughts on having from-import used on non-module
> objects in order to introduce attributes from a given object into the
> current scope?

[Tim Roberts]
> Unless I misunderstood the explanation, I believe this exact thing
> is one of the features being included in Python 2.1.

The description of this change is the NEWS file for 2.1a1 and 2.1a2 was
incomplete; it will be clearer in the next release.

Short course:  in

    from M import x

M no longer needs to be a true module object (which earlier NEWS said), but
the name "M" *must* be a key in sys.modules (which info was missing).  So
this isn't a hack aimed at letting people do bizarre things with
from-import, it's a hack aimed at letting people use "module-like objects".

Python has a long tradition of loosening type restrictions over time, and
this was a natural part of that tradition:  if an object supports getattr,
it supports the only thing the "import" part of from-import really needs.
The "from" part still needs "a module name", though, but the operational
definition of "a module name" is "a name that appears as a key in
sys.modules".  So meet those two requirements, and Python couldn't care less
whether it's "a real module" anymore.

Really bad idea (running under 2.1a2):

>>> somelist = [1, 2, 3, 2, 1]
>>> from somelist import count # blows up
Traceback (most recent call last):
  File "<pyshell#2>", line 1, in ?
    from somelist import count # blows up
ImportError: No module named somelist
>>> import sys
>>> sys.modules["somelist"] = somelist
>>> from somelist import count # works, for some meaning of "works"
>>> count(1), count(2), count(3), count(4)
(2, 2, 1, 0)
>>>

Don't try that at home; you'll be shot if you do <0.9 bang>.

useful-examples-left-to-your-fertile-imaginations-ly y'rs  - tim




From moshez at zadka.site.co.il  Sun Feb  4 12:19:34 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun,  4 Feb 2001 13:19:34 +0200 (IST)
Subject: from-import on non-module objects?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCMEGFINAA.tim.one@home.com>
References: <LNBBLJKPBEHFEDALKOLCMEGFINAA.tim.one@home.com>
Message-ID: <20010204111934.16EE7A840@darjeeling.zadka.site.co.il>

[Hamish Lawson]
> Have there been thoughts on having from-import used on non-module
> objects in order to introduce attributes from a given object into the
> current scope?

[Tim Roberts]
> Unless I misunderstood the explanation, I believe this exact thing
> is one of the features being included in Python 2.1.

[Tim Peters]
> The description of this change is the NEWS file for 2.1a1 and 2.1a2 was
> incomplete; it will be clearer in the next release.
<snip>
> Python has a long tradition of loosening type restrictions over time, and
> this was a natural part of that tradition:  if an object supports getattr,
> it supports the only thing the "import" part of from-import really needs.
> The "from" part still needs "a module name", though, but the operational
> definition of "a module name" is "a name that appears as a key in
> sys.modules".  So meet those two requirements, and Python couldn't care less
> whether it's "a real module" anymore.

Tim wrote a great explanation, and cute example, but he forgot to mention
one small thing: for
   from M import *
M, in addition to supporting getattr, must also either have an __all__
attribute which lists the names to be imported, or an __dict__ which
should be a dictionary whose keys are the names which will be important.

just-nitpicking-ly y'rs, Z.

-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From clarence at netlojix.com  Sun Feb  4 22:33:42 2001
From: clarence at netlojix.com (Clarence Gardner)
Date: Sun, 4 Feb 2001 13:33:42 -0800
Subject: from-import on non-module objects?
References: <mailman.981282736.25263.python-list@python.org>
Message-ID: <981322698.934790726@news.silcom.com>

On Sun, 04 Feb 2001, Tim Peters wrote:
>if an object supports getattr,
>it supports the only thing the "import" part of from-import really needs.
>
>Really bad idea (running under 2.1a2):
>
>>>> somelist = [1, 2, 3, 2, 1]
>>>> from somelist import count # blows up
>Traceback (most recent call last):
>  File "<pyshell#2>", line 1, in ?
>    from somelist import count # blows up
>ImportError: No module named somelist
>>>> import sys
>>>> sys.modules["somelist"] = somelist
>>>> from somelist import count # works, for some meaning of "works"
>>>> count(1), count(2), count(3), count(4)
>(2, 2, 1, 0)
>>>>
>

Boy, was this confusing to someone who'd never heard of the count
method of lists!  But firing up the friendly interactive Python and doing
a getattr([], 'count') showed it to me.  Then I had to print its docstring to
figure out how that result was (2,2,1,0)! 

Speaking of the friendly interactive Python, I saw recently that someone
finally made an interactive shell for Perl.  I always wondered why they
hadn't done it sooner.  [I'll bet it's not as good, though :)]


From thomas at xs4all.net  Mon Feb  5 00:14:53 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 5 Feb 2001 00:14:53 +0100
Subject: from-import on non-module objects?
In-Reply-To: <981322698.934790726@news.silcom.com>; from clarence@netlojix.com on Sun, Feb 04, 2001 at 01:33:42PM -0800
References: <mailman.981282736.25263.python-list@python.org> <981322698.934790726@news.silcom.com>
Message-ID: <20010205001453.H962@xs4all.nl>

On Sun, Feb 04, 2001 at 01:33:42PM -0800, Clarence Gardner wrote:

> Speaking of the friendly interactive Python, I saw recently that someone
> finally made an interactive shell for Perl.  I always wondered why they
> hadn't done it sooner.  [I'll bet it's not as good, though :)]

Perl isn't as easy, interactively, as Python, due to the way it manages
variables and code (as lazy as possible, that is :) And for the people that
really wanted one, there has been the Perl debugger for a while. The way to
get an interactive prompt was: 'perl -de1'. Quite recently (in 5.004 I think
?) they even added readline support in it. Not sure if that's the
interactive interpreter you were talking about or not, though; my sole
contribution to Perl has been the detection of a bug in the 'require'
statement :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From dan at eevolved.com  Mon Feb  5 16:13:32 2001
From: dan at eevolved.com (Dan Parisien)
Date: Mon, 05 Feb 2001 15:13:32 GMT
Subject: from-import on non-module objects?
References: <mailman.981282736.25263.python-list@python.org>
Message-ID: <wqzf6.192186$JT5.7035338@news20.bellglobal.com>

> useful-examples-left-to-your-fertile-imaginations-ly y'rs  - tim

your sig is so weird :)



From pj at sgi.com  Thu Feb  1 00:46:59 2001
From: pj at sgi.com (Paul Jackson)
Date: 31 Jan 2001 23:46:59 GMT
Subject: Perl-to-Python converter/translator?
References: <959jsd$njn$1@nnrp1.deja.com> <3A7879FC.EA879C9B@bigfoot.de>
Message-ID: <95a85j$p1ce$1@fido.engr.sgi.com>

Gerhard wrote:
|> Maybe the Perl/Python phrasebook at
|> http://starship.python.net/~da/jak/cookbook.html is helpful
|> to you.  Intra-document links don't work. Does anybody know a
|> better version of this document?

Try a google.com search for "perl python phrasebook" for pointers
to such Perl-Python translators, and to other comparisons of
Perl and Python.
-- 
                          I won't rest till it's the best ...
                          Manager, Linux System Software
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From tim.one at home.com  Fri Feb  2 07:57:31 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 2 Feb 2001 01:57:31 -0500
Subject: Perl-to-Python converter/translator?
In-Reply-To: <959uuh01qd3@drn.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIELKIMAA.tim.one@home.com>

[Grant Griffin]
> ...
> To rigorously and correctly translate any-and-all Perl to Python
> would surely take a heroic effort, but to do it at about the
> 90% level can't be all that big of a job for the language
> hobbyists out there (anybody know any? <wink>); ...

Sorry, Grant:  simply *parsing* Perl is beyond most peoples' pain tolerance;
search CPAN for an equivalent to Python's tokenize.py and report back on
your successes <wink>.

perhaps-you'd-make-better-progress-if-you-ran-it-through-an-fft-
    first-ly y'rs  - tim




From hamish_lawson at yahoo.co.uk  Fri Feb  2 15:56:39 2001
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Fri, 02 Feb 2001 14:56:39 GMT
Subject: Perl-to-Python converter/translator?
References: <959uuh01qd3@drn.newsguy.com> <mailman.981100210.20770.python-list@python.org>
Message-ID: <95ehr1$9d$1@nnrp1.deja.com>

Tim Peters wrote:

> Sorry, Grant:  simply *parsing* Perl is beyond most peoples' pain
> tolerance; search CPAN for an equivalent to Python's tokenize.py and
> report back on your successes <wink>.

To overcome the expected language incompatibilities between Perl 5 and
the future Perl 6, the language developers are intending to
provide a tool that will migrate existing Perl 5 code to Perl 6. I'm
not sure of the exact details, but I understand that it will make use
of the fact that Perl now has a compiler that can compile to and from
some intermediate format - I presume the scheme being to use the Perl 6
compiler to reconstitute as source code the intermediate code generated
by the Perl 5 compiler. Perhaps it may be possible to reconstitute this
intermediate code as Python source code instead.

Hamish Lawson


Sent via Deja.com
http://www.deja.com/


From not.this at seebelow.org  Fri Feb  2 19:31:13 2001
From: not.this at seebelow.org (Grant Griffin)
Date: 2 Feb 2001 10:31:13 -0800
Subject: Perl-to-Python converter/translator?
References: <mailman.981100210.20770.python-list@python.org>
Message-ID: <95eudh010c1@drn.newsguy.com>

In article <mailman.981100210.20770.python-list at python.org>, "Tim says...
>
>[Grant Griffin]
>> ...
>> To rigorously and correctly translate any-and-all Perl to Python
>> would surely take a heroic effort, but to do it at about the
>> 90% level can't be all that big of a job for the language
>> hobbyists out there (anybody know any? <wink>); ...
>
>Sorry, Grant:  simply *parsing* Perl is beyond most peoples' pain tolerance;

It sure was beyond _mine_, Tim!  Luckily, just about the time I realized I
couldn't parse my own Perl code, I discovered Python <wink>.

But this is a classic Catch-22: you switch to Python because you found yourself
unable to maintain your own Perl; you want to translate your Perl to Python so
you can maintain it; but there aren't any automatic translators out there
because nobody _else_ can figure out how to parse Perl--which is more-or-less
why you switched in the first place!

Then again, haven't Great Minds already solved the parsing part of automatic
translation?  I mean, doesn't Perl come with its own lexer, parser, and even
grammar (though last I checked, they had to doctor the parser that was
automatically generated from the grammar...)

Which reminds me...I really think Open Source Software is gonna become a big
thing one day--just for cases like this <wink>.

>search CPAN for an equivalent to Python's tokenize.py and report back on
>your successes <wink>.

I see what you mean.  (Though to be fair, I never found tokenize.py on PYPAN
either <wink>.)

But did I say "translate at about the 90% level"?--I really meant translate at
about the _7%_ level.  If somebody came up with something that didn't really
"parse" Perl but just did the simple operation of "clutter-removal" (eliminating
";", "{", "@", "my", etc. outside of strings) that would help.

i'd-write-one-myself-if-perl-hadn't-taught-me-to-be-lazy
   -<wink>-ly y'rs,

=g2

_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com



From sheila at spamcop.net  Thu Feb  1 06:15:29 2001
From: sheila at spamcop.net (Sheila King)
Date: Thu, 01 Feb 2001 05:15:29 GMT
Subject: Perl-to-Python converter/translator?
References: <959jsd$njn$1@nnrp1.deja.com>
Message-ID: <2urh7t0cmisg8ous1lfuo1s4f54c32q58e@4ax.com>

On Wed, 31 Jan 2001 18:00:51 GMT, Lance Sloan <lsloan00 at my-deja.com> wrote in
comp.lang.python in article <959jsd$njn$1 at nnrp1.deja.com>:

:I'm an experienced Perl programmer who has just been assigned
:a project that must be done in Python.  I'm looking for a
:Perl-to-Python translator to help me out.  I don't expect to
:write the whole project in Perl and translate it, though.  I'm
:just looking for some automation to help me see how my old work
:could be done in Python.

I've translated a single Perl script (written by someone else) to Python. If
it helps you at all, you can see it here:
http://www.aota.net/ubb/Forum3/HTML/001491-1.html

I would recommend that you read through the Python Tutorial as a start, and
then look at the Library Reference. You should be able to work through the
tutorial in short order, and then browse the Library Reference on an as-needed
basis. The tutorial may seem simple, but it will get you jump started and
won't take long..

http://www.python.org/doc/


--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/





From hgg9140 at cola.ca.boeing.com  Thu Feb  1 15:56:07 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Thu, 1 Feb 2001 14:56:07 GMT
Subject: Perl-to-Python converter/translator?
References: <959jsd$njn$1@nnrp1.deja.com>
Message-ID: <xqxofwmv4e0.fsf@cola.ca.boeing.com>

I've done some perl-->python, and found (at first) that I missed some
of the perl idioms.  So I did a python module to support perl-ish
idioms.  See "pyperl" at:

  http://www.seanet.com/~hgg9140/comp/index.html

It lets you pretty much read the perl code and write the python code
on the fly.  More tedious than true converter, but do-able.  

I agree with other posters that after you get into python's own
idioms, you won't miss perl.

Lance Sloan <lsloan00 at my-deja.com> writes:

> I'm an experienced Perl programmer who has just been assigned
> a project that must be done in Python.  I'm looking for a
> Perl-to-Python translator to help me out.  I don't expect to
> write the whole project in Perl and translate it, though.  I'm
> just looking for some automation to help me see how my old work
> could be done in Python.
> 
> I didn't find anything useful via Google or Vaults of Parnassus
> searches.  So, I would appreciate any pointers!
> 
> Thanks in advance.
> 
> --
> Lance Sloan - lsloan00 at my-deja.com
> 
> --
> Lance Sloan - lsloan00 at my-deja.com
> 
> 
> Sent via Deja.com
> http://www.deja.com/

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From Randy.L.Kemp at motorola.com  Thu Feb  1 18:00:03 2001
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Thu, 1 Feb 2001 11:00:03 -0600 
Subject: Perl-to-Python converter/translator?
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84DD6@il93exp01.css.mot.com>

Have we converted another to the fray?

-----Original Message-----
From: Gerhard H?ring [mailto:gerhard.nospam at bigfoot.de]
Sent: Wednesday, January 31, 2001 2:48 PM
To: python-list at python.org
Subject: Re: Perl-to-Python converter/translator?


Lance Sloan wrote:
> 
> I'm an experienced Perl programmer who has just been assigned
> a project that must be done in Python.  I'm looking for a
> Perl-to-Python translator to help me out.  I don't expect to
> write the whole project in Perl and translate it, though.  I'm
> just looking for some automation to help me see how my old work
> could be done in Python.
> 
> I didn't find anything useful via Google or Vaults of Parnassus
> searches.  So, I would appreciate any pointers!

Maybe the Perl/Python phrasebook at
http://starship.python.net/~da/jak/cookbook.html is helpful to you.
Intra-document links don't work. Does anybody know a better version of
this document?

Gerhard
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com





From loewis at informatik.hu-berlin.de  Fri Feb  2 22:37:02 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:37:02 +0100
Subject: Memory Leak Tools
References: <3A785D25.A26822@americasm01.nt.com>
Message-ID: <p6qwvb8ojgh.fsf@informatik.hu-berlin.de>

"Lee, Rick" <rickylee at americasm01.nt.com> writes:

> Any tool to help track down memory leaks?

In Python 2.0, it is quite hard to find a program that produces a
memory leak. In any case, just turn on debugging in the gc module -
that will report all the cycles it collects.

Regards,
Martin



From gtalvola at nameconnector.com  Thu Feb  1 20:40:48 2001
From: gtalvola at nameconnector.com (Geoff Talvola)
Date: Thu, 01 Feb 2001 14:40:48 -0500
Subject: executing JavaScript from Python
References: <959ovu$sqd$1@nnrp1.deja.com> <63E6AD93C013B993.0AA5FD315BC99418.FEB162B59977D6FB@lp.airnews.net>
Message-ID: <3A79BBC0.8F56A389@NameConnector.com>

> In article <959ovu$sqd$1 at nnrp1.deja.com>,  <echuck at mindspring.com> wrote:
> >urllib.urlopen() can be useful for grabbing a web page and checking its
> >contents. However, JavaScript generated content throws a monkey wrench
> >in this technique.
> >
> >Has anyone cobbled together a means to invoke JavaScript from Python,
> >possibly including the DOM and the substitution of output into the HTML
> >file?

On Windows, you can use Internet Explorer via COM:

>>> import urllib
>>> print urllib.urlopen('http://localhost/javascript.html').read()
<html>
<body>
<script>
document.write('Hello, World!')
</script>
</body>
</html>
>>> from win32com.client.gencache import EnsureDispatch
>>> ie = EnsureDispatch('InternetExplorer.Application')
>>> ie.Navigate('http://localhost/javascript.html')
>>> print ie.Document.documentElement.innerHTML
<HEAD></HEAD>
<BODY>
<SCRIPT>
document.write('Hello, World!')
</SCRIPT>
Hello, World! </BODY>
>>>

--


- Geoff Talvola
  Parlance Corporation
  gtalvola at NameConnector.com




From gtalvola at nameconnector.com  Fri Feb  2 01:14:01 2001
From: gtalvola at nameconnector.com (Geoff Talvola)
Date: Thu, 01 Feb 2001 19:14:01 -0500
Subject: executing JavaScript from Python
References: <959ovu$sqd$1@nnrp1.deja.com> <63E6AD93C013B993.0AA5FD315BC99418.FEB162B59977D6FB@lp.airnews.net> <3A79BBC0.8F56A389@NameConnector.com>
Message-ID: <3A79FBC9.249E8E8E@NameConnector.com>

(on the subject of getting web page contents _after_ the JavaScript has been executed, using Internet Explorer on Windows:)

I fiddled with this a bit more, and here's a class that actually waits for IE to be done before trying to get the HTML, and which lets you refresh to make sure you're not getting a cached page.  Note that this only works for GET operations -- I couldn't figure out the magic incantation to get POST to work.  (Navigate has a PostData argument, but whatever I put there, it just seems to ignore.)

I'm not going to spend any more time on this, but if someone can figure out how to get POST to work, I'm sure it would be appreciated...

############
import time
from win32com.client.gencache import EnsureDispatch
from win32com.client import constants

class InternetExplorerGetter:
    def __init__(self):
        self.ie = EnsureDispatch('InternetExplorer.Application')

    def get(self, url):
        self.ie.Navigate(url)
        return self.waitForResult()

    def refresh(self):
        self.ie.Refresh()
        return self.waitForResult()

    def waitForResult(self):
        while 1:
            state = self.ie.ReadyState
            if state == constants.READYSTATE_COMPLETE: break
            time.sleep(0.1)
        return self.ie.Document.documentElement.innerHTML

if __name__ == '__main__':
    print InternetExplorerGetter().get('http://www.python.org')
############

--


- Geoff Talvola
  Parlance Corporation
  gtalvola at NameConnector.com




From hamish_lawson at yahoo.co.uk  Thu Feb  8 18:06:28 2001
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Thu, 08 Feb 2001 17:06:28 GMT
Subject: executing JavaScript from Python
References: <959ovu$sqd$1@nnrp1.deja.com> <63E6AD93C013B993.0AA5FD315BC99418.FEB162B59977D6FB@lp.airnews.net> <mailman.981065169.10009.python-list@python.org>
Message-ID: <95ujmb$9j$1@nnrp1.deja.com>

Geoff Talvola <gtalvola at NameConnector.com> wrote:

> On Windows, you can use Internet Explorer via COM:
>
> >>> from win32com.client.gencache import EnsureDispatch
> >>> ie = EnsureDispatch('InternetExplorer.Application')
> >>> ie.Navigate('http://localhost/javascript.html')
> >>> print ie.Document.documentElement.innerHTML
> <HEAD></HEAD>
> <BODY>
> <SCRIPT>
> document.write('Hello, World!')
> </SCRIPT>
> Hello, World! </BODY>
> >>>

I tried that out myself and got the error below. I'm using Python 2.0
with win32all build 135.

Traceback (most recent call last):
  File "C:\temp\test2.py", line 11, in ?
    print ie.Document.documentElement.innerHTML
  File "C:\Python\win32com\client\__init__.py", line 340, in __getattr__
    return apply(self._ApplyTypes_, args)
  File "C:\Python\win32com\client\__init__.py", line 334, in
_ApplyTypes_
    return self._get_good_object_(apply(self._oleobj_.InvokeTypes,
(dispid, 0, wFlags, retType, argTypes) + args), user, resultCLSID)
pywintypes.com_error: (-2147352567, 'Exception occurred.', (0, None,
None, None, 0, -2147467259), None)


Hamish Lawson


Sent via Deja.com
http://www.deja.com/


From gtalvola at nameconnector.com  Thu Feb  8 20:04:26 2001
From: gtalvola at nameconnector.com (Geoff Talvola)
Date: Thu, 08 Feb 2001 14:04:26 -0500
Subject: executing JavaScript from Python
References: <959ovu$sqd$1@nnrp1.deja.com> <63E6AD93C013B993.0AA5FD315BC99418.FEB162B59977D6FB@lp.airnews.net> <mailman.981065169.10009.python-list@python.org> <95ujmb$9j$1@nnrp1.deja.com>
Message-ID: <3A82EDBA.19284CB9@NameConnector.com>

Hamish Lawson wrote:

> Geoff Talvola <gtalvola at NameConnector.com> wrote:
>
> > On Windows, you can use Internet Explorer via COM:
> >
> > >>> from win32com.client.gencache import EnsureDispatch
> > >>> ie = EnsureDispatch('InternetExplorer.Application')
> > >>> ie.Navigate('http://localhost/javascript.html')
> > >>> print ie.Document.documentElement.innerHTML
> > <HEAD></HEAD>
> > <BODY>
> > <SCRIPT>
> > document.write('Hello, World!')
> > </SCRIPT>
> > Hello, World! </BODY>
> > >>>
>
> I tried that out myself and got the error below. I'm using Python 2.0
> with win32all build 135.
>
> Traceback (most recent call last):
>   File "C:\temp\test2.py", line 11, in ?
>     print ie.Document.documentElement.innerHTML
>   File "C:\Python\win32com\client\__init__.py", line 340, in __getattr__
>     return apply(self._ApplyTypes_, args)
>   File "C:\Python\win32com\client\__init__.py", line 334, in
> _ApplyTypes_
>     return self._get_good_object_(apply(self._oleobj_.InvokeTypes,
> (dispid, 0, wFlags, retType, argTypes) + args), user, resultCLSID)
> pywintypes.com_error: (-2147352567, 'Exception occurred.', (0, None,
> None, None, 0, -2147467259), None)
>
> Hamish Lawson

You have to wait until the document is done loading before you can access the innerHTML -- if you try to access innerHTML too quickly you'll get the exception.  I sent out a followup message with code that does the waiting; if you can't find the message, let me know and I'll email it to you.

--


- Geoff Talvola
  Parlance Corporation
  gtalvola at NameConnector.com




From apederse at my-deja.com  Thu Feb  8 08:17:31 2001
From: apederse at my-deja.com (apederse at my-deja.com)
Date: Thu, 08 Feb 2001 07:17:31 GMT
Subject: executing JavaScript from Python
References: <959ovu$sqd$1@nnrp1.deja.com>
Message-ID: <95th69$36k$1@nnrp1.deja.com>

In article <959ovu$sqd$1 at nnrp1.deja.com>,
  echuck at mindspring.com wrote:
> urllib.urlopen() can be useful for grabbing a
web page and checking its
> contents. However, JavaScript generated content
throws a monkey wrench
> in this technique.
>
> Has anyone cobbled together a means to invoke
JavaScript from Python,
> possibly including the DOM and the substitution
of output into the HTML
> file?
>
> -Chuck
>
> Sent via Deja.com
> http://www.deja.com/
>

How about a combination of Mozilla RHINO and
JPython - just a though.

/Asle


Sent via Deja.com
http://www.deja.com/


From MarkH at ActiveState.com  Thu Feb  8 12:57:30 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 08 Feb 2001 11:57:30 GMT
Subject: executing JavaScript from Python
References: <959ovu$sqd$1@nnrp1.deja.com> <95th69$36k$1@nnrp1.deja.com>
Message-ID: <3A8288C2.6050905@ActiveState.com>

apederse at my-deja.com wrote:


> How about a combination of Mozilla RHINO and
> JPython - just a though.

If you are happy to use Mozilla specific technology, then PyXPCOM would also be a good answer.  See http://www.activestate.com/Products/Komodo/PyXPCOM/

Mark.



From slaurent at sonicfoundry.com  Tue Feb 13 19:13:58 2001
From: slaurent at sonicfoundry.com (Sean Laurent)
Date: Tue, 13 Feb 2001 12:13:58 -0600
Subject: os.popen: writing to...?
References: <959pqt$tgp$1@nnrp1.deja.com>
Message-ID: <vvti8t07p3pe5b7b1mrb5tq88knmt81g13@4ax.com>

On Wed, 31 Jan 2001 19:42:23 GMT, James Lockley
<poseidon_cma at my-deja.com> wrote:

>i now want to go to the next step and get rid of the batch files...
>i need to open a pipe to a command and then write the command arguments
>to it.
>
>as far as i could make out, this should work:
>
>>>> abaqus='c:\\abaqus\\5.8-14\\abaqus.exe'
>>>> abq=os.popen(abaqus,'w')
>>>> abq.write('post')
>
>but i get this:
>
>Traceback (innermost last):
>  File "<interactive input>", line 1, in ?
>IOError: [Errno 22] Invalid argument

Hrmmm.  I have been encountering the exact same problem under Windows
2000.

I have tried all the flavors of os.popen ... os.popen(), os.popen2(),
os.popen3() and os.popen4().  I have also tried the popen2 flavors eg.
popen2.popen2(), popen2.popen3(), and popen2.popen4().  None of them
seem to work...

>>> instream, outstream = popen2.popen2("p4 -x - edit")
>>> instream
<open file 'p4 -x edit', mode 'r' at 017A1C88>
>>> outstream
<open file 'p4 -x edit', mode 'w' at 017A1CB0>
>>> outstream.write('d:\\install\\proj\\projects.xml\n')
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
IOError: [Errno 22] Invalid argument

Does anyone have any bright ideas or suggestions?

-Sean-



From db3l at fitlinxx.com  Tue Feb 13 22:54:44 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 13 Feb 2001 16:54:44 -0500
Subject: os.popen: writing to...?
References: <959pqt$tgp$1@nnrp1.deja.com> <vvti8t07p3pe5b7b1mrb5tq88knmt81g13@4ax.com>
Message-ID: <ubss6w8nv.fsf@ctwd0143.fitlinxx.com>

Sean Laurent <slaurent at sonicfoundry.com> writes:

> >>> instream, outstream = popen2.popen2("p4 -x - edit")
> >>> instream
> <open file 'p4 -x edit', mode 'r' at 017A1C88>
> >>> outstream
> <open file 'p4 -x edit', mode 'w' at 017A1CB0>
> >>> outstream.write('d:\\install\\proj\\projects.xml\n')
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
> IOError: [Errno 22] Invalid argument
> 
> Does anyone have any bright ideas or suggestions?

One way I've been able to replicate this behavior (under NT, not 2000)
is if the process I've started exits before I get around to the
write().

Are you sure that "p4" is still running on your system at the point
when you try to do the write?  Could there be something wrong with the
options you are supplying or just something about running p4 in a
child process that is causing it to terminate early?

Or is it possible that the child shell can't find p4 in the path or
something?  In such a case, you might try reading first from instream
to see the pipe is closed, or if you use popen3, checking the stderr
handle to see if there is an error message from your shell.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From slaurent at sonicfoundry.com  Wed Feb 21 22:57:44 2001
From: slaurent at sonicfoundry.com (Sean Laurent)
Date: Wed, 21 Feb 2001 15:57:44 -0600
Subject: os.popen: writing to...?
References: <959pqt$tgp$1@nnrp1.deja.com> <vvti8t07p3pe5b7b1mrb5tq88knmt81g13@4ax.com> <ubss6w8nv.fsf@ctwd0143.fitlinxx.com>
Message-ID: <42e89t0cgbsta8i1kal9i0s0qrk0vl4jdg@4ax.com>

On 13 Feb 2001 16:54:44 -0500, David Bolen <db3l at fitlinxx.com> wrote:

>Sean Laurent <slaurent at sonicfoundry.com> writes:
>
>> >>> instream, outstream = popen2.popen2("p4 -x - edit")
>> >>> instream
>> <open file 'p4 -x edit', mode 'r' at 017A1C88>
>> >>> outstream
>> <open file 'p4 -x edit', mode 'w' at 017A1CB0>
>> >>> outstream.write('d:\\install\\proj\\projects.xml\n')
>> Traceback (innermost last):
>>   File "<interactive input>", line 1, in ?
>> IOError: [Errno 22] Invalid argument
>> 
>> Does anyone have any bright ideas or suggestions?
<TEXT DELETED>
>Or is it possible that the child shell can't find p4 in the path or
>something?  In such a case, you might try reading first from instream
>to see the pipe is closed, or if you use popen3, checking the stderr
>handle to see if there is an error message from your shell.


Well... I'm very glad you mentioned this... *laugh*

I'm extremely embarassed to admit this, but you were right.  The child
shell could _not_ find "p4" in the path.  Plain and simple.  Staring
me right in the face.

Once I saw your suggestion to check stderr, I couldn't believe I
hadn't done that.  Lo and behold... file not found.  *grin*  The big
problem was that most of my work is done from a command prompt, where
the "p4" command is always in the path.  Unfortunately, my testing
efforts were done from PythonWin, where "p4" was _not_ in the path.

*ugh*

I greatly appreciate your advice... Thanks again.

-Sean-


From dsh8290 at rit.edu  Fri Feb  2 01:35:29 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 1 Feb 2001 19:35:29 -0500
Subject: Design for slices
In-Reply-To: <j85a59.kd5.ln@mail>; from fakeaddress@nospammail.com on Thu, Feb 01, 2001 at 09:54:55AM +1100
References: <959uue$k02$1@cubacola.tninet.se> <j85a59.kd5.ln@mail>
Message-ID: <20010201193529.B6542@harmony.cs.rit.edu>

On Thu, Feb 01, 2001 at 09:54:55AM +1100, Ben Catanzariti wrote:
| I am fairly sure this came about due to the fact you are defining a position
| of a pointer, not an object ie. a letter in a string. So position x[0] means
| the pointer has not moved or has a null value and will occur directly before
| the first letter in a string....
| 
| Some one please correct me if i am wrong ... I am sure you will ;)

I think you are correct, for C.  In C a string can be declared as
either:

char[] /* array of characters */
char*  /* pointer to a character */

The char* can either be a pointer to a single character, or a pointer
to an array of characters.  Thus the following equivalences:

char[10] str1 ;
char*    str2 ;
/* string init not shown */

str1[0] == *str2
str[n] == *(str2 + n)

It also makes the following loop rather short and cute (with the NULL
terminator, of course) :

/* start off 1 element before the array */
char* ptr = str2-1 ; /* or str1 if you prefer */

/* advance the pointer to the end of the string */
while( *(++ptr) ) ;


Along the lines of Alex' example, during my beginning Comp. Sci.
courses I was given the exercise of implementing a 2d array (table).
This was done in Eiffel.  Following the Eiffel principles, the client
would provide both the lower and upper indices.  Not requiring the
client to use zero-based indexing created more work in my ARRAY2 class
because of the mathematical reasons already stated.

-D


| 
| "Gustaf Liljegren" <gustafl at algonet.se> wrote in message
| news:959uue$k02$1 at cubacola.tninet.se...
| > I'm learning Python and have a question about defining ranges/slices in
| > strings. I can't find any defence for what looks like a poor design, so I
| > have to ask. Why is the first character in the string x defined as x[0]
| and
| > not x[1]? This looks just like the typical geek thinking that I'm trying
| to
| > avoid with Python.
| >
| > Gustaf
| 



From scarblac at pino.selwerd.nl  Thu Feb  1 01:50:02 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 1 Feb 2001 00:50:02 GMT
Subject: Design for slices
References: <959uue$k02$1@cubacola.tninet.se>
Message-ID: <slrn97hc7t.ht5.scarblac@pino.selwerd.nl>

Gustaf Liljegren <gustafl at algonet.se> wrote in comp.lang.python:
> I'm learning Python and have a question about defining ranges/slices in 
> strings. I can't find any defence for what looks like a poor design, so I 
> have to ask. Why is the first character in the string x defined as x[0] and 
> not x[1]? This looks just like the typical geek thinking that I'm trying to 
> avoid with Python.

Because experience has taught generations of programmers that this is the
right thing to do.

For a string x of length n, the following simple equations hold:
x == x[0:i]+x[i:n]   for any i (even negative i)
len(x[i:n]) == n-i   for 0 <= i <= n
x[i:i] == []         for any i

This means that it's always simple to split strings. This is also the reason
that a string slice x[i:j] *does not* include x[j]. It really simplifies
situations that would otherwise call for "-1" and "+1" everywhere.
"Off-by-one errors" have caused lots of problems in the past. The Python
choice avoids these problems for the common cases.

People often say that you have to think of the indexes as marks *between*
the actual values. Say, you have the string "Python", then think of it as

 P y t h o n
0 1 2 3 4 5 6

And now it's obvious that "Python"[2:5] == "tho", a string of length 3,
which is expected since 5-2 = 3.

There are more reasons, but you'll come to appreciate them over time.

-- 
Remco Gerlich


From dnew at san.rr.com  Thu Feb  1 05:47:58 2001
From: dnew at san.rr.com (Darren New)
Date: Thu, 01 Feb 2001 04:47:58 GMT
Subject: Design for slices
References: <959uue$k02$1@cubacola.tninet.se>
Message-ID: <3A78EA7E.F0B864B1@san.rr.com>

> Why is the first character in the string x defined as x[0] and
> not x[1]? 

Along with all the mathematical answers, there's also the answer that (most)
computer languages only do one thing at a time. If you're counting and
processing the elements of a string (or list, or whatever), you do it in
three steps:

1) Am I done yet?
2) Process the current item.
3) Move to the next item.

The first step is really "Does the number of items I processed equal the
number of items in the list/string?" And before you process any, the number
of items you've processed is zero.

Try counting out 10 pennies from your hand to the desk, but don't say the
number at the same time your hands are moving. What did you say after you
moved one penny? "One." What did you say *before* you moved any pennies?

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
                 Ignorance can be cured. Naivety cures itself.


From djmoore at uh.edu  Thu Feb  1 00:39:12 2001
From: djmoore at uh.edu (DaveMoore)
Date: Wed, 31 Jan 2001 17:39:12 -0600
Subject: Any practice exercises for beginner???
References: <qz%d6.29887$KP3.8385706@news3.rdc1.on.home.com>
Message-ID: <a12h7tsq9o7bkgdm649cnh5ld49g0op5st@4ax.com>

On Wed, 31 Jan 2001 21:36:22 GMT, "shaka" <fabrice.n at home.com> wrote:
>
> I was just wondering beside the tutorial if there was any practice
>exercises to reinforce beginners strength in python?
>I can recall that when learning Java, some Java website offered some
>exercises for beginners.
>
First of all, check out the Useless Python Page:
<http://www.lowerstandard.com/python/pythonsource.html>.
Download and run the code, see what it does, then try to duplicate it.

There is a link there to the Association for Computing Machinery
Collegiate Programming Contest Problem Set Archive, but it's
been updated:
http://acm.uva.es/problemset/
(Note that contest entries must be in C/C++ or Pascal.)
(Great Grotty Ghu, this thing has an automated on-line
judge! I bet that's a pretty piece of code!)

Here's a problem I recently had fun with:

Many of the Win32 Network Resource enumeration tools return a list of 
dictionaries:
[{'name':'comp1','comment':'room 100','type':WKSTATION},
 {'name':'comp2','comment':'room 101','type':SERVER}]

Try converting this to a dictionary of lists:
{'comp1':['room100',WKSTATION],'comp2':['room 101',SERVER]}

(Typed by hand, hope I got the punctuation right.)

Now step through the dictionary, printing out the characteristics
associated with each computer in a neatly-formatted table.

Hm, didn't come out in order, did it? Fix that.

It's interesting to do this with a large (15-20 entries), sorted
list(the Win32 functions return a list sorted by computer name); if
you print the result dictionary out as each entry is added, you can
watch the dictionary reorganize itself as it grows.

(This is as far as I got, I'm interested in the enumerated resources,
and this turns out not to be the best way to do what I want. But it
was an educational exercise.)

If you have a lot of time to waste, er, invest, write a wrapper that
solicits input to fill the source list, and then lets you search the
resulting dictionary for a particular entry. Duplicates of
already-existing entries should be handled gracefully. If you're
really ambitious, provide a mechanism to edit and delete entries as
well.

OK, now re-write the interface so it uses Tk/Tcl dialogs instead of a
text-based interface.

-- 
Dave Moore == djmoore at uh.edu == I speak for me.
In the wrong hands, sanity is a dangerous weapon.


From mspiggie at my-deja.com  Thu Feb  1 00:45:37 2001
From: mspiggie at my-deja.com (mspiggie at my-deja.com)
Date: Wed, 31 Jan 2001 23:45:37 GMT
Subject: Any practice exercises for beginner???
References: <qz%d6.29887$KP3.8385706@news3.rdc1.on.home.com>
Message-ID: <95a82u$b6r$1@nnrp1.deja.com>

You may find some interesting ideas and source code at:

http://www.lowerstandard.com/python/pythonsource.html

Rob Andrews

In article <qz%d6.29887$KP3.8385706 at news3.rdc1.on.home.com>,
  "shaka" <fabrice.n at home.com> wrote:
>
>
>  I was just wondering beside the tutorial if there was any practice
> exercises to reinforce beginners strength in python?
> I can recall that when learning Java, some Java website offered some
> exercises for beginners.
>
>


Sent via Deja.com
http://www.deja.com/


From scarblac at pino.selwerd.nl  Thu Feb  1 01:36:52 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 1 Feb 2001 00:36:52 GMT
Subject: Any practice exercises for beginner???
References: <qz%d6.29887$KP3.8385706@news3.rdc1.on.home.com>
Message-ID: <slrn97hbf8.ht5.scarblac@pino.selwerd.nl>

shaka <fabrice.n at home.com> wrote in comp.lang.python:
>  I was just wondering beside the tutorial if there was any practice
> exercises to reinforce beginners strength in python?
> I can recall that when learning Java, some Java website offered some
> exercises for beginners.

I'm interested in this sort of thing. Can you tell us what the Java exercise
site was? Probably a good start would be to do the same things in Python.

-- 
Remco Gerlich


From sholden at holdenweb.com  Thu Feb  1 01:18:59 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 31 Jan 2001 19:18:59 -0500
Subject: redirecting output
References: <3A789343.48002B2C@ic.sunysb.edu>
Message-ID: <n92e6.622$j4.34767@e420r-atl2.usenetserver.com>

"Jp Calderone" <jcaldero at ic.sunysb.edu> wrote in message
news:3A789343.48002B2C at ic.sunysb.edu...
>
>
>   I've attached the python interpreter to a server,
> essentially letting people connect to the given address
> and port and pump stuff into it.  What I can't figure out
> is how I might go about redirecting the output from some
> of the builtin functions - dir(), print, etc, to their
> socket.  Is this possible?
>
Interesting idea, with some interesting security aspects!
The dir() function doesn't do any output: it just produces a list.

"print", of course, isn't a statement... but you should be able to
redirect its output by binding sys.stdout to the socket for the duration
of the session.

>  Thanks in advance,
>    Jp
>
regards
Steve




From Bill.Scherer at VerizonWireless.com  Fri Feb  2 06:46:15 2001
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Fri, 02 Feb 2001 00:46:15 -0500
Subject: redirecting output
References: <3A789343.48002B2C@ic.sunysb.edu>
Message-ID: <3A7A49A7.CA67D2DC@VerizonWireless.com>

look at the docs for the sys module.  You can redirect sys.stdout (and
sys.stderr and sys.stdin) to any sutable object...

Jp Calderone wrote:
> 
>   I've attached the python interpreter to a server,
> essentially letting people connect to the given address
> and port and pump stuff into it.  What I can't figure out
> is how I might go about redirecting the output from some
> of the builtin functions - dir(), print, etc, to their
> socket.  Is this possible?
> 
>  Thanks in advance,
>    Jp
> --
> http://mail.python.org/mailman/listinfo/python-list

-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From rjroy at takingcontrol.com  Thu Feb  1 01:17:09 2001
From: rjroy at takingcontrol.com (Robert Roy)
Date: Thu, 01 Feb 2001 00:17:09 GMT
Subject: Best way to match everything between tags
References: <DA0e6.94$o3.4354@news.world-online.no>
Message-ID: <3a78a853.1151132812@news1.on.sympatico.ca>

On Wed, 31 Jan 2001 23:46:17 +0100, "Henning VON ROSEN"
<hvrosen at world-online.no> wrote:

>Hi!
>I am learning regular expressions.
>
>What is thenatural way to match everything that is not "something"
>fx i want to maipulate all the text of a html document, but none of the tags
>
>
>This matches all the tags;
>
>s="text<body>some text<tag>more text</tag>text"
>r = re.compile(r"<.*?>")
> print r.sub("WWW",s)
>
>this works fine if I want to manipulate the tags, but I want to maipulate
>the text between tags
>
>I haven't found a one-expression way. Thankful for any indications!
>
>/Henning von Rosen
>
>
Some ideas:

You should define groups in your re

>>> s="text<body>some text<tag>more text</tag>text"

Definining a group around the content

>>> m = re.search(r'<[^>]*>([^<]*)<', s)
>>> m.groups()
('some text',)

Adding a group around the tag
>>> m = re.search(r'<([^>]*)>([^<]*)<\1', s)
>>> m.groups()
('body', 'some text')

Uing a lookahead to find the matching tag, this will find an inner
pair ie one with no more tags in it

>>> m = re.search(r'<([^>]*)>([^<]*)</\1', s)
>>> m.groups()
('tag', 'more text')

Example of that
>>> s="text<body>some text<tag>more <b>bolded</b> text</tag>text"

>>> m = re.search(r'<([^>]*)>([^<]*)</\1', s)
>>> m.groups()
('b', 'bolded')

This will pick up outermost tag pair
>>> m = re.search(r'<([^>]*)>(.*)</\1', s)
>>> m.groups()
('tag', 'more <b>bolded</b> text')


>>> s="text<body>some text<tag>more <tag> stuff</tag><b>bolded</b> text</tag>text"
>>> m = re.search(r'<([^>]*)>(.*)</\1', s)
>>> m.groups()
('tag', 'more <tag> stuff</tag><b>bolded</b> text')

 do non greedy match, this will pickup leftmost tag and leftmost close
tag
>>> m = re.search(r'<([^>]*)>(.*?)</\1', s)
>>> m.groups()
('tag', 'more <tag> stuff')

Lots of ways to get in trouble here
Bob


From f8dy at my-deja.com  Thu Feb  1 02:16:20 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Thu, 01 Feb 2001 01:16:20 GMT
Subject: Best way to match everything between tags
References: <DA0e6.94$o3.4354@news.world-online.no>
Message-ID: <95adcs$fru$1@nnrp1.deja.com>

In article <DA0e6.94$o3.4354 at news.world-online.no>,
  "Henning VON ROSEN" <hvrosen at world-online.no> wrote:
> Hi!
> I am learning regular expressions.
>
> What is thenatural way to match everything that is not "something"
> fx i want to maipulate all the text of a html document, but none of the tags

Regular expressions may not be the best solution for this.  Try subclassing
SGMLParser and passing the document through the parser.  Specific class
methods will be called for each start tag, each end tag, and each block of
text in between.  This gives you free reign to manipulate the
text-between-the-tags any way you like, and just pass the start/end tags
through unaltered.

See http://www.faqts.com/knowledge_base/view.phtml/aid/4200/fid/549 for a
working example.  The method you'd want to do your manipulation in is
"handle_data".

-M
--
You're smart; why haven't you learned Python yet?  http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From phrxy at csv.warwick.ac.uk  Thu Feb  1 06:10:56 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Thu, 1 Feb 2001 05:10:56 +0000
Subject: Best way to match everything between tags
In-Reply-To: <95adcs$fru$1@nnrp1.deja.com>
References: <DA0e6.94$o3.4354@news.world-online.no> <95adcs$fru$1@nnrp1.deja.com>
Message-ID: <Pine.SOL.4.30.0102010507010.24200-100000@mimosa.csv.warwick.ac.uk>

On Thu, 1 Feb 2001, Mark Pilgrim wrote:

> In article <DA0e6.94$o3.4354 at news.world-online.no>,
>   "Henning VON ROSEN" <hvrosen at world-online.no> wrote:
> > Hi!
> > I am learning regular expressions.
> >
> > What is thenatural way to match everything that is not "something"
> > fx i want to maipulate all the text of a html document, but none of the tags
>
> Regular expressions may not be the best solution for this.  Try subclassing
> SGMLParser and passing the document through the parser.  Specific class
[...]

The perl FAQ has a useful discussion of this 'how do I match html-ish tags
with regexps' question, IIRC (but it's a long time since I looked at
anything related to perl of course).  The short answer is "don't" as Mark
says.


John



From baasad at qualitynet.net  Thu Feb  1 01:07:45 2001
From: baasad at qualitynet.net (bashar A. Asad)
Date: Thu, 01 Feb 2001 03:07:45 +0300
Subject: how about webware
Message-ID: <95au2d$65d2@news.qualitynet.net>

hello;
I would like to know what do you guys think of webware as an application
server.   can I compare it to for example webobjects by apple ?
thankx

bashar


From reborn at neozone.net  Thu Feb  1 01:18:04 2001
From: reborn at neozone.net (¾Æ¹«)
Date: Thu, 1 Feb 2001 09:18:04 +0900
Subject: How can I get the information for jpython?
Message-ID: <95a9va$9rk$1@news1.kornet.net>

I'm studying about jpython..
But.. inpormation is too lack for jpython..
Who do you know the site for jpython?
(except www.jpython.org, www.jython.org)




From Bill.Scherer at VerizonWireless.com  Thu Feb  1 19:59:34 2001
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Thu, 01 Feb 2001 13:59:34 -0500
Subject: How can I get the information for jpython?
References: <95a9va$9rk$1@news1.kornet.net>
Message-ID: <3A79B216.C589C5DC@VerizonWireless.com>

Try:

http://jython.sourceforge.net/

"????" wrote:
> 
> I'm studying about jpython..
> But.. inpormation is too lack for jpython..
> Who do you know the site for jpython?
> (except www.jpython.org, www.jython.org)
> 
> --
> http://mail.python.org/mailman/listinfo/python-list

-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From amritansh at mobilian.com  Thu Feb  1 01:30:06 2001
From: amritansh at mobilian.com (Amritansh Raghav)
Date: Wed, 31 Jan 2001 16:30:06 -0800
Subject: overloading ==
Message-ID: <t7hbgnlmu55l42@corp.supernews.com>

can it be done?
i looked at the operator documentation under Python Runtime Services - but
didnt find anything to suggest it could be done.




From quinn at hork.ugcs.caltech.edu  Thu Feb  1 03:33:25 2001
From: quinn at hork.ugcs.caltech.edu (Quinn Dunkan)
Date: 1 Feb 2001 02:33:25 GMT
Subject: overloading ==
References: <t7hbgnlmu55l42@corp.supernews.com>
Message-ID: <slrn97hinl.i2n.quinn@hork.ugcs.caltech.edu>

On Wed, 31 Jan 2001 16:30:06 -0800, Amritansh Raghav <amritansh at mobilian.com>
wrote:
>can it be done?
>i looked at the operator documentation under Python Runtime Services - but
>didnt find anything to suggest it could be done.

Yes, define __cmp__ or __rcmp__.  See 3.3.1 in the language reference.


From drewpc at my-deja.com  Thu Feb  1 21:55:05 2001
From: drewpc at my-deja.com (drewpc at my-deja.com)
Date: Thu, 01 Feb 2001 20:55:05 GMT
Subject: overloading ==
References: <t7hbgnlmu55l42@corp.supernews.com> <slrn97hinl.i2n.quinn@hork.ugcs.caltech.edu>
Message-ID: <95cif3$bi3$1@nnrp1.deja.com>

In article
<slrn97hinl.i2n.quinn at hork.ugcs.caltech.edu>,
  quinn at hork.ugcs.caltech.edu (Quinn Dunkan)
wrote:
> On Wed, 31 Jan 2001 16:30:06 -0800, Amritansh
Raghav <amritansh at mobilian.com>
> wrote:
> >can it be done?
> >i looked at the operator documentation under
Python Runtime Services - but
> >didnt find anything to suggest it could be
done.
>
> Yes, define __cmp__ or __rcmp__.  See 3.3.1 in
the language reference.
>

if you look at the release notes for Python 2.1,
they've also added the following functions:

Operation          Method name
     <                      __lt__
     <=                    __le__
     >                      __gt__
     >=                    __ge__
     ==                    __eq__
     !=                     __ne__

(http://www.amk.ca/python/2.1/index.html#SECTION000400000000000000000)


Sent via Deja.com
http://www.deja.com/


From dbrueck at edgix.com  Thu Feb  1 20:26:11 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Thu, 1 Feb 2001 12:26:11 -0700
Subject: overloading ==
In-Reply-To: <t7hbgnlmu55l42@corp.supernews.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHIECKPIAA.dbrueck@edgix.com>

Add a __cmp__ method to your class:

def __cmp__(self, obj):
  # return 0, a negative, or a positive number

(you can also define __rcmp__ to handle cases where the object on the left
side of the conversion doesn't have a cmp method)

Note that if you define __cmp__ but not a __hash__ then your objects can't
act as dictionary keys (which is correct if they are mutable).

-Dave

> [mailto:python-list-admin at python.org]On Behalf Of Amritansh Raghav
>
> can it be done?
> i looked at the operator documentation under Python Runtime Services - but
> didnt find anything to suggest it could be done.




From nas at arctrix.com  Fri Feb  2 04:19:42 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Thu, 1 Feb 2001 19:19:42 -0800
Subject: overloading ==
In-Reply-To: <BLEOJGEMBNBNEAKMIFDHIECKPIAA.dbrueck@edgix.com>; from dbrueck@edgix.com on Thu, Feb 01, 2001 at 12:26:11PM -0700
References: <t7hbgnlmu55l42@corp.supernews.com> <BLEOJGEMBNBNEAKMIFDHIECKPIAA.dbrueck@edgix.com>
Message-ID: <20010201191942.A27454@glacier.fnational.com>

On Thu, Feb 01, 2001 at 12:26:11PM -0700, Dave Brueck wrote:
> (you can also define __rcmp__ to handle cases where the object on the left
> side of the conversion doesn't have a cmp method)

Its a lot more complicated that that. Python 2.1 will never call
__rcmp__.  Versions of Python older than 2.1 will need __rcmp__
only in some rare cases (that I can't remember at the moment).
Just define __cmp__ and forget about __rcmp__.  

  Neil



From drewpc at my-deja.com  Thu Feb  1 21:56:54 2001
From: drewpc at my-deja.com (drewpc at my-deja.com)
Date: Thu, 01 Feb 2001 20:56:54 GMT
Subject: overloading ==
References: <t7hbgnlmu55l42@corp.supernews.com>
Message-ID: <95ciif$bl8$1@nnrp1.deja.com>

In article <t7hbgnlmu55l42 at corp.supernews.com>,
  "Amritansh Raghav" <amritansh at mobilian.com> wrote:
> can it be done?
> i looked at the operator documentation under Python Runtime Services -
but
> didnt find anything to suggest it could be done.
>
>

It's better in Python 2.1.  Checkout the following URL:

http://www.amk.ca/python/2.1/index.html#SECTION000400000000000000000


Sent via Deja.com
http://www.deja.com/


From mike at skyweb.ca  Thu Feb  1 03:22:47 2001
From: mike at skyweb.ca (mike)
Date: Wed, 31 Jan 2001 20:22:47 -0600
Subject: serial port
Message-ID: <3a78ca89_2@skyweb.ca>

is there any modules for serial port communication?




From news at dorb.com  Thu Feb  1 04:06:51 2001
From: news at dorb.com (Darrell)
Date: Thu, 01 Feb 2001 03:06:51 GMT
Subject: serial port
References: <3a78ca89_2@skyweb.ca>
Message-ID: <fp4e6.142909$ju6.13499270@typhoon.nyroc.rr.com>

For windows check this out.    win32comport_demo.py

--Darrell

"mike" <mike at skyweb.ca> wrote in message news:3a78ca89_2 at skyweb.ca...
> is there any modules for serial port communication?
>
>




From grante at visi.com  Fri Feb  9 20:34:15 2001
From: grante at visi.com (Grant Edwards)
Date: Fri, 09 Feb 2001 19:34:15 GMT
Subject: serial port
References: <3a78ca89_2@skyweb.ca> <fp4e6.142909$ju6.13499270@typhoon.nyroc.rr.com>
Message-ID: <XCXg6.1870$a7.64117@ruti.visi.com>

In article <fp4e6.142909$ju6.13499270 at typhoon.nyroc.rr.com>, Darrell wrote:

>> is there any modules for serial port communication?

>For windows check this out.    win32comport_demo.py

For Posix, you can have a whack at

  ftp://ftp.visi.com/users/grante/python/PosixSerial.py

There aren't any docs, but an example Python/GTK app that uses
the above module:

  ftp://ftp.visi.com/users/grante/python/lcom.py

-- 
Grant Edwards                   grante             Yow!  Did YOU find a
                                  at               DIGITAL WATCH in YOUR box
                               visi.com            of VELVEETA?


From peter at engcorp.com  Thu Feb  1 04:56:50 2001
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 31 Jan 2001 22:56:50 -0500
Subject: serial port
References: <3a78ca89_2@skyweb.ca>
Message-ID: <3A78DE82.1EE39E06@engcorp.com>

mike wrote:
> 
> is there any modules for serial port communication?

Yes.  No.  Depends on the platform.  Are you running
on Windows?  Linux?  Elsewhere?


From embed at geocities.com  Mon Feb 19 22:29:40 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 19 Feb 2001 16:29:40 -0500
Subject: serial port
References: <3a78ca89_2@skyweb.ca> <3A78DE82.1EE39E06@engcorp.com>
Message-ID: <%fgk6.99$TW.1009@tor-nn1.netcom.ca>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3A78DE82.1EE39E06 at engcorp.com...
> mike wrote:
> >
> > is there any modules for serial port communication?
>
> Yes.  No.  Depends on the platform.  Are you running
> on Windows?  Linux?  Elsewhere?

This seems a darn shame.

I think a single portable API should be developed. One that will enumerate
the serial devices, query their speed-range abilities, and so on, so that a
single source can handle unix /dev/xxx semantics, Dos/Windows style "COMx:"
labels, etcetera.

I have already written one for win32, which I think is quite nice, and which
doesn't do anything that can't be done on all platforms. (No overlapped IO
support on Win32 for instance.)

Somebody better do something or else.... or else.... or else I will do it
myself, gosh darn it! ;-)

Warren





From grante at visi.com  Tue Feb 20 04:16:59 2001
From: grante at visi.com (Grant Edwards)
Date: Tue, 20 Feb 2001 03:16:59 GMT
Subject: serial port
References: <3a78ca89_2@skyweb.ca>
Message-ID: <slrn993odb.e75.grante@isis.visi.com>

In article <3a78ca89_2 at skyweb.ca>, mike wrote:

>is there any modules for serial port communication?

Yes.  There is (or was) a Win32 module called Sio, but I can't seem to find
it at www.python.org.  There used to be a well-hidden link somewhere on that
website.

I wrote a Posix serial I/O class that is still waiting to be documented

  ftp://ftp.visi.com/users/grante/python/PosixSerial.py

-- 
Grant Edwards                   grante             Yow!  .. or were you
                                  at               driving the PONTIAC that
                               visi.com            HONKED at me in MIAMI last
                                                   Tuesday?


From fuess at att.net  Wed Feb 21 05:59:52 2001
From: fuess at att.net (David Fuess)
Date: Wed, 21 Feb 2001 04:59:52 GMT
Subject: serial port
References: <3a78ca89_2@skyweb.ca> <slrn993odb.e75.grante@isis.visi.com>
Message-ID: <1mi69ts78koo7oavqh4htfs9ku5u54uqij@4ax.com>

You might want to check out Roger Burnham's SioModule available at
http://starship.python.net/crew/roger/. It is a complete Serial class
for Python. According to Hammond's "Python Programming on Win32" p373,
it is the "preferred route for serial I/O with Python"

Dave

On Tue, 20 Feb 2001 03:16:59 GMT, grante at visi.com (Grant Edwards)
wrote:

>In article <3a78ca89_2 at skyweb.ca>, mike wrote:
>
>>is there any modules for serial port communication?
>
>Yes.  There is (or was) a Win32 module called Sio, but I can't seem to find
>it at www.python.org.  There used to be a well-hidden link somewhere on that
>website.
>
>I wrote a Posix serial I/O class that is still waiting to be documented
>
>  ftp://ftp.visi.com/users/grante/python/PosixSerial.py



From MarkH at ActiveState.com  Wed Feb 21 23:41:42 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 21 Feb 2001 22:41:42 GMT
Subject: serial port
References: <3a78ca89_2@skyweb.ca> <slrn993odb.e75.grante@isis.visi.com> <1mi69ts78koo7oavqh4htfs9ku5u54uqij@4ax.com>
Message-ID: <3A94434D.6020800@ActiveState.com>

David Fuess wrote:

> You might want to check out Roger Burnham's SioModule available at
> http://starship.python.net/crew/roger/. It is a complete Serial class
> for Python. According to Hammond's "Python Programming on Win32" p373,
> it is the "preferred route for serial I/O with Python"

Note that the second edition is likely to point out that the win32file module has "native" serial port access available, and doesnt rely on a 3rd party DLL ;-)

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From fuess at att.net  Thu Feb 22 02:04:39 2001
From: fuess at att.net (David Fuess)
Date: Thu, 22 Feb 2001 01:04:39 GMT
Subject: serial port
References: <3a78ca89_2@skyweb.ca> <slrn993odb.e75.grante@isis.visi.com> <1mi69ts78koo7oavqh4htfs9ku5u54uqij@4ax.com> <3A94434D.6020800@ActiveState.com>
Message-ID: <k4p89toa1rsqcdi0rgp8u0fagqfjka7is4@4ax.com>

Wow, nothing like getting it from the horse's mouth, so to speak! Any
chance of an update on that chapter for us owners of well worn first
editions?

BTW, I checked out the  SioModule class further and found that it only
works with Python 1.5.2. I was rather disappointed that I was unable
to use it with my ActivePython 2.0! Do you have a work around Mark?

Thanks for PP on Win32, keep up the great work
Dave

On Wed, 21 Feb 2001 22:41:42 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>David Fuess wrote:
>
>> You might want to check out Roger Burnham's SioModule available at
>> http://starship.python.net/crew/roger/. It is a complete Serial class
>> for Python. According to Hammond's "Python Programming on Win32" p373,
>> it is the "preferred route for serial I/O with Python"
>
>Note that the second edition is likely to point out that the win32file module has "native" serial port access available, and doesnt rely on a 3rd party DLL ;-)
>
>Mark.



From MarkH at ActiveState.com  Thu Feb 22 07:07:16 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 22 Feb 2001 06:07:16 GMT
Subject: serial port
References: <3a78ca89_2@skyweb.ca> <slrn993odb.e75.grante@isis.visi.com> <1mi69ts78koo7oavqh4htfs9ku5u54uqij@4ax.com> <3A94434D.6020800@ActiveState.com> <k4p89toa1rsqcdi0rgp8u0fagqfjka7is4@4ax.com>
Message-ID: <3A94ABBF.8040509@ActiveState.com>

David Fuess wrote:

> Wow, nothing like getting it from the horse's mouth, so to speak! Any
> chance of an update on that chapter for us owners of well worn first
> editions?

Not for a little while :)  There is no second edition in the works ATM!  Check out the win32comport_demo.py file - that is about all I would cover :)

> BTW, I checked out the  SioModule class further and found that it only
> works with Python 1.5.2. I was rather disappointed that I was unable
> to use it with my ActivePython 2.0! Do you have a work around Mark?

Hrm.  There is a version for Python 2.0, but I can't find it.  It has been mentioned here a few times, so check google.

Mark.



From knotwell at ix.netcom.com  Thu Feb  1 04:50:07 2001
From: knotwell at ix.netcom.com (Brad Knotwell)
Date: 31 Jan 2001 19:50:07 -0800
Subject: question on threading.py
Message-ID: <86k87brrio.fsf@localhost.my.domain>

Hello all--

Do the _Condition methods notify and wait really need to call 
currentThread?  Unless currentThread has some side-effect I couldn't see
I think these methods don't have any reason to call them.

Lines 182 and 215 of v2.0 threading.py.

--Brad




From loewis at informatik.hu-berlin.de  Fri Feb  2 22:42:15 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:42:15 +0100
Subject: question on threading.py
References: <86k87brrio.fsf@localhost.my.domain>
Message-ID: <p6qu26coj7s.fsf@informatik.hu-berlin.de>

Brad Knotwell <knotwell at ix.netcom.com> writes:

> Do the _Condition methods notify and wait really need to call
> currentThread?  Unless currentThread has some side-effect I couldn't
> see I think these methods don't have any reason to call them.

I agree. Want to submit a patch to sourceforge.net/projects/python?

Regards,
Martin


From bill at janssen.org  Thu Feb  1 05:24:48 2001
From: bill at janssen.org (Bill Janssen)
Date: Wed, 31 Jan 2001 20:24:48 PST
Subject: Palm extensions for PIL available
Message-ID: <01Jan31.202456pst."3448"@watson.parc.xerox.com>

Fredrik,

I've completed some work on extending PIL 1.1.1 to support raw Palm
image quantization and output.  You can find a gzipped tar file
containing a patch and a new ImagePlugin at
ftp://ftp.parc.xerox.com/transient/janssen/PILpalm.tgz.

The README in that file explains how to apply the support to a vanilla
PIL 1.1.1 source distribution.

The most interesting thing to you, I think, is the extension of the
quantize method, via a new keyword, to support quantizing the colors
of an image to a specified palette.  That palette is specified by
passing an image containing the desired palette.  There were a number
of ways to add this functionality; this seemed to be the approach that
involved the minimal change to PIL.

If you'd like to discuss any philosophical or technical issues about
how I've done this, I'd be happy to do so.

I'd appreciate it if this got into the next release of PIL; of course
your release schedule is your own.

Regards,

Bill




From DSHIN at nc.rr.com  Thu Feb  1 06:43:33 2001
From: DSHIN at nc.rr.com (Daehyok Shin)
Date: Thu, 01 Feb 2001 05:43:33 GMT
Subject: [Q] ftplib: How to compare date?
Message-ID: <9I6e6.7505$p8.1228772@typhoon.southeast.rr.com>

I like to compare the creation dates of two files using ftplib.
But, I cannot find any function to get the date of remote files.
How can I do it?

--
Daehyok Shin (Peter)





From clgonsal at keeshah.penguinpowered.com  Thu Feb  1 08:27:02 2001
From: clgonsal at keeshah.penguinpowered.com (C.Laurence Gonsalves)
Date: Thu, 01 Feb 2001 07:27:02 GMT
Subject: [Q] ftplib: How to compare date?
References: <9I6e6.7505$p8.1228772@typhoon.southeast.rr.com>
Message-ID: <slrn97i4kh.abo.clgonsal@keeshah.penguinpowered.com>

On Thu, 01 Feb 2001 05:43:33 GMT, Daehyok Shin <DSHIN at nc.rr.com> wrote:
>I like to compare the creation dates of two files using ftplib.
>But, I cannot find any function to get the date of remote files.
>How can I do it?

I don't think there's any way to do this reliably. You can try and get a
list of the directory, but different FTP servers will return the listing
in different formats. If you only want it to work on known FTP servers,
you can parse the format yourself though (string.split and time.strptime
would probably come in handy at this point).

Also, if you want to compare the time stamps of files on the local
system with the time stamps of files on a remote FTP server, be aware
that the clocks probably aren't in sync, and might not even be in the
same time zone.

-- 
  C. Laurence Gonsalves                "Any sufficiently advanced
  clgonsal at kami.com                     technology is indistinguishable
  http://cryogen.com/clgonsal/          from magic." -- Arthur C. Clarke


From bsb at winnegan.de  Thu Feb  1 22:49:33 2001
From: bsb at winnegan.de (Siggy Brentrup)
Date: 01 Feb 2001 22:49:33 +0100
Subject: [Q] ftplib: How to compare date?
In-Reply-To: <9I6e6.7505$p8.1228772@typhoon.southeast.rr.com>
References: <9I6e6.7505$p8.1228772@typhoon.southeast.rr.com>
Message-ID: <87ae86nkeq.fsf@winnegan.de>

"Daehyok Shin" <DSHIN at nc.rr.com> writes:

> I like to compare the creation dates of two files using ftplib.
> But, I cannot find any function to get the date of remote files.
> How can I do it?

Try this:

>>> ftp=ftplib.FTP('localhost')
>>> print ftp.welcome
220 ProFTPD 1.2.0pre10 Server (Debian) [keuner.winnegan.de]
>>> welcome=ftp.login('me','xyzzy')
>>> ftp.sendcmd('MDTM /etc/passwd')
'213 20010131081337'

AFAIK, ftp servers return UTC.

HTH
  Siggy






From DSHIN at nc.rr.com  Fri Feb  2 07:00:43 2001
From: DSHIN at nc.rr.com (Daehyok Shin)
Date: Fri, 02 Feb 2001 06:00:43 GMT
Subject: [Q] ftplib: How to compare date?
References: <9I6e6.7505$p8.1228772@typhoon.southeast.rr.com> <mailman.981067505.10896.python-list@python.org>
Message-ID: <f2se6.10062$k8.2095707@typhoon.southeast.rr.com>

What I want is not the local time of systems,
but the creation time of files.


"Siggy Brentrup" <bsb at winnegan.de> wrote in message
news:mailman.981067505.10896.python-list at python.org...
> "Daehyok Shin" <DSHIN at nc.rr.com> writes:
>
> > I like to compare the creation dates of two files using ftplib.
> > But, I cannot find any function to get the date of remote files.
> > How can I do it?
>
> Try this:
>
> >>> ftp=ftplib.FTP('localhost')
> >>> print ftp.welcome
> 220 ProFTPD 1.2.0pre10 Server (Debian) [keuner.winnegan.de]
> >>> welcome=ftp.login('me','xyzzy')
> >>> ftp.sendcmd('MDTM /etc/passwd')
> '213 20010131081337'
>
> AFAIK, ftp servers return UTC.
>
> HTH
>   Siggy
>
>
>
>




From shredwheat at mediaone.net  Thu Feb  1 09:02:21 2001
From: shredwheat at mediaone.net (Pete Shinners)
Date: Thu, 01 Feb 2001 08:02:21 GMT
Subject: "in" operator for strings
Message-ID: <hK8e6.141262$y9.26215366@typhoon.we.rr.com>

with the back and forth going on about the "in" operator and
dictionaries, i was thinking it would be nice for the "in"
operator to also check for the presence of substrings in a
string. for example...


>>> "Waldo" in "Ralph Emerson"
0
>>> "Waldo" in "Ralph Waldo Emerson"
1


i find this much more intuitive and graceful than the current
syntax i am repeatedly using...

>>> "Ralph Emerson".find("Waldo") != -1
0


anyways, it may not stand a snowball's chance. but i'd hope
the 'powers that be' could at least take a second to think
about it and explain a yea or nea.

it's just an idea that came to me. i know i would like it, but
there is likely something i'm overlooking. i would think it
could have a serious chance of being accepted?





From hvrosen at world-online.no  Thu Feb  1 10:50:17 2001
From: hvrosen at world-online.no (Henning VON ROSEN)
Date: Thu, 1 Feb 2001 10:50:17 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com>
Message-ID: <2jae6.119$o3.4587@news.world-online.no>

"Pete Shinners" <shredwheat at mediaone.net> skrev i melding
news:hK8e6.141262$y9.26215366 at typhoon.we.rr.com...
> with the back and forth going on about the "in" operator and
> dictionaries, i was thinking it would be nice for the "in"
> operator to also check for the presence of substrings in a
> string. for example...
>
>
> >>> "Waldo" in "Ralph Emerson"
> 0
> >>> "Waldo" in "Ralph Waldo Emerson"
> 1

try this:
>>> "Waldo" in "Ralph Waldo Emerson".split()
1

as long as you are looking for a "word"

amike via Henning

> i find this much more intuitive and graceful than the current
> syntax i am repeatedly using...
>
> >>> "Ralph Emerson".find("Waldo") != -1
> 0
>
>
> anyways, it may not stand a snowball's chance. but i'd hope
> the 'powers that be' could at least take a second to think
> about it and explain a yea or nea.
>
> it's just an idea that came to me. i know i would like it, but
> there is likely something i'm overlooking. i would think it
> could have a serious chance of being accepted?
>
>
>




From mlh at idi.ntnu.no  Thu Feb  1 11:36:24 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 1 Feb 2001 11:36:24 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no>
Message-ID: <95beco$e0u$1@tyfon.itea.ntnu.no>

"Henning VON ROSEN" <hvrosen at world-online.no> wrote in message
news:2jae6.119$o3.4587 at news.world-online.no...
>
> "Pete Shinners" <shredwheat at mediaone.net> skrev i melding
[...]
> > >>> "Waldo" in "Ralph Emerson"
> > 0
> > >>> "Waldo" in "Ralph Waldo Emerson"
> > 1
>

This isn't quite logical... A string works like a sequence
of characters, and sequence membership only works on
single elements (in this case characters), not subsequences
(in this case, substrings).

So you could well say

  "W" in "Ralph Waldo Emerson"

(giving a 1), just as in

  1 in [1, 2, 3, 4, 5]

But you can't do what you ask for, just like you can't write

  [1, 2] in [1, 2, 3, 4]

> try this:
> >>> "Waldo" in "Ralph Waldo Emerson".split()
> 1

Probably a better idea. And for those of us who can't really
get used to calling methods directly on string literals,
you *could* write:

   "Waldo" in split("Ralph Waldo Emerson")

It might be old-fashioned, but... So what :-)

>
> as long as you are looking for a "word"

Right.

>
> amike via Henning

Ho - Esperantisto, Cxu? <palpebrumo>

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From aleaxit at yahoo.com  Thu Feb  1 12:55:32 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 1 Feb 2001 12:55:32 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no>
Message-ID: <95birk0r0d@news1.newsguy.com>

"Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
news:95beco$e0u$1 at tyfon.itea.ntnu.no...
    [snip]
> This isn't quite logical... A string works like a sequence
> of characters, and sequence membership only works on
> single elements (in this case characters), not subsequences
> (in this case, substrings).

Right, and an extension of this is basically what's being
asked for (though the original poster may not have thought
of this 'obvious' generalization, specialcasing string would
surely not be warranted).  Unfortunately, for general cases
it doesn't scale well -- i.e., now:

>>> print [1,2] in [6, 4, [1,2], 7]
1
>>> print [6,4] in [6, 4, [1,2], 7]
0

and having it return 1 in the second case too would be making
this 'in' very ambiguous and confusing, alas.

Also, of course, this would throw any parallel between
"x in y" and "for x in y" out of the windows unless the
latter starts looping on all *subsequences* -- eeep!-)

> But you can't do what you ask for, just like you can't write
>
>   [1, 2] in [1, 2, 3, 4]

Sure you can, it's a well-formed test and returns 0 since
[1,2] is not an item in the right-hand operand sequence.


> > try this:
> > >>> "Waldo" in "Ralph Waldo Emerson".split()
> > 1
>
> Probably a better idea.

Only if you're looking for words, not for any substring,
which is at least as frequent.

> And for those of us who can't really
> get used to calling methods directly on string literals,
> you *could* write:
>
>    "Waldo" in split("Ralph Waldo Emerson")
>
> It might be old-fashioned, but... So what :-)

So it doesn't work unless you "from string import *" (horrid
idea), "from string import split" (doubtful), or rewrite it
using an explicit string.split (probably best, but quite
verbose compared to using the split string-method -- and
you still have to add an import string to your module, which
is rarely needed now that we DO have string methods).


For general substring-matching, a class wrapper is not
too bad:

class subsOf:
    def __init__(self, seq):
        self.seq = seq
    def __contains__(self, subseq):
        return self.seq.find(subseq) != -1

this only works for strings, as written, AND only to
enable such idioms as

    if 'ald' in subsOf("Waldo"):
        print 'yep!'

but it's not too hard to generalize it to any sequence
type (at least if you're content to use elementary
algorithms for __contains__!-), implement __getitem__
to allow looping on all subsequences:-), etc.


Alex





From cjc26 at nospam.cornell.edu  Thu Feb  1 15:57:52 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Thu, 01 Feb 2001 14:57:52 GMT
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com>
Message-ID: <QPee6.1195$o91.132115@typhoon.nyroc.rr.com>

* Alex Martelli <aleaxit at yahoo.com> menulis:
| "Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
| news:95beco$e0u$1 at tyfon.itea.ntnu.no...
|     [snip]
| > This isn't quite logical... A string works like a sequence
| > of characters, and sequence membership only works on
| > single elements (in this case characters), not subsequences
| > (in this case, substrings).
| 
| Right, and an extension of this is basically what's being
| asked for (though the original poster may not have thought
| of this 'obvious' generalization, specialcasing string would
| surely not be warranted).

Strings seem to be special-cased already; for example, you don't have to
type

>>> print ''.join(['H','e','l','l','o',',',' ','w','o','r','l','d','!'])

if you don't want to <wink>.

Anyway, I think having 'in' look for substrings, when used with strings,
is more useful than its current behavior.  Having 'in' look for a single
element works well with lists and tuples, but when using strings I think
it's more common to search for a substring than to search for a single
character.


| Unfortunately, for general cases
| it doesn't scale well -- i.e., now:
| 
| >>> print [1,2] in [6, 4, [1,2], 7]
| 1
| >>> print [6,4] in [6, 4, [1,2], 7]
| 0
| 
| and having it return 1 in the second case too would be making
| this 'in' very ambiguous and confusing, alas.

Right, but we will never run into this problem with strings, because its
elements are always single characters.  So there wouldn't be any
ambiguity in extending the meaning of 'in' in this case.


| Also, of course, this would throw any parallel between
| "x in y" and "for x in y" out of the windows unless the
| latter starts looping on all *subsequences* -- eeep!-)

Again, I don't think this would be a problem if substring 'in' only
worked with strings and not all sequences, because I don't think
"for x in y", where y is a string, is a common construction.  In fact,
whenever this pops up in my code, it usually means there's a bug
somewhere (i.e. I forgot to do y.split() or something like that).
"for x in str" can be useful sometimes, but I think most of the time
string methods (or functions from the string module, if you're
old-fashioned ;) and indexing are used instead.


| [snip]
| 
| For general substring-matching, a class wrapper is not
| too bad:
| 
| class subsOf:
|     def __init__(self, seq):
|         self.seq = seq
|     def __contains__(self, subseq):
|         return self.seq.find(subseq) != -1
| 
| this only works for strings, as written, AND only to
| enable such idioms as
| 
|     if 'ald' in subsOf("Waldo"):
|         print 'yep!'
| 
| but it's not too hard to generalize it to any sequence
| type (at least if you're content to use elementary
| algorithms for __contains__!-), implement __getitem__
| to allow looping on all subsequences:-), etc.

This is probably the best solution, since I'm sure no one will go for
yet another change to the language <0.3wink>.


-- 
Cliff Crawford               http://www.people.cornell.edu/pages/cjc26/
                             print "Just another Python hacker"


From aleaxit at yahoo.com  Fri Feb  2 14:02:14 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 2 Feb 2001 14:02:14 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com>
Message-ID: <95eb4r0csk@news1.newsguy.com>

"Cliff Crawford" <cjc26 at nospam.cornell.edu> wrote in message
news:QPee6.1195$o91.132115 at typhoon.nyroc.rr.com...
> * Alex Martelli <aleaxit at yahoo.com> menulis:
    [snip]
> | of this 'obvious' generalization, specialcasing string would
> | surely not be warranted).
>
> Strings seem to be special-cased already; for example, you don't have to
> type
>
> >>> print ''.join(['H','e','l','l','o',',',' ','w','o','r','l','d','!'])
>
> if you don't want to <wink>.

Literals have different syntax sugar for various kinds of sequences
(tuples, lists, strings, Unicode strings -- that's it, others, such
as array.array's, are built without any peculiar syntax sugar).  That
is vastly different from having one operator, with identical syntax,
get completely different semantics for one sequence type (or two?)
versus all others; THAT one is a recipe for disaster & confusion.

> Anyway, I think having 'in' look for substrings, when used with strings,
> is more useful than its current behavior.

And maybe, since hardly anybody ever does modulo operations on
complex numbers, having '%' mean something utterly different
for them wrt any other kind of number would be "more useful" --
and somebody might argue that it's OK, because complex numbers
"seem to be special-cased already" (horribly confusing issues
of syntax sugar used for literals, with issues of uniformity
of operator semantics over various kinds of numbers).

And wouldn't it be "more useful" to have "foo"() look for a
function NAMED foo and call it, rather than complain about a
"call of non-function"?

Etc, etc.  This way Per^H^H^H madness lies.  Building a language
out of a collection of "more useful" (the usual euphemism for
such unspeakable horrors is "convenient") irregularities, ad
hoc "solutions", special cases, and second guessing of coder's
intent... ***thanks, but NO, THANKS***.  Python's *regularity and
simplicity* are its key strengths.

Python might perfectly well have decided that strings do not
expose a sequence interface, but rather are 'atoms' (one would
then explicitly call some suitable method, returning a list or
tuple of the string's characters, in those relatively rare
occasions where treating this as a sequence is needed).  That
would have some pluses, and some minuses, breaking roughly
even overall IMHO.  But if strings are sequences, and I see
little prospect for that to change (short of the mythical
"Python 3000", and I don't think such a change will in fact
happen even then), then having the 'in' operator mean
something utterly different for them, than it does for all
other sequences, would be nothing short of a disaster (I'd
_almost_ rule this out as a possibility, except that we have
one historical precedent -- if somebody in the core dev't
team feels strongly about it and does a patch, and the BDFL
unexpectedly agrees, then a horrid, unmitigated wart CAN be
added to Python; it's happened once, and now the team spirit
of the core developers turns them into apologists for the
wart -- oh well, on past performance we can expect about one
such occurrence per decade).

 for relatively rare
occasions where one wants the list or tuple of


> Again, I don't think this would be a problem if substring 'in' only
> worked with strings and not all sequences, because I don't think
> "for x in y", where y is a string, is a common construction.  In fact,

It's quite frequent in my code (mostly with strings built by
struct, or read from binary files, etc, and representing "an
array of bytes" rather than "a ``veritable'' string" in some
sense).  If strings were atomic, I would of course turn them
into buffers for such purposes (and buffers should probably
be used in many places where now strings are).  As strings are
sequences, then I expect them to behave just like any other
(non-mutable) sequence, and, in particular, to have a loose
parallel between 'if x in s' and 'for x in s'.  E.g., for
any sequences a and b of the same type:

def inboth1(a, b):
    return [x for x in a if x in b]

def inboth2(a, b):
    return [x for x in b if x in a]

switching the roles of a and b may affect the ordering and
repetitions in the result sequence, but NOT the membership
thereof -- a handy and useful axiom (which allows certain
refactoring operations to be applied without needing much
global context to check their correctness - one of the
greatest boons of such axioms, of course).


> This is probably the best solution, since I'm sure no one will go for
> yet another change to the language <0.3wink>.

The language keeps being changed (and there are grumblings
whenever that happens, of course).  A change to make strings
not sequences would probably be too pervasive to consider,
breaking zillions of lines of working code.  A change to
break the uniformity in the behavior of sequences of various
types is unfortunately more conceivable, but it would be an
very very bad idea if it ever happened.


Alex





From sholden at holdenweb.com  Fri Feb  2 23:21:26 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 2 Feb 2001 17:21:26 -0500
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com>
Message-ID: <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:95eb4r0csk at news1.newsguy.com...
> "Cliff Crawford" <cjc26 at nospam.cornell.edu> wrote in message
> news:QPee6.1195$o91.132115 at typhoon.nyroc.rr.com...
> > * Alex Martelli <aleaxit at yahoo.com> menulis:
>     [snip]
> > | of this 'obvious' generalization, specialcasing string would
> > | surely not be warranted).
> >
> > Strings seem to be special-cased already; for example, you don't have to
> > type
> >
> > >>> print ''.join(['H','e','l','l','o',',',' ','w','o','r','l','d','!'])
> >
> > if you don't want to <wink>.
>
The most anomalous thing about strings from my point of view is the lack of
a character dataype.  So

>>> lst = [1, 2, 3]
>>> type(lst)
<type 'list'>
>>> type(lst[1])
<type 'int'>
>>> strng = "123"
>>> type(strng)
<type 'string'>
>>> type(strng[1])
<type 'string'>

I wouldn't seriously argue that we _need_ a character type, but it's
certainly anomalous that an element of a sequence is itself a sequence.
Just the same, this kind of pragmatism is what keeps Python in the forefront
of popularity (in so far as it is).

[ ... ]

> > This is probably the best solution, since I'm sure no one will go for
> > yet another change to the language <0.3wink>.
>
> The language keeps being changed (and there are grumblings
> whenever that happens, of course).  A change to make strings
> not sequences would probably be too pervasive to consider,
> breaking zillions of lines of working code.  A change to
> break the uniformity in the behavior of sequences of various
> types is unfortunately more conceivable, but it would be an
> very very bad idea if it ever happened.
>
Python 2.0 came long enough after 1.5.2, and contained enough improvements,
that most users would take whatever pain there was in converting as the
price of gaining access to the new features ... all except "print >>," that
is, which was and remains an abomination on the syntax of the language
worthy of Bill Gates at his worst (he did write code, and I suspect that
much of it was awful). Execrescence. Mother of all warts. Martelli shall not
be forgotten!

But now here we are, less than six months after a major release, looking at
tweaks and fiddles and optimizations. It strikes me that the Python
community is in danger of falling victim to the "see how much more we can
stuff in without breaking backwards compatibility" syndrome.  Just because a
proposed change doesn't break existing code doesn't necessarily make it a
good thing. Subtle changes to the semantics of the language invalidate
teaching materials, for example.

The situation seemd to me to be better and more rational when we expected
Python 3K to be a major break from current traditions. At least this gave
the twiddlers and the tweakers some hope, without putting the existing
language at risk. There are undoubtledy arguments for many of the 2.1
changes, but I findsome of them unconvincing. Why not keep Python as it is
now, and prodce 3K about 998 years early?

There's a lot to be said for stability.

wishing-i-had-a-bit-more-myself-ly y'rs  - steve




From chris.gonnerman at usa.net  Sat Feb  3 16:57:32 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 3 Feb 2001 09:57:32 -0600
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com> <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com>
Message-ID: <006701c08dfa$08ed8c60$a100000a@local>

----- Original Message -----
From: "Steve Holden" <sholden at holdenweb.com>
Subject: Re: "in" operator for strings
[ hack ]

> But now here we are, less than six months after a major release, looking
at
> tweaks and fiddles and optimizations. It strikes me that the Python
> community is in danger of falling victim to the "see how much more we can
> stuff in without breaking backwards compatibility" syndrome.  Just because
a
> proposed change doesn't break existing code doesn't necessarily make it a
> good thing. Subtle changes to the semantics of the language invalidate
> teaching materials, for example.
>
> The situation seemd to me to be better and more rational when we expected
> Python 3K to be a major break from current traditions. At least this gave
> the twiddlers and the tweakers some hope, without putting the existing
> language at risk. There are undoubtledy arguments for many of the 2.1
> changes, but I findsome of them unconvincing. Why not keep Python as it is
> now, and prodce 3K about 998 years early?
>
> There's a lot to be said for stability.

DITTO.  This is my thought exactly.

I don't forsee upgrading to 2.1.  I still haven't gotten my main Linux
workstation up to 2.0
and frankly haven't missed it much, but I am using 2.0 on my Windows 98
system.

The "nested scopes" change (which as I understand it, creates namespaces for
each function
as it is called) is the only one I have heard about that I like, but frankly
it isn't important enough
to be worth putting up with all the other changes.

If it ain't broken, DON'T FIX IT!





From nas at arctrix.com  Sat Feb  3 18:47:07 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Sat, 3 Feb 2001 09:47:07 -0800
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
In-Reply-To: <006701c08dfa$08ed8c60$a100000a@local>; from chris.gonnerman@usa.net on Sat, Feb 03, 2001 at 09:57:32AM -0600
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com> <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com> <006701c08dfa$08ed8c60$a100000a@local>
Message-ID: <20010203094706.B30977@glacier.fnational.com>

On Sat, Feb 03, 2001 at 09:57:32AM -0600, Chris Gonnerman wrote:
> The "nested scopes" change (which as I understand it, creates
> namespaces for each function as it is called) is the only one I
> have heard about that I like, but frankly it isn't important
> enough to be worth putting up with all the other changes.

Can you give some examples of changes you have to "put up with"?
The only controversial changes in 2.1a2 that I know of are nested
scopes, function attributes and weak references.  The nested
scope changes are the only ones that affect existing code.  You
have to forgive me if I think your spreading FUD.

  Neil 



From chris.gonnerman at usa.net  Sun Feb  4 02:11:30 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 3 Feb 2001 19:11:30 -0600
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com> <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com> <006701c08dfa$08ed8c60$a100000a@local> <20010203094706.B30977@glacier.fnational.com>
Message-ID: <007701c08e47$696012a0$a100000a@local>

----- Original Message -----
From: "Neil Schemenauer" <nas at arctrix.com>
Subject: Re: Gratuitous Change (Was: Re: "in" operator for strings)


> Can you give some examples of changes you have to "put up with"?
> The only controversial changes in 2.1a2 that I know of are nested
> scopes, function attributes and weak references.  The nested
> scope changes are the only ones that affect existing code.  You
> have to forgive me if I think your spreading FUD.

Well, that hurts, but I guess I have it coming.  FUD is my enemy, let me
spread no FUD.

My primary complaint is against changes to the language semantics.  I am
against change in a language in general when the language is already IMHO
the finest I've ever worked with.

You are probably right, in that these changes in 2.1 are not so bad.  On the
other hand, I probably won't use them.  Heck, I still import string.  I'm
not
against the changes in 2.0, although I don't use most of them; I still have
servers
running 1.5.2 that would be a pain to upgrade.

One thing I see a lot in 2.0 code that is legal but UGLY is:

    nstr = ''.join((s,t,u,v))

(for instance) but I guess you can write ugly code in any language.

My only real complaint is the need to rebuild ALL MY EXTENSIONS
every time the version is upgraded.  The build process is painless under
Linux but a pain under Windows and I must live in both worlds.  Even
under Linux, I wind up rebuilding a LOT of software because I use a LOT
of Python.

Yeah, I'm rambling.  The point is (yes, I have one) that I think all changes
are potentially dangerous.  I do respect and trust GvR though, so perhaps
I'll try 2.1 when it's not alpha anymore.

BTW I've lost track... is 2.1 stackless or was that just a rumor?







From tim.one at home.com  Sun Feb  4 04:43:24 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 22:43:24 -0500
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
In-Reply-To: <007701c08e47$696012a0$a100000a@local>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEFMINAA.tim.one@home.com>

[Chris Gonnerman]
> ...
> BTW I've lost track... is 2.1 stackless or was that just a rumor?

Can't tell you whether it was a rumor (if it was, I had not heard it), but
can tell you 2.1 is not stackless.

2.1 is registerless instead.  We were getting too many complaints that
making Python faster was disturbing carefully tuned Python delay loops
<wink>.

working-on-making-2.2-ignore-ram-and-go-straight-to-disk-ly y'rs  - tim




From bowman at montana.com  Sat Feb  3 20:04:55 2001
From: bowman at montana.com (bowman)
Date: Sat, 3 Feb 2001 12:04:55 -0700
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com> <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com> <006701c08dfa$08ed8c60$a100000a@local> <mailman.981222485.28773.python-list@python.org>
Message-ID: <iCYe6.8606$qX5.6832@newsfeed.slurp.net>

Neil Schemenauer <nas at arctrix.com> wrote in message
>
> scope changes are the only ones that affect existing code.  You
> have to forgive me if I think your spreading FUD.

language changes that affect existing code have their own way of spreading
FUD. I'm not supporting a major software effort in Python at this time, just
some relatively simple tools for my own use. However, I've had a few
previously working tools fail or report errors since upgrading. This doesn't
do a lot to encourage me to undertake a project with a projected lifetime of
10 years or so in Python.

OTOH, the actual deliverable apps I am working on at present have already
passed the 10 year mark since their original release, have been constantly
enhanced and expanded, and ported to several platforms. They are written in
C, and not a line of code has had to be changed because of underlying
changes in the C language. I believe it is only in the last release of the
gnu toolchain that the decision was taken to drop support of the long
obsolete K&R declarations. It is this sort of stability that has made C,
despite its warts and blemishes, a widely used language.

Personally speaking, I am not a conservative, but there is something to be
said for conservative principals when dealing with a language. A parallel
would be when Java's event model changed -- altogether it was a good idea,
but forced one to make the choice of using the new model, or staying with
the old, knowing full well that many users were not going to upgrade their
jvm's until absolutely necessary.






From joconnor at cybermesa.com  Sat Feb  3 20:23:14 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 03 Feb 2001 19:23:14 +0000
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com> <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com> <006701c08dfa$08ed8c60$a100000a@local> <mailman.981222485.28773.python-list@python.org> <iCYe6.8606$qX5.6832@newsfeed.slurp.net>
Message-ID: <3A7C5AA2.A82C9B66@cybermesa.com>

bowman wrote:
> 
> Neil Schemenauer <nas at arctrix.com> wrote in message
> >
> > scope changes are the only ones that affect existing code.  You
> > have to forgive me if I think your spreading FUD.
> 
> language changes that affect existing code have their own way of spreading
> FUD. I'm not supporting a major software effort in Python at this time, just
> some relatively simple tools for my own use. However, I've had a few
> previously working tools fail or report errors since upgrading. This doesn't
> do a lot to encourage me to undertake a project with a projected lifetime of
> 10 years or so in Python.
> 
> OTOH, the actual deliverable apps I am working on at present have already
> passed the 10 year mark since their original release, have been constantly
> enhanced and expanded, and ported to several platforms. They are written in
> C, and not a line of code has had to be changed because of underlying
> changes in the C language. I believe it is only in the last release of the
> gnu toolchain that the decision was taken to drop support of the long
> obsolete K&R declarations. It is this sort of stability that has made C,
> despite its warts and blemishes, a widely used language.
> 
> Personally speaking, I am not a conservative, but there is something to be
> said for conservative principals when dealing with a language. A parallel
> would be when Java's event model changed -- altogether it was a good idea,
> but forced one to make the choice of using the new model, or staying with
> the old, knowing full well that many users were not going to upgrade their
> jvm's until absolutely necessary.

This is because C is a very old language and it's had time to mature and
figure out what works and what doesn't.  Python (and Java) is still
young and maturing and people are still playing with ideas to make it
work best.  And that can be painful at times.

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From bowman at montana.com  Sun Feb  4 06:40:03 2001
From: bowman at montana.com (bowman)
Date: Sat, 3 Feb 2001 22:40:03 -0700
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com> <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com> <006701c08dfa$08ed8c60$a100000a@local> <mailman.981222485.28773.python-list@python.org> <iCYe6.8606$qX5.6832@newsfeed.slurp.net> <3A7C5AA2.A82C9B66@cybermesa.com>
Message-ID: <QU5f6.9362$qX5.9335@newsfeed.slurp.net>

Jay O'Connor <joconnor at cybermesa.com> wrote in message
>
> This is because C is a very old language and it's had time to mature and
> figure out what works and what doesn't.  Python (and Java) is still
> young and maturing and people are still playing with ideas to make it
> work best.  And that can be painful at times.


agreed. Plus, C is a very simple language (not implying it is simple to
use). Python, out of the box, offers a higher level of abstraction and more
functionality --- and more for people to disagree on. Don't get me wrong, I
use Python just for that reason. The thrill of writing yet another linked
list wore off about 20 years ago. I guess I just get a little sad watching
nice little languages grow up :)





From mlh at idi.ntnu.no  Fri Feb  2 14:03:41 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 2 Feb 2001 14:03:41 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com>
Message-ID: <95ebcv$bn$1@tyfon.itea.ntnu.no>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:95birk0r0d at news1.newsguy.com...
> "Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
> news:95beco$e0u$1 at tyfon.itea.ntnu.no...
[...]
> Right, and an extension of this is basically what's being
> asked for (though the original poster may not have thought
> of this 'obvious' generalization, specialcasing string would
> surely not be warranted).  Unfortunately, for general cases
> it doesn't scale well -- i.e., now:

Well... Then "in" would test for both membership and
subsequenceness... Quite a strong case of ambiguity if you
ask me. In my opinion a really *bad* idea. (And would it
only allow contiguous subsequences, or any subsequence?
If we ever get built-in sets contiguity would be meningless,
as would it be if we get "in" tests for dictionaries.)

[...]
> and having it return 1 in the second case too would be making
> this 'in' very ambiguous and confusing, alas.

Right.

> Also, of course, this would throw any parallel between
> "x in y" and "for x in y" out of the windows unless the
> latter starts looping on all *subsequences* -- eeep!-)

Actually, that sounds very interesting ;-)

How about having a power-sequence-function (like the standard
power-set function in mathematics)? Then pow(seq) or seq.pow()
would return a (perhaps lazy) sequence containing all
subsequences... And one could have seq.pow(contiguous=1) if
one only wanted contiguous subsequences... Then one would
have:

  >>> "Waldo" in "Ralph Waldo Emerson".pow(contiguous=1)
  1

An ingenious implementation of the laziness would be needed
I guess... Or maybe not. A simple string-matching algorithm
would be needed for the contiguous case, and a O(n*m)
exhaustive search for the non-contiguous case. (Or perhaps
something better would be possible by putting the elements
in a hash table... O(n+m)? Might have quite some overhead,
perhaps... Oh, well)

>
> > But you can't do what you ask for, just like you can't write
> >
> >   [1, 2] in [1, 2, 3, 4]
>
> Sure you can, it's a well-formed test and returns 0 since
> [1,2] is not an item in the right-hand operand sequence.

Right. That's essentially what I meant. You can't write the
above to find out whether [1, 2] is a subsequence of
[1, 2, 3, 4].

> > Probably a better idea.
>
> Only if you're looking for words, not for any substring,
> which is at least as frequent.

You are right. But it's a better idea than doing something
that doesn't work, at least <0.8 wink>

> >    "Waldo" in split("Ralph Waldo Emerson")
> >
> > It might be old-fashioned, but... So what :-)
>
> So it doesn't work unless you "from string import *" (horrid
> idea), "from string import split" (doubtful), or rewrite it
> using an explicit string.split (probably best,

Why? This would in my opinion clearly depend on the size of
the script... If you don't expect another split function to
appear, I think it's quite OK to use

  from string import split

I mean -- I'm not against string methods. They're nice. I
just get a bit woozy seeing them called on literals. It
just gives me flashes of weird stuff like

  1.plus(2)

And... In my mind methods seem to be for modifying an
objects state first of all. Oh, well. Here we go -- another
pointless discussion. Sorry :)

> For general substring-matching, a class wrapper is not
> too bad:

[...]
> this only works for strings, as written, AND only to
> enable such idioms as
>
>     if 'ald' in subsOf("Waldo"):
>         print 'yep!'

I should have read the entire posting before replying, I
see. <wink>

This is the same as my power-sequence idea I guess.

> Alex

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From aleaxit at yahoo.com  Fri Feb  2 18:01:54 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 2 Feb 2001 18:01:54 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <95ebcv$bn$1@tyfon.itea.ntnu.no>
Message-ID: <95ep620rbt@news1.newsguy.com>

"Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
news:95ebcv$bn$1 at tyfon.itea.ntnu.no...
    [snip]
> > asked for (though the original poster may not have thought
> > of this 'obvious' generalization, specialcasing string would
> > surely not be warranted).  Unfortunately, for general cases
> > it doesn't scale well -- i.e., now:
>
> Well... Then "in" would test for both membership and
> subsequenceness... Quite a strong case of ambiguity if you
> ask me. In my opinion a really *bad* idea. (And would it

Right.

> only allow contiguous subsequences, or any subsequence?
> If we ever get built-in sets contiguity would be meningless,
> as would it be if we get "in" tests for dictionaries.)

Dictionaries aren't sequences, nor will sets be, so the
issue doesn't apply to either.  Still, you're right that,
*for sequences*, there are TWO definitions of subsequence
that are meaningful in different context -- contiguous
(aka compact) and generalized.

>>> print [1,2,4] in subseqOf(range(7), contiguous=1)
0
>>> print [1,2,4] in subseqOf(range(7), contiguous=0)
1

Implementation of class subseqOf is left as an
exercise to the reader:-).  Actually, a very simple
minded one IS simple to implement -- it's actually
simpler having subseqOf be a factory function, and:

def subseqOf(seq, contiguous=None):
    if contiguous: return contSubs(seq)
    else: return genrSubs(seq)

class genrSubs:
    def __init__(self,seq):
        self.seq=seq
    def __contains__(self,subs):
        start = 0
        for x in subs:
            try: where=self.seq[start:].index(x)
            except ValueError: return 0
            start = where+1
        return 1

class contSubs:
    def __init__(self,seq):
        self.seq=seq
        self.len=len(seq)
    def __contains__(self,subs):
        slen = len(subs)
        for i in range(self.len-slen):
            if self.seq[i:i+slen]==subs:
                return 1
        return 0

or thereabouts (warning, untested code!).
(One CAN, of course, do better than this!-).

Implementing __getitem__, to enumerate all
subsequences (contiguous and non) is also
of some interest -- a typical combinatorics
problem.  Generalized subsequences are a
picnic -- thanks to the natural isomorphism
with binary numbers in base self.len (I've
posted snippets using that to get "all
combinations" pretty recently).  Contiguous
subsequences are more interesting -- we
presumably still want to see the empty one
just once, so we'll single that out as
subsequence number 0 of any sequence, then
get to the real business of NON-empty subs.

Here, we want of course to count in a
_triangular_ way -- for a sequence of len
N, we'll have N subsequences that start
at 0, N-1 that start at 1, ..., N-x that
start at x, etc.  So, given a subsequence
index X, iterative computation of subseq
start and length is easy (again, untested
code...):
    if X==0: #special-case it
        return self.seq[0:0]
    X -= 1
    start = 0
    maxlen = N
    while start < N:
        if X < maxlen:
            return self.seq[start:start+X]
        X -= maxlen
        start += 1
        maxlen -= 1
    raise IndexError
but, that's lazy -- surely some closed form
exists (but there is no space in the margins
of this post for me to write it down...).

> > Also, of course, this would throw any parallel between
> > "x in y" and "for x in y" out of the windows unless the
> > latter starts looping on all *subsequences* -- eeep!-)
>
> Actually, that sounds very interesting ;-)
>
> How about having a power-sequence-function (like the standard
> power-set function in mathematics)? Then pow(seq) or seq.pow()
> would return a (perhaps lazy) sequence containing all
> subsequences... And one could have seq.pow(contiguous=1) if
> one only wanted contiguous subsequences... Then one would
> have:
>
>   >>> "Waldo" in "Ralph Waldo Emerson".pow(contiguous=1)
>   1
>
> An ingenious implementation of the laziness would be needed
> I guess... Or maybe not.

Not for the generalized case, unless you count the fact
of knowing that subsets of an N-element set, and binary
numbers of N bits, have a natural isomorphism, as in
some way "ingenious":-).  Of course one could cheat in
the enumeration for the more interesting contiguous case,
since the index being asked for only goes up by 1 at a
time in a for loop; but finding a closed-form expression
of "N-th contiguous subsequence" sounds much more fun,
and it might in fact have some use (e.g., "get a random
contiguous subsequence with all contiguous subsequences
having equal probability" -- if you can enumerate such
subsequences in closed form then that's a solved problem
too).

> A simple string-matching algorithm
> would be needed for the contiguous case, and a O(n*m)
> exhaustive search for the non-contiguous case. (Or perhaps
> something better would be possible by putting the elements
> in a hash table... O(n+m)? Might have quite some overhead,
> perhaps... Oh, well)

I suspect a modified Bayer-Moore could be very fast and
effective for contiguous-subsequence checks -- sublinear
just like "real" Bayer-Moore!-)

For the non-contiguous case, you could have a map
    'alphabet' element -> places where that element is
which is just O(N).  Now, you walk down the subsequence
you're checking, keeping only a counter of "farthest
left we could possibly be in the big sequence so far",
which you successively increase to the smallest item
in "places" that is larger than the current counter --
fail if you ever find an item in the subsequence that
is not in the alphabet, or if the restricted 'places'
is ever empty -- succeed if you get to the end.  This
looks a bit worse than O(M) in time, but not by much
unless there is a LOT of repetitiousness in the big
sequence; with bisection (binary search) over the
'places' sequences, O(M log N) appears like an easy
upperbound, and maybe it can be cut down further.

Of course, this only makes any sense for REALLY BIG
sequences.  The factory function could check for that,
returning a simpleminded wrapper class for sequences
up to some threshold length, and a sophisticated one
for longer sequences...!-)


> > >    "Waldo" in split("Ralph Waldo Emerson")
> > >
> > > It might be old-fashioned, but... So what :-)
> >
> > So it doesn't work unless you "from string import *" (horrid
> > idea), "from string import split" (doubtful), or rewrite it
> > using an explicit string.split (probably best,
>
> Why? This would in my opinion clearly depend on the size of
> the script... If you don't expect another split function to
> appear, I think it's quite OK to use
>
>   from string import split

In a small module, maybe.  Remember the status of local
import statements is in doubt... they may be disallowed
in Python 2.1 (hopefully they won't be, but, who knows),
so it's not clear that you could do this inside a function.
"doubtful" seems a fair assessment to me -- there ARE
doubts, depending on context; it's not necessarily "quite
OK" -- it may or may not be.

Why not just use string.split and play it safe...?\

Or, just as safe, the string-method...?

> I mean -- I'm not against string methods. They're nice. I
> just get a bit woozy seeing them called on literals. It

I think we'll all get used to that in time.

> just gives me flashes of weird stuff like
>
>   1.plus(2)
>
> And... In my mind methods seem to be for modifying an
> objects state first of all. Oh, well. Here we go -- another
> pointless discussion. Sorry :)

Why would mutators be any more worthy of syntax-sugar
support than accessors?  You've lost me here...


Alex





From mlh at idi.ntnu.no  Sat Feb  3 04:21:20 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 3 Feb 2001 04:21:20 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <95ebcv$bn$1@tyfon.itea.ntnu.no> <95ep620rbt@news1.newsguy.com>
Message-ID: <95ftl4$o6b$1@tyfon.itea.ntnu.no>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:95ep620rbt at news1.newsguy.com...
> "Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
> news:95ebcv$bn$1 at tyfon.itea.ntnu.no...

[...]

> > only allow contiguous subsequences, or any subsequence?
> > If we ever get built-in sets contiguity would be meningless,
> > as would it be if we get "in" tests for dictionaries.)
>
> Dictionaries aren't sequences, nor will sets be, so the
> issue doesn't apply to either.

Sure it does. It will then become the question of whether to
allow the same operator do check membership and subset-ness.

> Implementing __getitem__, to enumerate all
> subsequences (contiguous and non) is also
> of some interest -- a typical combinatorics
> problem.

[...]

> but, that's lazy -- surely some closed form
> exists (but there is no space in the margins
> of this post for me to write it down...).

[...]

> finding a closed-form expression
> of "N-th contiguous subsequence" sounds much more fun,
> and it might in fact have some use (e.g., "get a random
> contiguous subsequence with all contiguous subsequences
> having equal probability" -- if you can enumerate such
> subsequences in closed form then that's a solved problem
> too).

Interesting...

[...]

> I suspect a modified Bayer-Moore

(Boyer-Moore)

> could be very fast and
> effective for contiguous-subsequence checks -- sublinear
> just like "real" Bayer-Moore!-)

I'm sure :-)

What's the built-in string-matching algorithm in Python?
(Not very important to me, but...)

[...]

> with bisection (binary search) over the
> 'places' sequences, O(M log N) appears like an easy
> upperbound, and maybe it can be cut down further.

Well... If you could store an array of size N for each
letter you could easily do it in O(1) time... But then
you'd need O(M*N) time to initialize the stuff :-)

If you add a requirement that the matched letters must
be relatively close to each other (within a constant
distance limit) you can get a speedup. (A reasonable
requirement in some cases.) Actually, I know of some
harwdare that does this very quickly.

[...]

> > I think it's quite OK to use
> >
> >   from string import split
>
> In a small module, maybe.  Remember the status of local
> import statements is in doubt... they may be disallowed
> in Python 2.1 (hopefully they won't be, but, who knows),
> so it's not clear that you could do this inside a function.
> "doubtful" seems a fair assessment to me -- there ARE
> doubts, depending on context; it's not necessarily "quite
> OK" -- it may or may not be.

OK... I guess that's one way of interpreting the statement...
A rather mild one, IMO :-)

To me the module doesn't even have to be small. Just think
about all the other function and class names you define in
a module... Why should "split" specifically have to have
a module prefix? If I choose to import it that may be as
thought through as defining a function myself. Only if I
have another function called "split" would I feel the
need to add its module name. I do the same thing when
importing classes in Java, for instance. When I instantiate
a JFrame I find it more convenient to write

   mainFrame = new JFrame();

than

   mainFrame = new javax.swing.JFrame();

(OK... So I know you dislike the word "convenient". Well...
I find it prettier too ;-)


> Why not just use string.split and play it safe...?\
>
> Or, just as safe, the string-method...?

If in some project the other method seems unsafe, I certainly
will.

> > I mean -- I'm not against string methods. They're nice. I
> > just get a bit woozy seeing them called on literals. It
>
> I think we'll all get used to that in time.

I guess I'll have to. But I don't think they are all
logical...

For instance:

  " ".join(somesequence)

That means 'join the elements of somesequence by " "'

On the other hand,

  [1].extend(somesequence)

means 'extend [1] by somesequence'

Why not write somesequence.extend([1]) or
somesequence.join(" ") (not both, of course) so that they would
be consistent with each other? Oh, well. I'm sure it's just
that I don't see the Greatness of it all.

> Why would mutators be any more worthy of syntax-sugar
> support than accessors?  You've lost me here...

OK. Accessors are fine, but we hardly need those in Python
<wink>

The " ".join([1,2,3]) doesn't seem like an accessor to me.
It seems like a plain function with a weird syntax.

But, hey, since this is the way it *is*, I'm the one with
a problem here :-)

>
> Alex
>


--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick







From tim.one at home.com  Sat Feb  3 05:28:34 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 2 Feb 2001 23:28:34 -0500
Subject: "in" operator for strings
In-Reply-To: <95ftl4$o6b$1@tyfon.itea.ntnu.no>
Message-ID: <LNBBLJKPBEHFEDALKOLCGECDINAA.tim.one@home.com>

[Magnus Lie Hetland]
> ...
> What's the built-in string-matching algorithm in Python?
> (Not very important to me, but...)

Brute force, using memcmp.  Note that fancier algorithms that require
precomputing tables of size proportional to the alphabet cardinality aren't
very attractive in a Unicode world.

See mxTextTools for a Boyer-Moore 8-bit string searcher.




From aleaxit at yahoo.com  Sat Feb  3 10:12:20 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 3 Feb 2001 10:12:20 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <95ebcv$bn$1@tyfon.itea.ntnu.no> <95ep620rbt@news1.newsguy.com> <95ftl4$o6b$1@tyfon.itea.ntnu.no>
Message-ID: <95gkgg0fo7@news1.newsguy.com>

"Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
news:95ftl4$o6b$1 at tyfon.itea.ntnu.no...
    [snip]
> > > only allow contiguous subsequences, or any subsequence?
> > > If we ever get built-in sets contiguity would be meningless,
> > > as would it be if we get "in" tests for dictionaries.)
> >
> > Dictionaries aren't sequences, nor will sets be, so the
> > issue doesn't apply to either.
>
> Sure it does. It will then become the question of whether to
> allow the same operator do check membership and subset-ness.

Subsetness is one thing, subsequencehood another (actually
two others, given contiguity may be required or not); the
difference is ordering -- 'ab' is a subset but not a subsequence
of 'obak'.  The whole issue of subsequence is meaningless
where no sequence (ordering) exists.

> > with bisection (binary search) over the
> > 'places' sequences, O(M log N) appears like an easy
> > upperbound, and maybe it can be cut down further.
>
> Well... If you could store an array of size N for each
> letter you could easily do it in O(1) time... But then
> you'd need O(M*N) time to initialize the stuff :-)

You don't know M at initialization time -- it's the size
of the subsequence you're going to match (in the
normal notation I've seen used), while N is the size
of the big sequence in which you're looking for matches.

Alphabet size is generally thought of as fixed in most
published literature, though I suspect that O(N) is
more realistic for generic sequences.

You still can't do a match in O(1) time anyway, even
with the auxiliary structure you suggest; O(M) seems
the lower bound, except by precomputing ALL the
possible subsequences and indexing them into some
totally hypothetical O(1)-lookup-time structure (but
in fact computing the hash IS still O(M), so there is
no big-O advantage here).


> If you add a requirement that the matched letters must
> be relatively close to each other (within a constant
> distance limit) you can get a speedup. (A reasonable
> requirement in some cases.) Actually, I know of some
> harwdare that does this very quickly.

I'm curious about the algorithm and its auxiliary data
structures here.  The O(M) match still seems pretty
elusive to me.

> To me the module doesn't even have to be small. Just think
> about all the other function and class names you define in
> a module... Why should "split" specifically have to have
> a module prefix? If I choose to import it that may be as
> thought through as defining a function myself. Only if I

I like the clarity that comes from use of explicit prefixes --
no wondering 'WHAT split function is this one' any more
(string or re or ...?).  Calling a method of an object, as in
'foo bar'.split(), is clearest -- totally local context, zero
wondering-factor.  Calling a prefixed function is second
best.  Calling a non-prefixed function sends me looking for
it throughout the non-necessarily-small module -- and
I won't find the '^def split' either (so I have to look at the
from statements and the assignments -- sigh).

> (OK... So I know you dislike the word "convenient". Well...
> I find it prettier too ;-)

The (modest) convenience of the code author becomes
the (rather larger) IN-convenience of the maintainer.

When one has spent a few years doing maintenance on
large modules written by others, one's priorities tend to
change -- as does one's outlook on "prettiness" when it
interferes with clarity (the typical C++ scope obscurity
due to implicit-this, vs. Python's clarity with explicit
self, being a good example).

> Why not write somesequence.extend([1]) or
> somesequence.join(" ") (not both, of course) so that they would
> be consistent with each other? Oh, well. I'm sure it's just
> that I don't see the Greatness of it all.

I've discussed the "what object should join be a method
of" in a couple of lengthy postings 2 or 3 months ago --
it boils down to "where do I get the most leverage for
polymorphism", and the current architecture wins hands
down.  extend is similar -- which object needs the
polymorphism most, the one being modified or the
sequence just being 'read'?  And of course the answer
is the same.

So, the architecture DOES turn out to be fully consistent
here, because exactly the same forces are being
resolved!  In each case, the subsequence that is
just being read (for the purpose of building a string
out of it, or for that of extending another object) is
the argument -- all the polymorphism we need on it
is that it obeys the sequence-protocol, after all.  The
'joiner' (aka separator) object, and the object being
modified (extended), are in each case the target of
the method and get full benefit of polymorphism.


> The " ".join([1,2,3]) doesn't seem like an accessor to me.
> It seems like a plain function with a weird syntax.
>
> But, hey, since this is the way it *is*, I'm the one with
> a problem here :-)

And we're here to help you come to terms with it!-)

What I'm doing is asking a 'joiner object' "please read
this sequence of strings and join it up into one big
string, will you".  Function syntax does not give you
polymorphism directly in Python - it can at best be
syntax sugar for a method call that does do the poly
thing:

def myJoin(joiner, sequence):
    return joiner.join(sequence)

which is basically what string.join now does.  We do
not have multimethods in Python (one generic function
dispatched directly to the appropriate method), so
the main Python approach to polymorphism is calling
a method on an object (sometimes with weird syntax
sugar, such as a+b for a.__add__(b) or maybe for
b.__radd__(a), but that's another issue:-).  It only
remains to determine which object gets real benefit
from a given case of polymorphism.


Alex





From mlh at idi.ntnu.no  Sun Feb  4 15:55:04 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 4 Feb 2001 15:55:04 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <95ebcv$bn$1@tyfon.itea.ntnu.no> <95ep620rbt@news1.newsguy.com> <95ftl4$o6b$1@tyfon.itea.ntnu.no> <95gkgg0fo7@news1.newsguy.com>
Message-ID: <95jqlu$lcu$1@tyfon.itea.ntnu.no>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:95gkgg0fo7 at news1.newsguy.com...
> "Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
> news:95ftl4$o6b$1 at tyfon.itea.ntnu.no...
>     [snip]
> > > > only allow contiguous subsequences, or any subsequence?
> > > > If we ever get built-in sets contiguity would be meningless,
> > > > as would it be if we get "in" tests for dictionaries.)
> > >
> > > Dictionaries aren't sequences, nor will sets be, so the
> > > issue doesn't apply to either.
> >
> > Sure it does. It will then become the question of whether to
> > allow the same operator do check membership and subset-ness.
>
> Subsetness is one thing, subsequencehood another (actually
> two others, given contiguity may be required or not); the
> difference is ordering -- 'ab' is a subset but not a subsequence
> of 'obak'.  The whole issue of subsequence is meaningless
> where no sequence (ordering) exists.

Yes. I am well aware of that, and it seems that I mangled up
my original sentence a bit. My *real* point (as illustrated
by the examples and exmplanation of characters *in* strings
etc.) was the difference between elements and subsets
(or substrings, or subsequences). Order is irrelevant here,
and the ambiguity is very real. The contiguity is an added
problem. As I state above, in sets and dictionaries
contiguity would be meaningless, and thus not part of
the issue. It would seem to me that you agree on that
point? (So why object? <wink>)

> I like the clarity that comes from use of explicit prefixes --
> no wondering 'WHAT split function is this one' any more
> (string or re or ...?).  Calling a method of an object, as in
> 'foo bar'.split(), is clearest -- totally local context, zero
> wondering-factor.  Calling a prefixed function is second
> best.  Calling a non-prefixed function sends me looking for
> it throughout the non-necessarily-small module -- and
> I won't find the '^def split' either (so I have to look at the
> from statements and the assignments -- sigh).

Hm. OK.

> > The " ".join([1,2,3]) doesn't seem like an accessor to me.
> > It seems like a plain function with a weird syntax.
> >
> > But, hey, since this is the way it *is*, I'm the one with
> > a problem here :-)
>
> And we're here to help you come to terms with it!-)

And I thank you for it :)

> Alex

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From aleaxit at yahoo.com  Mon Feb  5 14:28:56 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 5 Feb 2001 14:28:56 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <95ebcv$bn$1@tyfon.itea.ntnu.no> <95ep620rbt@news1.newsguy.com> <95ftl4$o6b$1@tyfon.itea.ntnu.no> <95gkgg0fo7@news1.newsguy.com> <95jqlu$lcu$1@tyfon.itea.ntnu.no>
Message-ID: <95m9qo01emb@news1.newsguy.com>

"Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
news:95jqlu$lcu$1 at tyfon.itea.ntnu.no...
    [snip]
> the issue. It would seem to me that you agree on that
> point? (So why object? <wink>)

Because we ARE doing *object*-oriented design & coding
here...?

When you're dealing with subsets, there are no big
problems, because the fact that what you're modeling
is order-insensitive means it's OK to inject whatever
order you like behind the cover.  Call it a canonical
order, if you like.  The key issue is that it gives
you the isomorphism between "subsets of an N-elements
alphabet" and "binary numbers in base N", which I
used for the non-necessarily-contiguous-subsequences
case.  Weed out duplicates (at insertion time or on
the fly when needed), sort (ditto), and enjoy the
utter simplicity we already saw for n-n-c subsequences.
("The secret of life is finding the right isomorphisms",
to paraphrase the late C. Schulz -- his original quote
was actually "is getting the right side-effects", but
they aren't that far if you squint at them just right).

Sequences WITH contiguousness constraints (and, probably
even more complicated case, with *relaxed* contiguousness
constraints, as you hinted at -- "elements following
each other at linear distance no more than four", etc)
is a far more complicated case -- as are multisets in
the non-order case.  Heuristics can be found (e.g., if
in most cases your multisets have a multiplicity no more
than 27 in each given element, you can use base-29
numbers as the primary representation of any multiset,
with 28 standing in for "28 or more [rare case]", and
get MOST solutions almost as fast as for sets represented
as binary numbers), but the general-case is another thing
(not only is it hard to look for closed-formula enumerators
of such things, but proving "no closed formula exists" is
a real bear too).


Alex





From shredwheat at mediaone.net  Sat Feb  3 07:02:25 2001
From: shredwheat at mediaone.net (Pete Shinners)
Date: Sat, 03 Feb 2001 06:02:25 GMT
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <95ebcv$bn$1@tyfon.itea.ntnu.no> <95ep620rbt@news1.newsguy.com>
Message-ID: <R9Ne6.152539$y9.28122940@typhoon.we.rr.com>

> Well... Then "in" would test for both membership and
> subsequenceness... Quite a strong case of ambiguity if you
> ask me. In my opinion a really *bad* idea. (And would it

hmm, well when posting the idea this was really the only
drawback against such a feature. in my mind there is no problem
separating strings and lists. the two feature many similar
characteristics, but at no point have i ever had the two item
types confused.

i was pretty sure the idea would get shutdown, but i still
feel the functionality would be quite natural and useful.
anyways, there's obviously much more than one or two camps
of people using python, and python with or without an "in string"
operator is still going to be one of my preferred languages :]






From luthi at vaw.baug.ethz.ch  Thu Feb  1 13:21:48 2001
From: luthi at vaw.baug.ethz.ch (luthi at vaw.baug.ethz.ch)
Date: 01 Feb 2001 13:21:48 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no>
Message-ID: <yfmwvbad25f.fsf@collet.i-did-not-set--mail-host-address--so-shoot-me>

"Magnus Lie Hetland" <mlh at idi.ntnu.no> writes:

> But you can't do what you ask for, just like you can't write
> 
>   [1, 2] in [1, 2, 3, 4]
> 

This is exactly one of my current puzzles: How to can I compare that a set is
a subset of another, larger, set efficiently? Say I have to perform such a
comparison some 100000 times, it would come in very handy to write

(assume all_sets is a list of 100000 sets (i.e. lists))

for set in all_sets:
	if [1,2,50] in set:
		do_something

any good suggestion?

Thanx!

-- 
============================================================
Dr. Martin L?thi               Tel. +41 1 632 41 40
Glaciology Section             Fax. +41 1 632 11 92
VAW ETH Zuerich	
CH-8092 Zuerich                mail luthi at vaw.baug.ethz.ch
Switzerland
============================================================


From soerense at verwaltung.uni-mainz.de  Thu Feb  1 16:01:47 2001
From: soerense at verwaltung.uni-mainz.de (Ruediger Soerensen)
Date: 01 Feb 2001 16:01:47 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <yfmwvbad25f.fsf@collet.i-did-not-set--mail-host-address--so-shoot-me>
Message-ID: <ytwu26ela5g.fsf@verwaltung.uni-mainz.de>

>This is exactly one of my current puzzles: How to can I compare that a set is
>a subset of another, larger, set efficiently? Say I have to perform such a
>comparison some 100000 times, it would come in very handy to write
>
>(assume all_sets is a list of 100000 sets (i.e. lists))
>
>for set in all_sets:
>	if [1,2,50] in set:
>		do_something
>
>any good suggestion?


What about this:

def inset(subset,set):
    for u in subset:
        if not u in set:
            return 0
    return 1


print inset([1,3],[1,2,3,4])
print inset([1,8],[1,2,3,4])



-- 
R?diger "Black Hole" S?rensen || soerense at verwaltung.uni-mainz.de


From aleaxit at yahoo.com  Thu Feb  1 16:26:43 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 1 Feb 2001 16:26:43 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <yfmwvbad25f.fsf@collet.i-did-not-set--mail-host-address--so-shoot-me>
Message-ID: <95bvqt0141@news1.newsguy.com>

<luthi at vaw.baug.ethz.ch> wrote in message
news:yfmwvbad25f.fsf at collet.i-did-not-set--mail-host-address--so-shoot-me...
> "Magnus Lie Hetland" <mlh at idi.ntnu.no> writes:
>
> > But you can't do what you ask for, just like you can't write
> >
> >   [1, 2] in [1, 2, 3, 4]
>
> This is exactly one of my current puzzles: How to can I compare that a set
is
> a subset of another, larger, set efficiently? Say I have to perform such a

Are you actually dealing with sets, and subsets, or lists, lists of lists,
or maybe lists of sets?

> comparison some 100000 times, it would come in very handy to write
>
> (assume all_sets is a list of 100000 sets (i.e. lists))
>
> for set in all_sets:
> if [1,2,50] in set:
> do_something
>
> any good suggestion?

If you need map your sets as lists, then you aren't going to
get substantially better behavior than by:

def is_subset(small_set, big_set):
    for item in small_set:
        if item not in big_set: return 0
    return 1

That's gonna be SLOW -- but that depends on the representation
you've chosen for sets.

At the other extreme, if your sets were represented as
bit-vectors (feasible iff their items are or can be efficiently
mapped to small integers), then you might get VERY good
performance.  I've had occasion to use the GMP bit-oriented
functions (which gmpy exposes) for that -- sets of integers
in a reasonably compact range such as [0,1024), with each
set being a '1024-bits vector' (just 128 bytes); what gmpy
exposes could quite easily be wrapped up for such purposes
(right now it's somewhat un-handy because, sigh, "numbers
are immutable" has BIG performance implications for numbers
used as bit-vectors...!).

If it's unfeasible to map your sets' items to a reasonably
compact range of integers, and so a bit-vector approach is
out, then dictionaries are probably your best bet (although,
depending on exact usage patterns, the Standard C++ Library
approach of sets-as-red-black-trees will have occasional wins
in the performance arena, too).  Wrapping your dictionary
into a class, so you can have the latter implement
__contains__ etc and get nice syntax sugar, is not too hard
(but nice syntax sugar is rarely the key issue when one is
dealing with hundreds of thousands of sets -- survivable
performance tends to be higher-priority in such cases:-).

The subset-test approach doesn't change much with a dict
implementation, though -- it's just that each membership
test it uses is a bit faster.


Alex





From fredrik at effbot.org  Sat Feb  3 10:06:50 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 09:06:50 GMT
Subject: "in" operator for strings
References: <mailman.981176585.13606.python-list@python.org>
Message-ID: <KSPe6.11220$Qb7.1646316@newsb.telia.net>

Tim Peters wrote:
> See mxTextTools for a Boyer-Moore 8-bit string searcher.

or use re.compile/pattern.search for a Knuth-Morris-Pratt
style 8-bit/16-bit string searcher...

Cheers /F




From m.faassen at vet.uu.nl  Wed Feb  7 15:40:35 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 7 Feb 2001 14:40:35 GMT
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com>
Message-ID: <95rmp3$k1s$2@newshost.accu.uu.nl>

Pete Shinners <shredwheat at mediaone.net> wrote:
> with the back and forth going on about the "in" operator and
> dictionaries, i was thinking it would be nice for the "in"
> operator to also check for the presence of substrings in a
> string. for example...


>>>> "Waldo" in "Ralph Emerson"
> 0
>>>> "Waldo" in "Ralph Waldo Emerson"
> 1

'in' already has a meaning for strings, which are sequences of characters:

>>> 'f' in 'foo'
1
>>> 'b' in 'foo'
0

> i find this much more intuitive and graceful than the current
> syntax i am repeatedly using...

>>>> "Ralph Emerson".find("Waldo") != -1
> 0

Why I agree that this looks a bit odd, it isn't that terrible. I'll just dump 
my stance on 'in' here:

Let's not overload 'in' more than it is now. 'in' is already overloaded;
there's the for .. in.. meaning and the x in sequence meaning. 

Now at least these both have to do with sequences. They're not about dicts
and they don't have anything to do with string matching. I think it's not
a good idea to overload this meaning any further. It's okay if a custom
class has an explicit override __in__, though I'd expect this class then
to implement some form of the sequence interface overall.

I don't think all operators have to do something sensible for everything;
in the case of 'in' we obviously cannot even agree on what 'sensible' is.

'in' currently does 'something mumble mumble' with items in sequences. 
Not a very clear definition perhaps, but let's not muddy the waters
further either.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From gbell at uclink.berkeley.edu  Thu Feb  1 09:07:50 2001
From: gbell at uclink.berkeley.edu (gbell at uclink.berkeley.edu)
Date: Thu, 01 Feb 2001 08:07:50 GMT
Subject: IDLE & Gadfly frustration
Message-ID: <95b5gk$2gi$1@nnrp1.deja.com>

Two newbie questions:

1) I'm trying to write a script which creates, then makes use of, a
Gadfly database to be stored in my Python20 directory.  When I run the
script from the (MS-DOS) command line, all is well. When I run the
script using IDLE, the Gadfly data files are created in my TOOLS/IDLE
directory.  IDLE seems to have its own ideas about my preferred default
directory. Is there any way to change this?

2) A related frustration:  could someone tell me the proper syntax for
specifying directory paths in Python (or perhaps I'm asking for the
proper Gadfly syntax)?  The Gadfly documentation says:
  connection.startup("mydatabase", "mydirectory")
So far, I've only had luck setting the second variable to "" (perhaps
because of the problem mentioned in #1).  Should "mydirectory" begin
with c:?  Use double backward slashes?  forward slashes?

Many, many thanks.

Gregory Bell
gbell at uclink.berkeley.edu


Sent via Deja.com
http://www.deja.com/


From gregj at pobox.com  Thu Feb  1 10:07:17 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 01 Feb 2001 09:07:17 GMT
Subject: IDLE & Gadfly frustration
References: <95b5gk$2gi$1@nnrp1.deja.com>
Message-ID: <95b902$613$1@nnrp1.deja.com>

The directory you give to gadfly must exist. You can create a new
directory like this:

  import os
  os.mkdir(r"c:\mydirectory")

The r"..." string literal is a raw string; this prevents Python from
interpreting the backslashes as escapes. You can also do this:

  os.mkdir("c:\\mydirectory")

To create a new database from gadfly, after you have created the folder:

  from gadfly import gadfly
  cx = gadfly()
  cx.startup("databasename", r"c:\mydirectory")

  etc.

Hope this helps.

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com



In article <95b5gk$2gi$1 at nnrp1.deja.com>,
  gbell at uclink.berkeley.edu wrote:
> Two newbie questions:
>
> 1) I'm trying to write a script which creates, then makes use of, a
> Gadfly database to be stored in my Python20 directory.  When I run the
> script from the (MS-DOS) command line, all is well. When I run the
> script using IDLE, the Gadfly data files are created in my TOOLS/IDLE
> directory.  IDLE seems to have its own ideas about my preferred
default
> directory. Is there any way to change this?
>
> 2) A related frustration:  could someone tell me the proper syntax for
> specifying directory paths in Python (or perhaps I'm asking for the
> proper Gadfly syntax)?  The Gadfly documentation says:
>   connection.startup("mydatabase", "mydirectory")
> So far, I've only had luck setting the second variable to "" (perhaps
> because of the problem mentioned in #1).  Should "mydirectory" begin
> with c:?  Use double backward slashes?  forward slashes?
>
> Many, many thanks.
>
> Gregory Bell
> gbell at uclink.berkeley.edu


Sent via Deja.com
http://www.deja.com/


From gbell at uclink.berkeley.edu  Thu Feb  1 16:00:38 2001
From: gbell at uclink.berkeley.edu (gbell at uclink.berkeley.edu)
Date: Thu, 01 Feb 2001 15:00:38 GMT
Subject: IDLE & Gadfly frustration
References: <95b5gk$2gi$1@nnrp1.deja.com> <95b902$613$1@nnrp1.deja.com>
Message-ID: <95btmg$mep$1@nnrp1.deja.com>

Thank you -- the raw string & os.mkdir features should help me out.

But to reiterate my first question, does anyone know why running a
script via the ctrl-f5 function in IDLE (as opposed to the command
line) should affect the default directory where gadfly places data
files?  This is true when I set the "mydirectory" value (see original
question) to an empty string.  It seems that IDLE alters the default
path.  Am I right, and can I prevent this?

In article <95b902$613$1 at nnrp1.deja.com>,
  Greg Jorgensen <gregj at pobox.com> wrote:
> The directory you give to gadfly must exist. You can create a new
> directory like this:
>
>   import os
>   os.mkdir(r"c:\mydirectory")
>
> The r"..." string literal is a raw string; this prevents Python from
> interpreting the backslashes as escapes. You can also do this:
>
>   os.mkdir("c:\\mydirectory")
>
> To create a new database from gadfly, after you have created the
folder:
>
>   from gadfly import gadfly
>   cx = gadfly()
>   cx.startup("databasename", r"c:\mydirectory")
>
>   etc.
>
> Hope this helps.
>
> --
> Greg Jorgensen
> Portland, Oregon, USA
> gregj at pobox.com
>
> In article <95b5gk$2gi$1 at nnrp1.deja.com>,
>   gbell at uclink.berkeley.edu wrote:
> > Two newbie questions:
> >
> > 1) I'm trying to write a script which creates, then makes use of, a
> > Gadfly database to be stored in my Python20 directory.  When I run
the
> > script from the (MS-DOS) command line, all is well. When I run the
> > script using IDLE, the Gadfly data files are created in my
TOOLS/IDLE
> > directory.  IDLE seems to have its own ideas about my preferred
> default
> > directory. Is there any way to change this?
> >
> > 2) A related frustration:  could someone tell me the proper syntax
for
> > specifying directory paths in Python (or perhaps I'm asking for the
> > proper Gadfly syntax)?  The Gadfly documentation says:
> >   connection.startup("mydatabase", "mydirectory")
> > So far, I've only had luck setting the second variable to ""
(perhaps
> > because of the problem mentioned in #1).  Should "mydirectory"
begin
> > with c:?  Use double backward slashes?  forward slashes?
> >
> > Many, many thanks.
> >
> > Gregory Bell
> > gbell at uclink.berkeley.edu
>
> Sent via Deja.com
> http://www.deja.com/
>


Sent via Deja.com
http://www.deja.com/


From gregj at pobox.com  Thu Feb  1 20:52:52 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 01 Feb 2001 19:52:52 GMT
Subject: IDLE & Gadfly frustration
References: <95b5gk$2gi$1@nnrp1.deja.com> <95b902$613$1@nnrp1.deja.com> <95btmg$mep$1@nnrp1.deja.com>
Message-ID: <95ceqf$7m4$1@nnrp1.deja.com>

Python uses the current working directory, as returned by os.getcwd().
The Python interpreter will uses whatever directory it was invoked from
(where your DOS session is). IDLE uses the directory IDLE is in. You
can use os.chdir(path) to change the current working directory.


In article <95btmg$mep$1 at nnrp1.deja.com>,
  gbell at uclink.berkeley.edu wrote:
> Thank you -- the raw string & os.mkdir features should help me out.
>
> But to reiterate my first question, does anyone know why running a
> script via the ctrl-f5 function in IDLE (as opposed to the command
> line) should affect the default directory where gadfly places data
> files?  This is true when I set the "mydirectory" value (see original
> question) to an empty string.  It seems that IDLE alters the default
> path.  Am I right, and can I prevent this?

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From gbell at uclink.berkeley.edu  Thu Feb  1 22:44:32 2001
From: gbell at uclink.berkeley.edu (gbell at uclink.berkeley.edu)
Date: Thu, 01 Feb 2001 21:44:32 GMT
Subject: IDLE & Gadfly frustration
References: <95b5gk$2gi$1@nnrp1.deja.com> <95b902$613$1@nnrp1.deja.com> <95btmg$mep$1@nnrp1.deja.com> <95ceqf$7m4$1@nnrp1.deja.com>
Message-ID: <95clbt$ebe$1@nnrp1.deja.com>

Many thanks: that should help.


In article <95ceqf$7m4$1 at nnrp1.deja.com>,
  Greg Jorgensen <gregj at pobox.com> wrote:
> Python uses the current working directory, as returned by
os.getcwd().
> The Python interpreter will uses whatever directory it was invoked
from
> (where your DOS session is). IDLE uses the directory IDLE is in. You
> can use os.chdir(path) to change the current working directory.
>
> In article <95btmg$mep$1 at nnrp1.deja.com>,
>   gbell at uclink.berkeley.edu wrote:
> > Thank you -- the raw string & os.mkdir features should help me out.
> >
> > But to reiterate my first question, does anyone know why running a
> > script via the ctrl-f5 function in IDLE (as opposed to the command
> > line) should affect the default directory where gadfly places data
> > files?  This is true when I set the "mydirectory" value (see
original
> > question) to an empty string.  It seems that IDLE alters the
default
> > path.  Am I right, and can I prevent this?
>
> --
> Greg Jorgensen
> Portland, Oregon, USA
> gregj at pobox.com
>
> Sent via Deja.com
> http://www.deja.com/
>


Sent via Deja.com
http://www.deja.com/


From bobhicks at adelphia.net  Thu Feb  1 15:45:33 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Thu, 01 Feb 2001 14:45:33 GMT
Subject: IDLE & Gadfly frustration
References: <95b5gk$2gi$1@nnrp1.deja.com>
Message-ID: <B69EE0A4.9ADA%bobhicks@adelphia.net>

In Python, the forward slash character *can* be used as a separator in file
paths in both *NIX and Windows.

inp = open("C:/myfiles/mydoc.txt","r")


> From: gbell at uclink.berkeley.edu
> Organization: Deja.com
> Newsgroups: comp.lang.python
> Date: Thu, 01 Feb 2001 08:07:50 GMT
> Subject: IDLE & Gadfly frustration
> 
> Two newbie questions:
> 
> 1) I'm trying to write a script which creates, then makes use of, a
> Gadfly database to be stored in my Python20 directory.  When I run the
> script from the (MS-DOS) command line, all is well. When I run the
> script using IDLE, the Gadfly data files are created in my TOOLS/IDLE
> directory.  IDLE seems to have its own ideas about my preferred default
> directory. Is there any way to change this?
> 
> 2) A related frustration:  could someone tell me the proper syntax for
> specifying directory paths in Python (or perhaps I'm asking for the
> proper Gadfly syntax)?  The Gadfly documentation says:
> connection.startup("mydatabase", "mydirectory")
> So far, I've only had luck setting the second variable to "" (perhaps
> because of the problem mentioned in #1).  Should "mydirectory" begin
> with c:?  Use double backward slashes?  forward slashes?
> 
> Many, many thanks.
> 
> Gregory Bell
> gbell at uclink.berkeley.edu
> 
> 
> Sent via Deja.com
> http://www.deja.com/



From tushar at hursley.ibm.com  Thu Feb  1 12:05:24 2001
From: tushar at hursley.ibm.com (Tushar Wagle)
Date: Thu, 01 Feb 2001 11:05:24 +0000
Subject: xml install problem - No module named Node
Message-ID: <3A7942F4.291E3D6C@hursley.ibm.com>

I'm trying to use the new XML library in Python 2.0, but cannot
get the examples to work.

The error I get when running the iterator1.py example is:

   File: "iterator1.py", line 5, in ?
      from xml.dom.Node import Node
   ImportError: No module named Node

which looks like a basic installation error. However, some of the other examples

work, and the directories look populated. I looked in
/usr/local/lib/python2.0/site-packages/_xmlplus/dom and there is
no Node.py file....

I'm running RedHat 6.2, Python 2.0 and PyXML 0.6.3 (I tried the rpm and
rebuilding from source with no luck either way).

What's wrong?

All help appreciated! :)

- Tushar.




From hgg9140 at cola.ca.boeing.com  Thu Feb  1 15:59:17 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Thu, 1 Feb 2001 14:59:17 GMT
Subject: xml install problem - No module named Node
References: <3A7942F4.291E3D6C@hursley.ibm.com>
Message-ID: <xqxlmrqv48q.fsf@cola.ca.boeing.com>

I wonder if it was a typo in the example.  Seems to me it should read:

  from xml.dom.minidom import Node

or maybe

  from xml.dom.minidom import *

Tushar Wagle <tushar at hursley.ibm.com> writes:

> I'm trying to use the new XML library in Python 2.0, but cannot
> get the examples to work.
> 
> The error I get when running the iterator1.py example is:
> 
>    File: "iterator1.py", line 5, in ?
>       from xml.dom.Node import Node
>    ImportError: No module named Node
> 
> which looks like a basic installation error. However, some of the other examples
> 
> work, and the directories look populated. I looked in
> /usr/local/lib/python2.0/site-packages/_xmlplus/dom and there is
> no Node.py file....
> 
> I'm running RedHat 6.2, Python 2.0 and PyXML 0.6.3 (I tried the rpm and
> rebuilding from source with no luck either way).
> 
> What's wrong?
> 
> All help appreciated! :)
> 
> - Tushar.
> 
> 

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From kvthan at wm.edu  Thu Feb  1 18:24:18 2001
From: kvthan at wm.edu (Kapil Thangavelu)
Date: Thu, 01 Feb 2001 17:24:18 GMT
Subject: xml install problem - No module named Node
References: <3A7942F4.291E3D6C@hursley.ibm.com>
Message-ID: <6Zge6.2424$ln1.197197@newsread2.prod.itd.earthlink.net>

Tushar Wagle wrote:

> I'm trying to use the new XML library in Python 2.0, but cannot
> get the examples to work.
> 
> The error I get when running the iterator1.py example is:
> 
>    File: "iterator1.py", line 5, in ?
>       from xml.dom.Node import Node
>    ImportError: No module named Node
> 
> which looks like a basic installation error. However, some of the other
> examples
> 

Node is now defined in __init__.py of xml.dom

try instead
from xml.dom import Node

kapil



From tushar at hursley.ibm.com  Thu Feb  1 20:04:52 2001
From: tushar at hursley.ibm.com (Tushar Wagle)
Date: Thu, 01 Feb 2001 19:04:52 +0000
Subject: xml install problem - No module named Node
References: <3A7942F4.291E3D6C@hursley.ibm.com>
Message-ID: <3A79B353.69D5E104@hursley.ibm.com>

For anyone else wondering about this, the answer's simple:

>    File: "iterator1.py", line 5, in ?
>       from xml.dom.Node import Node
>    ImportError: No module named Node

Node is now defined in __init__ of xml.dom

try instead
from xml.dom import Node

Many thanks to Kapil for his reply :)

..And can someone please fix the demo code!... 8)





From loewis at informatik.hu-berlin.de  Fri Feb  2 22:45:49 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:45:49 +0100
Subject: xml install problem - No module named Node
References: <3A7942F4.291E3D6C@hursley.ibm.com> <3A79B353.69D5E104@hursley.ibm.com>
Message-ID: <p6qr91goj1u.fsf@informatik.hu-berlin.de>

Tushar Wagle <tushar at hursley.ibm.com> writes:

> ..And can someone please fix the demo code!... 8)

Fixed. Thanks! Please report further bugs to
sourceforge.net/projects/pyxml, though.

Regards,
Martin



From robin at jessikat.fsnet.co.uk  Thu Feb  1 12:22:39 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 1 Feb 2001 11:22:39 +0000
Subject: http://www.python.org/pipermail/python-dev/?
Message-ID: <OQciTXA$bUe6EwAT@jessikat.fsnet.co.uk>

I get a 404 at the above; has the archive moved?
-- 
Robin Becker


From mwh21 at cam.ac.uk  Thu Feb  1 13:30:18 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Feb 2001 12:30:18 +0000
Subject: http://www.python.org/pipermail/python-dev/?
References: <OQciTXA$bUe6EwAT@jessikat.fsnet.co.uk>
Message-ID: <m3ae867fhh.fsf@atrus.jesus.cam.ac.uk>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> I get a 404 at the above; has the archive moved?

Yes; it's on mail.python.org now.  Except both it & www.python.org
seem to be down ATM...

Cheers,
M.

-- 
  In case you're not a computer person, I should probably point out
  that "Real Soon Now" is a technical term meaning "sometime before
  the heat-death of the universe, maybe". 
                                     -- Scott Fahlman <sef at cs.cmu.edu>


From sbrunning at bigfoot.com  Thu Feb  1 13:47:46 2001
From: sbrunning at bigfoot.com (Simon B.)
Date: Thu, 01 Feb 2001 12:47:46 GMT
Subject: http://www.python.org/pipermail/python-dev/?
References: <OQciTXA$bUe6EwAT@jessikat.fsnet.co.uk>
Message-ID: <95bltg$ff1$1@nnrp1.deja.com>

In article <OQciTXA$bUe6EwAT at jessikat.fsnet.co.uk>,
  Robin Becker <robin at jessikat.fsnet.co.uk> wrote:

I get a 404 at <http://www.python.org/>, too. I suppose that it must be
down.

--
Simon Brunning
I cheated in the finals of my metaphysics exam. I looked into the soul
of the boy sitting next to me. - Woody Allen


Sent via Deja.com
http://www.deja.com/


From fabrice.n at home.com  Thu Feb  1 13:03:58 2001
From: fabrice.n at home.com (shaka)
Date: Thu, 01 Feb 2001 12:03:58 GMT
Subject: Reply to Remco Gerlich
Message-ID: <Ogce6.35628$KP3.9705844@news3.rdc1.on.home.com>

Well I run Java on a special IDE called codewarrior, so if you go on
codewarrior(metrowerk company) website, they have a tutorial on java+some
assignments to do. But you have to sign up for an account.
http://www.codewarrioru.com/CodeWarriorU




From spbridge at my-deja.com  Thu Feb  1 13:33:08 2001
From: spbridge at my-deja.com (spbridge at my-deja.com)
Date: Thu, 01 Feb 2001 12:33:08 GMT
Subject: Python, ASP and global.asa
Message-ID: <95bl24$ese$1@nnrp1.deja.com>

As a relative newcomer to Python, I have to
Develop for ASP during a
transition to Apache / Linux.
My Problem is that Python Scripts in the
global.asa always bring an
"indentation error" although there is definately
none there.
Does anyone have any suggestions?

Steve Bridge
sbridge at workways-gmbh.de


Sent via Deja.com
http://www.deja.com/


From syver at NOSPAMcyberwatcher.com  Fri Feb  2 21:41:37 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 2 Feb 2001 21:41:37 +0100
Subject: Python, ASP and global.asa
References: <95bl24$ese$1@nnrp1.deja.com>
Message-ID: <95f66j$h66$1@troll.powertech.no>

Hi Steve, I've found a couple of solutions to your problem.

It seems that IIS won't tolerate python script in global.asa if the
applications scripting language isn't set to python.

There are however a way to get python code to execute in global.asa without
switching the default language of the application and that is to provide
a script block for the default language in global.asa.

example global.asa if the default scripting language for the application was
Jscript


<script language=Jscript runat=server>
function dummy()
{ 
}
</script>


<script language=Python runat=server>
def Session_OnStart():
  import time
  Session.Contents['OnStart'] =  time.ctime(time.time())
</script>

If the default language was VB you'll have to have a Vb script block there instead. So to be safe you'll have to have both. There seems to be no difference in the order the script blocks appear in.






<spbridge at my-deja.com> wrote in message news:95bl24$ese$1 at nnrp1.deja.com...
> As a relative newcomer to Python, I have to
> Develop for ASP during a
> transition to Apache / Linux.
> My Problem is that Python Scripts in the
> global.asa always bring an
> "indentation error" although there is definately
> none there.
> Does anyone have any suggestions?
> 
> Steve Bridge
> sbridge at workways-gmbh.de
> 
> 
> Sent via Deja.com
> http://www.deja.com/





From jl.berliet at triade.tm.fr  Thu Feb  1 14:15:17 2001
From: jl.berliet at triade.tm.fr (Jean-Louis BERLIET)
Date: Thu, 1 Feb 2001 14:15:17 +0100
Subject: Quelles sociétés développent en PYTHON en FRANCE
Message-ID: <95bno3$483$1@s1.read.news.oleane.net>

Pour convaincre des clients de passer ? PYTHON, les arguments techniques
seuls ne suffisent pas, il faut ?galement donner des exemples de projets ou
de soci?t?s utilisant cette technologie.

Il est toujours possible de donner des exemples am?ricains d'utilisation de
PYTHON (moteur de recherche GOOGLE, site web YAHOO, infracture e-speak de
HP, ...).

Mais QU'EN EST-IL EN FRANCE ????

Il serait tr?s int?ressant de faire un point !

En ce qui me concerne, je peux donner un exemple tr?s r?ussi dans le monde
bancaire (traitement des ch?ques en euro) ou le choix de PYTHON nous a
permis d'?tre moins cher que les concurrents (temps de d?v infinimiment
moins importants qu'avec des solutions de type JAVA ou C) et pour notre
client de tenir les d?lais !

D'autres t?moignages ?????

Jean-Louis BERLIET

P.S. : si certains PYTHONISTES sont int?ress?s par une embauche sur LYON,
qu'ils me contactent, les projets foisonnent (PYTHON, ZOPE, ...) !






From J.Jonkers at acriter.com  Thu Feb  1 14:31:46 2001
From: J.Jonkers at acriter.com (Johan Jonkers)
Date: Thu, 01 Feb 2001 14:31:46 +0100
Subject: Quelles =?iso-8859-1?Q?soci=E9t=E9s=20d=E9veloppent?= en PYTHON en 
	FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net>
Message-ID: <3A796542.C036DCD4@acriter.com>

Making postings in a non-English language will probably result in not
many people answering. Bad English is usually more understood then
French.

Johan

-- 
        -- Programmers don't die; they just GOSUB with no RETURN --


From fredrik at effbot.org  Thu Feb  1 14:44:57 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 01 Feb 2001 13:44:57 GMT
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net> <3A796542.C036DCD4@acriter.com>
Message-ID: <tLde6.10586$AH6.1742183@newsc.telia.net>

Johan Jonkers wrote:
> Making postings in a non-English language will probably result in not
> many people answering. Bad English is usually more understood then
> French.

Det beror v?l iofs p? m?lgruppen.  Jag ?r n?stan helt s?ker
p? att franska Python-anv?ndare faktiskt kan l?sa Jean-Louis
meddelande ;-)

mvh /F

PS. f?r dom som kan l?sa mitt meddelande:
http://groups.yahoo.com/group/python-se




From martin.skott at propylon.com  Thu Feb  1 17:10:51 2001
From: martin.skott at propylon.com (Martin Skott)
Date: Thu, 01 Feb 2001 16:10:51 GMT
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net> <3A796542.C036DCD4@acriter.com> <tLde6.10586$AH6.1742183@newsc.telia.net>
Message-ID: <8znqo13n.fsf@propylon.com>

"Fredrik Lundh" <fredrik at effbot.org> writes:

> Johan Jonkers wrote:
> > Making postings in a non-English language will probably result in not
> > many people answering. Bad English is usually more understood then
> > French.
> 
> Det beror v?l iofs p? m?lgruppen.  Jag ?r n?stan helt s?ker
> p? att franska Python-anv?ndare faktiskt kan l?sa Jean-Louis
> meddelande ;-)

Well I feel both of you are right, but it would be a pain to read
groups where you could only understand some of the messages because
some elite group are discussing in some language you don't know.
The question from the original poster (as far as my two years of
French can help me) was basicly:
We always hear about Python being used in the US (mentions Google and
a  few other), but what about France? There was a lot more, but this
is the core of the question.
I would like to extend this to Europe (or Asia or Anarctica or any
other part of the world outside US). Where are people using Python?
I can stat by telling that I am Danish, but working with Python in
Ireland (in a job I found through this group). 

-- 
Martin Skott
Software Developer
Propylon - Enabling Universal Mobility
http://www.propylon.com
Tel: +353 096 37245
Mobile: +353 087 9680370


From moshez at zadka.site.co.il  Fri Feb  2 01:32:58 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri,  2 Feb 2001 02:32:58 +0200 (IST)
Subject: Quelles sociétés développent en PYTHON en FRANCE
In-Reply-To: <8znqo13n.fsf@propylon.com>
References: <8znqo13n.fsf@propylon.com>, <95bno3$483$1@s1.read.news.oleane.net> <3A796542.C036DCD4@acriter.com> <tLde6.10586$AH6.1742183@newsc.telia.net>
Message-ID: <20010202003258.EB777A840@darjeeling.zadka.site.co.il>

On Thu, 01 Feb 2001, Martin Skott <martin.skott at propylon.com> wrote:

> I would like to extend this to Europe (or Asia or Anarctica or any
> other part of the world outside US). Where are people using Python?

I'm using it in Israel -- and I'm not the only one. There are a couple
of Zope-shops here too.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From J.Jonkers at acriter.com  Fri Feb  2 07:37:18 2001
From: J.Jonkers at acriter.com (Johan Jonkers)
Date: Fri, 02 Feb 2001 07:37:18 +0100
Subject: Quelles =?iso-8859-1?Q?soci=E9t=E9s=20d=E9veloppent?= en PYTHON en 
	FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net> <3A796542.C036DCD4@acriter.com> <tLde6.10586$AH6.1742183@newsc.telia.net> <8znqo13n.fsf@propylon.com>
Message-ID: <3A7A559E.ACD9F54C@acriter.com>

Been using Python for like 6 months now here in Holland.


Martin Skott wrote:
> 
> "Fredrik Lundh" <fredrik at effbot.org> writes:
> 
> > Johan Jonkers wrote:
> > > Making postings in a non-English language will probably result in not
> > > many people answering. Bad English is usually more understood then
> > > French.
> >
> > Det beror v?l iofs p? m?lgruppen.  Jag ?r n?stan helt s?ker
> > p? att franska Python-anv?ndare faktiskt kan l?sa Jean-Louis
> > meddelande ;-)
> 
> Well I feel both of you are right, but it would be a pain to read
> groups where you could only understand some of the messages because
> some elite group are discussing in some language you don't know.
> The question from the original poster (as far as my two years of
> French can help me) was basicly:
> We always hear about Python being used in the US (mentions Google and
> a  few other), but what about France? There was a lot more, but this
> is the core of the question.
> I would like to extend this to Europe (or Asia or Anarctica or any
> other part of the world outside US). Where are people using Python?
> I can stat by telling that I am Danish, but working with Python in
> Ireland (in a job I found through this group).
> 
> --
> Martin Skott
> Software Developer
> Propylon - Enabling Universal Mobility
> http://www.propylon.com
> Tel: +353 096 37245
> Mobile: +353 087 9680370

-- 
        -- Programmers don't die; they just GOSUB with no RETURN --


From Marc.Poinot at onera.fr  Fri Feb  2 09:35:11 2001
From: Marc.Poinot at onera.fr (Marc Poinot)
Date: Fri, 02 Feb 2001 09:35:11 +0100
Subject: Quelles =?iso-8859-1?Q?soci=E9t=E9s=20d=E9veloppent?= en PYTHON en 
	FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net> <3A796542.C036DCD4@acriter.com> <tLde6.10586$AH6.1742183@newsc.telia.net> <8znqo13n.fsf@propylon.com>
Message-ID: <3A7A713F.A047ECC8@onera.fr>

Well... yes, we have computers too here...
Then, an extension to computer is operating system (such as VMS).
And we also use to put some application software on top os OS (such as
Addition-v4.6). And some of the best of us have knowledge in
programming (COBOL-32).
I've heard about some Xinul (or Nuxil, I can't remember...), a kind
of graphical computer-human interface.
And finally, yes we have some Python too. But today, there's some
snow on Paris, and I guess they are sleeping in a warm place of 
the Zoo.

*** End of joke namespace:

We are trying hard to organize here an European meeting for
Pythoners. But it's difficult, because there is at least one major
Python event in the US every 6 months...
Have a look at the non-english page on Python home:

http://www.python.org/doc/NonEnglish.html

Marcvs [alias Mais sur le Francaise liste tu veux ecrire
        le Francaise avec aussi]



From fernstrom at geocities.com  Fri Feb  2 15:19:06 2001
From: fernstrom at geocities.com (christer fernstrom)
Date: Fri, 02 Feb 2001 14:19:06 GMT
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net> <3A796542.C036DCD4@acriter.com> <tLde6.10586$AH6.1742183@newsc.telia.net> <8znqo13n.fsf@propylon.com>
Message-ID: <95efkh$u2n$1@nnrp1.deja.com>

Being a swede in France I had the privilege to understand both the
original question and Fredrik's comment. And to answer the question: at
Xerox research (XRCE in Grenoble) we do quite a lot of prototyping work
in Python. We find it hard to convince Python developers to move to
Java, and ususally they're back to Python before long.

  christer


Sent via Deja.com
http://www.deja.com/


From jasonic at nomadicsltd.com  Sat Feb  3 05:30:28 2001
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Fri, 2 Feb 2001 23:30:28 -0500
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net> <3A796542.C036DCD4@acriter.com>
Message-ID: <t7n23u39o43302@corp.supernews.com>

Johan Jonkers <J.Jonkers at acriter.com> wrote in message
news:3A796542.C036DCD4 at acriter.com...
> Making postings in a non-English language will probably result in not
> many people answering. Bad English is usually more understood then
> French.

Nonsense.. imho this is a narrow point of view.

If you are French and want to find French speaking Python programmers it
makes perfect sense to post here in you own language. If YOU cannot rad it,
then true, you will pass over it. 'more understood' ---by whom?

Only those who can read/write French will likely be suitable. The same
applies for any language.
Thre are so many progammers around the world, for whom English is not therir
mother tongue. I am forever impressed a the additioanl obstacles which must
be over come.. first to learn english then to learn computer programming.

There are many smart people in the world can READ english, especially
computer or science-oriented text, but perhaps cannot write or speak it
well. There are others who cannot conduct a normal converstation, but can
write better englsh than many native speakers.

For Asians, there is an even greater set of hurdle to surmount, because
writing systems and langauge roots have almost phonetic, structural,
sytntactic or etymological connections whatever. For Asians going to UK or
USA, it is very expensive, so mostly on the rich and priviledged can afford
a extended trip whihc allows them to develop fluent language skills. Ther is
a very high awareness and so local english langauge schools are big business
[I don't know about Asians visiting Australia but imagine that is more and
more attractive option]

For example there might be many people right now in Korea or China or Japan
reading this newgsgroup who do not find it easy to contribute, even though
they may be proficient Python programmers.

European languagse are really are very close together. Dutch, Germans and
Scandanavians are famously brilliant for fluent and very correct English.
This is partly linguistic, since Olde English has inherited much from them,
but also cultural.

France on the other hand, is notoriously proud/fearful of losing the
distinction of language. Thus movies and TV are almsot all dubbed. People
lose out because they do not have exposure to even hear english much.
Result: catch-22 [attrappement-22] many young people find it much harder to
learn, unless they are fortunate enough to travel and stay in English
speaking countries, or where there is abundant english use.


- Jason




From loewis at informatik.hu-berlin.de  Fri Feb  2 22:54:19 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:54:19 +0100
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net>
Message-ID: <p6qofwkoino.fsf@informatik.hu-berlin.de>

"Jean-Louis BERLIET" <jl.berliet at triade.tm.fr> writes:

> Pour convaincre des clients de passer ? PYTHON, les arguments
> techniques seuls ne suffisent pas, il faut ?galement donner des
> exemples de projets ou de soci?t?s utilisant cette technologie.

Ich finde, dass B?cher in der jeweiligen Landessprache auch sehr
wichtig sind. Menschen lesen nunmal lieber "schwierige" Materie in
ihrer Muttersprache, selbst wenn sie Englisch k?nnen.

Gibt es eigentlich franz?sische Python-B?cher? Auf

http://www.amk.ca/bookstore/non-english.html

sind jedenfalls keine aufgef?hrt... Ist irgendjemand an einer
?bersetzung eines deutschen Buchs interessiert :-?

Ciao,
Martin


From hinsen at cnrs-orleans.fr  Mon Feb  5 18:05:18 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 05 Feb 2001 18:05:18 +0100
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net> <p6qofwkoino.fsf@informatik.hu-berlin.de>
Message-ID: <m3lmrl3vsh.fsf@chinon.cnrs-orleans.fr>

Martin von Loewis <loewis at informatik.hu-berlin.de> writes:

> Gibt es eigentlich franz?sische Python-B?cher? Auf

Es gibt eine franzoesische Uebersetzung von "Learning Python" und von
"Python Pocket Reference". Aber der Markt kann sicher mehr vertragen.

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From jasonic at nomadicsltd.com  Sat Feb  3 05:02:25 2001
From: jasonic at nomadicsltd.com (Jason Cunliffe)
Date: Fri, 2 Feb 2001 23:02:25 -0500
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net>
Message-ID: <t7n0fb53irbq34@corp.supernews.com>

Jean-Louis BERLIET <jl.berliet at triade.tm.fr> wrote in message
news:95bno3$483$1 at s1.read.news.oleane.net...
> Pour convaincre des clients de passer ? PYTHON, les arguments techniques
> seuls ne suffisent pas, il faut ?galement donner des exemples de projets
ou
> de soci?t?s utilisant cette technologie.

To convince clients to switch to Python, technical arguuments are not
enough. One must also provide examples of projects or companies using this
technology.

> Il est toujours possible de donner des exemples am?ricains d'utilisation
de
> PYTHON (moteur de recherche GOOGLE, site web YAHOO, infracture e-speak de
> HP, ...)

It is always possible to give American examples of Python use (Google's
search emgine, YAHOO web site, HP's E-Speak infrastructure)

> Mais QU'EN EST-IL EN FRANCE ????

But WHERE IS IT IN FRANCE ???

> Il serait tr?s int?ressant de faire un point !

It will be very interesting to make a point ["""idiom - sorry,not sure how
to translate this one correctly"""]

> En ce qui me concerne, je peux donner un exemple tr?s r?ussi dans le monde
> bancaire (traitement des ch?ques en euro) ou le choix de PYTHON nous a
> permis d'?tre moins cher que les concurrents (temps de d?v infinimiment
> moins importants qu'avec des solutions de type JAVA ou C) et pour notre
> client de tenir les d?lais !

As for me, I can give an very successful example in the world of banking
(handling checks in EUROS) where the choice of PYTHON allowed us to be
cheaper than the competition (develoment time is infinitely less
important[significant?] than with JAVA or C solutions) and to reduce dealys
for clients

> D'autres t?moignages ?????

Any other witnesses [testimonials/stories/case studies]

> Jean-Louis BERLIET
>
> P.S. : si certains PYTHONISTES sont int?ress?s par une embauche sur LYON,
> qu'ils me contactent, les projets foisonnent (PYTHON, ZOPE, ...) !

P.S. : if any PYHONISTAS are itersted in a gig in LYON contact me. Projects
are brewing (PYTHON, ZOPE, ...)


hth
JASON
___________________________________________________________
Jason CUNLIFFE = NOMADICS['Interactive Art and Technology']




From hinsen at cnrs-orleans.fr  Mon Feb  5 18:08:00 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 05 Feb 2001 18:08:00 +0100
Subject: Quelles sociétés développent en PYTHON en FRANCE
References: <95bno3$483$1@s1.read.news.oleane.net>
Message-ID: <m3itmp3vnz.fsf@chinon.cnrs-orleans.fr>

"Jean-Louis BERLIET" <jl.berliet at triade.tm.fr> writes:

> Pour convaincre des clients de passer ? PYTHON, les arguments techniques
> seuls ne suffisent pas, il faut ?galement donner des exemples de projets ou
> de soci?t?s utilisant cette technologie.

Bien sur, mais...

> Il est toujours possible de donner des exemples am?ricains d'utilisation de
> PYTHON (moteur de recherche GOOGLE, site web YAHOO, infracture e-speak de
> HP, ...).
> 
> Mais QU'EN EST-IL EN FRANCE ????

Je ne le sais pas non plus, mais je ne vois pas l'interet quand il
s'agit d'argumenter pour Python. Enfin, si les americains ont des
bonnes experiences, ca compte pour la France aussi, l'informatique
n'est pas fondamentalement different des deux cotes de l'Atlantique.

Ceci dit, par pure curiosite j'aimerais aussi connaitre la reponse!

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From vonehr at ira.uka.de  Thu Feb  1 14:47:57 2001
From: vonehr at ira.uka.de (Markus von Ehr)
Date: Thu, 01 Feb 2001 14:47:57 +0100
Subject: win32com: passing object to threads
Message-ID: <3A79690D.563A7248@ira.uka.de>

Hi,

I play a mp3 File with MediaPlayer using COM like this:
(Globals. are my global variables)

  Globals.ie =
win32com.client.dynamic.Dispatch("InternetExplorer.Application")
  Globals.ie.Navigate(r'about:blank')
  Globals.doc = Globals.ie.Document
  Globals.doc.body.insertAdjacentHTML("afterBegin",
  """
  <OBJECT ID="MediaPlayer1"
  CLASSID="CLSID:22D6f312-B0F6-11D0-94AB-0080C74C7E95">
  </OBJECT>
  """)
    
  Globals.mp = Globals.doc.body.children("MediaPlayer1").object

  Globals.mp.FileName = "MySoundFile.mp3"

In another thread I want to set the New SoundFile:

  Globals.mp.FileName = "null.wav"

and I have the error:

  Pythoncomerror, (-2147221008, 'CoInitialize hasn't been called', None,
None)

I know that I have to use the: pythoncom.CoInitialize() and 
pythoncom.CoUninitialize() in a way, but don't know why. 
Can anybody help me?

Thanks a lot for any answer,

Markus


From aleaxit at yahoo.com  Thu Feb  1 15:55:46 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 1 Feb 2001 15:55:46 +0100
Subject: win32com: passing object to threads
References: <3A79690D.563A7248@ira.uka.de>
Message-ID: <95btdl016u0@news1.newsguy.com>

"Markus von Ehr" <vonehr at ira.uka.de> wrote in message
news:3A79690D.563A7248 at ira.uka.de...
    [snip]
> In another thread I want to set the New SoundFile:
>
>   Globals.mp.FileName = "null.wav"
>
> and I have the error:
>
>   Pythoncomerror, (-2147221008, 'CoInitialize hasn't been called', None,
> None)
>
> I know that I have to use the: pythoncom.CoInitialize() and
> pythoncom.CoUninitialize() in a way, but don't know why.
> Can anybody help me?

Each thread needs to call CoInitialize for itself -- that's
the COM rule.  Are you asking the "why" of this COM rule, or
the "why" it gets reflected down to your Python code for
all threads except the main one?

Assuming you accept both of these design decisions as
reasonable, the remaining meanings of "why" seem to follow:
as you are doing a COM operation from this other thread, the
thread needs to be COM-enabled.

A thread being "COM enabled" means it having told COM
explicitly exactly how it intends to operate within it -- in
what "apartment", in COM parlance.  This specific "explicit
is better than implicit" design decision by the Microsoft
COM team answers the first "why": COM refuses to guess
exactly HOW a given thread might want to operate with it.

As for the second "why", I don't think Python lets a package
(e.g. Pythoncom) "hook" some generic function on the event
"creation of an arbitrary new thread", so, even if implicitly
and magically COM-enabling all threads were a good architecture
(and I have my doubts), I don't see how Pythoncom might
implement that architecture.


Alex





From MarkH at ActiveState.com  Fri Feb  2 00:21:39 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 01 Feb 2001 23:21:39 GMT
Subject: win32com: passing object to threads
References: <3A79690D.563A7248@ira.uka.de>
Message-ID: <3A79EE92.2000908@ActiveState.com>

Markus von Ehr wrote:


 > I know that I have to use the: pythoncom.CoInitialize() and
 > pythoncom.CoUninitialize()  in a way, but don't know
 > why.   Can anybody help me?

COM has different behaviour depending on the COM threading model.  This 
threading model is controlled by CoInitialize/CoInitializeEx(), hence 
you have to understand the different behaviours and make an appropriaate 
choice.  Just to scare you even more, if you pass objects between 
different "threading apartments", you have to use the scarily named 
pythoncom.CoMarshalInterThreadInterfaceInStream() to pass COM objects 
between threads.

The only reasonable documentation on this for Python is "Python 
Programming on Win32" (by me and Andy, published by OReilly.)  However, 
generic documentation on COM's threading models will also help.

Mark.



From gmunsey at adobe.com  Thu Feb  1 15:46:06 2001
From: gmunsey at adobe.com (Grant Munsey)
Date: Thu, 1 Feb 2001 06:46:06 -0800
Subject: Registering LOTS of C level functions
Message-ID: <OEee6.80706$Wq1.34166930@nnrp5-w.sbc.net>

I am covering several libraries ... each of which has a couple of thousand
calls (don't ask!).
I would really like some way to be able to register all of them to Python
with one actual
C function and an ID so that inside the C function I can generate the actual
call by using
a data structure that tells the actual functions address and what parameters
it requires.

As it stands I don't see any way to do this without an excessive amount of
trickery. Looks
to me like there is no way to specify an extra parameter to the low level
registration process
for each function ... There is the "self" parameter but that would require
me to allocate a Python
object for each procedure I wanted to cover.

The only thing I can think of at the moment is to generate a bunch of
re-direction procedures ....
Tiny procedures that just get called from Python, add the ID code ... and
then pass on to the
"real" command handler.

Anyone have any suggestions?




From vonehr at ira.uka.de  Thu Feb  1 16:17:42 2001
From: vonehr at ira.uka.de (Markus von Ehr)
Date: Thu, 01 Feb 2001 16:17:42 +0100
Subject: Registering LOTS of C level functions
References: <OEee6.80706$Wq1.34166930@nnrp5-w.sbc.net>
Message-ID: <3A797E16.F89555CB@ira.uka.de>

Grant Munsey wrote:
> 
> I am covering several libraries ... each of which has a couple of thousand
> calls (don't ask!).
> I would really like some way to be able to register all of them to Python
> with one actual
> C function and an ID so that inside the C function I can generate the actual
> call by using
> a data structure that tells the actual functions address and what parameters
> it requires.
> 
> As it stands I don't see any way to do this without an excessive amount of
> trickery. Looks
> to me like there is no way to specify an extra parameter to the low level
> registration process
> for each function ... There is the "self" parameter but that would require
> me to allocate a Python
> object for each procedure I wanted to cover.
> 
> The only thing I can think of at the moment is to generate a bunch of
> re-direction procedures ....
> Tiny procedures that just get called from Python, add the ID code ... and
> then pass on to the
> "real" command handler.
> 
> Anyone have any suggestions?

You can use the calldll module by Sam Rushing which can call 
both __stdcall and __cdecl functions.

ftp://squirl.nightmare.com/pub/python/python-ext/

build a dictionary with your function IDs and the function addresses to
call 
them for each lib.

Markus


From donn at u.washington.edu  Thu Feb  1 18:39:05 2001
From: donn at u.washington.edu (Donn Cave)
Date: 1 Feb 2001 17:39:05 GMT
Subject: Registering LOTS of C level functions
References: <OEee6.80706$Wq1.34166930@nnrp5-w.sbc.net>
Message-ID: <95c6vp$gvg$1@nntp6.u.washington.edu>

Quoth "Grant Munsey" <gmunsey at adobe.com>:
| I am covering several libraries ... each of which has a couple of
| thousand calls (don't ask!).  I would really like some way to be
| able to register all of them to Python with one actual C function
| and an ID so that inside the C function I can generate the actual
| call by using a data structure that tells the actual functions
| address and what parameters it requires.

I do something like that for BeOS Python API modules, but I code
the functions separately as usual and put the argument table data
in static storage local to the function.  I have only about a
thousand functions, spread out over 30 odd modules, but I can't
see why it wouldn't work just fine at your scale.  (I didn't
use a tuple parse format, cf. PyArg_ParseTuple(), because I'm
dealing with C++ style function overloading.)  The key point is
to generate the code automatically.

| The only thing I can think of at the moment is to generate a
| bunch of re-direction procedures .... Tiny procedures that
| just get called from Python, add the ID code ... and then
| pass on to the "real" command handler.

Well, of course that would work.  There may not be a lot of
difference, one way or the other, however you do it - it would
be interesting to hear about results from trying a couple of
approaches.

	Donn Cave, donn at u.washington.edu


From danielk at aracnet.com  Thu Feb  1 16:00:57 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Thu, 01 Feb 2001 07:00:57 -0800
Subject: popen2 questions
Message-ID: <r5qi7tg81mldmic7tkd5ijj5v9v3stsdsc@4ax.com>

The documentation for the 'popen2' module states:

"""
... if bufsize is specified, it specifies the buffer size for the I/O pipes...
"""

however, on Windows, if I set this to anything but '-1', I get

Traceback (most recent call last):
  File "po.py", line 2, in ?
    fin,fout = popen2.popen2("XTEST",128,'t')
  File "c:\python20\lib\popen2.py", line 120, in popen2
    w, r = os.popen2(cmd, mode, bufsize)
ValueError: bufsize must be -1

Is 'bufsize' supported on Windows?

What does the '-1' setting change?

The bottom line what I'm trying to accomplish is to squeeze any optimizations
out of popen2 that I can so I thought to play with the 'bufsize' parameter.

I'm running Python 2.0 on Win2k.

Advanced thanks,
Daniel Klein
Portland,OR USA


From J.Jonkers at acriter.com  Fri Feb  2 07:38:26 2001
From: J.Jonkers at acriter.com (Johan Jonkers)
Date: Fri, 02 Feb 2001 07:38:26 +0100
Subject: popen2 questions
References: <r5qi7tg81mldmic7tkd5ijj5v9v3stsdsc@4ax.com>
Message-ID: <3A7A55E2.53DC2CFB@acriter.com>

-1 probably means that it's not supported. The data will be send
directly and not buffered. Anyways, that's what it looks like to me,
however I (gladly!) don't use windows :)

Johan
-- 
        -- Programmers don't die; they just GOSUB with no RETURN --


From danielk at aracnet.com  Fri Feb  2 14:53:55 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Fri, 02 Feb 2001 05:53:55 -0800
Subject: popen2 questions
References: <r5qi7tg81mldmic7tkd5ijj5v9v3stsdsc@4ax.com> <3A7A55E2.53DC2CFB@acriter.com>
Message-ID: <viel7tkf92jmcijke2u0gi14ouuqrtm5bu@4ax.com>

On Fri, 02 Feb 2001 07:38:26 +0100, Johan Jonkers <J.Jonkers at acriter.com>
wrote:

>-1 probably means that it's not supported. The data will be send
>directly and not buffered. Anyways, that's what it looks like to me,
>however I (gladly!) don't use windows :)

Thanks Johan, that's what I suspected.

I have to test these kinds of things on _all_ platforms that are required by my
users, and that includes Windows. In this industry, it is fatal to make value
judgements especially on a product as ubiquitous as Windows.

Have a nice day. :^)

Daniel Klein
Portland, OR USA



From ryu-jin at bigfoot.com  Thu Feb  1 16:34:43 2001
From: ryu-jin at bigfoot.com (Ryujin)
Date: Thu, 01 Feb 2001 15:34:43 GMT
Subject: Standalone applications
Message-ID: <95c000$8nc$1@ryu.it>

Hi, Im trying to understand which is the best way to create a Standalone application in Python.
I downloaded Standalone.py from http://starship.python.net/crew/gmcm/distribute.html
and it works with python1.5 but the executable must stay in the same directory of the .so files.
Then I tried to download Python2C and it generate from a file.py two files:
__file.c adn _file.py 
and then I cant figure out how to compile the file.c

Any suggestion???
thx alot
Matte

-- 
---------------------------------------
Memelli Matteo					
Could u imagine a world without Windoz?	 
ryu-jin at bigfoot.com			 
---------------------------------------



From loewis at informatik.hu-berlin.de  Fri Feb  2 22:56:23 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:56:23 +0100
Subject: Standalone applications
References: <95c000$8nc$1@ryu.it>
Message-ID: <p6qlmrooik8.fsf@informatik.hu-berlin.de>

"Ryujin" <ryu-jin at bigfoot.com> writes:

> Any suggestion???

I personally prefer freeze. It requires a C compiler, but gives you
best control and reliability. Plus it allows to create *truly*
stand-alone binaries, i.e. binaries that are not even dependent on
shared libraries.

Regards,
Martin


From ryu-jin at bigfoot.com  Sun Feb  4 14:11:36 2001
From: ryu-jin at bigfoot.com (Ryujin)
Date: Sun, 04 Feb 2001 13:11:36 GMT
Subject: Standalone applications
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de>
Message-ID: <95jept$251$1@ryu.it>

In article <p6qlmrooik8.fsf at informatik.hu-berlin.de>, Martin von Loewis
<loewis at informatik.hu-berlin.de> wrote:
As I can read in the Freeze Readme,  it is not possible to freeze and ship
Tkinter application...did u find any solution to this problem?
thx

matte

> I personally prefer freeze. It requires a C compiler, but gives you best
> control and reliability. Plus it allows to create *truly* stand-alone
> binaries, i.e. binaries that are not even dependent on shared libraries.
> 
> Regards, Martin


-- 
---------------------------------------
Memelli Matteo					
Could u imagine a world without Windoz?	 
ryu-jin at bigfoot.com			 
---------------------------------------



From loewis at informatik.hu-berlin.de  Fri Feb  9 18:41:47 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Feb 2001 18:41:47 +0100
Subject: Standalone applications
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de> <95jept$251$1@ryu.it>
Message-ID: <p6qy9vf3g9w.fsf@informatik.hu-berlin.de>

"Ryujin" <ryu-jin at bigfoot.com> writes:

> As I can read in the Freeze Readme,  it is not possible to freeze and ship
> Tkinter application...did u find any solution to this problem?

I always used Tix' SAM (stand-alone modules) approach. That compiles
Tcl and Tk's .tcl files into libraries, which can then be linked into
complete binaries. I understand TclPro offers a similar mechanism.

Regards,
Martin



From thomas.heller at ion-tof.com  Fri Feb  9 19:27:02 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Fri, 9 Feb 2001 19:27:02 +0100
Subject: Standalone applications
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de> <95jept$251$1@ryu.it> <p6qy9vf3g9w.fsf@informatik.hu-berlin.de>
Message-ID: <961cpr$it636$1@ID-59885.news.dfncis.de>

"Martin von Loewis" <loewis at informatik.hu-berlin.de> wrote in message news:p6qy9vf3g9w.fsf at informatik.hu-berlin.de...
> "Ryujin" <ryu-jin at bigfoot.com> writes:
>
> > As I can read in the Freeze Readme,  it is not possible to freeze and ship
> > Tkinter application...did u find any solution to this problem?
>
> I always used Tix' SAM (stand-alone modules) approach. That compiles
> Tcl and Tk's .tcl files into libraries, which can then be linked into
> complete binaries. I understand TclPro offers a similar mechanism.
Interesting.
Do you have any pointers?

Regards,

Thomas




From paulus132 at my-deja.com  Tue Feb  6 16:37:57 2001
From: paulus132 at my-deja.com (paulus132 at my-deja.com)
Date: Tue, 06 Feb 2001 15:37:57 GMT
Subject: Standalone applications
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de>
Message-ID: <95p5ok$7cu$1@nnrp1.deja.com>

In article <p6qlmrooik8.fsf at informatik.hu-berlin.de>,
  Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
> "Ryujin" <ryu-jin at bigfoot.com> writes:
>
> > Any suggestion???
>
> I personally prefer freeze. It requires a C compiler, but gives you
> best control and reliability. Plus it allows to create *truly*
> stand-alone binaries, i.e. binaries that are not even dependent on
> shared libraries.
>

I also have some good experience with freeze, but the binaries are big
(1.8M on Sun, 0.5M after strip), even for very simple *.py scripts !!

Do you have some tips to reduce it ??

Paul Perrin



> Regards,
> Martin
>


Sent via Deja.com
http://www.deja.com/


From dsh8290 at rit.edu  Tue Feb  6 19:27:12 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 6 Feb 2001 13:27:12 -0500
Subject: Standalone applications
In-Reply-To: <95p5ok$7cu$1@nnrp1.deja.com>; from paulus132@my-deja.com on Tue, Feb 06, 2001 at 03:37:57PM +0000
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de> <95p5ok$7cu$1@nnrp1.deja.com>
Message-ID: <20010206132712.B1814@harmony.cs.rit.edu>

On Tue, Feb 06, 2001 at 03:37:57PM +0000, paulus132 at my-deja.com wrote:
| In article <p6qlmrooik8.fsf at informatik.hu-berlin.de>,
|   Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
| > "Ryujin" <ryu-jin at bigfoot.com> writes:
| >
| > > Any suggestion???
| >
| > I personally prefer freeze. It requires a C compiler, but gives you
| > best control and reliability. Plus it allows to create *truly*
| > stand-alone binaries, i.e. binaries that are not even dependent on
| > shared libraries.
| >
| 
| I also have some good experience with freeze, but the binaries are big
| (1.8M on Sun, 0.5M after strip), even for very simple *.py scripts !!
| 
| Do you have some tips to reduce it ??

Tell the users to install the interpreter on their own.  Then your
package will only consist of your code and be much smaller. ;-)

-D




From grante at visi.com  Wed Feb  7 19:55:33 2001
From: grante at visi.com (Grant Edwards)
Date: Wed, 07 Feb 2001 18:55:33 GMT
Subject: Standalone applications
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de> <95p5ok$7cu$1@nnrp1.deja.com>
Message-ID: <FSgg6.6907$77.161161@ruti.visi.com>

In article <95p5ok$7cu$1 at nnrp1.deja.com>, paulus132 at my-deja.com wrote:

>I also have some good experience with freeze, but the binaries are big
>(1.8M on Sun, 0.5M after strip), even for very simple *.py scripts !!
>
>Do you have some tips to reduce it ??

Just write more code!  Eventually the 500k of overhead won't be
so noticable.  ;)

-- 
Grant Edwards                   grante             Yow!  My NOSE is NUMB!
                                  at               
                               visi.com            


From loewis at informatik.hu-berlin.de  Fri Feb  9 18:43:00 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Feb 2001 18:43:00 +0100
Subject: Standalone applications
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de> <95p5ok$7cu$1@nnrp1.deja.com>
Message-ID: <p6qvgqj3g7v.fsf@informatik.hu-berlin.de>

paulus132 at my-deja.com writes:

> I also have some good experience with freeze, but the binaries are big
> (1.8M on Sun, 0.5M after strip), even for very simple *.py scripts !!
> 
> Do you have some tips to reduce it ??

I'd check the config.c generated by freeze. It may be that it decides
to link-in builtin modules which are not needed in your application.

Regards,
Martin



From jurgen.defurne at philips.com  Wed Feb  7 08:42:51 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Wed, 7 Feb 2001 08:42:51 +0100
Subject: Standalone applications
Message-ID: <0056900015717749000002L092*@MHS>

Basically, what you are trying to do here is the same that Clipper did way back.
THe minumum program we could get with it was 100kB. If it took the database
library and the advanced stuff, we got executables between 350kB and 700 kB.
What it did have, was that once most libraries were in it, the programs expanded
less rapid.

Does freeze compile everything to C and then to binary ? Shouldn't there
be a way that you can create an executable, which consists of two parts :
- the first part being code and libraries
- the second part a data part, which consists of all byte codes needed

Jurgen




paulus132 at my-deja.com@SMTP at python.org on 06/02/2001 17:03:11
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	Re: Standalone applications
Classification:	

In article <p6qlmrooik8.fsf at informatik.hu-berlin.de>,
  Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
> "Ryujin" <ryu-jin at bigfoot.com> writes:
>
> > Any suggestion???
>
> I personally prefer freeze. It requires a C compiler, but gives you
> best control and reliability. Plus it allows to create *truly*
> stand-alone binaries, i.e. binaries that are not even dependent on
> shared libraries.
>

I also have some good experience with freeze, but the binaries are big
(1.8M on Sun, 0.5M after strip), even for very simple *.py scripts !!

Do you have some tips to reduce it ??

Paul Perrin



> Regards,
> Martin
>


Sent via Deja.com
http://www.deja.com/
--
http://mail.python.org/mailman/listinfo/python-list





From ngps at madcap.dyndns.org  Thu Feb  1 18:06:07 2001
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 1 Feb 2001 17:06:07 GMT
Subject: [ANN] M2Crypto 0.06 snapshot #2 -- ZServerSSL
Message-ID: <95c51v$3oa$1@coco.singnet.com.sg>

Hi,

A new snapshot of M2Crypto is now available. The major change:

 ZServerSSL Rides Again
------------------------

Whew! It's been a year! ZServerSSL was last released on 1 Feb 2000 for
Zope 2.1.3.

It has now been dusted off and re-released for Zope 2.3.0.

Tested with Zope-2.3.0-win32-x86 with M2Crypto for Python1.

Tested with Zope-2.3.0-src on FreeBSD with M2Crypto for 
both Python1 and Python2.

Interoperated with Netscape, IE and Opera on Win32, and also Netscape
on FreeBSD.

Beware browsers with broken HTTP/1.1 - they drive ZServerSSL to 100%
CPU utilisation: An example is Netscape Communicator 4.72 on my
FreeBSD boxen.

(Or maybe it's ZServer's HTTP/1.1 that's broken. ;-)

X.509 certificate-based client authentication for HTTPS and the
encrypting monitor, both features of the ZServerSSL released a year
ago, to come RSN!


As usual, get M2Crypto here:

    http://www.post1.com/home/ngps/m2


Feedback is appreciated. Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From victor at prodigy.net  Thu Feb  1 18:18:38 2001
From: victor at prodigy.net (Victor Muslin)
Date: Thu, 01 Feb 2001 17:18:38 GMT
Subject: sys.exit and exceptions (newbie question)
Message-ID: <3a7998cf.869169328@localhost>

What is a pythonic idiom for forcing an immediate program termination?
sys.exit() does not do it, because it throws a SystemExit exception,
so if it is called within a try block, a generic except block is still
executed. Of course it is possible to specifically catch SystemExit,
as in the following example, but it is a pain. Moreover, one may want
to exit in a code that plugs into somebody else's code framed by try:
except: blocks.

import sys
try:
	print 'Exiting...'
	sys.exit(0)
except SystemExit:
	print 'System exit...'
except:
	print 'Oops... except'


From gregj at pobox.com  Thu Feb  1 20:28:43 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 01 Feb 2001 19:28:43 GMT
Subject: sys.exit and exceptions (newbie question)
References: <3a7998cf.869169328@localhost>
Message-ID: <95cdcs$6b9$1@nnrp1.deja.com>

Functions that exit without giving the caller a chance to catch the
exception are a very bad idea. Exception handling is good Python
practice; forcing an un-catchable exit is not.


In article <3a7998cf.869169328 at localhost>,
  victor at prodigy.net (Victor Muslin) wrote:
> What is a pythonic idiom for forcing an immediate program termination?
> sys.exit() does not do it, because it throws a SystemExit exception,
> so if it is called within a try block, a generic except block is still
> executed. Of course it is possible to specifically catch SystemExit,
> as in the following example, but it is a pain. Moreover, one may want
> to exit in a code that plugs into somebody else's code framed by try:
> except: blocks.
>
> import sys
> try:
> 	print 'Exiting...'
> 	sys.exit(0)
> except SystemExit:
> 	print 'System exit...'
> except:
> 	print 'Oops... except'
>

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From fredrik at effbot.org  Thu Feb  1 20:54:07 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Thu, 01 Feb 2001 19:54:07 GMT
Subject: sys.exit and exceptions (newbie question)
References: <3a7998cf.869169328@localhost>
Message-ID: <z9je6.10648$AH6.1755149@newsc.telia.net>

Victor Muslin wrote:
> What is a pythonic idiom for forcing an immediate program termination?

not sure there is one, but if you insist, you
can use:

    # terminate without cleaning up
    os._exit(errcode)

or

    # force coredump
    os.abort()

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From ben.hutchings at roundpoint.com  Thu Feb  1 22:41:31 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 01 Feb 2001 13:41:31 -0800
Subject: sys.exit and exceptions (newbie question)
References: <3a7998cf.869169328@localhost>
Message-ID: <uu26eaxo4.fsf@roundpoint.com>

victor at prodigy.net (Victor Muslin) writes:

> What is a pythonic idiom for forcing an immediate program termination?
> sys.exit() does not do it, because it throws a SystemExit exception,
> so if it is called within a try block, a generic except block is still
> executed.

I'm not sure that there is a Pythonic idiom for forcing immediate
termination, since Python (a) has a pretty good exception system and
(b) does not allow you to get into the sort of mess that you can do
with C, where a clean exit becomes impossible.

An except block that catches all exceptions should always re-raise the
exceptioneal with.  If it fails to re-raise
KeyboardInterrupt, for example, the program won't be interruptible by
the user.  In the more general case, failing to re-raise exceptions
can be a good way to hide bugs.

You may find the os.abort() or the os._exit() function useful.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From thomas at xs4all.net  Fri Feb  2 00:25:51 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 2 Feb 2001 00:25:51 +0100
Subject: sys.exit and exceptions (newbie question)
In-Reply-To: <3a7998cf.869169328@localhost>; from victor@prodigy.net on Thu, Feb 01, 2001 at 05:18:38PM +0000
References: <3a7998cf.869169328@localhost>
Message-ID: <20010202002551.S962@xs4all.nl>

On Thu, Feb 01, 2001 at 05:18:38PM +0000, Victor Muslin wrote:

> What is a pythonic idiom for forcing an immediate program termination?
> sys.exit() does not do it, because it throws a SystemExit exception,
> so if it is called within a try block, a generic except block is still
> executed. 

You can use sys._exit(). This exits in a very abortive way -- right away,
dropping everything it's doing, without doing any kind of cleanup. In
general it's better to avoid unqualified except: clauses, and avoid code
that uses it (unless, of course, they reraise the exception if they can't
handle it.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From moshez at zadka.site.co.il  Fri Feb  2 14:01:29 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri,  2 Feb 2001 15:01:29 +0200 (IST)
Subject: sys.exit and exceptions (newbie question)
In-Reply-To: <20010202002551.S962@xs4all.nl>
References: <20010202002551.S962@xs4all.nl>, <3a7998cf.869169328@localhost>
Message-ID: <20010202130129.7059DA840@darjeeling.zadka.site.co.il>

On Fri, 2 Feb 2001 00:25:51 +0100, Thomas Wouters <thomas at xs4all.net> wrote:

> You can use sys._exit(). 

No you can't -- but you can use os._exit().
Everything else Thomas said is true...
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From thomas at xs4all.net  Fri Feb  2 15:03:00 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 2 Feb 2001 15:03:00 +0100
Subject: sys.exit and exceptions (newbie question)
In-Reply-To: <20010202130129.7059DA840@darjeeling.zadka.site.co.il>; from moshez@zadka.site.co.il on Fri, Feb 02, 2001 at 03:01:29PM +0200
References: <20010202002551.S962@xs4all.nl>, <3a7998cf.869169328@localhost> <20010202002551.S962@xs4all.nl> <20010202130129.7059DA840@darjeeling.zadka.site.co.il>
Message-ID: <20010202150300.V962@xs4all.nl>

On Fri, Feb 02, 2001 at 03:01:29PM +0200, Moshe Zadka wrote:
> On Fri, 2 Feb 2001 00:25:51 +0100, Thomas Wouters <thomas at xs4all.net> wrote:

> > You can use sys._exit(). 

> No you can't -- but you can use os._exit().
> Everything else Thomas said is true...

D'oh! I really, really need vacation. I've made *four* errors this week! And
one of them was even a sysadmin error (added an emailalias in the wrong
file, so it didn't work) -- that hasn't happened in years... I think I may
be going senile ;P

Maybe-I-should-just-move-to-management-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From max at alcyone.com  Fri Feb  2 18:46:08 2001
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 02 Feb 2001 09:46:08 -0800
Subject: sys.exit and exceptions (newbie question)
References: <20010202002551.S962@xs4all.nl>, <3a7998cf.869169328@localhost> <20010202002551.S962@xs4all.nl> <20010202130129.7059DA840@darjeeling.zadka.site.co.il> <mailman.981122703.28106.python-list@python.org>
Message-ID: <3A7AF260.250FAC03@alcyone.com>

Thomas Wouters wrote:

> D'oh! I really, really need vacation. I've made *four* errors this
> week! And
> one of them was even a sysadmin error (added an emailalias in the
> wrong
> file, so it didn't work) -- that hasn't happened in years... I think I
> may
> be going senile ;P

So you don't feel so bad, I remembered it as being sys._exit too.  :-)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Conversation is the enemy of good wine and food.
\__/ Alfred Hitchcock
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From fredrik at effbot.org  Fri Feb  2 09:10:25 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 08:10:25 GMT
Subject: sys.exit and exceptions (newbie question)
References: <3a7998cf.869169328@localhost> <mailman.981087063.16577.python-list@python.org>
Message-ID: <RXte6.11068$AH6.1772830@newsc.telia.net>

Thomas Wouters wrote:
> You can use sys._exit(). This exits in a very abortive way -- right away,
> dropping everything it's doing, without doing any kind of cleanup.

only if you're not catching AttributeError, that is ;-)

>>> import sys
>>> sys._exit()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: _exit

(it's "os._exit(code)", not "sys._exit()")

Cheers /F




From mj at muc.das-werk.de  Thu Feb  1 18:50:33 2001
From: mj at muc.das-werk.de (Michael Jonas)
Date: Thu, 1 Feb 2001 18:50:33 +0100
Subject: Win32 CreateProcess with In/out Redirection and a Way to kill the job?
Message-ID: <NDBBICIKFJMPFCCMBLNIGEGJCDAA.mj@muc.das-werk.de>

Seems like I need people with win32 geek factor....

What I want:
I'd like to code a extension module with a function popen3 (same as in os)
and a function is_alive() and terminate().

What I allready have:
a module which has an popen3 method which does exactly the same like the
posixmodule popen3 except that I extracted it to get hold of the process
Handle (from the normally inaccessible dict). The redirection works fine, I
can check if the job is still alive with my new is_alive(<file>). But I'm
really struggling to terminate the job.

Where my proble is:
The CreateProcess in the popen3 not only starts the job specified, it first
starts a cmd.exe which then starts the job. Ergo: the process handle I save
is the handle of the console app. The started app does not care at all if I
do a TerminateProcess on the console. The cmd.exe dies and the started job
stays alive. I tried to find a way to get hold of the Window Handle of the
console, but did not find a way. (The doughter process seems to die if I
close the Console Window. So I thought I could send a WM_Close to the
Window....) Nothing really worked till here...


Any idea how to implement something like this on Win32?!?


Thanks Michel

PS. Please cc the mail also to mj at muc.das-werk.de, cause I'm having trouble
getting the mail from the list.

-----------------------------------------
Michael Jonas
Head of R&D
Das Werk Digitale Bildverarbeitungs GmbH
mail:  mj at muc.das-werk.de
phone: +49(89)368148-700
fax:   +49(89)368148-111
-----------------------------------------




From MarkH at ActiveState.com  Fri Feb  2 00:23:31 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 01 Feb 2001 23:23:31 GMT
Subject: Win32 CreateProcess with In/out Redirection and a Way to kill the job?
References: <mailman.981064983.9908.python-list@python.org>
Message-ID: <3A79EF02.6040104@ActiveState.com>

Michael Jonas wrote:


> Any idea how to implement something like this on Win32?!?

Try _not_ using "cmd.exe" as an intermediate shim.  You should be able to execute the process directory.  popen does it this way for maximum compatibility.

Mark.



From vasvir at iit.demokritos.gr  Sat Feb  3 15:13:24 2001
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Sat, 03 Feb 2001 16:13:24 +0200
Subject: Win32 CreateProcess with In/out Redirection and a Way to kill the 
 job?
References: <mailman.981064983.9908.python-list@python.org>
Message-ID: <3A7C1204.A226B21E@iit.demokritos.gr>

Michael Jonas wrote:
> 
> Seems like I need people with win32 geek factor....
> 
> What I want:
> I'd like to code a extension module with a function popen3 (same as in os)
> and a function is_alive() and terminate().
> 
> What I allready have:
> a module which has an popen3 method which does exactly the same like the
> posixmodule popen3 except that I extracted it to get hold of the process
> Handle (from the normally inaccessible dict). The redirection works fine, I
> can check if the job is still alive with my new is_alive(<file>). But I'm
> really struggling to terminate the job.
> 
> Where my proble is:
> The CreateProcess in the popen3 not only starts the job specified, it first
> starts a cmd.exe which then starts the job. Ergo: the process handle I save
> is the handle of the console app. The started app does not care at all if I
> do a TerminateProcess on the console. The cmd.exe dies and the started job
> stays alive. I tried to find a way to get hold of the Window Handle of the
> console, but did not find a way. (The doughter process seems to die if I
> close the Console Window. So I thought I could send a WM_Close to the
> Window....) Nothing really worked till here...
> 
> Any idea how to implement something like this on Win32?!?
> 
> Thanks Michel
> 
> PS. Please cc the mail also to mj at muc.das-werk.de, cause I'm having trouble
> getting the mail from the list.

I don't know how exactly the popen* functions are implemented but I
think it's better to use CreateProcess/DuplicateHandle to mimic the
popen* behavior in windows and fork/dup in UNIX.

PS
In UNIX a libc popen invocation fires a shell also (bin/sh).

	.Bill


From tmealy at my-deja.com  Thu Feb  1 18:57:03 2001
From: tmealy at my-deja.com (tmealy at my-deja.com)
Date: Thu, 01 Feb 2001 17:57:03 GMT
Subject: Confused re: popen()
Message-ID: <95c81a$vf$1@nnrp1.deja.com>

I am having one heck of a time figuring out how to run a dos program
interactively.

I want to:
1.  Issue a "CVS login" command, which when run in a dos responds with
a password prompt.
2.  Enter the password.
3.  Issue a "CVS update" command.

Can anyone give me an example of how to do this?  Or is there a package
out there that can hand CVS?

I am running ActivePython 2 on a Win2000 box.


Sent via Deja.com
http://www.deja.com/


From trace at reinventnow.com  Thu Feb  1 19:52:23 2001
From: trace at reinventnow.com (trace s. ruggles)
Date: Thu, 01 Feb 2001 12:52:23 -0600
Subject: (MySQLdb and FreeBSD) on a VPS?
Message-ID: <B69F0C87.6D9%trace@reinventnow.com>

Hi all,

Has anyone had any luck installing MySQLdb while running under FreeBSD on a
VPS (virtual private server)?  (specifically, FreeBSD version 4.1)

The distutils.py module can't seem to find the right directories to do the
install because of the oddities of the VPS.  It's looking for directories
like "/usr/local/lib/..." while it should be looking for
"/usr/home/axiomfire/usr/local/lib/..." (the "/usr/home/aximofir" is my
specific directory on the VPS).

I've tried to go in and muck around with the default directory settings in
the sysconfig.py module within distutils, but I just seem to be making
things worse...

Thanks,
Trace


-- 
Trace S. Ruggles => 512/845.9354 => tracer at axiomfire.com
"Ignite your marketing with interactive precision."
AxiomFire, Inc. => http://www.axiomfire.com



From h4rv3st.m00n at gmx.net  Thu Feb  1 20:00:50 2001
From: h4rv3st.m00n at gmx.net (Harvest T. Moon)
Date: Thu, 1 Feb 2001 20:00:50 +0100
Subject: Q: elements in a list [newbie]
Message-ID: <95cbfs$gu0iq$1@ID-22517.news.dfncis.de>

i hope this is not some kind of 'very stupid' question, how can i find out
how many elements are in a list?
(generated by file.readlines())

Harvest T. Moon




From aahz at panix.com  Thu Feb  1 20:14:28 2001
From: aahz at panix.com (Aahz Maruch)
Date: 1 Feb 2001 11:14:28 -0800
Subject: Q: elements in a list [newbie]
References: <95cbfs$gu0iq$1@ID-22517.news.dfncis.de>
Message-ID: <95ccik$t4o$1@panix2.panix.com>

In article <95cbfs$gu0iq$1 at ID-22517.news.dfncis.de>,
Harvest T. Moon <h4rv3st.m00n at gmx.net> wrote:
>
>i hope this is not some kind of 'very stupid' question, how can i find out
>how many elements are in a list?
>(generated by file.readlines())

len()

See the section on built-in functions in the Python docs.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From gregj at pobox.com  Thu Feb  1 20:23:05 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 01 Feb 2001 19:23:05 GMT
Subject: Q: elements in a list [newbie]
References: <95cbfs$gu0iq$1@ID-22517.news.dfncis.de>
Message-ID: <95cd2b$5uk$1@nnrp1.deja.com>

len(list)

In article <95cbfs$gu0iq$1 at ID-22517.news.dfncis.de>,
  "Harvest T. Moon" <h4rv3st.m00n at gmx.net> wrote:
> i hope this is not some kind of 'very stupid' question, how can i
find out
> how many elements are in a list?
> (generated by file.readlines())
>
> Harvest T. Moon
>
>

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From Bill.Scherer at VerizonWireless.com  Thu Feb  1 23:57:58 2001
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Thu, 01 Feb 2001 17:57:58 -0500
Subject: Q: elements in a list [newbie]
References: <95cbfs$gu0iq$1@ID-22517.news.dfncis.de>
Message-ID: <3A79E9F6.C70518B6@VerizonWireless.com>

len, as in:

l = file.readlines()
length = len(l)




"Harvest T. Moon" wrote:
> 
> i hope this is not some kind of 'very stupid' question, how can i find out
> how many elements are in a list?
> (generated by file.readlines())
> 
> Harvest T. Moon
> 
> --
> http://mail.python.org/mailman/listinfo/python-list

-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From chris.gonnerman at usa.net  Fri Feb  2 05:54:55 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 1 Feb 2001 22:54:55 -0600
Subject: elements in a list [newbie]
References: <95cbfs$gu0iq$1@ID-22517.news.dfncis.de>
Message-ID: <018301c08cd4$97b74660$a100000a@local>

----- Original Message -----
From: "Harvest T. Moon" <h4rv3st.m00n at gmx.net>
Subject: Q: elements in a list [newbie]


> i hope this is not some kind of 'very stupid' question, how can i find out
> how many elements are in a list?
> (generated by file.readlines())

f = file.readlines()
l = len(f)

Thar ya go.





From binnc at my-deja.com  Thu Feb  1 20:02:01 2001
From: binnc at my-deja.com (binnc at my-deja.com)
Date: Thu, 01 Feb 2001 19:02:01 GMT
Subject: Subroutine Equivalent in Python (newbie Q)
Message-ID: <95cbr2$4o2$1@nnrp1.deja.com>

Hello,

Another question from someone totally new to
Python. Is there an equivalent in Python to a sub-
routine, (e.g. gosub and return).  I want to
create a modular program with sub-routines to
perform distinct tasks wihin the program for
organizational and debugging purposes, etc.  Is
the only (or best) way to do this is with
modules?  A function works but the values
obtained within the function do not appear to be
valid outside of that function.  I guess I am
looking for the best approach to create the
subroutines for execution from the main flow of
the program.

Thanks for any assistance!


Sent via Deja.com
http://www.deja.com/


From aahz at panix.com  Thu Feb  1 20:16:26 2001
From: aahz at panix.com (Aahz Maruch)
Date: 1 Feb 2001 11:16:26 -0800
Subject: Subroutine Equivalent in Python (newbie Q)
References: <95cbr2$4o2$1@nnrp1.deja.com>
Message-ID: <95ccma$7d$1@panix2.panix.com>

In article <95cbr2$4o2$1 at nnrp1.deja.com>,  <binnc at my-deja.com> wrote:
>
>Another question from someone totally new to
>Python. Is there an equivalent in Python to a sub-
>routine, (e.g. gosub and return).  

I suggest you read the Python tutorial; it will answer that question and
many others.  See http://www.python.org/doc/current/tut/tut.html
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From gregj at pobox.com  Thu Feb  1 20:24:55 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 01 Feb 2001 19:24:55 GMT
Subject: Subroutine Equivalent in Python (newbie Q)
References: <95cbr2$4o2$1@nnrp1.deja.com>
Message-ID: <95cd5q$66u$1@nnrp1.deja.com>

Python has functions, modules, and packages, which support modular
programming at various levels. Python also has classes, which support
object-oriented programming. I strongly urge you to read the excellent
Python tutorial at www.python.org.


In article <95cbr2$4o2$1 at nnrp1.deja.com>,
  binnc at my-deja.com wrote:
> Hello,
>
> Another question from someone totally new to
> Python. Is there an equivalent in Python to a sub-
> routine, (e.g. gosub and return).  I want to
> create a modular program with sub-routines to
> perform distinct tasks wihin the program for
> organizational and debugging purposes, etc.  Is
> the only (or best) way to do this is with
> modules?  A function works but the values
> obtained within the function do not appear to be
> valid outside of that function.  I guess I am
> looking for the best approach to create the
> subroutines for execution from the main flow of
> the program.
>
> Thanks for any assistance!
>
> Sent via Deja.com
> http://www.deja.com/
>

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From cfelling at iae.nl  Fri Feb  2 05:44:26 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 2 Feb 2001 05:44:26 +0100
Subject: Subroutine Equivalent in Python (newbie Q)
References: <95cbr2$4o2$1@nnrp1.deja.com>
Message-ID: <95ddva$mq1$1@animus.fel.iae.nl>

binnc at my-deja.com wrote:
> Hello,

> Another question from someone totally new to
> Python. Is there an equivalent in Python to a sub-
> routine, (e.g. gosub and return).  I want to
> create a modular program with sub-routines to
> perform distinct tasks wihin the program for
> organizational and debugging purposes, etc.  Is
> the only (or best) way to do this is with
> modules?  A function works but the values
> obtained within the function do not appear to be
> valid outside of that function.  I guess I am

Ai, I fear your BASIC-abuse is shining through here:) IIRQ from the
one week I had to work with BASIC, it offered the most basic form of
subroutine calls possible, no context saving, no new var scope, only
a return to were you came from.  So basicly there was but one scope in
such a program, comparable to the global scope of Python.  In almost
any other programming language I know of a subroutine creates a new
namespace/scope.  The joy of this is that it prevents nasty, hard to
find nameclashes; ie in sub A you choose i for a local var, you go off
to sub B, sub C etc, and somewhere in that line som other sub also
choose to use i as a local var, but helas it isn't a local var, so
the program logic of sub A might break in unpredictable ways.

> looking for the best approach to create the
> subroutines for execution from the main flow of
> the program.

There are several tools available to lumb code together to perform
reoccuring tasks.  The most basic one is the function, like:

>>> def fun(args):
>>>     return result

In its pure form it performs some calculation on its args and returns
a result.  But Python is not a pure functional language:) so other
things are possible here to, like:

>>> a = "spam"
>>> def more_fun(args):
>>>     print a, args     # a is automatically looked up in the outer scope
... 
>>> more_fun("and spam")
spam and spam

And for yet some more trickery:

>>> a = "spam, spam"
>>> def more_fun(args):
>>>     global a         # this is needed to be able to assign to the global a
>>>     a = a + ', ' + args
... 
>>> more_fun("spam and spam")
>>> print a
spam, spam, spam and spam
>>> more_fun("spam and spam")
>>> print a
spam, spam, spam and spam, spam and spam

The trick is here that the global keyword tells Python to look a up in
the global scope and not in the local function scope.  But global scope
get's hairy quickly, and func within func still only know of two scopes,
their own and the module-global scope.  There are better ways of doing
things, like:

>>> class C:
>>>     a = 'spam'
>>>     def __init__(self, s='more spam'):
>>>         self.s = self.a + 'and ' + s
>>>     def prt(self):
>>>         print self.s
...
>>> c = C()
>>> cc = C('spam and eggs')
>>> c.prt()
spam and more spam
>>> cc.prt()
spam and spam and eggs
>>> cc.s = 'no more spam'
>>> cc.prt()
no more spam
>>> c.prt()
spam and more spam


So a class is ideally suited to lumb together some code (several funcs
in a single class) and some data to go along with that code.  Read all
about it in the FIne Manuals
-- 
groetjes, carel


From rtrocca at libero.it  Thu Feb  1 20:12:02 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Thu, 01 Feb 2001 19:12:02 GMT
Subject: wxPython on LinuxPPC
Message-ID: <1eo5bsr.3fhlxt1f6ehcaN%rtrocca@libero.it>

Hello, I've been trying to make wxPython compile on LinuxPPC, but I've
failed (evenif wxWindow compiled well).
Has somebody done it?
I've been using python 2.0 compiled by myself.

Riccardo


From robin at alldunn.com  Fri Feb  2 04:54:12 2001
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 1 Feb 2001 19:54:12 -0800
Subject: wxPython on LinuxPPC
References: <1eo5bsr.3fhlxt1f6ehcaN%rtrocca@libero.it>
Message-ID: <acqe6.54$wa5.19398793@news.randori.com>

"Riccardo Trocca" <rtrocca at libero.it> wrote in message news:1eo5bsr.3fhlxt1f6ehcaN%rtrocca at libero.it...
> Hello, I've been trying to make wxPython compile on LinuxPPC, but I've
> failed (evenif wxWindow compiled well).
> Has somebody done it?
> I've been using python 2.0 compiled by myself.
> 

I don't have any details, but I know it's been done by at least a couple people.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!




From bernhard at abnoba.intevation.de  Mon Feb  5 14:48:27 2001
From: bernhard at abnoba.intevation.de (Bernhard Reiter)
Date: 5 Feb 2001 13:48:27 GMT
Subject: wxPython on LinuxPPC
References: <1eo5bsr.3fhlxt1f6ehcaN%rtrocca@libero.it>
Message-ID: <95mavb$lsj$2@newsserver.rrzn.uni-hannover.de>

In article <1eo5bsr.3fhlxt1f6ehcaN%rtrocca at libero.it>,
	rtrocca at libero.it (Riccardo Trocca) writes:
> Hello, I've been trying to make wxPython compile on LinuxPPC, but I've
> failed (evenif wxWindow compiled well).
> Has somebody done it?

Yes. 
It works fine here, there were no problems.
Make sure that you have exactly the same version of wxGTK and wxPython.
	Bernhard

> I've been using python 2.0 compiled by myself.

-- 
Professional Service around Free Software                (intevation.net)  
The FreeGIS Project                                         (freegis.org)
Association for a Free Informational Infrastructure            (ffii.org)
FSF Europe                                            	  (fsfeurope.org)


From ccosse at psl.nmsu.edu  Thu Feb  1 20:54:49 2001
From: ccosse at psl.nmsu.edu (charles cosse)
Date: Thu, 01 Feb 2001 12:54:49 -0700
Subject: how to run script in windows via C system call w/o popup dos shell
Message-ID: <3A79BF08.5C4018C7@psl.nmsu.edu>

Hello,
just spent too much time trying to get NT not to pop up a window
everytime
my C function made a "system" call  to run a python script.  I think the

solution is merely to rename the script ".pyw" rather than ".py".
This info was found in David Beasley's (excellent) book: Python
Essential
Reference, p. 85. It sure beats creating a new "NT Service" -- which
is what i thought was gonna have to do. Hope this helps someone,
someday.

Charlie



From bobhicks at adelphia.net  Thu Feb  1 21:08:11 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Thu, 01 Feb 2001 20:08:11 GMT
Subject: MacPython
Message-ID: <B69F2C58.9C72%bobhicks@adelphia.net>

Is there a MacPython tutorial anywhere?

<bh />



From shwang5 at students.uiuc.edu  Thu Feb  1 21:27:49 2001
From: shwang5 at students.uiuc.edu (seung-won hwang)
Date: Thu, 1 Feb 2001 14:27:49 -0600
Subject: Python CGI programming
Message-ID: <Pine.GSO.4.10.10102011423280.2324-100000@ux13.cso.uiuc.edu>

Hello, I tried a python CGI programming example in 'learning python'
and the msg as follows says that the address is incorrect though
I specified the address of .py file correctly. Any help would be
greatly appreciated. Thanks! (I attached my .html file just in case,
after the error mesg)


---
The page cannot be displayed 
The page you are looking for cannot be displayed because the address is
incorrect.  

--------------------------------------------------------------------------------

Please try the following:

If you typed the page address in the Address bar, check that it is entered
correctly.

Open the www-sal.cs.uiuc.edu home page and then look for links to the
information you want. 
Click  Search to look for information on the Internet. 



HTTP 405 - Resource not allowed
Internet Explorer  

--
<FORM METHOD=POST
ACTION="http://www-sal.cs.uiuc.edu/~shwang5/feedback.py">
<UL><I>Please fill out the entire form:</I></UL>
<CENTER><TABLE WIDTH="100%" >
<TR><TD ALIGN=RIGHT WIDTH="20%">Name:</TD>
    <TD><INPUT TYPE=text NAME=name SIZE=50 VALUE=""></TD></TR>
<TR><TD ALIGN=RIGHT>Email Address:</TD>
    <TD><INPUT TYPE=text NAME=email SIZE=50 VALUE=""></TD></TR>
<TR><TD ALIGN=RIGHT>Mailing Address:</TD>
    <TD><INPUT TYPE=text NAME=address SIZE=50 VALUE=""></TD></TR>
<TR><TD ALIGN=RIGHT>Type of Message:</TD>
<TD><INPUT TYPE=radio NAME=type CHECKED VALUE=comment>comment&nbsp;
    <INPUT TYPE=radio NAME=type VALUE=complaint>complaint</TD></TR>
<TR><TD ALIGN=RIGHT VALIGN=TOP>Enter the text in here:</TD>
    <TD><TEXTAREA NAME=text ROWS=5, COLS=50 VALUE="">
        </TEXTAREA></TD></TR>
<TR><TD></TD>
<TD><INPUT type=submit name=send value="Send the feedback!"></TD></TR>
</TABLE></CENTER>
</FORM>



From shwang5 at students.uiuc.edu  Thu Feb  1 22:26:05 2001
From: shwang5 at students.uiuc.edu (seung-won hwang)
Date: Thu, 1 Feb 2001 15:26:05 -0600
Subject: Python CGI programming
In-Reply-To: <Pine.GSO.4.10.10102011423280.2324-100000@ux13.cso.uiuc.edu>
References: <Pine.GSO.4.10.10102011423280.2324-100000@ux13.cso.uiuc.edu>
Message-ID: <Pine.GSO.4.10.10102011525320.23558-100000@ux13.cso.uiuc.edu>

FYI, when I type the following example, web browser just shows
the source code..not the test string... How can I fix this?

#!/usr/bin/python

# Tell the browser how to render the text
print "Content-Type: text/plain\n\n"

print "Hello, Python!"  # print a test string





From gerhard.nospam at bigfoot.de  Thu Feb  1 23:36:57 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Thu, 01 Feb 2001 23:36:57 +0100
Subject: Python CGI programming
References: <Pine.GSO.4.10.10102011423280.2324-100000@ux13.cso.uiuc.edu> <Pine.GSO.4.10.10102011525320.23558-100000@ux13.cso.uiuc.edu>
Message-ID: <3A79E509.31F99F12@bigfoot.de>

Your Python code is ok. Assuming the Python interpreter really is at
/usr/bin/python. If not, you must adjust the first line of the script to
point to the location of the python interpreter. Another common place
would be /usr/local/bin/python.

First you should find out if the webserver supports user CGI scripts at
all. If it does, these must normally be put in a directory named
cgi-bin, like ~/home_html/cgi-bin. The script must also have the
executable bit set (assuming you are on Unix, of course).

chmod +x feedback.py    sets the executable bit of your script.

Gerhard

seung-won hwang wrote:
> 
> FYI, when I type the following example, web browser just shows
> the source code..not the test string... How can I fix this?
> 
> #!/usr/bin/python
> 
> # Tell the browser how to render the text
> print "Content-Type: text/plain\n\n"
> 
> print "Hello, Python!"  # print a test string

-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From joconnor at cybermesa.com  Fri Feb  2 00:49:21 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 01 Feb 2001 23:49:21 +0000
Subject: Python CGI programming
References: <Pine.GSO.4.10.10102011423280.2324-100000@ux13.cso.uiuc.edu> <Pine.GSO.4.10.10102011525320.23558-100000@ux13.cso.uiuc.edu>
Message-ID: <3A79F601.424B7F3F@cybermesa.com>

seung-won hwang wrote:
> 
> FYI, when I type the following example, web browser just shows
> the source code..not the test string... How can I fix this?
> 
> #!/usr/bin/python
> 
> # Tell the browser how to render the text
> print "Content-Type: text/plain\n\n"
> 
> print "Hello, Python!"  # print a test string


Make sure that 

a) the script is in the cgi-bin directory for the web server.  If it's
not, the server will display the file rather than execute it
b) the script has execute permission set (although I think this raises a
500 error instead)

Take care,

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From jcm at bigskytel.com  Fri Feb  2 02:39:55 2001
From: jcm at bigskytel.com (David Porter)
Date: Thu, 1 Feb 2001 18:39:55 -0700
Subject: Python CGI programming
In-Reply-To: <Pine.GSO.4.10.10102011525320.23558-100000@ux13.cso.uiuc.edu>; from shwang5@students.uiuc.edu on Thu, Feb 01, 2001 at 03:26:05PM -0600
References: <Pine.GSO.4.10.10102011423280.2324-100000@ux13.cso.uiuc.edu> <Pine.GSO.4.10.10102011525320.23558-100000@ux13.cso.uiuc.edu>
Message-ID: <20010201183955.A7746@bigskytel.com>

* seung-won hwang <shwang5 at students.uiuc.edu>:
> 
> FYI, when I type the following example, web browser just shows
> the source code..not the test string... How can I fix this?

Make the file executable and put it in your web server's CGI bin, or some
other place where the web server will execute programs. 

David



From chris.gonnerman at usa.net  Fri Feb  2 06:01:39 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 1 Feb 2001 23:01:39 -0600
Subject: Python CGI programming
References: <Pine.GSO.4.10.10102011423280.2324-100000@ux13.cso.uiuc.edu> <Pine.GSO.4.10.10102011525320.23558-100000@ux13.cso.uiuc.edu>
Message-ID: <019901c08cd5$3baba040$a100000a@local>

----- Original Message -----
From: "seung-won hwang" <shwang5 at students.uiuc.edu>
Subject: Re: Python CGI programming


> FYI, when I type the following example, web browser just shows
> the source code..not the test string... How can I fix this?
>
> #!/usr/bin/python
>
> # Tell the browser how to render the text
> print "Content-Type: text/plain\n\n"
>
> print "Hello, Python!"  # print a test string

The webserver does not recognize your script as a cgi script.  Depending on
the webserver,
you may be required to put the script in a special subdirectory (cgi-bin or
scripts for instance), make registry or config file changes on the server
(unlikely to be allowed if you don't own the server yourself), etc.

The 405 error you are getting is because METHOD=POST is not allowed for
non-cgi documents by your webserver, and as I said, it doesn't know it is a
cgi script.

Can you tell us anything else about your problem?  (What host OS/webserver,
do you have root or administrative access, etc.)?





From burkhard.kayser at freenet.de  Thu Feb  1 21:37:29 2001
From: burkhard.kayser at freenet.de (Burkhard Kayser)
Date: Thu, 1 Feb 2001 21:37:29 +0100
Subject: A Telnet question
Message-ID: <95chjn$18n$02$1@news.t-online.com>

Hi,

I'm a newbie in python. My first application is based on telnetlib and I've
a problem with unreachable hosts.

When I try to connect to a host, wich is currently not reachable,
socket.connect(), wich is invoked by telnet.open(), times out after some
minutes. How can I change this timeout to some seconds ?
I tried it under SuSE 7.0.

Has onyone an idea ?

Thanks





From Bill.Scherer at VerizonWireless.com  Fri Feb  2 01:27:04 2001
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Thu, 01 Feb 2001 19:27:04 -0500
Subject: A Telnet question
References: <95chjn$18n$02$1@news.t-online.com>
Message-ID: <3A79FED8.F6A3C43C@VerizonWireless.com>

Get timeoutsocket from:

http://www.timo-tasi.org/python/timeoutsocket.py


Burkhard Kayser wrote:
> 
> Hi,
> 
> I'm a newbie in python. My first application is based on telnetlib and I've
> a problem with unreachable hosts.
> 
> When I try to connect to a host, wich is currently not reachable,
> socket.connect(), wich is invoked by telnet.open(), times out after some
> minutes. How can I change this timeout to some seconds ?
> I tried it under SuSE 7.0.
> 
> Has onyone an idea ?
> 
> Thanks
> 
> --
> http://mail.python.org/mailman/listinfo/python-list

-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From gregj at pobox.com  Fri Feb  2 01:54:34 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Fri, 02 Feb 2001 00:54:34 GMT
Subject: A Telnet question
References: <95chjn$18n$02$1@news.t-online.com>
Message-ID: <95d0g7$oql$1@nnrp1.deja.com>

timeout_socket is in the Vaults of Parnassus, here:

http://www.vex.net/parnassus/apyllo.py/126307487.528281449

and here:

http://www.cs.usyd.edu.au/~piers/python/

Other implementations have been discussed in this forum; search
for "socket timeout" to find them.


In article <95chjn$18n$02$1 at news.t-online.com>,
  "Burkhard Kayser" <burkhard.kayser at freenet.de> wrote:
> Hi,
>
> I'm a newbie in python. My first application is based on telnetlib
and I've
> a problem with unreachable hosts.
>
> When I try to connect to a host, wich is currently not reachable,
> socket.connect(), wich is invoked by telnet.open(), times out after
some
> minutes. How can I change this timeout to some seconds ?
> I tried it under SuSE 7.0.


--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From gerhard.nospam at bigfoot.de  Fri Feb  2 02:15:10 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 02 Feb 2001 02:15:10 +0100
Subject: A Telnet question
References: <95chjn$18n$02$1@news.t-online.com>
Message-ID: <3A7A0A1E.D5C7FDCA@bigfoot.de>

Burkhard Kayser wrote:
> 
> Hi,
> 
> I'm a newbie in python. My first application is based on telnetlib and I've
> a problem with unreachable hosts.
> 
> When I try to connect to a host, wich is currently not reachable,
> socket.connect(), wich is invoked by telnet.open(), times out after some
> minutes. How can I change this timeout to some seconds ?
> I tried it under SuSE 7.0.

Sounds like a typical application for the timeoutsocket module,
available at http://www.timo-tasi.org/python/timeoutsocket.py , which I
downloaded, but never used myself ;-) It's a wrapper around the standard
Python socket module with the additional possibility of setting a
timeout. So it might be just what you need.

Gerhard
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From joconnor at cybermesa.com  Thu Feb  1 22:18:03 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 01 Feb 2001 21:18:03 +0000
Subject: Python/C++ interface
Message-ID: <3A79D28B.1BEB46A0@cybermesa.com>

All,

I was doing some testing with the Python->C++ interface and ran into
something odd.

The example shows calling a C function from python with 

	import spam
	spam.system ("ls -l")

and on the C++ side... the signature reads
	PyObject* spam_system(PyObject *self, PyObject * args) {
		...
	}

Now.  On the C++ side, args is considered, and parsed, as a Tuple. 
However, what I found was that a Tuple was not being passed, just a
String.  The only time a Tuple was passed was if there was more than one
argument.  The whole argument list would be converted to a Tuple, then.

Is this an error in the documentation?  Or has the passing method
changed between 1.5.x and 2.0 (I'm using 2.0) and the documentation is
just out of date?

Also,  is there a good way of printing the name of the type of a
PyObject?

Thanks
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From mwh21 at cam.ac.uk  Thu Feb  1 22:23:43 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Feb 2001 21:23:43 +0000
Subject: Python/C++ interface
References: <3A79D28B.1BEB46A0@cybermesa.com>
Message-ID: <m33ddy6qsg.fsf@atrus.jesus.cam.ac.uk>

Jay O'Connor <joconnor at cybermesa.com> writes:

> Now.  On the C++ side, args is considered, and parsed, as a Tuple. 
> However, what I found was that a Tuple was not being passed, just a
> String.  The only time a Tuple was passed was if there was more than one
> argument.  The whole argument list would be converted to a Tuple, then.

You missed a METH_VARARGS somewhere.

> Is this an error in the documentation?  Or has the passing method
> changed between 1.5.x and 2.0 (I'm using 2.0) and the documentation is
> just out of date?
> 
> Also,  is there a good way of printing the name of the type of a
> PyObject?

Something very like printf(ob->ob_type->tp_name).
Or print type(object).name in Python.

Cheers,
M.

-- 
  For their next act, they'll no doubt be buying a firewall running
  under NT, which makes about as much sense as building a prison out
  of meringue.                                           -- -:Tanuki:-
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From joconnor at cybermesa.com  Thu Feb  1 22:28:23 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Thu, 01 Feb 2001 21:28:23 +0000
Subject: Python/C++ interface
References: <3A79D28B.1BEB46A0@cybermesa.com> <m33ddy6qsg.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3A79D4F7.CF63C817@cybermesa.com>

Michael Hudson wrote:

> You missed a METH_VARARGS somewhere.

Ah, yes...I see it now...in my static declaration of the exported method
list

> > Also,  is there a good way of printing the name of the type of a
> > PyObject?
> 
> Something very like printf(ob->ob_type->tp_name).


Doh!  I was so close to that a few times :)

Thanks

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From loewis at informatik.hu-berlin.de  Fri Feb  2 22:59:35 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 22:59:35 +0100
Subject: Python/C++ interface
References: <3A79D28B.1BEB46A0@cybermesa.com> <m33ddy6qsg.fsf@atrus.jesus.cam.ac.uk> <3A79D4F7.CF63C817@cybermesa.com>
Message-ID: <p6qitmsoiew.fsf@informatik.hu-berlin.de>

Jay O'Connor <joconnor at cybermesa.com> writes:

> > Something very like printf(ob->ob_type->tp_name).
> 
> 
> Doh!  I was so close to that a few times :)

Or, call _PyObject_Dump(ob). Works nicely from gdb as well (I wish
Misc/gdbinit had been around much longer - I only learned about
_PyObject_Dump from that, either).

Regards,
Martin



From idfx at my-deja.com  Fri Feb  9 20:29:14 2001
From: idfx at my-deja.com (idfx)
Date: Fri, 09 Feb 2001 19:29:14 GMT
Subject: Python/C++ interface
References: <3A79D28B.1BEB46A0@cybermesa.com>
Message-ID: <961gdt$fm4$1@nnrp1.deja.com>

In article <3A79D28B.1BEB46A0 at cybermesa.com>,
  Jay O'Connor <joconnor at cybermesa.com> wrote:
> All,
>
> I was doing some testing with the Python->C++ interface and ran into
> something odd.
>
> The example shows calling a C function from python with
>
> 	import spam
> 	spam.system ("ls -l")
>
> and on the C++ side... the signature reads
> 	PyObject* spam_system(PyObject *self, PyObject * args) {
> 		...
> 	}
>
> Now.  On the C++ side, args is considered, and parsed, as a Tuple.
> However, what I found was that a Tuple was not being passed, just a
> String.  The only time a Tuple was passed was if there was more than
one
> argument.  The whole argument list would be converted to a Tuple,
then.
>
> Is this an error in the documentation?  Or has the passing method
> changed between 1.5.x and 2.0 (I'm using 2.0) and the documentation is
> just out of date?
>
> Also,  is there a good way of printing the name of the type of a
> PyObject?

     I may be wrong, or over-simplifying, but I think you've gotten
confused by the terms, understandably.  Yes, args is read into a tuple,
always, but if there's only one argument, then it becomes a tuple of
length 1.

ie:    tpl = ("yourstring",)

that way, you can refer to as many arguments as there are,

( numargs=len(tpl) )

instead of worrying about special cases for 1 or 0 arguments.
Your args is still a string, it's just that it's a string that happens
to be an element of a tuple.


   OTOH, I may be totally mis-reading your problem.  The other
possibility is that, for a system call, (e.g. 'ls -l'), C++ makes a
string of the argument(s) to preserve the command you are passing to
the system, which should be a string anyway, or so I'm told.


    As for the printing of type names, this is something I had trouble
with, too.  Eventually, I came up with this, although I'm sure there
are more elegant ways to address the problem:

def printype( obj ):
	naym = `type(obj)`       #    string it,
	i = naym.find("'")
	naym = naym[(i+1):]      #    trim it,
	i = naym.find("'")
	naym = naym[:i]
	return naym              #    return it!   (as a string)

    Hey, what can I say? It works. Usually.    ; )

Hope that helps.

--
yrs,
in_sanity,

idfx


Sent via Deja.com
http://www.deja.com/


From drewpc at colorado.edu  Thu Feb  1 22:31:13 2001
From: drewpc at colorado.edu (drewpc at colorado.edu)
Date: Thu, 01 Feb 2001 21:31:13 GMT
Subject: interfaces?
Message-ID: <95ckj2$dlc$1@nnrp1.deja.com>

Let me preface this posting by saying that I'm not a fan of Java.  A lot
of people may argue with me, but that's not the point.  The point is
that I think that, one of the only useful features of Java is the idea
of an interface.

If you know what a Java interface is, skip this paragraph.  For those of
you that don't know, here's how I understand it.  Imagine you're writing
an API for a database.  You want every database module (MySQL,
PostgreSQL, Oracle, etc) to implement the connect(), query(), insert(),
update(), and drop() methods.  What if you could create a pseudo class
that all of your database modules would "implement".  Meaning that the
pseudo class would define the methods (and variables too, I think)
connect(), query(), insert(), update(), and drop() methods.  It wouldn't
implement them at all, but just define them.  Then, inside each module
you'd "implement" (which would be a little different then "inherit") the
interface that you defined above.  All that means is that the compiler
(or interpreter) checks to make sure the the class that implements the
interface does in fact use the functions that were defined in the
interface.

Now, my question.  Is this something that people would be interested in
seeing in Python?  If so, then I think I'll write up a PEP (Python
Enhancement Proposal, http://python.sourceforge.net/peps/) about this.
Comments?  Cheap shots?  One-liners?


Sent via Deja.com
http://www.deja.com/


From cjc26 at nospam.cornell.edu  Fri Feb  2 03:18:07 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Fri, 02 Feb 2001 02:18:07 GMT
Subject: interfaces?
References: <95ckj2$dlc$1@nnrp1.deja.com>
Message-ID: <zNoe6.10481$o91.812818@typhoon.nyroc.rr.com>

* drewpc at colorado.edu <drewpc at colorado.edu> menulis:
| 
| Now, my question.  Is this something that people would be interested in
| seeing in Python?  If so, then I think I'll write up a PEP (Python
| Enhancement Proposal, http://python.sourceforge.net/peps/) about this.
| Comments?  Cheap shots?  One-liners?

Well, the usual way to indicate that a class implements, say, the file
interface, is to write a comment saying "this class implements the file
interface". ;)  You probably don't need anything more complicated than
that, since Python is dynamically typed, blah blah, etc. (where "blah
blah, etc." is meant to sum up the many discussions on c.l.py about this
very topic ;)  If you were really paranoid, you could always do
something like this:

def func(f):
    if not hasattr(f, "readlines"):
        raise NeedFileObjectError
    for line in f.readlines():
        # rest of function goes here

try:
    func(suspectedFileObject)
except NeedFileObjectError:
    print >>sys.stderr, "You must use a file object!"

Of course, most people would just leave out the hasattr() check, and let
func() raise an AttributeError if it's passed something which doesn't
act like a file object.

and-we-don't-need-to-declare-which-exceptions-our-functions-raise-
    either-<wink>-ly y'rs, Cliff


-- 
Cliff Crawford               http://www.people.cornell.edu/pages/cjc26/
                             print "Just another Python hacker"


From aleaxit at yahoo.com  Fri Feb  2 09:50:33 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 2 Feb 2001 09:50:33 +0100
Subject: interfaces?
References: <95ckj2$dlc$1@nnrp1.deja.com>
Message-ID: <95dscp02us3@news1.newsguy.com>

<drewpc at colorado.edu> wrote in message news:95ckj2$dlc$1 at nnrp1.deja.com...
> Let me preface this posting by saying that I'm not a fan of Java.  A lot
> of people may argue with me, but that's not the point.  The point is
> that I think that, one of the only useful features of Java is the idea
> of an interface.

It's a good idea, but COM did it better, before Java ever existed:-).

E.g., one think I love about COM interfaces vs Java's: a COM object
is *informed* when one of its interfaces is requested of it -- it
gets to 'intercept' the interface-request, if you will, and thus can
do whatever work it may need to prepare itself for "actually HAVING"
the interface it advertises -- acquire resources, etc; explicitness
of navigation between interfaces is just lovely -- as usual, "explicit
is better than implicit".

In Java, like in C++, the native protocol for "navigating between the
interfaces of a given object" is a cast (dynamic_cast, in C++), and
the object doesn't even get _informed_ that somebody's trying to cast
it in a such-and-such way.  So, when you need 'hooks' upon interface
navigation, you can't have them within the native protocol -- you
must redundantly check at each method-entry, or use other non-native
protocols whereby your "secondary interfaces" are requested by some
method-call.

It's a bit better in Python, since __getattr__ does provide a natural
place to hang a hook for interface-navigation -- whenever a method in
interface X is requested, and you don't yet have it in your dict, your
__getattr__ get called, and you can diagnose the situation, do whatever
you require to prepare to actually support interface X, and add X's
methods to your __dict__ (or switch your runtime class to add some
X implementation to its bases, etc) so your __getattr__ won't be
bothered again about them.  Still, it's somewhat dark magic in Python,
while it's clear as crystal in COM (if you understand COM's semantics
rules about identity &tc; Don Box's great "Essential COM" is highly
recommended to foster such understanding!-).


Anyway, "interfaces are a good concept" is a widely shared opinion,
among Pythonistas as well as among others -- including users of
other languages that don't model interfaces *explicitly*, but still
let you use them one way or another.  You could say of Python (and
Smalltalk, etc) that interfaces are there, just not 'bundled' --
each method-cum-signature (that you can conceptually group into an
interface) is on its own.  You could say of C++ (and Eiffel, etc)
that interfaces are there, just not named as such -- you can model
them, if you wish, by using special-cases of classes (with multiple
inheritance), and good C++/Eiffel designers do that a LOT -- see,
e.g., John Lakos' great book "Large Scale C++ Software Design"
(Addison-Wesley, 1996) -- what you and I and COM and Java call
"interfaces", Lakos calls "protocol classes", but, terminology
apart, the concept IS very much there, and it's developed solidly
and convincingly in the context of a splendid discussion of the
issues of large-scale software design (dependency management, &c).

(I can't mention Lakos without a nod to Robert Martin, who writes
about very similar concerns -- interfaces, components, dependency
managements -- in VERY clear ways, having a gift for wrapping up
the wisdom he imparts into easily digestible little morsels -- if
you can stomach reading a little bit of C++ code, about all of his
articles at http://www.objectmentor.com/publications/articlesByAuthor.html
are great reading, IMHO; Lakos' book is more organic and thorough,
but by that same token it requires a substantial investment of time
and energy to study it in depth and digest it as it deserves to
be -- Martin's articles, focusing on one concern at a time, may
prove much more palatable fare for most of us over-worked, chronically
short of time, short-attention-span, any-unbillable-minute-had-
BETTER-give-great-returns, typical software professionals).


Anyway, back to Python -- see http://www.prescod.net/pytypes/, and
http://www.zope.org/Wikis/Interfaces/FrontPage, for some of the
previous thoughts on interfaces in Pyhon.


Alex





From tshumway at transafari.com  Fri Feb  2 15:28:53 2001
From: tshumway at transafari.com (Terrel Shumway)
Date: Fri, 02 Feb 2001 06:28:53 -0800
Subject: interfaces?
References: <95ckj2$dlc$1@nnrp1.deja.com>
Message-ID: <3A7AC425.33EBD58F@transafari.com>

drewpc at colorado.edu wrote:

> that I think that, one of the only useful features of Java is the idea
> of an interface.
>

Interfaces are good, but they are only a first step.  In Java, you can
declare an interface, and you can claim that a class implements that
interface, but that is all the support that the language provides.

The question is: as a client, do you believe that claim, just because
it has the right method signatures?  The most effective way I have
seen to know this is to test it.

See http://junit.org/ and http://pyunit.sourceforge.net/  for starting
points.

BTW: Unit testing even works for assembly language. :-)

    if-it-is-worth-implementing-it-is-worth-testing-ly yours,
        Terrel





From m.faassen at vet.uu.nl  Sun Feb  4 02:21:06 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 4 Feb 2001 01:21:06 GMT
Subject: interfaces?
References: <95ckj2$dlc$1@nnrp1.deja.com>
Message-ID: <95iaq2$gbm$1@newshost.accu.uu.nl>

drewpc at colorado.edu wrote:
> Let me preface this posting by saying that I'm not a fan of Java.  A lot
> of people may argue with me, but that's not the point.  The point is
> that I think that, one of the only useful features of Java is the idea
> of an interface.

[snip]
> Now, my question.  Is this something that people would be interested in
> seeing in Python?  If so, then I think I'll write up a PEP (Python
> Enhancement Proposal, http://python.sourceforge.net/peps/) about this.
> Comments?  Cheap shots?  One-liners?

One-liner: http://www.zope.org/Members/michel/InterfacesPEP/PEP.txt

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From ctavares at develop.com  Sun Feb  4 11:38:24 2001
From: ctavares at develop.com (Chris Tavares)
Date: Sun, 04 Feb 2001 10:38:24 GMT
Subject: interfaces?
References: <95ckj2$dlc$1@nnrp1.deja.com>
Message-ID: <Aiaf6.702$lC2.151062@newsread1.prod.itd.earthlink.net>

<drewpc at colorado.edu> wrote in message news:95ckj2$dlc$1 at nnrp1.deja.com...
> Let me preface this posting by saying that I'm not a fan of Java.  A lot
> of people may argue with me, but that's not the point.  The point is
> that I think that, one of the only useful features of Java is the idea
> of an interface.
>

[... snip ...]

I think that the Smalltalk approach to interfaces would be more appropriate
for Python than the Java one - dynamic vs. static and all.

The equivalent code in Python would be:

class IDontDoAnything:
    def UselessMethod(self):
        return self.SubclassResponsibility()

And then you don't define SubclassResponsibility. That way, when you try and
call UselessMethod on a class that doesn't implement it, you'll get an
attribute error.

Static verification doesn't really belong in Python to my mind. I *like*
just being able to only implement the part of the protocol that I need.

-Chris





From syver at NOSPAMcyberwatcher.com  Sun Feb  4 21:50:07 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sun, 4 Feb 2001 21:50:07 +0100
Subject: interfaces?
References: <95ckj2$dlc$1@nnrp1.deja.com> <Aiaf6.702$lC2.151062@newsread1.prod.itd.earthlink.net>
Message-ID: <95kfeg$fpt$1@troll.powertech.no>

"Chris Tavares" <ctavares at develop.com> wrote > I think that the Smalltalk
approach to interfaces would be more appropriate
> for Python than the Java one - dynamic vs. static and all.
>
> The equivalent code in Python would be:
>
> class IDontDoAnything:
>     def UselessMethod(self):
>         return self.SubclassResponsibility()

Wouldn't you rather do it like this in Python:

class IDontDoAnything:
    def UselessMethod(self):
        raise NotImplementedError()




From fernstrom at geocities.com  Thu Feb  1 22:58:04 2001
From: fernstrom at geocities.com (fernstrom at geocities.com)
Date: Thu, 01 Feb 2001 21:58:04 GMT
Subject: Programmatic control of scroll position in PmwScrolledFrame
Message-ID: <95cm56$f9i$1@nnrp1.deja.com>

I would like to set the scrolling in a PmwScrolledFrame programatically
(in order to retain the same scroll position every time the same
information is displayed in the frame). I have tried two ways:

 - by using get() and set() in the scrollbar of the scrolledframe object.
get() returns plausible values, but set() has no effect

 - by using the private function in PmwScrolledFrame. _yview(), with
first parameter set to 'moveto'. Doesn't work either.

Does anybody have a clue?

    christer


Sent via Deja.com
http://www.deja.com/


From btheld2 at my-deja.com  Thu Feb  1 23:42:31 2001
From: btheld2 at my-deja.com (btheld2 at my-deja.com)
Date: Thu, 01 Feb 2001 22:42:31 GMT
Subject: Using WITH_FREE_THREAD to avoid PyEval_AcquireThread error
Message-ID: <95cool$hr5$1@nnrp1.deja.com>

I am having some problems with a PyEval_AcquireThread error.  I have an
extension DLL (created using SWIG).  A Python callback function in this
dll is calling a COM function.  This COM function is creating a new
Document in my MDI MFC Application.  When this occurs I get an
PyEval_AcquireThread error (Null thread?).

After digging through the code, it appears that a pre-processor
directive (WITH_FREE_THREAD) may solve this problem.  I just tried
defining this and rebuilding python, but the system cannot find
threadstate.h.

So, Is this a solution?  If not, what is?  If so, where is the
threadstate.h file?

Ben


Sent via Deja.com
http://www.deja.com/


From RoD at qnet20.com  Thu Feb  1 23:24:00 2001
From: RoD at qnet20.com (Rod)
Date: Thu, 1 Feb 2001 23:24:00
Subject: Diamond x Jungle Carpet Python
Message-ID: <20010202072423.184C9F4EA@mail.python.org>

I have several Diamond x Jungle Capret Pythons for SALE.

Make me an offer....

Go to: www.qnet20.com




From peter at engcorp.com  Fri Feb  2 14:31:51 2001
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 02 Feb 2001 08:31:51 -0500
Subject: Diamond x Jungle Carpet Python
References: <mailman.981100230.20774.python-list@python.org>
Message-ID: <3A7AB6C7.386D14D@engcorp.com>

Rod wrote:
> 
> I have several Diamond x Jungle Capret Pythons for SALE.
> 
> Make me an offer....
> 
> Go to: www.qnet20.com

Which platform?  Version 1.5.2 or earlier?  Doesn't
ActiveState already make these available as a free
download provided you don't need to redistribute them?

Jeez, why don't people ever provide enough detail in 
their posts?!

:-)


From jhauser at ifm.uni-kiel.de  Fri Feb  2 14:43:23 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 02 Feb 2001 14:43:23 +0100
Subject: Diamond x Jungle Carpet Python
References: <mailman.981100230.20774.python-list@python.org> <3A7AB6C7.386D14D@engcorp.com>
Message-ID: <877l399p50.fsf@lisboa.ifm.uni-kiel.de>

Peter Hansen <peter at engcorp.com> writes:

> Rod wrote:
> > 
> > I have several Diamond x Jungle Capret Pythons for SALE.
> > 
> > Make me an offer....
> > 
> > Go to: www.qnet20.com
> 
> Which platform?  Version 1.5.2 or earlier?  Doesn't
> ActiveState already make these available as a free
> download provided you don't need to redistribute them?
> 
> Jeez, why don't people ever provide enough detail in 
> their posts?!
> 
> :-)

What? You do not take this serious? I have just ordered two of them
for the sides of my home entrance. And Rod has given me a written
guarantee that these babies will not move and only roll their eyes, as
long as the the outside temperature is below zero degree (hope he
meant celsius). So I think I'm safe for the winter :-)

__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From dougfort at downright.com  Fri Feb  2 02:28:49 2001
From: dougfort at downright.com (Doug Fort)
Date: Thu, 1 Feb 2001 20:28:49 -0500
Subject: httplib: response.read() vs response.read(size)
Message-ID: <p3oe6.14610$Sl.630650@iad-read.news.verio.net>

We're developing a website load testing application
(http://www.stressmy.com).  We give httplib a pretty good workout.  And
generally get excellent results.  Recently, we ran into a pair of unrelated
sites that use some form of graphics server (Netscape-Enterprise/4.1).
There, our simulated users hung up waiting for a read on the socket in the
httplib.HTTPResponse.read().  The code looks like this (with a lot of
deletions)

self._client = httplib.HTTPConnection("", None)
...
response = self._client.getresponse()
...
self._data = response.read()

I was able to work around this with the following:

if response.length:
    self._data = response.read(response.length)
else:
    self._data = response.read()

My question is:  should this be moved up into httplib.HTTPResponse.read()?
I'm posting this here, instead of sending a bug or a diff to SourceForge
because I really don't understand the new HTTP 1.1 state machine in httplib,
and I don't want to aggravate the developers with bogus bugs or dumb patches
while they're working on 2.1.
--
Doug Fort
Senior Meat Manager
Downright Software LLC
http://www.dougfort.net





From jobs at sliceware.com  Fri Feb  2 02:51:45 2001
From: jobs at sliceware.com (Adam Feuer)
Date: Fri, 02 Feb 2001 01:51:45 GMT
Subject: Python/Zope Software Architect/Developer wanted
Message-ID: <Rooe6.3061$vu6.484085@news.uswest.net>

VSA Manager Software Architect/Lead Developer

Software Architect/Developer

Sliceware, Inc.
http://www.sliceware.com/

Sliceware has an open position for a Software Architect/Developer located in 
Portland, Oregon. You will be responsible for the design and implementation of 
one of Sliceware's core software products, the server end of an Internet-based 
client-server system. The server will be implemented in Zope. You will work with 
our Chief Architect, the lead architect of the client end project, and other 
team members.

Must Have Experience:
* Previous work as a Software Architect
* Good communication skills
* Shipping software to end-users (As opposed to building websites or 
web-applications that don't ship to users.)
* Object-oriented development (Python, Java, C++, Smalltalk, etc.)
* Scalable, multi-tier web applications
* Enterprise Databases (Oracle, Sybase, PostgreSQL, MySQL, etc.)
* Distributed computing
* Open Source software and development methodologies
* Unix, Linux
* XML
* W. Richard Stevens :-)
* BS - Computer Science (or equivalent)

Nice To Have Experience:
* Python
* Zope
* XML-RPC or SOAP
* Project Management
* Public Key Infrastructure, X.509 Certificates, SSL and TLS
* LDAP/Directory programming
* SNMP
* CIM

About Sliceware, Inc.:
Sliceware Inc. develops, markets and sells software for office networks and 
software tools to administer them remotely.

Please send your resume to jobs at sliceware.com

Contact:
Adam Feuer
jobs at sliceware.com
503-977-2800

6932 SW Macadam Ave. Suite C
Portland OR 97219





From g_will at cyberus.ca  Fri Feb  2 07:18:05 2001
From: g_will at cyberus.ca (Gordon Williams)
Date: Fri, 2 Feb 2001 01:18:05 -0500
Subject: Four problems with Gordon McMillan's Installer
Message-ID: <95djga$2a9s$1@news2.ottawa.cyberus.ca>

Hi All,

I'm having big problems with Gordon  McMillan installer (3i) to make a
standalone package.  I'm trying to put together a package using wxPython
2.2.2, Python 2. under W95.  I have used earlier versions of the installer,
but this is the first time for 3i.  Here is what is happening:

1) Minor problem is that "wxPython.wxc.pyd" is in the standalone directory
but it should be named "wxc.pyd".

2) wx22_2.dll is not found so I am putting it in under "misc".  This is the
notice that I am getting (whatever found and not found means):
analyzing multiarray
 found python20.dll
 found MSVCRT.dll
 found KERNEL32.dll
analyzing wxPython.wxc
 found wx22_2.dll
 lib not found: wx22_2.dll
 found python20.dll
 found MSVCRT.dll
analyzing _numpy
 found python20.dll
 found MSVCRT.dll
 found KERNEL32.dll


3) I'm getting notices about fixcase for all the files that it is
recompiling.  The names before and after are the same!!

Creating BS_Main06B.pyz
 name before fixcase e:\program files\python\lib\threading.py
 name after fixcase e:\program files\python\lib\threading.py


4) This is the major problem.  It is not finding the standard time module.
I am importing time into a number of modules and it cant find it all the
time.  If I move the import statement up or down a few lines it sometimes
fixes the problem, only to appear in another module.  Here is what I am
getting during a run of the standalone.  I cant see anything in the .log
file or the screen output during the build process.

 File "e:\program files\python\lib\imputil.py", line 252, in _import_one
    module = self._process_result(result, fqname)
  File "e:\program files\python\lib\imputil.py", line 281, in
_process_result
    exec code in module.__dict__
  File "e:\program files\python\junk_files\Main_thread.py", line 1, in ?
    import threading, time, Registers, DataSep
  File "e:\program files\python\lib\imputil.py", line 88, in _import_hook
    top_module = self._import_top_module(parts[0])
  File "e:\program files\python\lib\imputil.py", line 170, in
_import_top_module

    module = item.import_top(name)
  File "e:\program files\python\lib\imputil.py", line 197, in import_top
    return self._import_one(None, name, name)
  File "e:\program files\python\lib\imputil.py", line 252, in _import_one
    module = self._process_result(result, fqname)
  File "e:\program files\python\lib\imputil.py", line 281, in
_process_result
    exec code in module.__dict__
  File "e:\program files\python\lib\threading.py", line 4, in ?
    import time
  File "e:\program files\python\lib\imputil.py", line 91, in _import_hook
    raise ImportError, 'No module named ' + fqname
ImportError: No module named time

I have a web bookmark to Gordon's installer pages that are out of date as
they lead me to version 3f with no indication that there is a new installer
3i available.  Maybe these old pages could be deleted so others are not also
mislead.

Needless to say I'm going crazy.  Any help or ideas would be appreciated.

Gordon Williams




From thomas.heller at ion-tof.com  Fri Feb  2 10:36:07 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Fri, 2 Feb 2001 10:36:07 +0100
Subject: Four problems with Gordon McMillan's Installer
References: <95djga$2a9s$1@news2.ottawa.cyberus.ca>
Message-ID: <95dv28$guuo7$1@ID-59885.news.dfncis.de>

You should probably try out py2exe, which is still beta, but moving fast.
It has been used to create wxPython programs easily.

http://starship.python.net/crew/theller/py2exe

If you insist using Gordon's installer, search c.l.p. Solutions
to your problems have been posted AFAIK.

Regards,

Thomas




From g_will at cyberus.ca  Fri Feb  2 16:45:10 2001
From: g_will at cyberus.ca (Gordon Williams)
Date: Fri, 2 Feb 2001 10:45:10 -0500
Subject: Four problems with Gordon McMillan's Installer
References: <95djga$2a9s$1@news2.ottawa.cyberus.ca> <95dv28$guuo7$1@ID-59885.news.dfncis.de>
Message-ID: <95eklf$o50$1@news2.ottawa.cyberus.ca>

Yes, I used your program a couple of days ago when it was hot off the press.
It worked nicely and very smoothly.  The problem that I had with it was that
you pack all the source .pyc files into an easily expandable zip file.
These can be easily extracted and decompiled back to .py for all the world
to see.  I dont want people messing with the source so I was looking for
something that was not quite as easy to hack.

Maybe there is something that can be done with your program to at least hide
the python files inside the zip file to provide at least some level of
security?  I dont know how others handle this for commercial apps.

Regards,

Gordon Williams

"Thomas Heller" <thomas.heller at ion-tof.com> wrote in message
news:95dv28$guuo7$1 at ID-59885.news.dfncis.de...
> You should probably try out py2exe, which is still beta, but moving fast.
> It has been used to create wxPython programs easily.
>
> http://starship.python.net/crew/theller/py2exe
>
>




From thomas.heller at ion-tof.com  Fri Feb  2 16:57:54 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Fri, 2 Feb 2001 16:57:54 +0100
Subject: Four problems with Gordon McMillan's Installer
References: <95djga$2a9s$1@news2.ottawa.cyberus.ca> <95dv28$guuo7$1@ID-59885.news.dfncis.de> <95eklf$o50$1@news2.ottawa.cyberus.ca>
Message-ID: <95ele5$gqs15$1@ID-59885.news.dfncis.de>

> Yes, I used your program a couple of days ago when it was hot off the press.
> It worked nicely and very smoothly.
Glad to hear this!

  The problem that I had with it was that
> you pack all the source .pyc files into an easily expandable zip file.
> These can be easily extracted and decompiled back to .py for all the world
> to see.  I dont want people messing with the source so I was looking for
> something that was not quite as easy to hack.
>
> Maybe there is something that can be done with your program to at least hide
> the python files inside the zip file to provide at least some level of
> security?  I dont know how others handle this for commercial apps.
Well, this is on the to-do list for later.
I want to do this in two stages:
- first, use a notstandard zip-header so that winzip does
not recognize it
- second, maybe crypt the .pyc files in the archive somehow
and decrypt them in the exe.

Do you think this would be apppropriate?

I have no idea yet for the .pyd files.

Thomas





From dbrueck at edgix.com  Fri Feb  2 18:24:26 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Fri, 2 Feb 2001 10:24:26 -0700
Subject: Four problems with Gordon McMillan's Installer
In-Reply-To: <95ele5$gqs15$1@ID-59885.news.dfncis.de>
Message-ID: <BLEOJGEMBNBNEAKMIFDHKEEOPIAA.dbrueck@edgix.com>

> > Maybe there is something that can be done with your program to
> at least hide
> > the python files inside the zip file to provide at least some level of
> > security?  I dont know how others handle this for commercial apps.
> Well, this is on the to-do list for later.
> I want to do this in two stages:
> - first, use a notstandard zip-header so that winzip does
> not recognize it
> - second, maybe crypt the .pyc files in the archive somehow
> and decrypt them in the exe.
>
> Do you think this would be apppropriate?

Yes, that sounds like a good idea... you can use the imp/ihooks (is ihooks
obsolete?) modules to hook into the normal module importer and decrypt them
on the fly. For the zipfile you could still use a normal format but save it
to disk in a munged-up format (something simple like the rotor module would
keep most people out anyway) and then unmunge before reading it.

-Dave

P.S. - I just tried the latest version of py2exe on a wxPython program and
it worked _flawlessly_ and took little time to set up and get going. Very
cool...




From loewis at informatik.hu-berlin.de  Fri Feb  2 23:02:16 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 23:02:16 +0100
Subject: Four problems with Gordon McMillan's Installer
References: <95djga$2a9s$1@news2.ottawa.cyberus.ca> <95dv28$guuo7$1@ID-59885.news.dfncis.de> <95eklf$o50$1@news2.ottawa.cyberus.ca>
Message-ID: <p6qg0hwoiaf.fsf@informatik.hu-berlin.de>

"Gordon Williams" <g_will at cyberus.ca> writes:

> I dont want people messing with the source so I was looking for
> something that was not quite as easy to hack.

In that case, I recommend freezing. It also allows to integrate all
DLLs (provided you have or can produce static libraries for them).

Regards,
Martin



From fgeiger at datec.at  Sun Feb  4 10:37:43 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Sun, 4 Feb 2001 10:37:43 +0100
Subject: Four problems with Gordon McMillan's Installer
References: <95djga$2a9s$1@news2.ottawa.cyberus.ca> <95dv28$guuo7$1@ID-59885.news.dfncis.de>
Message-ID: <95j7si$24f$1@newsreaderm1.core.theplanet.net>

Had similar problems. Tried your py2exe and it worked - cool!

Regards
Franz GEIGER


"Thomas Heller" <thomas.heller at ion-tof.com> wrote in message
news:95dv28$guuo7$1 at ID-59885.news.dfncis.de...
> You should probably try out py2exe, which is still beta, but moving fast.
> It has been used to create wxPython programs easily.
>
> http://starship.python.net/crew/theller/py2exe
>
> If you insist using Gordon's installer, search c.l.p. Solutions
> to your problems have been posted AFAIK.
>
> Regards,
>
> Thomas
>
>




From robin at alldunn.com  Fri Feb  2 17:56:22 2001
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 2 Feb 2001 08:56:22 -0800
Subject: Four problems with Gordon McMillan's Installer
References: <95djga$2a9s$1@news2.ottawa.cyberus.ca>
Message-ID: <uFBe6.1748$wa5.54198721@news.randori.com>

> 
> 4) This is the major problem.  It is not finding the standard time module.
> I am importing time into a number of modules and it cant find it all the
> time.  

In the Installer sources in support/archive_rt.py after "import imputil" add this line:

sys.path.append(imputil.BuiltinImporter())


-- 
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!




From Dan.Rolander at marriott.com  Fri Feb  2 18:41:01 2001
From: Dan.Rolander at marriott.com (Rolander, Dan)
Date: Fri, 2 Feb 2001 12:41:01 -0500 
Subject: Four problems with Gordon McMillan's Installer
Message-ID: <6176E3D8E36FD111B58900805FA7E0F80CCF6360@mcnc-mdm1-ex01>

Gordon (Williams):

I am using Installer with wxPython 2.2.2, and Python 2.0 under Win2K. I had
some initial problems but everything is fine now. Here are my comments:

1) Yes, the problem with wxPython.wxc.pyd is known. Apparently this started
happening with Python 2.0. You'll see this often-- files will be named with
the full package path name. As far as I know the only workaround is to
rename them after the build.

2) Yes, I also include wx22_2.dll on the misc line, like this--
misc= MYSTANDALONE, wxPython.wx22_2

3) I see the fixcase messages also but they don't seem to affect anything.

4) I have an application that is using wxPython and the time module and I
haven't had a problem with it. BUT... I do have a couple of updated files
from Gordon McMillan that are newer than the 3i distribution and did fix A
LOT of problems I was having before. He's advised they will be part of the
3j distribution whenever that is released. I'm attaching them. I recommend
you replace your existing versions with these.

5) Yes, Gordon's starship pages are out of date and I don't know why they
are still there. Be sure to use http://www.mcmillan-inc.com instead.

I hope this helps,
Dan

-------------- MEInc\dist\bindepend.py ----------------
# copyright 1999 McMillan Enterprises, Inc.
# license: use as you please. No warranty.
#
# use dumpbin.exe (if present) to find the binary
# dependencies of an extension module.
# if dumpbin not available, pick apart the PE hdr of the binary
# while this appears to work well, it is complex and subject to
# problems with changes to PE hdrs (ie, this works only on 32 bit Intel
# Windows format binaries)
#
# Note also that you should check the results to make sure that the
# dlls are redistributable. I've listed most of the common MS dlls
# under "excludes" below; add to this list as necessary (or use the
# "excludes" option in the INSTALL section of the config file).

import os
import time
import string
import sys
import tempfile
import finder

seen = {}
excludes = {'KERNEL32.DLL':1, 
      'ADVAPI.DLL':1, 
      'MSVCRT.DLL':1,
      'ADVAPI32.DLL':1,
      'COMCTL32.DLL':1,
      'CRTDLL.DLL':1,
      'GDI32.DLL':1,
      'MFC42.DLL':1,
      'NTDLL.DLL':1,
      'OLE32.DLL':1,
      'OLEAUT32.DLL':1,
      'RPCRT4.DLL':1,
      'SHELL32.DLL':1,
      'USER32.DLL':1,
      'WINSPOOL.DRV':1,
      'WS2HELP.DLL':1,
      'WS2_32.DLL':1,
      'WSOCK32.DLL':1,
      'WINMM.DLL':1,
      'COMDLG32.DLL':1,
      'ZLIB.DLL':1,
      'ODBC32.DLL':1,
      'VERSION.DLL':1}     

def getfullnameof(mod, xtrapath = None):
  """Return the full path name of MOD.
  
      MOD is the basename of a dll or pyd.
      XTRAPATH is a path or list of paths to search first.
      Return the full path name of MOD.
      Will search the full Windows search path, as well as sys.path"""
  epath = finder.getpath()
  if mod[-4:] in ('.pyd', '.PYD'):
    epath = epath + sys.path
  if xtrapath is not None:
    if type(xtrapath) == type(''):
      epath.insert(0, xtrapath)
    else:
      epath = xtrapath + epath
  for p in epath:
    npth = os.path.join(p, mod)
    if os.path.exists(npth):
      return npth
  return ''
  
def getImports1(pth):
    """Find the binary dependencies of PTH.
    
        This implementation (not used right now) uses the MSVC utility
dumpbin"""
    rslt = []
    tmpf = tempfile.mktemp()
    os.system('dumpbin /IMPORTS "%s" >%s' %(pth, tmpf))
    time.sleep(0.1)
    txt = open(tmpf,'r').readlines()
    os.remove(tmpf)
    i = 0
    while i < len(txt):
        tokens = string.split(txt[i])
        if len(tokens) == 1 and string.find(tokens[0], '.') > 0:
            rslt.append(string.strip(tokens[0]))
        i = i + 1
    return rslt
    
def getImports2(pth):
    """Find the binary dependencies of PTH.
    
        This implementation walks through the PE header"""
    import struct
    rslt = []
    try:
      f = open(pth, 'rb').read()
      pehdrd = struct.unpack('l', f[60:64])[0]
      magic = struct.unpack('l', f[pehdrd:pehdrd+4])[0]
      numsecs = struct.unpack('h', f[pehdrd+6:pehdrd+8])[0]
      numdirs = struct.unpack('l', f[pehdrd+116:pehdrd+120])[0]
      idata = ''
      if magic == 17744:
          importsec, sz = struct.unpack('2l', f[pehdrd+128:pehdrd+136])
          secttbl = pehdrd + 120 + 8*numdirs
          secttblfmt = '8s7l2h'
          seclist = []
          for i in range(numsecs):
              seclist.append(struct.unpack(secttblfmt,
f[secttbl+i*40:secttbl+(i+1)*40]))
              #nm, vsz, va, rsz, praw, preloc, plnnums, qrelocs, qlnnums,
flags \
              # = seclist[-1]
          for i in range(len(seclist)-1):
              if seclist[i][2] <= importsec < seclist[i+1][2]:
                  break
          vbase = seclist[i][2]
          raw = seclist[i][4]
          idatastart = raw + importsec - vbase
          idata = f[idatastart:idatastart+seclist[i][1]]
          i = 0
          while 1:
              vsa =  struct.unpack('5l', idata[i*20:i*20+20])[3]
              if vsa == 0:
                  break
              sa = raw + vsa - vbase
              end = string.find(f, '\000', sa)
              nm = f[sa:end]
              if nm:
                  rslt.append(nm)
              i = i + 1
    except IOError:
      print "bindepend cannot analyze %s - file not found!" % pth
    except struct.error:
      print "bindepend cannot analyze %s - error walking thru pehdr" % pth
    return rslt
    
def Dependencies(lTOC):
  """Expand LTOC to include all the closure of binary dependencies.
  
     LTOC is a logical table of contents, ie, a seq of tuples (name, path).
     Return LTOC expanded by all the binary dependencies of the entries
     in LTOC, except those listed in the module global EXCLUDES"""
  for (nm, pth) in lTOC:
    fullnm = string.upper(os.path.basename(pth))
    if seen.get(string.upper(nm),0):
      continue
    print "analyzing", nm
    seen[string.upper(nm)] = 1
    dlls = getImports(pth)
    for lib in dlls:
        print " found", lib
        if excludes.get(string.upper(lib),0):
          continue
        if seen.get(string.upper(lib),0):
          continue
        npth = getfullnameof(lib)
        if npth:
          lTOC.append((lib, npth))
        else:
          print " lib not found:", lib, "dependency of", 
  return lTOC
  
        
##if getfullnameof('dumpbin.exe') == '':
##    def getImports(pth):
##        return getImports2(pth)
##else:
##    def getImports(pth):
##        return getImports1(pth)
        
def getImports(pth):
    """Forwards to either getImports1 or getImports2
    """
    return getImports2(pth)



-------------- support\archive_rt.py ----------------
#
# Gordon McMillan (as inspired and influenced by Greg Stein)
#

# subclasses may not need marshal or struct, but since they're
# builtin, importing is safe.
#
# While an Archive is really an abstraction for any "filesystem
# within a file", it is tuned for use with imputil.FuncImporter.
# This assumes it contains python code objects, indexed by the
# the internal name (ie, no '.py').
# See carchive.py for a more general archive (contains anything)
# that can be understood by a C program.

#archive_rt is a stripped down version of MEInc.Dist.archive.
#It has had all building logic removed.
#It's purpose is to bootstrap the Python installation.

_verbose = 0
_listdir = None
import marshal
import struct
import imp
import sys
import imputil

_c_suffixes = filter(lambda x: x[2] == imp.C_EXTENSION, imp.get_suffixes())

for nm in ('nt', 'posix', 'dos', 'os2', 'mac'):
  if nm in sys.builtin_module_names:
    mod = __import__(nm)
    _listdir = mod.listdir

class Archive(imputil.Importer):
  """ A base class for a repository of python code objects.
      The extract method is used by imputil.ArchiveImporter
      to get code objects by name (fully qualified name), so
      an enduser "import a.b" would become
        extract('a.__init__')
        extract('a.b')
  """
  MAGIC = 'PYL\0'
  HDRLEN = 12        # default is MAGIC followed by python's magic, int pos
of toc
  TOCPOS = 8
  TRLLEN = 0        # default - no trailer
  TOCTMPLT = {}     #
  os = None
  _bincache = None
  def __init__(self, path=None, start=0):
    "Initialize an Archive. If path is omitted, it will be an empty
Archive."
    self.toc = None
    self.path = path
    self.start = start
    import imp
    self.pymagic = imp.get_magic()
    if path is not None:
      self.lib = open(self.path, 'rb')
      self.checkmagic()
      self.loadtoc()
      if Archive._bincache is None:
        Archive._bincache = {}
        for fnm in _listdir(sys.path[0]):
          for ext, mode, typ in _c_suffixes:
            sz = len(ext)
            if fnm[-sz:] == ext:
              Archive._bincache[fnm[:-sz]] = sys.path[0] + '/' + fnm, (ext,
mode, typ)
              if _verbose:
                print "Archive caching", fnm[:-sz]

  ####### Sub-methods of __init__ - override as needed #############
  def checkmagic(self):
    """ Overridable.
        Check to see if the file object self.lib actually has a file
        we understand.
    """
    self.lib.seek(self.start) #default - magic is at start of file
    if self.lib.read(len(self.MAGIC)) != self.MAGIC:
      raise RuntimeError, "%s is not a valid %s archive file" \
% (self.path, self.__class__.__name__)
    if self.lib.read(len(self.pymagic)) != self.pymagic:
      raise RuntimeError, "%s has version mismatch to dll" % (self.path)

  def loadtoc(self):
    """ Overridable.
        Default: After magic comes an int (4 byte native) giving the
        position of the TOC within self.lib.
        Default: The TOC is a marshal-able string.
    """
    self.lib.seek(self.start + self.TOCPOS)
    (offset,) = struct.unpack('=i', self.lib.read(4))
    self.lib.seek(self.start + offset)
    self.toc = marshal.load(self.lib)

  ######## This is what is called by FuncImporter #######
  ## Since an Archive is flat, we ignore parent and modname.

  def get_code(self, parent, modname, fqname):
    if _verbose:
      print "get_code(%s, %s, %s, %s)" % (self, parent, modname, fqname)
    rslt = self.extract(fqname) # None if not found, (ispkg, code) otherwise
    if rslt is None:
      if parent:
        importer = getattr(parent, "__importer__", None)
        if importer:
          func = getattr(importer, "func", None)
          if func == self.get_code:
            file, desc = Archive._bincache.get(fqname, (None, None))
            if file:
              try:
                fp = open(file, desc[1])
              except IOError:
                pass
              else:
                module = imp.load_module(fqname, fp, file, desc)
                if _verbose:
                  print "#found", file
                return 0, module, {'__file__':file}
      if _verbose:
        print "#not found"
      return None
    ispkg, code = rslt
    if ispkg:
      return ispkg, code, {'__path__' : [fqname]}
    return ispkg, code, {}

  ####### Core method - Override as needed  #########
  def extract(self, name):
    """ Get the object corresponding to name, or None.
        For use with imputil ArchiveImporter, object is a python code
object.
        'name' is the name as specified in an 'import name'.
        'import a.b' will become:
        extract('a') (return None because 'a' is not a code object)
        extract('a.__init__') (return a code object)
        extract('a.b') (return a code object)
        Default implementation:
          self.toc is a dict
          self.toc[name] is pos
          self.lib has the code object marshal-ed at pos
    """
    ispkg, pos = self.toc.get(name, (0,None))
    if pos is None:
      return None
    self.lib.seek(self.start + pos)
    return ispkg, marshal.load(self.lib)

  ########################################################################
  # Informational methods

  def contents(self):
    """Return a list of the contents
       Default implementation assumes self.toc is a dict like object.
       Not required by ArchiveImporter.
    """
    return self.toc.keys()

  ########################################################################
  # Building
  
  ####### Top level method - shouldn't need overriding #######
##  def build(self, path, lTOC):
##    """Create an archive file of name 'path'.
##       lTOC is a 'logical TOC' - a list of (name, path, ...)
##       where name is the internal name, eg 'a'
##       and path is a file to get the object from, eg './a.pyc'.
##    """
##    self.path = path
##    self.lib = open(path, 'wb')
##    #reserve space for the header
##    if self.HDRLEN:
##      self.lib.write('\0'*self.HDRLEN)
##
##    #create an empty toc
##
##    if type(self.TOCTMPLT) == type({}):
##      self.toc = {}
##    else:       # assume callable  
##      self.toc = self.TOCTMPLT()
##
##    for tocentry in lTOC:
##      self.add(tocentry)   # the guts of the archive
##
##    tocpos = self.lib.tell() 
##    self.save_toc(tocpos)
##    if self.TRLLEN:
##      self.save_trailer(tocpos)
##    if self.HDRLEN:
##      self.update_headers(tocpos) 
##    self.lib.close()
##
##
##  ####### manages keeping the internal TOC and the guts in sync #######
##  def add(self, entry):
##    """Override this to influence the mechanics of the Archive.
##       Assumes entry is a seq beginning with (nm, pth, ...) where
##       nm is the key by which we'll be asked for the object.
##       pth is the name of where we find the object. Overrides of
##       get_obj_from can make use of further elements in entry.
##    """
##    if self.os is None:
##      import os
##      self.os = os
##    nm = entry[0]
##    pth = entry[1]
##    ispkg = self.os.path.splitext(self.os.path.basename(pth))[0] ==
'__init__'
##    self.toc[nm] = (ispkg, self.lib.tell())
##    f = open(entry[1], 'rb')
##    f.seek(8) #skip magic and timestamp
##    self.lib.write(f.read())
##
##  def save_toc(self, tocpos):
##    """Default - toc is a dict
##       Gets marshaled to self.lib
##    """
##    marshal.dump(self.toc, self.lib)
##
##  def save_trailer(self, tocpos):
##    """Default - not used"""
##    pass
##
##  def update_headers(self, tocpos):
##    """Default - MAGIC + Python's magic + tocpos"""
##    self.lib.seek(self.start)
##    self.lib.write(self.MAGIC)
##    self.lib.write(self.pymagic)
##    self.lib.write(struct.pack('=i', tocpos))
   
##############################################################
#
# ZlibArchive - an archive with compressed entries
#

class ZlibArchive(Archive):
  MAGIC = 'PYZ\0'
  TOCPOS = 8
  HDRLEN = 12
  TRLLEN = 0
  TOCTMPLT = {}
  LEVEL = 9

  def __init__(self, path=None, offset=0):
    Archive.__init__(self, path, offset)
    # dynamic import so not imported if not needed
    global zlib
    import zlib
   
  def extract(self, name):
    (ispkg, pos, lngth) = self.toc.get(name, (0, None, 0))
    if pos is None:
      return None
    self.lib.seek(self.start + pos)
    return ispkg, marshal.loads(zlib.decompress(self.lib.read(lngth)))

##  def add(self, entry):
##    if self.os is None:
##      import os
##      self.os = os
##    nm = entry[0]
##    pth = entry[1]
##    ispkg = self.os.path.splitext(self.os.path.basename(pth))[0] ==
'__init__'
##    f = open(pth, 'rb')
##    f.seek(8) #skip magic and timestamp
##    obj = zlib.compress(f.read(), self.LEVEL)
##    self.toc[nm] = (ispkg, self.lib.tell(), len(obj))
##    self.lib.write(obj)
## 

if type(__import__) == type(open):
    imputil.ImportManager().install()
    sys.path.append(imputil.BuiltinImporter())




From thomas.heller at ion-tof.com  Fri Feb  2 19:11:59 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Fri, 2 Feb 2001 19:11:59 +0100
Subject: Four problems with Gordon McMillan's Installer
References: <mailman.981135004.31623.python-list@python.org>
Message-ID: <95et9h$ge831$1@ID-59885.news.dfncis.de>

"Dave Brueck" <dbrueck at edgix.com> wrote in message news:mailman.981135004.31623.python-list at python.org...
> > > Maybe there is something that can be done with your program to
> > at least hide
> > > the python files inside the zip file to provide at least some level of
> > > security?  I dont know how others handle this for commercial apps.
> > Well, this is on the to-do list for later.
> > I want to do this in two stages:
> > - first, use a notstandard zip-header so that winzip does
> > not recognize it
> > - second, maybe crypt the .pyc files in the archive somehow
> > and decrypt them in the exe.
> >
> > Do you think this would be apppropriate?
>
> Yes, that sounds like a good idea... you can use the imp/ihooks (is ihooks
> obsolete?) modules to hook into the normal module importer and decrypt them
> on the fly.

This is what I had planned.

> For the zipfile you could still use a normal format but save it
> to disk in a munged-up format (something simple like the rotor module would
> keep most people out anyway) and then unmunge before reading it.
>
> -Dave
>
> P.S. - I just tried the latest version of py2exe on a wxPython program and
> it worked _flawlessly_ and took little time to set up and get going. Very
> cool...
>
>

Thanks for trying it,

Thomas




From lisowski.tomasz at sssa.NOSPAM.com.pl  Fri Feb  2 08:17:03 2001
From: lisowski.tomasz at sssa.NOSPAM.com.pl (Tomasz Lisowski)
Date: Fri, 2 Feb 2001 08:17:03 +0100
Subject: Q: import statement within a function
Message-ID: <95dmp6$n4s$1@news.tpi.pl>

This is perhaps rather a basic question, but ...

I have seen often import statements included within a function's body, not
in the global scope, as it is most common.

Approach 1:
-------------
def fun(s):
    ...
    import string
    words=string.split(s)

as compared to:

Approach 2:
-------------
import string

def fun(s):
    ...
    words=string.split(s)


What are the advantages of the first approach over the second. Won't Python
do the importing job many times, each time the function fun() is called?

Regards,
Tomasz Lisowski





From fredrik at effbot.org  Fri Feb  2 09:10:37 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 08:10:37 GMT
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl>
Message-ID: <1Yte6.11069$AH6.1772843@newsc.telia.net>

Tomasz Lisowski wrote:
> I have seen often import statements included within a function's body, not
> in the global scope, as it is most common.
> ...
> What are the advantages of the first approach over the second. Won't Python
> do the importing job many times, each time the function fun() is called?

Depends on what you mean with "importing job" -- Python
caches imported modules in sys.modules.  The "real" import
is only done the first time you need a module.

also see:
http://effbot.org/guides/import-confusion.htm
and the language reference.

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From lisowski.tomasz at sssa.NOSPAM.com.pl  Fri Feb  2 09:33:51 2001
From: lisowski.tomasz at sssa.NOSPAM.com.pl (Tomasz Lisowski)
Date: Fri, 2 Feb 2001 09:33:51 +0100
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net>
Message-ID: <95dr93$dcs$1@news.tpi.pl>

U?ytkownik "Fredrik Lundh" <fredrik at effbot.org> napisa? w wiadomo?ci
news:1Yte6.11069$AH6.1772843 at newsc.telia.net...
> Tomasz Lisowski wrote:
> > I have seen often import statements included within a function's body,
not
> > in the global scope, as it is most common.
> > ...
> > What are the advantages of the first approach over the second. Won't
Python
> > do the importing job many times, each time the function fun() is called?
>
> Depends on what you mean with "importing job" -- Python
> caches imported modules in sys.modules.  The "real" import
> is only done the first time you need a module.
>
> also see:
> http://effbot.org/guides/import-confusion.htm
> and the language reference.

OK, I have read your mini-guide, and the language reference, and now I see,
that the penalty for executing import statement after it has been imported
for the first time is minimal. What are then advantages of importing a
module WITHIN a function. Why not place this statement in the global scope?

If a module is imported within a function, the module's name is added to the
CURRENT namespace, right. It means that other functions in the module have
no access to this module, even if it is listed in sys.modules. IMHO, this is
rather a drawback of this approach. Let me ask once more: what are the
advantages?

Tomasz Lisowski





From fredrik at effbot.org  Fri Feb  2 10:40:39 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 09:40:39 GMT
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl>
Message-ID: <rgve6.11079$AH6.1774088@newsc.telia.net>

Tomasz Lisowski wrote:
> OK, I have read your mini-guide, and the language reference, and now I see,
> that the penalty for executing import statement after it has been imported
> for the first time is minimal. What are then advantages of importing a
> module WITHIN a function.

- lazy importing: if you place it within a function, the module
is only imported if your program really needs it.

- import dependencies (see the recursive import section in the
mini-guide).  if you use from-import on the module level, you
can end up in situations where the things you import doesn't
yet exist.

- performance: local lookup is faster than global lookup.

- readability: it can be easier to grok your code if you import
things (i.e. define names) near the place you're using them.

(etc)

Cheers /F




From lisowski.tomasz at sssa.NOSPAM.com.pl  Fri Feb  2 13:34:15 2001
From: lisowski.tomasz at sssa.NOSPAM.com.pl (Tomasz Lisowski)
Date: Fri, 2 Feb 2001 13:34:15 +0100
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl> <rgve6.11079$AH6.1774088@newsc.telia.net>
Message-ID: <95e9c2$rmm$1@news.tpi.pl>

U?ytkownik "Fredrik Lundh" <fredrik at effbot.org> napisa? w wiadomo?ci
news:rgve6.11079$AH6.1774088 at newsc.telia.net...
> Tomasz Lisowski wrote:
> > OK, I have read your mini-guide, and the language reference, and now I
see,
> > that the penalty for executing import statement after it has been
imported
> > for the first time is minimal. What are then advantages of importing a
> > module WITHIN a function.
>
> - lazy importing: if you place it within a function, the module
> is only imported if your program really needs it.
>
> - import dependencies (see the recursive import section in the
> mini-guide).  if you use from-import on the module level, you
> can end up in situations where the things you import doesn't
> yet exist.
>
> - performance: local lookup is faster than global lookup.

 Yes, that's really a good point!

> - readability: it can be easier to grok your code if you import
> things (i.e. define names) near the place you're using them.

Thanks, Frederik, that is really a good explanation! BTW, I really
appreciate your "Introduction to Tkinter". I am using it alot.

Tomasz




From peter at engcorp.com  Fri Feb  2 14:29:56 2001
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 02 Feb 2001 08:29:56 -0500
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl> <rgve6.11079$AH6.1774088@newsc.telia.net> <95e9c2$rmm$1@news.tpi.pl>
Message-ID: <3A7AB654.D6C16746@engcorp.com>

Tomasz Lisowski wrote:
> 
> "Fredrik Lundh" <fredrik at effbot.org>:
> > Tomasz Lisowski wrote:
> > > for the first time is minimal. What are then advantages of importing a
> > > module WITHIN a function.
> >
> > - lazy importing: if you place it within a function, the module
> > is only imported if your program really needs it.
> >
> > - import dependencies (see the recursive import section in the
> > mini-guide).  if you use from-import on the module level, you
> > can end up in situations where the things you import doesn't
> > yet exist.
> >
> > - performance: local lookup is faster than global lookup.
> 
>  Yes, that's really a good point!
> 
> > - readability: it can be easier to grok your code if you import
> > things (i.e. define names) near the place you're using them.
> 
> Thanks, Frederik, that is really a good explanation! BTW, I really
> appreciate your "Introduction to Tkinter". I am using it alot.

The one you highlight, "performance", should probably be
considered the _least_ important of the points Fredrik raised.

The last one, which implies improved maintainability, is
far more important (IMHO), and that alone is a good reason
to put import within a function instead of 'without'
(provided the imported module is not needed throughout
the importing module of course).


From lisowski.tomasz at sssa.NOSPAM.com.pl  Sat Feb  3 08:43:47 2001
From: lisowski.tomasz at sssa.NOSPAM.com.pl (Tomasz Lisowski)
Date: Sat, 3 Feb 2001 08:43:47 +0100
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl> <rgve6.11079$AH6.1774088@newsc.telia.net> <95e9c2$rmm$1@news.tpi.pl> <3A7AB654.D6C16746@engcorp.com>
Message-ID: <95gcn6$imb$1@news.tpi.pl>

Uzytkownik "Peter Hansen" <peter at engcorp.com> napisal w wiadomosci
news:3A7AB654.D6C16746 at engcorp.com...
<.>
> > > - lazy importing: if you place it within a function, the module
> > > is only imported if your program really needs it.
> > >
> > > - import dependencies (see the recursive import section in the
> > > mini-guide).  if you use from-import on the module level, you
> > > can end up in situations where the things you import doesn't
> > > yet exist.
> > >
> > > - performance: local lookup is faster than global lookup.
> >
> >  Yes, that's really a good point!
> >
> > > - readability: it can be easier to grok your code if you import
> > > things (i.e. define names) near the place you're using them.
> >
> > Thanks, Frederik, that is really a good explanation! BTW, I really
> > appreciate your "Introduction to Tkinter". I am using it alot.
>
> The one you highlight, "performance", should probably be
> considered the _least_ important of the points Fredrik raised.

That depends, on what the given function does, and how often it is called,
of course.

> The last one, which implies improved maintainability, is
> far more important (IMHO), and that alone is a good reason
> to put import within a function instead of 'without'
> (provided the imported module is not needed throughout
> the importing module of course).

I dare to disagree with you, as I find it very convenient to find all import
statements at the beginning of the source file, without having to scan the
file, if there are any other modules, that I have requested. Perhaps my
opinion is influenced by my Delphi programming habits, where I could find
all modules imported in the 'uses' clause, easily locateable in the source
file.

Moreover, it is a matter of taste, whether I put the import statement at the
beginning of the file, even if the given module is imported only by one
function, since this function will be eventually called (there was surely a
reason for writing it, wasn't it?). Therefore I have asked Frederik about
REAL advantages, that could persuade me to put import statements within a
function. Performance, IMHO, is one of them.

Tomasz Lisowski




From peter at engcorp.com  Sat Feb  3 15:34:16 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 03 Feb 2001 09:34:16 -0500
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl> <rgve6.11079$AH6.1774088@newsc.telia.net> <95e9c2$rmm$1@news.tpi.pl> <3A7AB654.D6C16746@engcorp.com> <95gcn6$imb$1@news.tpi.pl>
Message-ID: <3A7C16E8.1805AF24@engcorp.com>

Tomasz Lisowski wrote:
> 
> "Peter Hansen" <peter at engcorp.com>:
> > The one you highlight, "performance", should probably be
> > considered the _least_ important of the points Fredrik raised.
> 
> That depends, on what the given function does, and how often it is called,
> of course.
> 
> > The last one, which implies improved maintainability, is
> > far more important (IMHO) ...
> 
> Moreover, it is a matter of taste, whether I put the import statement at the
> beginning of the file, even if the given module is imported only by one
> function, since this function will be eventually called (there was surely a
> reason for writing it, wasn't it?). Therefore I have asked Frederik about
> REAL advantages, that could persuade me to put import statements within a
> function. Performance, IMHO, is one of them.

We're definitely talking from different backgrounds, but there's
room for that.  In my world, performance is NEVER an issue ahead
of time, but getting things to work is.  Performance becomes an
issue only when it is insufficient for the purpose, and in 
that case it would be good to know that by putting the import
inside a "hot spot" function, its performance may be improved.
After profiling and seeing unacceptable performance.

On the other hand, for me maintainability is ALWAYS an issue
ahead of time and throughout the life of the code.  To make
the code more maintainable, it may be necessary to comment
why a module is imported (if it is imported at the top far
away from the function using it).  By placing the import
within the single function which uses it, the code becomes
effectively self-documenting and needs no additional comment
regarding the module.  I can always 'grep' the code to find
all import statements, if that is necessary (which it rarely 
is, for me).

So to me this is very much a REAL advantage, but I can see
you have a different (and perfectly valid, for you) perspective.

:)


From danielk at aracnet.com  Fri Feb  2 15:00:01 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Fri, 02 Feb 2001 06:00:01 -0800
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl> <rgve6.11079$AH6.1774088@newsc.telia.net> <95e9c2$rmm$1@news.tpi.pl>
Message-ID: <i6fl7tg2hdbtqqcv2e5195lc3atq811ma4@4ax.com>

On Fri, 2 Feb 2001 13:34:15 +0100, "Tomasz Lisowski"
<lisowski.tomasz at sssa.NOSPAM.com.pl> wrote:

>> > OK, I have read your mini-guide, and the language reference, and now I

Where can I find this 'mini-guide'. I already have the 'The Standard Python
Library' and of course the language reference.

Thanks,
Daniel Klein


From fredrik at effbot.org  Fri Feb  2 15:51:42 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 14:51:42 GMT
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl> <rgve6.11079$AH6.1774088@newsc.telia.net> <95e9c2$rmm$1@news.tpi.pl> <i6fl7tg2hdbtqqcv2e5195lc3atq811ma4@4ax.com>
Message-ID: <2Qze6.11108$AH6.1780895@newsc.telia.net>

Daniel Klein wrote:
> Where can I find this 'mini-guide'. I already have the 'The Standard Python
> Library' and of course the language reference.

    http://effbot.org/guides

Cheers /F




From mmueller at dgfz.de  Fri Feb  2 08:56:08 2001
From: mmueller at dgfz.de (Mike Müller)
Date: Fri, 2 Feb 2001 08:56:08 +0100
Subject: Naming conventions for path?
Message-ID: <95dope$o4k@micky.ibh-dd.de>

Setting up my first bigger Python program I find myself al little bit
confused with the use of path. That is, using path as a variable I don?t
exactly know what I am referring to when joining and spliting path names. I
would like to use self explaining variable names for those path parts. Lets
look an example:

.../dir1/dir2/filename.ext
-----------++++++.***

The whole thing is usually referred to as path. But also the part underlined
with ---- can be addressed with path. Alternatively the whole thing can be
named filename (or any thing else if want to but I talk about conventions).

+++.*** can also be filename

furthermore

+++ filenameRoot ?

and

*** fileExtension ?

also

---- directory or dir ?

Are there any conventions in Python (or any other language) that are more or
less widely used? Or should I come up with my own naming scheme and just
stay consistent throughout may project(s)?

Thanks for advise.

Mike







From zope at thewebsons.com  Fri Feb  2 09:41:18 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Fri, 02 Feb 2001 00:41:18 -0800
Subject: Chokes on GETPPID
Message-ID: <5.0.2.1.0.20010202003611.009e95d0@thewebsons.com>

Hi;
I'm trying to install Zope and it chokes in this block:
 >>>
if not READ_ONLY:
     pf = open(PID_FILE, 'w')
     pid=str(os.getpid())
     try: pid=str(os.getppid())+' '+pid
     except: pass
     pf.write(pid)
     pf.close()
<<<
It specifically identified the *try* statement as the choke-point. How do I 
trouble-shoot this?
TIA,
BenO




From fredrik at effbot.org  Fri Feb  2 10:59:45 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 09:59:45 GMT
Subject: Chokes on GETPPID
References: <mailman.981104043.22088.python-list@python.org>
Message-ID: <lyve6.11082$AH6.1774924@newsc.telia.net>

Ben Ocean wrote:
> I'm trying to install Zope and it chokes in this block:
>  >>>
> if not READ_ONLY:
>      pf = open(PID_FILE, 'w')
>      pid=str(os.getpid())
>      try: pid=str(os.getppid())+' '+pid
>      except: pass
>      pf.write(pid)
>      pf.close()
> <<<
> It specifically identified the *try* statement as the choke-point.
> How do I trouble-shoot this?

I assume "choke" means that it stopped and printed a
traceback message?  how about posting that traceback?

</F>




From zope at thewebsons.com  Fri Feb  2 15:18:11 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Fri, 02 Feb 2001 06:18:11 -0800
Subject: Chokes on GETPPID
In-Reply-To: <lyve6.11082$AH6.1774924@newsc.telia.net>
References: <mailman.981104043.22088.python-list@python.org>
Message-ID: <5.0.2.1.0.20010202061728.00a4fab0@thewebsons.com>

At 09:59 AM 2/2/2001 +0000, you wrote:
>Ben Ocean wrote:
> > I'm trying to install Zope and it chokes in this block:
> >  >>>
> > if not READ_ONLY:
> >      pf = open(PID_FILE, 'w')
> >      pid=str(os.getpid())
> >      try: pid=str(os.getppid())+' '+pid
> >      except: pass
> >      pf.write(pid)
> >      pf.close()
> > <<<
> > It specifically identified the *try* statement as the choke-point.
> > How do I trouble-shoot this?
>
>I assume "choke" means that it stopped and printed a
>traceback message?  how about posting that traceback?

...with traceback...
 >>>
Traceback (innermost last):
   File "/apache/Zope-2.2.4-linux2-x86/z2.py", line 710, in ?
     pf = open(PID_FILE, 'w')
IOError: [Errno 13] Permission denied: 
'/apache/Zope-2.2.4-linux2-x86/var/Z2.pid
'<<<

></F>
>
>
>--
>http://mail.python.org/mailman/listinfo/python-list




From fredrik at effbot.org  Fri Feb  2 15:49:43 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 14:49:43 GMT
Subject: Chokes on GETPPID
References: <mailman.981104043.22088.python-list@python.org> <mailman.981123363.28330.python-list@python.org>
Message-ID: <bOze6.11107$AH6.1780860@newsc.telia.net>

Ben Ocean wrote:
> > >  >>>
> > > if not READ_ONLY:
> > >      pf = open(PID_FILE, 'w')
> > >      pid=str(os.getpid())
> > >      try: pid=str(os.getppid())+' '+pid
> > >      except: pass
> > >      pf.write(pid)
> > >      pf.close()
> > > <<<
> > > It specifically identified the *try* statement as the choke-point.
> > > How do I trouble-shoot this?
> >
> >I assume "choke" means that it stopped and printed a
> >traceback message?  how about posting that traceback?
>
> ...with traceback...
>  >>>
> Traceback (innermost last):
>    File "/apache/Zope-2.2.4-linux2-x86/z2.py", line 710, in ?
>      pf = open(PID_FILE, 'w')
> IOError: [Errno 13] Permission denied:
> '/apache/Zope-2.2.4-linux2-x86/var/Z2.pid
> '<<<

I'd say it specifically identified the *open* statement as
the choke-point...

Do you have write privileges in that directory?  If the file
already exists, is it writable?

Cheers /F




From robin at jessikat.fsnet.co.uk  Fri Feb  2 11:33:55 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 2 Feb 2001 10:33:55 +0000
Subject: nested scopes
Message-ID: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>

Is it really true that J Hylton's nested scopes will cause code like

def bingo():
  from string import *
  ....

to be declared bad? I'm already against nested scopes, but requiring
this to be wrong seems awful. We seem to be replacing a simple 2-level
system by a more complex one & restrictions.
-- 
Robin Becker


From mwh21 at cam.ac.uk  Fri Feb  2 13:57:52 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 02 Feb 2001 12:57:52 +0000
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
Message-ID: <m3itmt5jjj.fsf@atrus.jesus.cam.ac.uk>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> Is it really true that J Hylton's nested scopes will cause code like
> 
> def bingo():
>   from string import *
>   ....
> 
> to be declared bad?

I'm not sure about how hard a requirement it is.  It is illegal in
current CVS and probably will be in 2.1a2.  Subsequent enormous howls
of protest might result in a more gentle solution.  We shall see.

Cheers,
M.

-- 
  I have a cat, so I know that when she digs her very sharp claws into
  my chest or stomach it's really a sign of affection, but I don't see
  any reason for programming languages to show affection with pain.
                                        -- Erik Naggum, comp.lang.lisp


From robin at jessikat.fsnet.co.uk  Fri Feb  2 14:15:42 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 2 Feb 2001 13:15:42 +0000
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
 <m3itmt5jjj.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <Z$sXYLA+Lre6Ewxt@jessikat.fsnet.co.uk>

In article <m3itmt5jjj.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
<mwh21 at cam.ac.uk> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> Is it really true that J Hylton's nested scopes will cause code like
>> 
>> def bingo():
>>   from string import *
>>   ....
>> 
>> to be declared bad?
>
>I'm not sure about how hard a requirement it is.  It is illegal in
>current CVS and probably will be in 2.1a2.  Subsequent enormous howls
>of protest might result in a more gentle solution.  We shall see.
>
>Cheers,
>M.
>
HOWL HOWL HOWL, gratuitous changes for no very obvious benefit and
mysterious side effects are bad mumble, howl, whine.
-- 
Robin Becker


From phrxy at csv.warwick.ac.uk  Fri Feb  2 16:52:45 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Fri, 2 Feb 2001 15:52:45 +0000
Subject: nested scopes
In-Reply-To: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
Message-ID: <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk>

On Fri, 2 Feb 2001, Robin Becker wrote:

> Is it really true that J Hylton's nested scopes will cause code like
>
> def bingo():
>   from string import *
>   ....
>
> to be declared bad? I'm already against nested scopes, but requiring
> this to be wrong seems awful. We seem to be replacing a simple 2-level
> system by a more complex one & restrictions.
[...]

Are nested scopes actually going to end up in Python proper, then?  As
opposed to remaining as an optional patch?

And why should the above become bad rather than just import into the
appropriate namespace?


John



From robin at jessikat.fsnet.co.uk  Fri Feb  2 17:05:58 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 2 Feb 2001 16:05:58 +0000
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
 <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk>

In article <Pine.SOL.4.30.0102021550090.28216-100000 at mimosa.csv.warwick.
ac.uk>, John J. Lee <phrxy at csv.warwick.ac.uk> writes
>On Fri, 2 Feb 2001, Robin Becker wrote:
>
>> Is it really true that J Hylton's nested scopes will cause code like
>>
>> def bingo():
>>   from string import *
>>   ....
>>
>> to be declared bad? I'm already against nested scopes, but requiring
>> this to be wrong seems awful. We seem to be replacing a simple 2-level
>> system by a more complex one & restrictions.
>[...]
>
>Are nested scopes actually going to end up in Python proper, then?  As
>opposed to remaining as an optional patch?
>

don't know about that, apparently 2.1a has it

>And why should the above become bad rather than just import into the
>appropriate namespace?
>
...
I have a feeling that in order to make nested scopes workable/acceptable
the byte compiler is being hacked and this has had some unfortunate side
effects. It took me a while to get used to the two level (or three if
you split builtins from globals) scoping rule, but I don't miss Algol
style nesting a lot and I find the explicit lambda stuff a help rather
than a hindrance. I would drop this real soon if I could.
-- 
Robin Becker


From thomas at xs4all.net  Fri Feb  2 19:46:09 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 2 Feb 2001 19:46:09 +0100
Subject: nested scopes
In-Reply-To: <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk>; from robin@jessikat.fsnet.co.uk on Fri, Feb 02, 2001 at 04:05:58PM +0000
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk> <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk>
Message-ID: <20010202194609.W962@xs4all.nl>

On Fri, Feb 02, 2001 at 04:05:58PM +0000, Robin Becker wrote:
> In article <Pine.SOL.4.30.0102021550090.28216-100000 at mimosa.csv.warwick.
> ac.uk>, John J. Lee <phrxy at csv.warwick.ac.uk> writes
> >On Fri, 2 Feb 2001, Robin Becker wrote:
> >
> >> Is it really true that J Hylton's nested scopes will cause code like
> >>
> >> def bingo():
> >>   from string import *
> >>   ....
> >>
> >> to be declared bad? I'm already against nested scopes, but requiring
> >> this to be wrong seems awful. We seem to be replacing a simple 2-level
> >> system by a more complex one & restrictions.
> >[...]
> >
> >Are nested scopes actually going to end up in Python proper, then?  As
> >opposed to remaining as an optional patch?

> don't know about that, apparently 2.1a has it

It won't be optional, it'll be part of Python. I guess that if there is
a *really* *really* *really* lot of code out there that will break, it could
be withheld until Python 2.2, and the practice of 'from foo import *' in
function/class scopes be made to issue a warning in Python 2.1, but frankly
I doubt there'll be that much code breakage. At least it's not a *silent*
breakage ! It's a compiler error, so the first run of a script under the new
Python will detect it. It's not like it's a new kind of exception being
raised.

> >And why should the above become bad rather than just import into the
> >appropriate namespace?
> >
> ...
> I have a feeling that in order to make nested scopes workable/acceptable
> the byte compiler is being hacked and this has had some unfortunate side
> effects.

Now, now, 'hacked' is a bit strong. Jeremy did a fairly extensive rewrite of
the compiler, complete and decent enough not to be called a 'hack' for sure.
Besides, developers never 'hack', they 'extend' :-) The problem with 'from
foo import *' in a function or class scope, though, is that it makes it
impossible for the compiler to see what names will be defined in that scope,
and hence what names a nested scope uses will resolve to that. I don't like
it at all, either. In fact, I was the first to whine about it, incessantly,
on python-dev. I think I have to buy Jeremy a beer (or whatever's his
poison) on the Python conference to make up ;)

The situation does lend itself to a workaround (a *true* hack, mind
you:) but I haven't looked into implementing it, yet. (Jeremy might be doing
that, but I'm not sure): The compiler could try and live with 'from foo
import *' as long as the function doesn't define a scope nested in that
function. Unfortunately, this isn't going to fix every instance of the
problem: a single lambda statement inside the function will make the import
illegal again. Disabling nested-scoping in the presense of 'from foo import
*' might be technically possible, but would create enormous amounts of
confusion, I'm sure.

And lastly, I'd like to point out (as Jeremy does :) that the reference
manual already states that 'from foo import *' in anything but a module
scope is illegal. Even if the nested scopes patch was withdrawn from Python
2.1, 'from foo import *' in local scopes *will* generate a warning (warnings
are also new in 2.1) so your best bet, either way, is sanitizing your code. 

(Don't get me wrong: I seriously think 'from foo import *' is overused. The
language reference makes it pretty clear what it was intended for, and I
think even the common 'from Tkinter import *' is wrong use -- but I don't
like code breakage, regardless.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From aahz at panix.com  Fri Feb  2 20:08:07 2001
From: aahz at panix.com (Aahz Maruch)
Date: 2 Feb 2001 11:08:07 -0800
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk> <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk> <mailman.981139624.1198.python-list@python.org>
Message-ID: <95f0in$907$1@panix3.panix.com>

In article <mailman.981139624.1198.python-list at python.org>,
Thomas Wouters  <thomas at xs4all.net> wrote:
>
>(Don't get me wrong: I seriously think 'from foo import *' is overused. The
>language reference makes it pretty clear what it was intended for, and I
>think even the common 'from Tkinter import *' is wrong use -- but I don't
>like code breakage, regardless.)

Yup.  I even have one slide in my upcoming Thread Tutorial at IPC9
devoted to editorializing about import.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From robin at jessikat.fsnet.co.uk  Fri Feb  2 20:51:41 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 2 Feb 2001 19:51:41 +0000
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
 <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk>
 <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk>
 <mailman.981139624.1198.python-list@python.org>
Message-ID: <qUQrSUAN$we6EwEu@jessikat.fsnet.co.uk>

In article <mailman.981139624.1198.python-list at python.org>, Thomas
Wouters <thomas at xs4all.net> writes
.....
>Now, now, 'hacked' is a bit strong. Jeremy did a fairly extensive rewrite of
>the compiler, complete and decent enough not to be called a 'hack' for sure.
>Besides, developers never 'hack', they 'extend' :-) The problem with 'from
>foo import *' in a function or class scope, though, is that it makes it
>impossible for the compiler to see what names will be defined in that scope,
>and hence what names a nested scope uses will resolve to that. I don't like
>it at all, either. In fact, I was the first to whine about it, incessantly,
>on python-dev. I think I have to buy Jeremy a beer (or whatever's his
>poison) on the Python conference to make up ;)
>

I really don't understand why this short cut should be considered such
an abomination. Not using it leads to bad maintenance problems ie I
either flood my module with names by importing right at the top or I am
forced to explicitly import in the target scope. Any name changes in the
origin module must then be pursued over the whole world.

As for the compiler statically determining what's in scope I thought the
whole point of python was that it was dynamic and did late binding.

I can't figure out how any compiler is going to figure out

that

def outer(name):
 exec name + '=1'
 def inner():
  return g
 return inner()

can work if I call outer('g') so this must only be static scoping; if
it's static why should dynamic events like imports be clashing with it.
I suppose there's some attempt at code optimisation being pursued. 

Optimality is the enemy of stability. 
-- 
Robin Becker


From jeremy at alum.mit.edu  Sat Feb  3 00:07:10 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 02 Feb 2001 23:07:10 GMT
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk> <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk> <mailman.981139624.1198.python-list@python.org> <qUQrSUAN$we6EwEu@jessikat.fsnet.co.uk>
Message-ID: <95fein$shj$1@nnrp1.deja.com>

In article <qUQrSUAN$we6EwEu at jessikat.fsnet.co.uk>,
  Robin Becker <robin at jessikat.fsnet.co.uk> wrote:
  [reflecting on the virtues of 'from ... import *':]
> I really don't understand why this short cut should be considered such
> an abomination. Not using it leads to bad maintenance problems ie I
> either flood my module with names by importing right at the top or I
> am forced to explicitly import in the target scope. Any name changes
> in the origin module must then be pursued over the whole world.

I don't understand your complaint.  The first part I follow: You don't
want to import names into the module scope.  Then you say that explicit
imports will cause you to pursue changes over the whole world.  I don't
get that.

If you have a function that does "from mymodule import *", you
presumably use some names defined in mymodule like eggs or spam.  If you
change mymodule so that eggs is replaced with beans, you've got to track
down any client of mymodule and update it.  If you used "from mymodule
import eggs", then you could search for all import lines that mention
mymodule and eggs and find all the cases.  (This is pursuing it over the
whole world, but it's what you get when you change the interface of a
module.)

If you used "from mymodule import *", you'll need to search for all
scopes that import mymodule and then see if they also use the name
eggs.  It's much harder to find all the cases, because you aren't being
explicit about which names you're using from mymodule.

On the third hand, if you change the name foo in mymodule to bar, but
foo was not used, then it's immaterial whether you used import * or an
explicit import.

Unless I'm missing something, your maintenance problems are reduced by
never using import *.

> As for the compiler statically determining what's in scope I thought
> the whole point of python was that it was dynamic and did late
> binding.

I think that's a bit of an exageration :-).  I don't think there's a
"whole point" to Python.

> I can't figure out how any compiler is going to figure out
> that
>
> def outer(name):
>     exec name + '=1'
>     def inner():
>         return g
>     return inner()
>
> can work if I call outer('g') so this must only be static scoping; if
> it's static why should dynamic events like imports be clashing with
> it. I suppose there's some attempt at code optimisation being pursued.

Interesting point you raise.  The compiler is supposed to flag this as
an error, just like import * and for exactly the same reason.  Dynamic
name binding via exec or import * are at odds with static scoping -- and
you shouldn't try to mix the two.  I'll have to fix this in the next
release.

--
-- Jeremy Hylton, <http://www.python.org/~jeremy/>


Sent via Deja.com
http://www.deja.com/


From robin at jessikat.fsnet.co.uk  Sat Feb  3 01:46:55 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 3 Feb 2001 00:46:55 +0000
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
 <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk>
 <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk>
 <mailman.981139624.1198.python-list@python.org>
 <qUQrSUAN$we6EwEu@jessikat.fsnet.co.uk> <95fein$shj$1@nnrp1.deja.com>
Message-ID: <BxsxbLA$T1e6EwSf@jessikat.fsnet.co.uk>

In article <95fein$shj$1 at nnrp1.deja.com>, Jeremy Hylton
<jeremy at alum.mit.edu> writes
...
probaly correct analysis of maintainance elided.

The only difference on a name change would be the import as well as the
actual reference.

>I think that's a bit of an exageration :-).  I don't think there's a
>"whole point" to Python.
>
>> I can't figure out how any compiler is going to figure out
>> that
>>
>> def outer(name):
>>     exec name + '=1'
>>     def inner():
>>         return g
>>     return inner()
>>
>> can work if I call outer('g') so this must only be static scoping; if
>> it's static why should dynamic events like imports be clashing with
>> it. I suppose there's some attempt at code optimisation being pursued.
>
>Interesting point you raise.  The compiler is supposed to flag this as
>an error, just like import * and for exactly the same reason.  Dynamic
>name binding via exec or import * are at odds with static scoping -- and
>you shouldn't try to mix the two.  I'll have to fix this in the next
>release.

The implication of this is that we're forbidden to use exec if it
clashes with a statically scoped name or am I still being stupid?

Python is a joy to use because it lacks such pedantic restrictions.

>--
>-- Jeremy Hylton, <http://www.python.org/~jeremy/>
>
>
>Sent via Deja.com
>http://www.deja.com/

-- 
Robin Becker


From tim.one at home.com  Sat Feb  3 07:30:12 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 01:30:12 -0500
Subject: nested scopes
In-Reply-To: <qUQrSUAN$we6EwEu@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCEECNINAA.tim.one@home.com>

[Robin Becker]
> ...
> As for the compiler statically determining what's in scope I
> thought the whole point of python was that it was dynamic and
> did late binding.

Na, the whole point of Python was to make Guido filthy rich.  It was
actually a bet between him and Bill Gates, and, despite what Microsoft may
tell you, the final outcome is still uncertain <ahem>.

WRT scopes, the determination of which names are local is entirely done at
compile-time.  That wasn't always true, but has been true since version
0.9.9, and was the single biggest speedup in Python's history.  The
determination of which names are local is still done at compile-time in the
presence of "import *" and "exec", but then the results can't be trusted,
and the compiler generates different code that does all sorts of crap under
the covers to give "the expected" results -- most of the time.  I doubt that
crap will get extended to live nice with deeply nested scopes, though.

> ...
> Optimality is the enemy of stability.

in-america-they-tell-us-it's-drugs<wink>-ly y'rs  - tim




From robin at jessikat.fsnet.co.uk  Sat Feb  3 12:25:23 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 3 Feb 2001 11:25:23 +0000
Subject: nested scopes
References: <qUQrSUAN$we6EwEu@jessikat.fsnet.co.uk>
 <mailman.981181863.15223.python-list@python.org>
Message-ID: <467PpKAjq+e6EwFJ@jessikat.fsnet.co.uk>

In article <mailman.981181863.15223.python-list at python.org>, Tim Peters
<tim.one at home.com> writes
>[Robin Becker]
....
>WRT scopes, the determination of which names are local is entirely done at
>compile-time.  That wasn't always true, but has been true since version
>0.9.9, and was the single biggest speedup in Python's history.  The
>determination of which names are local is still done at compile-time in the
>presence of "import *" and "exec", but then the results can't be trusted,
>and the compiler generates different code that does all sorts of crap under
>the covers to give "the expected" results -- most of the time.  I doubt that
>crap will get extended to live nice with deeply nested scopes, though.
>
>> ...
>> Optimality is the enemy of stability.
>
>in-america-they-tell-us-it's-drugs<wink>-ly y'rs  - tim
>
>
as I suspected this is about optimality; speed vs clarity/simplicity is
the argument here. I guess that the nested scopes stuff has just brought
it to the fore. Luckily we're not forced to use nested scopes in python,
pascal experience makes me despise them because of all that paging up
and down to find the defining declaration. 
-- 
Robin Becker


From tim.one at home.com  Sat Feb  3 20:57:51 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 14:57:51 -0500
Subject: nested scopes
In-Reply-To: <467PpKAjq+e6EwFJ@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEELINAA.tim.one@home.com>

[Robin Becker]
> as I suspected this is about optimality;

Since the local vrbl optimization happened 7 years ago, it really is too
late to cry "foul!" on that one <wink>.  BTW, the *rules* for when a name is
local did *not* change then; they simply got exploited; and Python actually
got more predictable because of it (although only in obscure endcases, none
of which I saw that weren't contrived).

> speed vs clarity/simplicity is the argument here.

Sorry, but there isn't, and never was, great clarity about what "import *"
and "exec" do in all cases; that's why the Reference Manual hedged about
them even before the local vrbl optimization was conceived.  You have always
had to be a bona fide expert in Python's internals to predict their behavior
in cases where interaction with the namespaces is anything less than bloody
straightforward.  That was true before the local vrbl optimization too.

> I guess that the nested scopes stuff has just brought it to the fore.

It's still in the background for me <wink>.

> Luckily we're not forced to use nested scopes in python, pascal
> experience makes me despise them because of all that paging up
> and down to find the defining declaration.

I'll repeat something I've said dozens of times over the years in the
ongoing scope debate:  it's going to be much harder in Python, because there
aren't any "defining declarations" in Python.  You have to search up and
search down, looking at all assignments, all imports, all "for" loops
(whether or not nested in listcomps) ... everywhere a binding occurs.

Python's "2-level" scheme was a deliberate counterpoint to the abuses of
deep lexical nesting in Pascal programs of the time (according to Guido),
and it's something I've grown to like a lot.  But newbies eternally stumble
over that functions nest textually but not lexically in Python, and this
will make life simpler for them.  People mucking with excruciatingly subtle
"exec" and "import" tricks arguably deserved to be shot anyway <wink>.

not-that-a-plain-"import-*"-at-function-scope-is-excruciatingly-
    subtle-ly y'rs  - tim




From fredrik at effbot.org  Sat Feb  3 21:55:12 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 20:55:12 GMT
Subject: nested scopes
References: <mailman.981230290.31442.python-list@python.org>
Message-ID: <Qe_e6.11370$AH6.1837949@newsc.telia.net>

Tim Peters wrote:
> Python's "2-level" scheme was a deliberate counterpoint to the abuses of
> deep lexical nesting in Pascal programs of the time (according to Guido),
> and it's something I've grown to like a lot.  But newbies eternally stumble
> over that functions nest textually but not lexically in Python, and this
> will make life simpler for them.

I'm beginning to suspect that a new "sorry, you can no longer
nest def statements" might have been an easier sell...

Cheers /F




From aahz at panix.com  Sat Feb  3 22:00:12 2001
From: aahz at panix.com (Aahz Maruch)
Date: 3 Feb 2001 13:00:12 -0800
Subject: nested scopes
References: <mailman.981230290.31442.python-list@python.org> <Qe_e6.11370$AH6.1837949@newsc.telia.net>
Message-ID: <95hrgs$3hu$1@panix6.panix.com>

In article <Qe_e6.11370$AH6.1837949 at newsc.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
>Tim Peters wrote:
>>
>> Python's "2-level" scheme was a deliberate counterpoint to the abuses of
>> deep lexical nesting in Pascal programs of the time (according to Guido),
>> and it's something I've grown to like a lot.  But newbies eternally stumble
>> over that functions nest textually but not lexically in Python, and this
>> will make life simpler for them.
>
>I'm beginning to suspect that a new "sorry, you can no longer
>nest def statements" might have been an easier sell...

Taking that seriously, that would cause problems for classes.  Not to
mention the lambda issue, which I personally would be just as happy to
see go bye-bye.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From tim.one at home.com  Sun Feb  4 05:53:47 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 23:53:47 -0500
Subject: nested scopes
In-Reply-To: <95hrgs$3hu$1@panix6.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEFOINAA.tim.one@home.com>

[/F]
> I'm beginning to suspect that a new "sorry, you can no longer
> nest def statements" might have been an easier sell...

Not a chance.  You can't stop people from complaining, you can only
influence *what* they complain about.  If Python never ever changed anything
again, there would be howls of protest against that too.  It's much easier
to bear complaints about enforcing something the Reference Manual always
said you shouldn't do (while it's far from certain things will stay this way
for the beta release, if they do then "sorry, but tough luck" is a
defensible response).

[Aahz, Copyright 2001 by aahz at pobox.com, reproduced without permission]
> Taking that seriously, that would cause problems for classes.

/F meant "a def block can no longer contain a def block".  It's curious how
few have ever complained that a def inside a class doesn't have automagic
access to the class locals.  But do the same thing wrt a def in a def, and
it's like Guido is trying to use the time machine to go back and kill John
McCarthy's mother before he was born <wink>.

> Not to mention the lambda issue, which I personally would be just
> as happy to see go bye-bye.

Well, it *has*, but in the way people expect:

>>> def f(n):
...     return lambda i: i + n
...
>>> f(3)(4)
7
>>>

No more "n=n" abuse needed.  Politically impossible to remove lambda at this
stage anyway.  lambda is the print>> of Pytnon functions <wink>.




From clee at thalamus.wustl.edu  Sat Feb  3 23:35:31 2001
From: clee at thalamus.wustl.edu (clee at thalamus.wustl.edu)
Date: Sat, 03 Feb 2001 22:35:31 GMT
Subject: nested scopes
References: <467PpKAjq+e6EwFJ@jessikat.fsnet.co.uk> <mailman.981230290.31442.python-list@python.org>
Message-ID: <95i13k$pfg$1@nnrp1.deja.com>

In article <mailman.981230290.31442.python-list at python.org>,
  "Tim Peters" <tim.one at home.com> wrote:

> Python's "2-level" scheme was a deliberate counterpoint to the abuses of
> deep lexical nesting in Pascal programs of the time (according to Guido),
> and it's something I've grown to like a lot.  But newbies eternally
stumble
> over that functions nest textually but not lexically in Python, and this
> will make life simpler for them.

Out of curiosity, were explicit declarations for whether scopes nest
ever considered?  Something like:

def f():
    A = 1
    closure: # all functions declared in this block can access f's
             # variables
        def g():
            A = 2

Alternative names for closure might be be "bindlocals" or something.  I
guess this doesn't minimize a beginner's surprise, but it would prevent
code breakage.

  -chris


Sent via Deja.com
http://www.deja.com/


From robin at jessikat.fsnet.co.uk  Sun Feb  4 00:59:59 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 3 Feb 2001 23:59:59 +0000
Subject: nested scopes
References: <467PpKAjq+e6EwFJ@jessikat.fsnet.co.uk>
 <mailman.981230290.31442.python-list@python.org>
Message-ID: <$NN+ELA$tJf6Ewfn@jessikat.fsnet.co.uk>

In article <mailman.981230290.31442.python-list at python.org>, Tim Peters
<tim.one at home.com> writes
>[Robin Becker]
>> as I suspected this is about optimality;
>
>Since the local vrbl optimization happened 7 years ago, it really is too
>late to cry "foul!" on that one <wink>.  BTW, the *rules* for when a name is
>local did *not* change then; they simply got exploited; and Python actually
>got more predictable because of it (although only in obscure endcases, none
>of which I saw that weren't contrived).
>
...
it's never too late to cry foul
-- 
Robin Becker


From root at rainerdeyke.com  Sat Feb  3 18:24:51 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 03 Feb 2001 17:24:51 GMT
Subject: nested scopes
References: <mailman.981181863.15223.python-list@python.org>
Message-ID: <D9Xe6.160071$ge4.56951006@news2.rdc2.tx.home.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.981181863.15223.python-list at python.org...
> WRT scopes, the determination of which names are local is entirely done at
> compile-time.  That wasn't always true, but has been true since version
> 0.9.9, and was the single biggest speedup in Python's history.  The
> determination of which names are local is still done at compile-time in
the
> presence of "import *" and "exec", but then the results can't be trusted,
> and the compiler generates different code that does all sorts of crap
under
> the covers to give "the expected" results -- most of the time.  I doubt
that
> crap will get extended to live nice with deeply nested scopes, though.

"Most of the time" is significant here.  There are at least two
inconsistencies in functions with 'import *' as of 2.0.  Can you say "source
of subtle bugs"?


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From tim.one at home.com  Sat Feb  3 20:57:50 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 14:57:50 -0500
Subject: nested scopes
In-Reply-To: <D9Xe6.160071$ge4.56951006@news2.rdc2.tx.home.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEELINAA.tim.one@home.com>

[Tim]
> ...
> The determination of which names are local is still done at
> compile-time in the presence of "import *" and "exec", but then
> the results can't be trusted, and the compiler generates
> different code that does all sorts of crap under the covers to
> give "the expected" results -- most of the time. ....

[Rainer Deyke]
> "Most of the time" is significant here.

I used to lose about 100 hairs a day.  Now I think it's closer to 101.  Much
more significant to me <wink>.

> There are at least two inconsistencies in functions with 'import *'
> as of 2.0.  Can you say "source of subtle bugs"?

In this case, no.  Yes, "import *" is a notorious source of subtle bugs, but
all of them reported in real life are due to its advertised behavior (like
"hey, 'open()' is broken!", "works for me", "no, it's broken", "what did you
change?", "nothing!", "did it used to work?", "well, yes, but then it broke
for no reason at all!", "and you changed nothing", "well, nothing
relevant -- it just broke!", "so you did change something -- what (exactly)
did you change?", "I added a global counter, so maybe the compiler generates
bad code now ... and, OK, sure, I added 'from os import *'", ... and that
one is so frequent the replybots automatically respond  "from os import *"
whenever somebody complains "hey, 'open()' is broken!" now).

I posted some cases after the 0.9.9 release that showed endcase differences
in "import *" and "exec" semantics due to the then-new local variable
optimization.  All of them were contrived.  Guido didn't care; neither did
I.  In the approximately seven years since then, to the best of my knowledge
no bug report on "import *" has ever been filed, and perhaps two were filed
on "exec" but in non-contrived cases that were clearly broken.

Nevertheless, if you fear the import* "bugs" *you* contrived, SourceForge is
open 24 hours a day for patches <wink>.

never-use-import-*-myself-except-in-an-interactive-shell-ly y'rs  - tim




From scarblac at pino.selwerd.nl  Sun Feb  4 15:50:31 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 4 Feb 2001 14:50:31 GMT
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk> <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk> <mailman.981139624.1198.python-list@python.org>
Message-ID: <slrn97qqjl.lc4.scarblac@pino.selwerd.nl>

Thomas Wouters <thomas at xs4all.net> wrote in comp.lang.python:
> It won't be optional, it'll be part of Python. I guess that if there is
> a *really* *really* *really* lot of code out there that will break, it could
> be withheld until Python 2.2, and the practice of 'from foo import *' in
> function/class scopes be made to issue a warning in Python 2.1, but frankly
> I doubt there'll be that much code breakage. At least it's not a *silent*
> breakage ! It's a compiler error, so the first run of a script under the new
> Python will detect it. It's not like it's a new kind of exception being
> raised.

Argh. I use this quite a lot. To simulate enums, I have a module that
defines a lot of constants, and these are used in lots of little functions
all over the place, but I wouldn't want those constants in every module's
global namespace. So I use from import *... I'm not even sure if I can
switch to importing them at the start of the module because there might be
name clashes.

So because of the Pascal fans, we can't use import * anymore, and always
have to type full module names? I'm starting to agree with the people who
say that it's all going downhill... This sort of thing was meant for 
Python 3000! For a reason!!

-- 
Remco Gerlich




From scarblac at pino.selwerd.nl  Sun Feb  4 16:34:46 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 4 Feb 2001 15:34:46 GMT
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk> <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk> <mailman.981139624.1198.python-list@python.org> <slrn97qqjl.lc4.scarblac@pino.selwerd.nl>
Message-ID: <slrn97qt6k.lc4.scarblac@pino.selwerd.nl>

Remco Gerlich <scarblac at pino.selwerd.nl> wrote in comp.lang.python:
> So because of the Pascal fans, we can't use import * anymore, and always
> have to type full module names? I'm starting to agree with the people who
> say that it's all going downhill... This sort of thing was meant for 
> Python 3000! For a reason!!

Actually, I was reading some things in the PEP wrong, about masking local
variables and so on. It's ok now, I can do without the from import thing,
I've tried alpha 2 out, and most things I tried worked exactly how I hoped
and not how I feared, so now I like the patch :-).

-- 
Remco Gerlich



From tim.one at home.com  Sat Feb  3 05:19:19 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 2 Feb 2001 23:19:19 -0500
Subject: nested scopes
In-Reply-To: <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCOECCINAA.tim.one@home.com>

>> def bingo():
>>   from string import *
> >    ....

[John J. Lee]
> Are nested scopes actually going to end up in Python proper,
> then?  As opposed to remaining as an optional patch?

They're *in* 2.1a2, released earlier today.

> And why should the above become bad rather than just import
> into the appropriate namespace?

Whether a name is local to a given scope is, in Python, determined by
whether the name is *bound* somewhere in that scope.  "import *" counts as a
binding, but the compiler has no idea which names are being bound.  So when
compiling code for a nested function N, referencing a non-local name (one
that is not bound to within N's body) X, the compiler has no idea which
scope X belongs to if an "import *" (or "exec" stmt) appears in an enclosing
function E.  Any name whatsoever *may* become local in E then, but it's
impossible to know anything about that at compile-time.

If you still think it's easy, submit a patch <wink>.

invitation-or-threat-is-in-the-eye-of-the-beholder-ly y'rs  - tim




From loewis at informatik.hu-berlin.de  Fri Feb  2 23:14:04 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 23:14:04 +0100
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
Message-ID: <p6qbsskohqr.fsf@informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> Is it really true that J Hylton's nested scopes will cause code like
> 
> def bingo():
>   from string import *
>   ....
> 
> to be declared bad? 

That code has been bad for a long time. See

http://www.python.org/doc/current/ref/import.html

which says

# The from form with "*" may only occur in a module scope.

It is only now that this restriction is enforced by CPython; again,
the code was incorrect for a long time (see

http://www.python.org/doc/1.4/ref/ref6.html#REF17806

to get my point; it was also in the 1.1 version of the language
reference).

Regards,
Martin


From robin at jessikat.fsnet.co.uk  Sat Feb  3 01:47:55 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 3 Feb 2001 00:47:55 +0000
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
 <p6qbsskohqr.fsf@informatik.hu-berlin.de>
Message-ID: <BxtxLXA7U1e6EwyI@jessikat.fsnet.co.uk>

In article <p6qbsskohqr.fsf at informatik.hu-berlin.de>, Martin von Loewis
<loewis at informatik.hu-berlin.de> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> Is it really true that J Hylton's nested scopes will cause code like
>> 
>> def bingo():
>>   from string import *
>>   ....
>> 
>> to be declared bad? 
>
>That code has been bad for a long time. See
>
>http://www.python.org/doc/current/ref/import.html
>
>which says
>
># The from form with "*" may only occur in a module scope.
>
>It is only now that this restriction is enforced by CPython; again,
>the code was incorrect for a long time (see
>
>http://www.python.org/doc/1.4/ref/ref6.html#REF17806
>
>to get my point; it was also in the 1.1 version of the language
>reference).
>
>Regards,
>Martin
imho it's the restriction which is bad. It's just a bunch of imports and
assignments.
-- 
Robin Becker


From mwh21 at cam.ac.uk  Tue Feb  6 16:55:53 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Feb 2001 15:55:53 +0000
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <p6qbsskohqr.fsf@informatik.hu-berlin.de>
Message-ID: <m3y9vj3iwm.fsf@atrus.jesus.cam.ac.uk>

Martin von Loewis <loewis at informatik.hu-berlin.de> writes:

> Robin Becker <robin at jessikat.fsnet.co.uk> writes:
> 
> > Is it really true that J Hylton's nested scopes will cause code like
> > 
> > def bingo():
> >   from string import *
> >   ....
> > 
> > to be declared bad? 
> 
> That code has been bad for a long time. See
> 
> http://www.python.org/doc/current/ref/import.html
> 
> which says
> 
> # The from form with "*" may only occur in a module scope.
> 
> It is only now that this restriction is enforced by CPython; again,
> the code was incorrect for a long time (see
> 
> http://www.python.org/doc/1.4/ref/ref6.html#REF17806
> 
> to get my point; it was also in the 1.1 version of the language
> reference).

I've been meaning to say this for a few days, but finally I get round
to saying that this particular argument bothers me; Python isn't a
language you program with a reference manual open at all times
(unlike, say, C++) - and this IMHO is a good thing.  This particular
change probably isn't too stressful, but I slightly fear more onerous
restrictions might be lurking in the ref manual to be whipped out when
we next get comfortable with the language.

Cheers,
M.

-- 
  (Of course SML does have its weaknesses, but by comparison, a
  discussion of C++'s strengths and flaws always sounds like an
  argument about whether one should face north or east when one
  is sacrificing one's goat to the rain god.)         -- Thant Tessman


From loewis at informatik.hu-berlin.de  Fri Feb  9 18:46:34 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Feb 2001 18:46:34 +0100
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <p6qbsskohqr.fsf@informatik.hu-berlin.de> <m3y9vj3iwm.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <p6qsnln3g1x.fsf@informatik.hu-berlin.de>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> I've been meaning to say this for a few days, but finally I get round
> to saying that this particular argument bothers me; Python isn't a
> language you program with a reference manual open at all times
> (unlike, say, C++) - and this IMHO is a good thing.  This particular
> change probably isn't too stressful, but I slightly fear more onerous
> restrictions might be lurking in the ref manual to be whipped out when
> we next get comfortable with the language.

It certainly must be decided on a case-by-case bases, evaluating the
impacts in each case. If (like in this case), the ruled-out code uses
what many consider bad style, and if the code contradicts with a new
desirable feature, then it seems acceptable to break that code. If
code is broken, there should be clear instructions for people affected
how to modify their code.

In other cases, it may be better to remove the restriction from the
documentation.

Regards,
Martin



From jeremy at alum.mit.edu  Fri Feb  2 23:48:24 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 02 Feb 2001 22:48:24 GMT
Subject: nested scopes
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
Message-ID: <95fdfk$rfc$1@nnrp1.deja.com>

In article <msXosJAT0oe6Ewmn at jessikat.fsnet.co.uk>,
  Robin Becker <robin at jessikat.fsnet.co.uk> wrote:
> Is it really true that J Hylton's nested scopes will cause code like
>
> def bingo():
>   from string import *
>   ....
>
> to be declared bad? I'm already against nested scopes, but requiring
> this to be wrong seems awful. We seem to be replacing a simple 2-level
> system by a more complex one & restrictions.

This code will not work in 2.1 alpha 2.  It will raise a SyntaxError
exception.  I expect the restriction to be eased somewhat for 2.1b1, so
that it will be allowed as long as the function doesn't contain other
function definitions.

This is partly the result of the nested scopes change, and partly the
result of some confusion about the reference manual.  I'll explain the
real problem first.

def f():
    from string import strip
    def g(s):
        return strip(s)
    return g

In 2.1a2, the name strip in g() will use the binding for strip in f().
The resolution of names is determined at compile time.  If f() didn't
have a binding for strip, then the use in g() would check in globals
and builtins.  I think this is straightforward.

The 'from ... import *' form is a problem, because the compiler can't
tell what names will be present in a function scope at compile time.

def f():
    from string import *
    def g(s):
        return strip(s)
    return g

The compiler has know way to know whether 'from string import *' will
create a binding for strip in f().  It could ignore the possibility
that a binding would be created at runtime and treat strip as a global,
but that would be a confusing special case.  Instead, the compiler flags
it as an error and forces the programmer to be explicit about the
binding.

While I was working on nested scopes, I noticed that the 2.0 reference
manual says that "from ... import *" is illegal anywhere except at the
module level (and noted that the compiler did not enforce this rule).
Since I was modifying the compiler anyway, I decided to enforce this
rule.  It meant I didn't have to work about the case I just described
above, because any use of 'from ... import *' would be illegal.

No one seems to remember why the reference manual outlawed this form of
import.  (It also outlawed assigning to a name declared global, e.g.
"def f(): global time; import time" and we're not sure about that
either.)  This form seems to be used often enough that we should allow
it when there is no ambiguity.  I plan to fix the compiler for the
next release so that it only complains when 'from ... import *' is used
in the presence of nested scopes.  As Thomas noted in another message,
use of a lambda statement counts as "in the presence of nested scopes."

--
-- Jeremy Hylton, <http://www.python.org/~jeremy/>


Sent via Deja.com
http://www.deja.com/


From geoff at homegain.com  Sat Feb  3 03:38:54 2001
From: geoff at homegain.com (Geoffrey Gerrietts)
Date: Fri, 2 Feb 2001 18:38:54 -0800 
Subject: nested scopes
Message-ID: <393D961EA4FCD311B03F00D0B746D65802625F27@HQSXCH01>

Jeremy writes:
> I don't understand your complaint.  The first part I follow: You don't
> want to import names into the module scope.  Then you say 
> that explicit
> imports will cause you to pursue changes over the whole 
> world.  I don't
> get that.

I think that part of your confusion might be based on your conception
of what a module is for. The premise upon which you base your argument
is that you will not be using the names you acquire in a dynamic way.

For example, if I have a large file, or a large collection of files, 
all of which contain a large number of variable definitions, I can suck 
those up into a function and use dir() to grab the list and do things
with them.

Harder now, without polluting your global namespace, but you can presumably
load all the names into a local dictionary, assuming __import__ will not 
restrict us.

An alternative situation where this might be used is in fact highlighted
by the very case that the limitation allowed 2.1 to ignore; if we are 
trying to exploit the dynamism of the language by executing one piece of
code
while changing out the names it imports from underneath it.

While I can't say I don't find nested scopes interesting, I confess that 
I find them just as troublesome, and I'm really kinda sad to see dynamic
features yanked from the language (even if they're unintentional features)
just so we can have nested scoping.

---
Geoff Gerrietts <geoff at homegain.com>



From tim.one at home.com  Sat Feb  3 05:11:50 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 2 Feb 2001 23:11:50 -0500
Subject: nested scopes
In-Reply-To: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCIECBINAA.tim.one@home.com>

[Robin Becker]
> Is it really true that J Hylton's nested scopes will cause code like
>
> def bingo():
>   from string import *
>   ....
>
> to be declared bad?

As has been pointed out, the "*" form of import has always been documented
as unsupported at any but module scope.  Sometimes Guido wishes people would
read the manuals he worked so hard to write <wink>.

But as the 2.1a2 NEWS file says,

    In a future release, the compiler may allow this form when
    there is no possibility of ambiguity.

I guess this depends on how loudly people scream.  Before screaming too
loudly, though, realize that in 2.0 and before, the mere presence of "import
*" in a function forces the compiler to generate worse (slower) code for the
entire function body, since it no longer has any idea which names are local,
and so has to treat all names as if they "might be" local.  "exec" stmts
have the same effect on the scope in which they appear.

> I'm already against nested scopes, but requiring this to be wrong
> seems awful.

Was already wrong, and it's unclear how making your code clearer and faster
is "awful".

> We seem to be replacing a simple 2-level system by a more complex
> one & restrictions.

Exactly right, except that it has been a 3-level system (which Guido called
a 2-level system in a successful attempt to make people think it's simpler
than it is <wink>).

I don't think I know of any other language trying to mix arbitrarily deeply
nested scopes with lack of declarations.  The schemes for "import *" and
"exec" were hanging by a thread before; it's doubtful they'll ever work well
with deep nesting.

I may have missed one, but I counted three checkins to repair nested use of
"import *" in the CVS tree for the 2.1a2 release.  That's not much out of
aobut ~6Mb of Python source code.  "from Tkinter import *" is the leader for
worst offender.  2/3 of the offending lines were written by Guido, 1/3 by
Barry.

so-i-suppose-guido-is-twice-as-outraged-ly y'rs  - tim




From loewis at informatik.hu-berlin.de  Sat Feb  3 12:17:17 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 03 Feb 2001 12:17:17 +0100
Subject: nested scopes
References: <mailman.981171654.11907.python-list@python.org>
Message-ID: <p6qhf2crp6q.fsf@informatik.hu-berlin.de>

Geoffrey Gerrietts <geoff at homegain.com> writes:

> I think that part of your confusion might be based on your conception
> of what a module is for. The premise upon which you base your argument
> is that you will not be using the names you acquire in a dynamic way.

I think this is a fair assumption. You give a counter-example;
unfortunately, I cannot understand that example:

> For example, if I have a large file, or a large collection of files,
> all of which contain a large number of variable definitions, I can
> suck those up into a function and use dir() to grab the list and do
> things with them.

How do you exactly do this? If you have a list of module names, it
appears that you will need to do

def foo(module_names):
  for m in module_names:
    exec "from %s import *" % m
  return dir()

or something like that. That continues to work in the current CVS
Python, since the check for import is a static check, not a dynamic
one. Or did you have a different algorithm in mind which would stop
working?

In any case, it might be better to rewrite this function as

def foo(module_names):
  result = []
  module = {}
  for m in module_names:
    exec "from %s import *" % m in module
    result += dir(module[m])
  return result

Even better might be

def foo(module_names):
  result = {}
  for m in module_names:
    exec "from %s import *" % m in result
  return dir(x)

With that, you execute the import neither in local nor in global
namespace; you do so in a "fresh" namespace.

> While I can't say I don't find nested scopes interesting, I confess
> that I find them just as troublesome, and I'm really kinda sad to
> see dynamic features yanked from the language (even if they're
> unintentional features) just so we can have nested scoping.

I can't see any dynamic features being yanked - I can see that of the
many ways to do a certain dynamic thing, one way is removed.
There's Only One Way To Do It, anyway.

Regards,
Martin



From root at rainerdeyke.com  Sat Feb  3 18:24:51 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 03 Feb 2001 17:24:51 GMT
Subject: nested scopes
References: <mailman.981171654.11907.python-list@python.org> <p6qhf2crp6q.fsf@informatik.hu-berlin.de>
Message-ID: <D9Xe6.160072$ge4.56951782@news2.rdc2.tx.home.com>

"Martin von Loewis" <loewis at informatik.hu-berlin.de> wrote in message
news:p6qhf2crp6q.fsf at informatik.hu-berlin.de...
> def foo(module_names):
>   for m in module_names:
>     exec "from %s import *" % m
>   return dir()

Yuck.

def foo(*names):
  dict = {}
  for name in names:
    dict.update(__import__(name).__dict__)
  return dict

Note the absensce of local variables 'm' and 'module_names' in the returned
dictionary.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From loewis at informatik.hu-berlin.de  Fri Feb  9 18:48:14 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Feb 2001 18:48:14 +0100
Subject: nested scopes
References: <mailman.981171654.11907.python-list@python.org> <p6qhf2crp6q.fsf@informatik.hu-berlin.de> <D9Xe6.160072$ge4.56951782@news2.rdc2.tx.home.com>
Message-ID: <p6qpugr3fz5.fsf@informatik.hu-berlin.de>

"Rainer Deyke" <root at rainerdeyke.com> writes:

> "Martin von Loewis" <loewis at informatik.hu-berlin.de> wrote in message
> news:p6qhf2crp6q.fsf at informatik.hu-berlin.de...
> > def foo(module_names):
> >   for m in module_names:
> >     exec "from %s import *" % m
> >   return dir()
> 
> Yuck.
> 
> def foo(*names):
>   dict = {}
>   for name in names:
>     dict.update(__import__(name).__dict__)
>   return dict
> 
> Note the absensce of local variables 'm' and 'module_names' in the returned
> dictionary.

OTOH, note the presence of names starting with an underscore in the
returned dictionary.

Regards,
Martin



From n8gray at caltech.edu.is.my.email.address  Mon Feb  5 11:01:58 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Mon, 5 Feb 2001 02:01:58 -0800
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org>
Message-ID: <95lu1i$ebs@gap.cco.caltech.edu>

Tim Peters wrote:

> Whether a name is local to a given scope is, in Python, determined by
> whether the name is *bound* somewhere in that scope.  "import *" counts as
> a
> binding, but the compiler has no idea which names are being bound.  So
> when compiling code for a nested function N, referencing a non-local name
> (one that is not bound to within N's body) X, the compiler has no idea
> which scope X belongs to if an "import *" (or "exec" stmt) appears in an
> enclosing
> function E.  Any name whatsoever *may* become local in E then, but it's
> impossible to know anything about that at compile-time.

Interesting issue.  What if the following was allowed:
>>> from tim_bot_module import tb*

telling the compiler that only names beginning with "tb" are imported.  
This way you could still design libraries to use from ... import ..* 
without totally throwing efficiency out the window.  I know import * is 
considered harmful, but occasionally (think tkinter) it's a useful 
construct, assuming the library is designed to prevent name conflicts.

Just a thought,
-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From dsh8290 at rit.edu  Mon Feb  5 18:42:42 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 5 Feb 2001 12:42:42 -0500
Subject: nested scopes
In-Reply-To: <95lu1i$ebs@gap.cco.caltech.edu>; from n8gray@caltech.edu.is.my.email.address on Mon, Feb 05, 2001 at 02:01:58AM -0800
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu>
Message-ID: <20010205124241.A27402@harmony.cs.rit.edu>

On Mon, Feb 05, 2001 at 02:01:58AM -0800, Nathaniel Gray wrote:
| Tim Peters wrote:
| 
| > Whether a name is local to a given scope is, in Python, determined by
| > whether the name is *bound* somewhere in that scope.  "import *" counts as
| > a
| > binding, but the compiler has no idea which names are being bound.  So
| > when compiling code for a nested function N, referencing a non-local name
| > (one that is not bound to within N's body) X, the compiler has no idea
| > which scope X belongs to if an "import *" (or "exec" stmt) appears in an
| > enclosing
| > function E.  Any name whatsoever *may* become local in E then, but it's
| > impossible to know anything about that at compile-time.
| 
| Interesting issue.  What if the following was allowed:
| >>> from tim_bot_module import tb*
| 

Hmm, wouldn't it be better to use :

>>> import tim_bot_module as tb
>>> tb.name_you_want_to_use

?

IMO prefixing names with strings like "tb" is for namespace-less
languages like C.  Since python has namespaces, why not use them?


What I wonder about is whether names defined in a try block or a for
block will still exist outside the block.

Take this Java code for example:

String s ;
try
{
	s = some_func_that_might_throw() ;
}
catch ( ThrownException ex )
{
	s = "Some other value" ;
}


So you look at it and whether the function throws an exception or not,
s has been initialized appropriately.  javac won't compile this since
"s might not be initialized".

In Python, it is much nicer :

try :
	s = some_func_that_might_throw()
except ThrownException :
	s = "Some other value"

and everything works as expected.  If the exception was thrown, s will
have the one value, if not it will have the other.


Will the new nested scopes require me to write:

s = None # make the name exist in this scope first
try :
	...

?


I think it is acceptable if variables like loop counters and temps
inside a loop don't exist in the function's namespace. ie:

for i in range( 10 ) :
	pass
print i # this gives NameError


but :

i = 1
for i in range( 10 ) :
	pass
print i # prints '9' since the name i already existed and was rebound



What I wonder about is how important/useful nested scopes really are.
When I first learned python, I had an Eiffel/C++/C background with
maybe a little Java.  Each of those has nested lexical scopes.  I
thought python was weird and lacking.  I have since learned that
python is quite natural and doesn't require various ugly idioms
(declaring the variable at the top) to make the assignment in the
nested scope persist outside the loop/if/whatever. 

Last quarter I took a class called "Programming Language Concepts"
that discussed the design of programming languages.  It was the first
time I had heard of dynamic scoping.  (Actually, I think I saw it
mentioned in a lisp context, but I didn't know what it meant)  The
professor didn't know of any situations where it would be
advantageous, and I couldn't contrive any either.  I was kind of
surprised when I heard python described as "dynamically scoped"
because its scoping was really quite natural for me.  The [23]-level
scoping makes the dynamic scoping almost static and I don't see that
as a weakness.

I really don't know, at this moment, whether adding nested static
scopes will strengthen or weaken the language.

just-a-few-thoughts-ly y'rs -D




From dnew at san.rr.com  Mon Feb  5 19:39:07 2001
From: dnew at san.rr.com (Darren New)
Date: Mon, 05 Feb 2001 18:39:07 GMT
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org>
Message-ID: <3A7EF349.407C71E1@san.rr.com>

D-Man wrote:
> professor didn't know of any situations where it would be
> advantageous, and I couldn't contrive any either.  I was kind of
> surprised when I heard python described as "dynamically scoped"
> because its scoping was really quite natural for me.

That's because your prof doesn't know what "dynamic scoping" is.


def alpha():
  a = 5
  gamma()
  beta()
  gamma()

def beta():
  a = 17
  gamma()

def gamma():
  print a

alpha()

With dynamic scoping, this would print "5" then "17" then "5" again, as
"gamma" would basically walk up the run-time call stack looking for the most
recent "a" to print. (Note that "a" is supposed to be local everywhere, not
a global.) Obviously, this is *not* how Python works.  

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
                 Ignorance can be cured. Naivety cures itself.


From dsh8290 at rit.edu  Tue Feb  6 15:38:09 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 6 Feb 2001 09:38:09 -0500
Subject: nested scopes
In-Reply-To: <3A7EF349.407C71E1@san.rr.com>; from dnew@san.rr.com on Mon, Feb 05, 2001 at 06:39:07PM +0000
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com>
Message-ID: <20010206093809.A1360@harmony.cs.rit.edu>

On Mon, Feb 05, 2001 at 06:39:07PM +0000, Darren New wrote:
| D-Man wrote:
| > professor didn't know of any situations where it would be
| > advantageous, and I couldn't contrive any either.  I was kind of
| > surprised when I heard python described as "dynamically scoped"
| > because its scoping was really quite natural for me.
| 
| That's because your prof doesn't know what "dynamic scoping" is.

Maybe not,  it seemed very confusing.  I still have the book, though I
haven't read it much (maybe I should ...)

Do you know of any situations where it would be advantageous to have
dynamic scoping instead of static scoping?  (Note that I'm not asking
for situations where dynamic scoping exists, but where it would be
better)

| 
| 
| def alpha():
|   a = 5
|   gamma()
|   beta()
|   gamma()
| 
| def beta():
|   a = 17
|   gamma()
| 
| def gamma():
|   print a
| 
| alpha()
| 
| With dynamic scoping, this would print "5" then "17" then "5" again, as
| "gamma" would basically walk up the run-time call stack looking for the most
| recent "a" to print. (Note that "a" is supposed to be local everywhere, not

Yeah, this is what I thought dynamic scoping would do.

| a global.) Obviously, this is *not* how Python works.  

Right, in Python, gamma() would complain that it can't find 'a'.  This
is why I said that python's scope rules are really not hard to
understand.

-D




From dnew at san.rr.com  Tue Feb  6 19:19:16 2001
From: dnew at san.rr.com (Darren New)
Date: Tue, 06 Feb 2001 18:19:16 GMT
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org>
Message-ID: <3A804026.AD3D1540@san.rr.com>

D-Man wrote:
> Do you know of any situations where it would be advantageous to have
> dynamic scoping instead of static scoping?  (Note that I'm not asking
> for situations where dynamic scoping exists, but where it would be
> better)

My understanding is that it was more an artifact of the implementation of
LISP than it was a desirous effect. It might be better if you're trying to
form closures or do funky macro stuff or something, but I'm not sure.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
                 Ignorance can be cured. Naivety cures itself.


From mwh21 at cam.ac.uk  Tue Feb  6 19:43:28 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Feb 2001 18:43:28 +0000
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org> <3A804026.AD3D1540@san.rr.com>
Message-ID: <m3r91b3b5b.fsf@atrus.jesus.cam.ac.uk>

Darren New <dnew at san.rr.com> writes:

> D-Man wrote:
> > Do you know of any situations where it would be advantageous to have
> > dynamic scoping instead of static scoping?  (Note that I'm not asking
> > for situations where dynamic scoping exists, but where it would be
> > better)
> 
> My understanding is that it was more an artifact of the implementation of
> LISP than it was a desirous effect. It might be better if you're trying to
> form closures or do funky macro stuff or something, but I'm not sure.

It's good when you're calling a function (call it A) that calls
another function (call this one B) that has parameterizable behaviour,
because A doesn't have to put all of B's parameters in its argument
list - you get B to read these paramters from special (ie. dynamically
scoped) variables and bind them around your call to A.  Can't think of
a real good example now, I'm afraid.  The Common Lisp printer uses
special variables extensively.

I'm not sure this could be made to work in Python, not least because
it has a rather different notion of binding to Python.

Cheers,
M.

-- 
  Roll on a game of competetive offence-taking. 
                                            -- Dan Sheppard, ucam.chat


From root at rainerdeyke.com  Tue Feb  6 21:10:03 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 06 Feb 2001 20:10:03 GMT
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org> <3A804026.AD3D1540@san.rr.com> <m3r91b3b5b.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <vSYf6.182681$ge4.63154971@news2.rdc2.tx.home.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3r91b3b5b.fsf at atrus.jesus.cam.ac.uk...
> It's good when you're calling a function (call it A) that calls
> another function (call this one B) that has parameterizable behaviour,
> because A doesn't have to put all of B's parameters in its argument
> list - you get B to read these paramters from special (ie. dynamically
> scoped) variables and bind them around your call to A.  Can't think of
> a real good example now, I'm afraid.  The Common Lisp printer uses
> special variables extensively.

Some Python functions (including 'globals' and 'locals') have behavior
similar to that of dynamic scoping.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From mwh21 at cam.ac.uk  Tue Feb  6 21:35:09 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Feb 2001 20:35:09 +0000
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org> <3A804026.AD3D1540@san.rr.com> <m3r91b3b5b.fsf@atrus.jesus.cam.ac.uk> <vSYf6.182681$ge4.63154971@news2.rdc2.tx.home.com>
Message-ID: <m3lmrj35z6.fsf@atrus.jesus.cam.ac.uk>

"Rainer Deyke" <root at rainerdeyke.com> writes:

> "Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
> news:m3r91b3b5b.fsf at atrus.jesus.cam.ac.uk...
> > It's good when you're calling a function (call it A) that calls
> > another function (call this one B) that has parameterizable behaviour,
> > because A doesn't have to put all of B's parameters in its argument
> > list - you get B to read these paramters from special (ie. dynamically
> > scoped) variables and bind them around your call to A.  Can't think of
> > a real good example now, I'm afraid.  The Common Lisp printer uses
> > special variables extensively.
> 
> Some Python functions (including 'globals' and 'locals') have behavior
> similar to that of dynamic scoping.

How so?  Note I'm using "dynamic scope" in a pretty technical sense
here - "indefinite scope and dynamic extent".  Python variables have
definite scope and dynamic extent (although the objects they point to
have indefinite extent).

Here's a reference:

http://ringer.cs.utsa.edu/research/AI/cltl/clm/node43.html

Cheers,
M.

-- 
  ARTHUR:  Why should he want to know where his towel is?
    FORD:  Everybody should know where his towel is.
  ARTHUR:  I think your head's come undone.
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 7


From root at rainerdeyke.com  Tue Feb  6 22:45:39 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 06 Feb 2001 21:45:39 GMT
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org> <3A804026.AD3D1540@san.rr.com> <m3r91b3b5b.fsf@atrus.jesus.cam.ac.uk> <vSYf6.182681$ge4.63154971@news2.rdc2.tx.home.com> <m3lmrj35z6.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <7g_f6.182817$ge4.63254223@news2.rdc2.tx.home.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3lmrj35z6.fsf at atrus.jesus.cam.ac.uk...
> "Rainer Deyke" <root at rainerdeyke.com> writes:
> > Some Python functions (including 'globals' and 'locals') have behavior
> > similar to that of dynamic scoping.
>
> How so?  Note I'm using "dynamic scope" in a pretty technical sense
> here - "indefinite scope and dynamic extent".  Python variables have
> definite scope and dynamic extent (although the objects they point to
> have indefinite extent).

I am referring to the fact that these functions access the namespace of the
calling function.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From mwh21 at cam.ac.uk  Tue Feb  6 22:59:03 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Feb 2001 21:59:03 +0000
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org> <3A804026.AD3D1540@san.rr.com> <m3r91b3b5b.fsf@atrus.jesus.cam.ac.uk> <vSYf6.182681$ge4.63154971@news2.rdc2.tx.home.com> <m3lmrj35z6.fsf@atrus.jesus.cam.ac.uk> <7g_f6.182817$ge4.63254223@news2.rdc2.tx.home.com>
Message-ID: <m3hf27323c.fsf@atrus.jesus.cam.ac.uk>

"Rainer Deyke" <root at rainerdeyke.com> writes:

> "Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
> news:m3lmrj35z6.fsf at atrus.jesus.cam.ac.uk...
> > "Rainer Deyke" <root at rainerdeyke.com> writes:
> > > Some Python functions (including 'globals' and 'locals') have behavior
> > > similar to that of dynamic scoping.
> >
> > How so?  Note I'm using "dynamic scope" in a pretty technical sense
> > here - "indefinite scope and dynamic extent".  Python variables have
> > definite scope and dynamic extent (although the objects they point to
> > have indefinite extent).
> 
> I am referring to the fact that these functions access the namespace of the
> calling function.

Oh, I see (I think).  I wouldn't have said that was an instructive
example myself.

Cheers,
M.

-- 
  All parts should go together without forcing. You must remember that
  the parts you are  reassembling were disassembled by you.  Therefore,
  if you can't get them together again, there must be a reason. By all
  means, do not use a hammer.          -- IBM maintenance manual, 1925


From ben.hutchings at roundpoint.com  Tue Feb  6 21:36:43 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 06 Feb 2001 12:36:43 -0800
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org> <3A804026.AD3D1540@san.rr.com> <m3r91b3b5b.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <ur91ba6qs.fsf@roundpoint.com>

Michael Hudson <mwh21 at cam.ac.uk> writes:
<snip>
[about dynamic scoping]
> It's good when you're calling a function (call it A) that calls
> another function (call this one B) that has parameterizable behaviour,
> because A doesn't have to put all of B's parameters in its argument
> list - you get B to read these paramters from special (ie. dynamically
> scoped) variables and bind them around your call to A.  Can't think of
> a real good example now, I'm afraid.  The Common Lisp printer uses
> special variables extensively.
<snip>

If you consider Unix processes as functions, then a process's
environment is its dynamic scope.  There are many interesting uses for
environmental variables.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From johann at physics.berkeley.edu  Tue Feb  6 22:08:40 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 06 Feb 2001 13:08:40 -0800
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org>
Message-ID: <mtd7cvh63r.fsf@astron.berkeley.edu>

D-Man  writes:

> Do you know of any situations where it would be advantageous to have
> dynamic scoping instead of static scoping?  (Note that I'm not asking
> for situations where dynamic scoping exists, but where it would be
> better)

Well, it's very useful for output.  You can write all of your
functions to output to the standard output stream, and then redirect
that output stream with something like:

;;
;; Pseudo-lisp.  All of my function names are wrong, but I'm lazy.
;;
(let ((*standard-output-stream* (open-output-file "file.dat")))
  (generate-big-report))

That way, output will be redirected while you're inside
generate-big-report, but will automatically be switched back.
In python, a similar thing would be

try:
  temp = sys.stdout                   # redirect output.
  file = open('file.dat', 'w')
  sys.stdout = file
  generate_big_report()
finally:
  file.close()
  sys.stdout = temp                   # make sure that you fix it.

-- 
Johann Hibschman                           johann at physics.berkeley.edu



From jhylton at my-deja.com  Mon Feb  5 19:57:56 2001
From: jhylton at my-deja.com (Jeremy Hylton)
Date: Mon, 05 Feb 2001 18:57:56 GMT
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org>
Message-ID: <95mt3e$9pe$1@nnrp1.deja.com>

In article <mailman.981395193.31631.python-list at python.org>,
  D-Man <dsh8290 at rit.edu> wrote:
> What I wonder about is whether names defined in a try block or a for
> block will still exist outside the block.
>
> Take this Java code for example:
>
> String s ;
> try
> {
> 	s = some_func_that_might_throw() ;
> }
> catch ( ThrownException ex )
> {
> 	s = "Some other value" ;
> }
>
> So you look at it and whether the function throws an exception or not,
> s has been initialized appropriately.  javac won't compile this since
> "s might not be initialized".
>
> In Python, it is much nicer :
>
> try :
> 	s = some_func_that_might_throw()
> except ThrownException :
> 	s = "Some other value"
>
> and everything works as expected.  If the exception was thrown, s will
> have the one value, if not it will have the other.
>
> Will the new nested scopes require me to write:
>
> s = None # make the name exist in this scope first
> try :
> 	...
>
> ?

No.  Nested scopes don't require you to do anything different unless you
actually use nested scopes.  In places where Java would raise a
compile-time error because it can't tell if a name will be used
unitialized, Python will not complain at comile time.  Instead, it will
raise a NameError if an error occurs at run time.  I won't offer any
opinions about which is better.

> I think it is acceptable if variables like loop counters and temps
> inside a loop don't exist in the function's namespace. ie:

Regardless of whether you find it acceptable, the language doesn't work
that way.

> Last quarter I took a class called "Programming Language Concepts"
> that discussed the design of programming languages.  It was the first
> time I had heard of dynamic scoping.  (Actually, I think I saw it
> mentioned in a lisp context, but I didn't know what it meant)  The
> professor didn't know of any situations where it would be
> advantageous, and I couldn't contrive any either.  I was kind of
> surprised when I heard python described as "dynamically scoped"
> because its scoping was really quite natural for me.  The [23]-level
> scoping makes the dynamic scoping almost static and I don't see that
> as a weakness.

Python's scoping rules aren't dynamic, never have been.  (At least not
as long as I've been using the language.)  It used to be that scopes
didn't nest, but that doesn't make scoping dynamic.

>
> I really don't know, at this moment, whether adding nested static
> scopes will strengthen or weaken the language.

PEP 227 has a brief discussion of some benefits, among them that lambdas
are easier to use.  http://python.sourceforge.net/peps/

--
-- Jeremy Hylton, <http://www.python.org/~jeremy/>


Sent via Deja.com
http://www.deja.com/


From dsh8290 at rit.edu  Tue Feb  6 15:44:53 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 6 Feb 2001 09:44:53 -0500
Subject: nested scopes
In-Reply-To: <95mt3e$9pe$1@nnrp1.deja.com>; from jhylton@my-deja.com on Mon, Feb 05, 2001 at 06:57:56PM +0000
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <95mt3e$9pe$1@nnrp1.deja.com>
Message-ID: <20010206094453.B1360@harmony.cs.rit.edu>

On Mon, Feb 05, 2001 at 06:57:56PM +0000, Jeremy Hylton wrote:
| In article <mailman.981395193.31631.python-list at python.org>,
|   D-Man <dsh8290 at rit.edu> wrote:
[snip]
| 
| No.  Nested scopes don't require you to do anything different unless you
| actually use nested scopes.  In places where Java would raise a
| compile-time error because it can't tell if a name will be used
| unitialized, Python will not complain at comile time.  Instead, it will
| raise a NameError if an error occurs at run time.  I won't offer any
| opinions about which is better.

Hmm, ok.  I think I need to learn more details of how this nested
scope change affects the rules/semantics of python.

| 
| > I think it is acceptable if variables like loop counters and temps
| > inside a loop don't exist in the function's namespace. ie:
| 
| Regardless of whether you find it acceptable, the language doesn't work
| that way.

I just thought that that might have been something the nested scopes
would have changed.  I was letting you know I won't complain if that
change happens, though I'm not requesting it either :-).

I think that maybe my experience with C/C++/Java isn't helping me
here.  Those languages have nested scopes and variables declared in
such a scope disappear outside that scope.  This makes sense.  I guess
the difference is that a loop/if/try block defines a scope in
C/C++/Java but doesn't in Python.  This is probably where I got
confused.

| 
| > Last quarter I took a class called "Programming Language Concepts"
| > that discussed the design of programming languages.  It was the first
| > time I had heard of dynamic scoping.  (Actually, I think I saw it
| > mentioned in a lisp context, but I didn't know what it meant)  The
| > professor didn't know of any situations where it would be
| > advantageous, and I couldn't contrive any either.  I was kind of
| > surprised when I heard python described as "dynamically scoped"
| > because its scoping was really quite natural for me.  The [23]-level
| > scoping makes the dynamic scoping almost static and I don't see that
| > as a weakness.
| 
| Python's scoping rules aren't dynamic, never have been.  (At least not
| as long as I've been using the language.)  It used to be that scopes
| didn't nest, but that doesn't make scoping dynamic.

I thought I heard someone say that python's scoping rules are (prior
to 2.1a1, that is) dynamic.

| 
| >
| > I really don't know, at this moment, whether adding nested static
| > scopes will strengthen or weaken the language.
| 
| PEP 227 has a brief discussion of some benefits, among them that lambdas
| are easier to use.  http://python.sourceforge.net/peps/

I think I ought to read that in more detail, and probably read up on
the accepted definitions of some terms.

Thanks for the reply.

-D




From n8gray at caltech.edu.is.my.email.address  Tue Feb  6 09:00:47 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Tue, 6 Feb 2001 00:00:47 -0800
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org>
Message-ID: <95obae$oc2@gap.cco.caltech.edu>

D-Man wrote:

> On Mon, Feb 05, 2001 at 02:01:58AM -0800, Nathaniel Gray wrote:
> | 
> | Interesting issue.  What if the following was allowed:
> | >>> from tim_bot_module import tb*
> | 
> 
> Hmm, wouldn't it be better to use :
> 
> >>> import tim_bot_module as tb
> >>> tb.name_you_want_to_use
> 
> ?
> 
> IMO prefixing names with strings like "tb" is for namespace-less
> languages like C.  Since python has namespaces, why not use them?
> 

Yes, that would be the preferred method, but there are modules (TKInter, 
Numpy, wxPython) that are designed to be used with the "import *" idiom.  
This proposal would just allow you to use those modules without sacrificing 
efficiency.

I agree, though, that "import .. as .." goes a long way towards making 
"import *" unnecessary.

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From n8gray at caltech.edu.is.my.email.address  Tue Feb  6 09:00:47 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Tue, 6 Feb 2001 00:00:47 -0800
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org>
Message-ID: <95obd1$ocb@gap.cco.caltech.edu>

D-Man wrote:

> On Mon, Feb 05, 2001 at 02:01:58AM -0800, Nathaniel Gray wrote:
> | 
> | Interesting issue.  What if the following was allowed:
> | >>> from tim_bot_module import tb*
> | 
> 
> Hmm, wouldn't it be better to use :
> 
> >>> import tim_bot_module as tb
> >>> tb.name_you_want_to_use
> 
> ?
> 
> IMO prefixing names with strings like "tb" is for namespace-less
> languages like C.  Since python has namespaces, why not use them?
> 

Yes, that would be the preferred method, but there are modules (TKInter, 
Numpy, wxPython) that are designed to be used with the "import *" idiom.  
This proposal would just allow you to use those modules without sacrificing 
efficiency.

I agree, though, that "import .. as .." goes a long way towards making 
"import *" unnecessary.

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From moshez at zadka.site.co.il  Tue Feb  6 13:48:16 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue,  6 Feb 2001 14:48:16 +0200 (IST)
Subject: nested scopes
In-Reply-To: <95lu1i$ebs@gap.cco.caltech.edu>
References: <95lu1i$ebs@gap.cco.caltech.edu>, <mailman.981176465.13517.python-list@python.org>
Message-ID: <20010206124816.A2EA5A840@darjeeling.zadka.site.co.il>

On Mon, 5 Feb 2001, Nathaniel Gray <n8gray at caltech.edu> wrote:

> Interesting issue.  What if the following was allowed:
> >>> from tim_bot_module import tb*

Why wouldn't

import tim_bot_module as tb

And losing the "tb" prefix in tim_bot_module work?

-- 
For public key: finger moshez at debian.org | gpg --import
<doogie> Debian - All the power, without the silly hat.
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From jschmitt at vmlabs.com  Tue Feb  6 07:42:29 2001
From: jschmitt at vmlabs.com (John Schmitt)
Date: Mon, 5 Feb 2001 22:42:29 -0800 
Subject: nested scopes
Message-ID: <008F0A63472BD311AF9800104BCD102561CB4B@minirex.vmlabs.com>


> -----Original Message-----
> From: Nathaniel Gray [mailto:n8gray at caltech.edu.is.my.email.address]
> Sent: Monday, February 05, 2001 2:02 AM
> To: python-list at python.org
> Subject: RE: nested scopes
[snip]
> telling the compiler that only names beginning with "tb" are 
> imported.  
> This way you could still design libraries to use from ... import ..* 
> without totally throwing efficiency out the window.  I know 
> import * is 
> considered harmful, but occasionally (think tkinter) it's a useful 
> construct, assuming the library is designed to prevent name conflicts.
[snip]

When I first picked up Python, I typed in a couple of examples from
Grayson's book.  I would always say:

import Tkinter
tk = Tkinter
root = tk.Tk()
# etc

I tacked on a 'tk.' in front of tk calls.  I wanted to know which function
came from tk and which came from the standard python library etc. 

For what it's worth, I'd like to say that I'm happy that from ... import..*
is discouraged and if it disappeard from the language completely, I wouldn't
miss it.  Only slightly related is the fact anything that looks like
Pascal's 'with' would be completely wasted on me. :-)   I'd rather prepend
class name.  Explicit is better than implicit, right?

John



From greg at cosc.canterbury.ac.nz  Fri Feb  9 01:24:10 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 09 Feb 2001 13:24:10 +1300
Subject: nested scopes
References: <mailman.981262456.13388.python-list@python.org>
Message-ID: <3A8338AA.D713D628@cosc.canterbury.ac.nz>

Tim Peters wrote:
> 
> It's curious how
> few have ever complained that a def inside a class doesn't have automagic
> access to the class locals.

But they *do* complain about having to make explicit
references to 'self', which is the analogous thing there.

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From jmarshal at mathworks.com  Fri Feb  9 08:12:45 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 9 Feb 2001 07:12:45 GMT
Subject: nested scopes
References: <mailman.981262456.13388.python-list@python.org> <3A8338AA.D713D628@cosc.canterbury.ac.nz>
Message-ID: <96059d$din$1@news.mathworks.com>

Greg Ewing <greg at cosc.canterbury.ac.nz> wrote:
> Tim Peters wrote:
>> 
>> It's curious how
>> few have ever complained that a def inside a class doesn't have automagic
>> access to the class locals.

> But they *do* complain about having to make explicit
> references to 'self', which is the analogous thing there.

Still this is a different issue than lexical scoping.  Consider:

   class T:
      def m(self):
         print self.x

   T.x = 'foo'

   T().m()

Here, T().m() prints "foo".  Lexically, x is not defined in m's scope.
I believe it was a good decision to require going through self to
access class/object fields.  This is very different than something like:

   def f():
      x = 1
      def g():
         print x
      g()

where x is lexically in g's scope.


From aahz at panix.com  Fri Feb  9 16:40:21 2001
From: aahz at panix.com (Aahz Maruch)
Date: 9 Feb 2001 07:40:21 -0800
Subject: nested scopes
References: <mailman.981262456.13388.python-list@python.org> <3A8338AA.D713D628@cosc.canterbury.ac.nz> <96059d$din$1@news.mathworks.com>
Message-ID: <961315$l60$1@panix2.panix.com>

In article <96059d$din$1 at news.mathworks.com>,
Joshua Marshall  <jmarshal at mathworks.com> wrote:
>
>Still this is a different issue than lexical scoping.  Consider:
>
>   class T:
>      def m(self):
>         print self.x
>
>   T.x = 'foo'
>
>   T().m()
>
>Here, T().m() prints "foo".  Lexically, x is not defined in m's scope.
>I believe it was a good decision to require going through self to
>access class/object fields.  

Note that you don't necessarily want to go through self to access class
fields.  Here you're relying on the search through the class hierarchy
to access T.x with self.x.  (I'm pretty sure you know this, but I don't
want any newcomers getting confused.)
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From jeremy at alum.mit.edu  Fri Feb  9 18:55:12 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 09 Feb 2001 17:55:12 GMT
Subject: nested scopes
References: <mailman.981262456.13388.python-list@python.org> <3A8338AA.D713D628@cosc.canterbury.ac.nz> <96059d$din$1@news.mathworks.com>
Message-ID: <961atr$a9v$1@nnrp1.deja.com>

In article <96059d$din$1 at news.mathworks.com>,
  Joshua Marshall <jmarshal at mathworks.com> wrote:

> > But they *do* complain about having to make explicit
> > references to 'self', which is the analogous thing there.
>
> Still this is a different issue than lexical scoping.  Consider:
>
>    class T:
>       def m(self):
>          print self.x
>
>    T.x = 'foo'
>
>    T().m()
>
> Here, T().m() prints "foo".  Lexically, x is not defined in m's scope.
> I believe it was a good decision to require going through self to
> access class/object fields.  This is very different than something
> like:
>
>    def f():
>       x = 1
>       def g():
>          print x
>       g()
>
> where x is lexically in g's scope.
>

I agree that these examples are very different.  In the first, the
variable name is self and x is an attribute.  In the second, x is itself
a variable.  Python has very different rules for name lookup and
attribute lookup.

The few complaints I have seen about nested scopes on this list all
describe dynamic behavior that is probably better done with objects and
attributes.  While nested scopes are useful, many of the uses people put
them to in other languages are probably best done with instances in
Python.

--
-- Jeremy Hylton, <http://www.python.org/~jeremy/>


Sent via Deja.com
http://www.deja.com/


From paolo.salvan.amniup at libero.it  Fri Feb  2 11:50:39 2001
From: paolo.salvan.amniup at libero.it (paolo.salvan.amniup at libero.it)
Date: Fri, 02 Feb 2001 10:50:39 GMT
Subject: wish: "struct" module extension proposal
Message-ID: <95e3dt$kq0$1@nnrp1.deja.com>

I'm using the "struct" module to pack/unpack binary data...

I often pack strings rappresenting raw binary data putting a 4-bytes
integer indicating the string length, followed by the string
data...(it's likes the Pascal string coding schema, but with 4-bytes
prefix instead of the 1 byte prefix)

I think it would be generally useful if the struct format string was
extended to directly support this packing schema...what you think?

                          Paolo


Sent via Deja.com
http://www.deja.com/


From spam at spam.com  Fri Feb  2 23:16:56 2001
From: spam at spam.com (Penfold)
Date: Fri, 2 Feb 2001 22:16:56 -0000
Subject: "struct" module extension proposal
References: <95e3dt$kq0$1@nnrp1.deja.com>
Message-ID: <t7mclpti1s39f6@corp.supernews.co.uk>

I often pack strings in reverse using an 8 byte number at the end of the
string to tell the user how long it was
I too think that the struct module needs extending to directly support my
arbitrrary packing structure.  What
do you think ?

Des :-)

<paolo.salvan.amniup at libero.it> wrote in message
news:95e3dt$kq0$1 at nnrp1.deja.com...
> I'm using the "struct" module to pack/unpack binary data...
>
> I often pack strings rappresenting raw binary data putting a 4-bytes
> integer indicating the string length, followed by the string
> data...(it's likes the Pascal string coding schema, but with 4-bytes
> prefix instead of the 1 byte prefix)
>
> I think it would be generally useful if the struct format string was
> extended to directly support this packing schema...what you think?
>
>                           Paolo
>
>
> Sent via Deja.com
> http://www.deja.com/




From kirschh at lionbioscience.com  Fri Feb  2 14:08:20 2001
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 02 Feb 2001 14:08:20 +0100
Subject: execfile('bla.py'), can bla.py know its full path
Message-ID: <yv2g0hxb5bv.fsf@lionsp093.lion-ag.de>

When a script is called by execfile, can it find out the full absolute
path name by which it was called?


  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s


From jcm at bigskytel.com  Fri Feb  2 14:36:55 2001
From: jcm at bigskytel.com (David Porter)
Date: Fri, 2 Feb 2001 06:36:55 -0700
Subject: execfile('bla.py'), can bla.py know its full path
In-Reply-To: <yv2g0hxb5bv.fsf@lionsp093.lion-ag.de>; from kirschh@lionbioscience.com on Fri, Feb 02, 2001 at 02:08:20PM +0100
References: <yv2g0hxb5bv.fsf@lionsp093.lion-ag.de>
Message-ID: <20010202063655.A12537@bigskytel.com>

* Harald Kirsch <kirschh at lionbioscience.com>:

> When a script is called by execfile, can it find out the full absolute
> path name by which it was called?

You could tell it by passing a variable to it containing the value:

execfile('/home/dap/bla.py', {'path': '/home/dap/bla.py'})


David



From tshumway at transafari.com  Fri Feb  2 14:42:32 2001
From: tshumway at transafari.com (Terrel Shumway)
Date: Fri, 02 Feb 2001 05:42:32 -0800
Subject: How can I get Memory/CPU stats for the python VM?
Message-ID: <3A7AB948.DEA24C31@transafari.com>

I am running Python 2.0 on a Linux box that has an protected /proc/.

How can I find out how big the python process is and how much CPU time
it is taking?

The profile module is not enough:
    1) I need to get info from a deployed process (non-debug)
    2) It doesn't  give memory stats

Rebuilding the python interpreter or an extension module *is* feasible.

    TIA,
    -- Terrel





From loewis at informatik.hu-berlin.de  Fri Feb  2 23:25:15 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 23:25:15 +0100
Subject: How can I get Memory/CPU stats for the python VM?
References: <3A7AB948.DEA24C31@transafari.com>
Message-ID: <p6q8znooh84.fsf@informatik.hu-berlin.de>

Terrel Shumway <tshumway at transafari.com> writes:

> I am running Python 2.0 on a Linux box that has an protected /proc/.
> 
> How can I find out how big the python process is and how much CPU time
> it is taking?

time python foo.py

gives you the execution time.

> The profile module is not enough:
>     1) I need to get info from a deployed process (non-debug)
>     2) It doesn't  give memory stats
>
> Rebuilding the python interpreter or an extension module *is*
> feasible.

With the gc module, you can get the number of living objects. With the
resource module, you can get shared and unshared memory sizes etc, if
your operating system supports that. Unfortunately, Linux reports
zeroes in many of these fields. Reading /proc/self/maps or
/proc/self/stat* will provide the proper detail on Linux - you only
need to find the kernel documentation to interpret these numbers :-)
Hint: /usr/src/linux/fs/proc/array.c has the code that generates these
lists.

Regards,
Martin


From robin at jessikat.fsnet.co.uk  Fri Feb  2 15:49:50 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 2 Feb 2001 14:49:50 +0000
Subject: zope.org?
Message-ID: <24QsZIAOkse6Ewl$@jessikat.fsnet.co.uk>

Anyone know what's up with zope.org, seems off line and a lot of
problems with the mailing lists.
-- 
Robin Becker


From mikael at isy.liu.se  Fri Feb  2 16:06:16 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 02 Feb 2001 16:06:16 +0100 (MET)
Subject: zope.org?
In-Reply-To: <24QsZIAOkse6Ewl$@jessikat.fsnet.co.uk>
Message-ID: <XFMail.010202160616.mikael@isy.liu.se>

On 02-Feb-01 Robin Becker wrote:
 >  Anyone know what's up with zope.org, seems off line and a lot of
 >  problems with the mailing lists.

Well, I don't know about the mailing lists, but the web service seems
to be up. Accessing www.zope.org causes me no problems.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    02-Feb-01
Time:    16:05:07

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From chrisw at nipltd.com  Fri Feb  2 16:50:57 2001
From: chrisw at nipltd.com (Chris Withers)
Date: Fri, 02 Feb 2001 15:50:57 +0000
Subject: zope.org?
References: <XFMail.010202160616.mikael@isy.liu.se>
Message-ID: <3A7AD761.BB2A884D@nipltd.com>

Mikael Olofsson wrote:
> 
> On 02-Feb-01 Robin Becker wrote:
>  >  Anyone know what's up with zope.org, seems off line and a lot of
>  >  problems with the mailing lists.
> 
> Well, I don't know about the mailing lists, but the web service seems
> to be up. Accessing www.zope.org causes me no problems.

zope.org came up about 1/2 hour ago. The lists are still down, for over a day
now :-(

*sigh* Some kindof BIND problem and disk failure, apparently...

cheers,

Chris



From binnc at my-deja.com  Fri Feb  2 16:08:31 2001
From: binnc at my-deja.com (binnc at my-deja.com)
Date: Fri, 02 Feb 2001 15:08:31 GMT
Subject: String as a Variable?
Message-ID: <95eih7$si$1@nnrp1.deja.com>

Another newbie question...

Is there a way to have a string represent a variable.

For eaxample, I have a the following:

<snip>

serial_number = '3'
print 'ip_address_serial_' + serial_number
ip_address_serial_3

Now I want to use the string ip_address_serial_3 to pull
the value assigned to the variable of the same name.

How can I represent the string as a variable?

Thanks in advance!


Sent via Deja.com
http://www.deja.com/


From DSHIN at nc.rr.com  Fri Feb  2 16:31:03 2001
From: DSHIN at nc.rr.com (Daehyok Shin)
Date: Fri, 02 Feb 2001 15:31:03 GMT
Subject: String as a Variable?
References: <95eih7$si$1@nnrp1.deja.com>
Message-ID: <XoAe6.11365$k8.2384026@typhoon.southeast.rr.com>

Use __dict__.
For example, in a class,

str='ip_address_serial'
self.__dict__[str]=str
print self.ip_address_serial

<binnc at my-deja.com> wrote in message news:95eih7$si$1 at nnrp1.deja.com...
> Another newbie question...
>
> Is there a way to have a string represent a variable.
>
> For eaxample, I have a the following:
>
> <snip>
>
> serial_number = '3'
> print 'ip_address_serial_' + serial_number
> ip_address_serial_3
>
> Now I want to use the string ip_address_serial_3 to pull
> the value assigned to the variable of the same name.
>
> How can I represent the string as a variable?
>
> Thanks in advance!
>
>
> Sent via Deja.com
> http://www.deja.com/




From fredrik at effbot.org  Fri Feb  2 16:32:52 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 15:32:52 GMT
Subject: String as a Variable?
References: <95eih7$si$1@nnrp1.deja.com>
Message-ID: <EqAe6.11113$AH6.1781694@newsc.telia.net>

binnc at my-deja.com wrote:
> Is there a way to have a string represent a variable.

yes, but Python's not Tcl, and there are better
ways to do it.

start here:

    http://www.python.org/doc/current/tut/node7.html
    => 5.4 Dictionaries

Cheers /F




From phd at phd.pp.ru  Fri Feb  2 16:57:01 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 2 Feb 2001 18:57:01 +0300 (MSK)
Subject: String as a Variable?
In-Reply-To: <95eih7$si$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.30.0102021855430.13648-100000@fep132.fep.ru>

On Fri, 2 Feb 2001 binnc at my-deja.com wrote:
> Is there a way to have a string represent a variable.

   Although there are ways to do it, the recommnded way is to awoid it at
all. Better, collect all neccessary variables in a dictionary:

   x = mydict["ip_addr_" + value]

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From mikael at isy.liu.se  Fri Feb  2 17:05:04 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 02 Feb 2001 17:05:04 +0100 (MET)
Subject: String as a Variable?
In-Reply-To: <95eih7$si$1@nnrp1.deja.com>
Message-ID: <XFMail.010202170504.mikael@isy.liu.se>

On 02-Feb-01 binnc at my-deja.com wrote:
 >  Now I want to use the string ip_address_serial_3 to pull
 >  the value assigned to the variable of the same name.


>>> exec('a=0')
>>> a
0

HTH

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    02-Feb-01
Time:    17:04:09

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From aleaxit at yahoo.com  Fri Feb  2 17:05:39 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 2 Feb 2001 17:05:39 +0100
Subject: String as a Variable?
References: <95eih7$si$1@nnrp1.deja.com>
Message-ID: <95elsm0nuj@news1.newsguy.com>

<binnc at my-deja.com> wrote in message news:95eih7$si$1 at nnrp1.deja.com...
> Another newbie question...
>
> Is there a way to have a string represent a variable.

Yes, there are quite a few.  It's an *exceedingly* rare
occurrence that using any of those ways is actually the
best approach to anything, though; you would most likely
be best advised to refactor your code to work otherwise.


> For eaxample, I have a the following:
>
> <snip>
>
> serial_number = '3'
> print 'ip_address_serial_' + serial_number
> ip_address_serial_3
>
> Now I want to use the string ip_address_serial_3 to pull
> the value assigned to the variable of the same name.

The built-in function vars() returns a mapping that
represents the variable-to-value correspondence at
this time.  Indexing that mapping with a string will
return the value currently corresponding to the
variable whose name is that string -- if no such
variable exists, of course, a KeyError will be raised.

I.e.,

    print vars()['ip_address_serial_' + serial_number]

will print the value you require.

But it's a bad ideas to use variables in this way.

Why not have ip_addresses be a dictionary, and
index it with '3' directly?  This is much more
direct, explicit, simple, and well-performing, than
hiding the 'pieces' of a mapping as part of the
vars-dictionary -- which is what you're doing when
you use variables (rather than dictionary entries)
to refer to values, but then want to access them by
string-index.

Further, vars() and friends do not, in general, let
you RE-BIND variables to new values -- only access
the value a variable is currently bound to.

You can bypass these issues with the exec statement
(and, as you're at it, you might use the eval
function to fetch your variables -- it's overkill,
though).  But, MUCH better, you can bypass them by
using a dictionary when a dictionary is what you
really want.


Alex





From bogus@does.not.exist.com  Sat Feb  3 21:34:41 2001
From: bogus@does.not.exist.com (Jonathan Polley)
Date: Sat, 03 Feb 2001 14:34:41 -0600
Subject: String as a Variable?
References: <95eih7$si$1@nnrp1.deja.com>
Message-ID: <3A7C6B61.641D75E0@collins.rockwell.com>

I actually use the exec() and eval() functions.  If I need to assign to
the string/variable, I will use exec():

My_Variable = "This_Thing"
My_Expression = "= 15.75"

exec (My_Variable + My_Expression)

To read the variable, use eval():

print eval (My_Variable)

Jon

p.s.  While it may not be an accpeted practice to do this, I have found
it quite useful.  At work, we are able to quickly develop automated
tests that will do bounds checking on various python classes and
methods.

binnc at my-deja.com wrote:

> Another newbie question...
>
> Is there a way to have a string represent a variable.
>
> For eaxample, I have a the following:
>
> <snip>
>
> serial_number = '3'
> print 'ip_address_serial_' + serial_number
> ip_address_serial_3
>
> Now I want to use the string ip_address_serial_3 to pull
> the value assigned to the variable of the same name.
>
> How can I represent the string as a variable?
>
> Thanks in advance!
>
> Sent via Deja.com
> http://www.deja.com/



From dan at eevolved.com  Fri Feb  2 18:15:57 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 02 Feb 2001 17:15:57 GMT
Subject: expanding a list and dict to a *args and **dict
Message-ID: <hXBe6.172472$JT5.6263159@news20.bellglobal.com>

Is it possible to 'expand' a list or dictionary into its components to call 
a function expecting many parameters?

I want to be able to call a function without accessing each individual 
element manually (list[0], list[1], list[2])

I'm sending random parameters over a network (*args and **dict). When I 
unpickle them on the other side, I get a python list and dictionary. I then 
call an arbitrary function (actually a reference to a function that was 
bound at runtime a la 
self.randomfunction = obj.specificfunction
)

That function is not expecting a list and a dictionary, instead it is 
expecting specific parameters(that i don't know at runtime) and I want to 
take the list and turn it into a bunch of parameters like list[0], list[1], 
list[2] (but as stated above, I can't do that)

Ha. I hope someone understood me :)

Dan


From mwh21 at cam.ac.uk  Fri Feb  2 18:21:55 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 02 Feb 2001 17:21:55 +0000
Subject: expanding a list and dict to a *args and **dict
References: <hXBe6.172472$JT5.6263159@news20.bellglobal.com>
Message-ID: <m3ae8557bg.fsf@atrus.jesus.cam.ac.uk>

Dan Parisien <dan at eevolved.com> writes:

> Is it possible to 'expand' a list or dictionary into its components to call 
> a function expecting many parameters?
> 
> I want to be able to call a function without accessing each individual 
> element manually (list[0], list[1], list[2])
> 
> I'm sending random parameters over a network (*args and **dict). When I 
> unpickle them on the other side, I get a python list and dictionary. I then 
> call an arbitrary function (actually a reference to a function that was 
> bound at runtime a la 
> self.randomfunction = obj.specificfunction
> )
> 
> That function is not expecting a list and a dictionary, instead it is 
> expecting specific parameters(that i don't know at runtime) and I want to 
> take the list and turn it into a bunch of parameters like list[0], list[1], 
> list[2] (but as stated above, I can't do that)
> 
> Ha. I hope someone understood me :)

Only a bit, but do you know about apply (a builtin function) and/or
this sort of thing:

>>> def f(a,b,c):
...  print a,b,c
... 
>>> f(*(1,2),**{'c':3})
1 2 3

I *think* that's what your asking for...

Cheers,
M.

-- 
  I have gathered a posie of other men's flowers, and nothing but the
  thread that binds them is my own.                       -- Montaigne


From phd at phd.pp.ru  Fri Feb  2 18:39:33 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 2 Feb 2001 20:39:33 +0300 (MSK)
Subject: expanding a list and dict to a *args and **dict
In-Reply-To: <hXBe6.172472$JT5.6263159@news20.bellglobal.com>
Message-ID: <Pine.LNX.4.30.0102022039050.13648-100000@fep132.fep.ru>

On Fri, 2 Feb 2001, Dan Parisien wrote:
> Is it possible to 'expand' a list or dictionary into its components to call
> a function expecting many parameters?

   Use builtin function "apply"! Read docs...

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From tiddlerdeja at my-deja.com  Fri Feb  2 18:35:56 2001
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Fri, 02 Feb 2001 17:35:56 GMT
Subject: jython 2 java class - help?????
Message-ID: <95er5s$9od$1@nnrp1.deja.com>

Can anyone tell me if they have had much success with this?

I can follow the Graph.py demo and I get this to work. With this demo
you get a PythonGraph.java file to javac.

When I attempt this for MyEg.py, I get the MyEg.class files, but I
don't get the PythonMyEg.java file to compile.

My exmaple does derive from java.lang.Object, it does work from
jythjon.bat and has a if __main...... thingy.

I also want to be able to create a jar that I can deply on another VM
that does not have jython-2.0 installed.

Any help appreciated.


Sent via Deja.com
http://www.deja.com/


From dsh8290 at rit.edu  Fri Feb  2 19:24:07 2001
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 2 Feb 2001 13:24:07 -0500
Subject: jython 2 java class - help?????
In-Reply-To: <95er5s$9od$1@nnrp1.deja.com>; from tiddlerdeja@my-deja.com on Fri, Feb 02, 2001 at 05:35:56PM +0000
References: <95er5s$9od$1@nnrp1.deja.com>
Message-ID: <20010202132407.B8584@harmony.cs.rit.edu>

(these things are probably better for the jython list)

$ jythonc --all --jar MyEg.jar  MyEg.py

Will create a jar file MyEg.jar that will contain all the .class files
you need (from your .py source and the jython.jar file).

By default jythonc does its work in a directory ./jpywork .  This is
where the .java and .class files are output.  If you specify a jar
file, jythonc will take the .class files it output to ./jpywork and
put them in the jar.

HTH,
-D

On Fri, Feb 02, 2001 at 05:35:56PM +0000, tiddlerdeja at my-deja.com wrote:
| Can anyone tell me if they have had much success with this?
| 
| I can follow the Graph.py demo and I get this to work. With this demo
| you get a PythonGraph.java file to javac.
| 
| When I attempt this for MyEg.py, I get the MyEg.class files, but I
| don't get the PythonMyEg.java file to compile.
| 
| My exmaple does derive from java.lang.Object, it does work from
| jythjon.bat and has a if __main...... thingy.
| 
| I also want to be able to create a jar that I can deply on another VM
| that does not have jython-2.0 installed.
| 
| Any help appreciated.
| 



From reborn at neozone.net  Fri Feb  2 23:28:22 2001
From: reborn at neozone.net (syKim)
Date: Sat, 3 Feb 2001 07:28:22 +0900
Subject: jython 2 java class - help?????
References: <95er5s$9od$1@nnrp1.deja.com>
Message-ID: <95fc9d$kad$1@news1.kornet.net>

If you just want run & test in jython..

now, you could try like that : (attach bottom of your code)

if __name__ == '__main__'
  import pawt
    pawt.test(yourclassname(),size=(number,number))

# yourclassname, number - anything you want

you can get the result like that

jython filename.py

In webrower, I'm discussing yet..








From tiddlerdeja at my-deja.com  Mon Feb  5 12:32:59 2001
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Mon, 05 Feb 2001 11:32:59 GMT
Subject: jython 2 java class - help?????
References: <95er5s$9od$1@nnrp1.deja.com> <mailman.981138302.585.python-list@python.org>
Message-ID: <95m31b$hfi$1@nnrp1.deja.com>

Thanks for that, I'll give it a go, but I think I'll run into my
original problems.

I did have a look for a jython group to post to, but didn't find one.
Would you be able to tell me what it is?

Thanks again.

In article <mailman.981138302.585.python-list at python.org>,
  D-Man <dsh8290 at rit.edu> wrote:
>
> (these things are probably better for the jython list)
>
> $ jythonc --all --jar MyEg.jar  MyEg.py
>
> Will create a jar file MyEg.jar that will contain all the .class files
> you need (from your .py source and the jython.jar file).
>
> By default jythonc does its work in a directory ./jpywork .  This is
> where the .java and .class files are output.  If you specify a jar
> file, jythonc will take the .class files it output to ./jpywork and
> put them in the jar.
>
> HTH,
> -D
>
> On Fri, Feb 02, 2001 at 05:35:56PM +0000, tiddlerdeja at my-deja.com wrote:
> | Can anyone tell me if they have had much success with this?
> |
> | I can follow the Graph.py demo and I get this to work. With this demo
> | you get a PythonGraph.java file to javac.
> |
> | When I attempt this for MyEg.py, I get the MyEg.class files, but I
> | don't get the PythonMyEg.java file to compile.
> |
> | My exmaple does derive from java.lang.Object, it does work from
> | jythjon.bat and has a if __main...... thingy.
> |
> | I also want to be able to create a jar that I can deply on another VM
> | that does not have jython-2.0 installed.
> |
> | Any help appreciated.
> |
>
>


Sent via Deja.com
http://www.deja.com/


From dsh8290 at rit.edu  Mon Feb  5 16:49:50 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 5 Feb 2001 10:49:50 -0500
Subject: jython 2 java class - help?????
In-Reply-To: <95m31b$hfi$1@nnrp1.deja.com>; from tiddlerdeja@my-deja.com on Mon, Feb 05, 2001 at 11:32:59AM +0000
References: <95er5s$9od$1@nnrp1.deja.com> <mailman.981138302.585.python-list@python.org> <95m31b$hfi$1@nnrp1.deja.com>
Message-ID: <20010205104950.A27254@harmony.cs.rit.edu>

On Mon, Feb 05, 2001 at 11:32:59AM +0000, tiddlerdeja at my-deja.com wrote:
| Thanks for that, I'll give it a go, but I think I'll run into my
| original problems.
| 
| I did have a look for a jython group to post to, but didn't find one.
| Would you be able to tell me what it is?

It is mentioned on the web site, jython.sourceforge.net.

jython-users at lists.sourceforge.net

-D




From echuck at mindspring.com  Fri Feb  2 19:53:27 2001
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Fri, 02 Feb 2001 13:53:27 -0500
Subject: Comparison oddities
Message-ID: <5.0.2.1.0.20010202134528.05141ad0@mail.mindspring.com>

It appears that tuples and strings are "greater than" lists.

 >python
ActivePython 2.0, build 202 (ActiveState Tool Corp.)
based on Python 2.0 (#8, Oct 19 2000, 11:30:05) [MSC 32 bit (Intel)] on win32
 >>> a = (1, 1)
 >>> b = [1, 1]
 >>> a < b
0
 >>> a > b
1
 >>> a==b
0
 >>> a = '1'
 >>> a < b
0
 >>> a > b
1


Does this seem a little odd to anyone?

Since tuples and lists are both "sequences of anything", you would hope 
that they would compare in "the natural fashion".

Since comparing a string and a list doesn't make sense, would an exception 
be more appropriate?

And yes, I know the manual says "Otherwise, objects of different types 
*always* compare unequal, and are ordered consistently but arbitrarily. ... 
In the future, the comparison rules for objects of different types are 
likely to change."

My question is more of a design question.


-Chuck




From aahz at panix.com  Fri Feb  2 20:10:46 2001
From: aahz at panix.com (Aahz Maruch)
Date: 2 Feb 2001 11:10:46 -0800
Subject: Comparison oddities
References: <mailman.981140164.1422.python-list@python.org>
Message-ID: <95f0nm$9am$1@panix3.panix.com>

In article <mailman.981140164.1422.python-list at python.org>,
Chuck Esterbrook  <echuck at mindspring.com> wrote:
>
>It appears that tuples and strings are "greater than" lists.
>Does this seem a little odd to anyone?
>
>Since tuples and lists are both "sequences of anything", you would hope 
>that they would compare in "the natural fashion".
>
>Since comparing a string and a list doesn't make sense, would an exception 
>be more appropriate?
>
>And yes, I know the manual says "Otherwise, objects of different types 
>*always* compare unequal, and are ordered consistently but arbitrarily. ... 
>In the future, the comparison rules for objects of different types are 
>likely to change."
>
>My question is more of a design question.

I'm only going to answer your second question: this ordering is required
by the fact that lists are *not* homogenous.  A list can contain any
other Python type, including lists, strings, dicts, and numbers.
list.sort() is required to work well in all of this.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From jmg at ecs.soton.ac.uk  Fri Feb  2 20:38:17 2001
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: 02 Feb 2001 19:38:17 +0000
Subject: Command line editing in python shell
Message-ID: <g0elxgub86.fsf@scumbag.ecs.soton.ac.uk>

I have compiled python 2.0 on a Mandrake 7.2 installation. Command
line editing is impossible: The up-arrow prints ^[[A and so on.

I guess that some libraries are not being found, as I have previously
compiled python 2.0 on a RedHat 6.2 installation and everything worked
fine.

Any ideas how I can fix this one ?

Thanks,

Jacek


From thomas at xs4all.net  Fri Feb  2 21:14:33 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 2 Feb 2001 21:14:33 +0100
Subject: Command line editing in python shell
In-Reply-To: <g0elxgub86.fsf@scumbag.ecs.soton.ac.uk>; from jmg@ecs.soton.ac.uk on Fri, Feb 02, 2001 at 07:38:17PM +0000
References: <g0elxgub86.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <20010202211433.X962@xs4all.nl>

On Fri, Feb 02, 2001 at 07:38:17PM +0000, Jacek Generowicz wrote:
> I have compiled python 2.0 on a Mandrake 7.2 installation. Command
> line editing is impossible: The up-arrow prints ^[[A and so on.
> 
> I guess that some libraries are not being found, as I have previously
> compiled python 2.0 on a RedHat 6.2 installation and everything worked
> fine.
> Any ideas how I can fix this one ?

You have to compile in the 'readline' module. In Python 2.0 that means
editing 'Modules/Setup' and uncommenting the line that mentions readline.
You'll probably also want to uncomment some other lines.

If that doesn't work, you might lack the headerfiles for readline, or they
might be in an odd location. If you don't have them installed, you have to
find an RPM that provides them (probably readline-devel.<something>.rpm.) If
they are in an odd location, you have to tweak the -I and -L arguments
behind readline. 

This will almost all be unnecessary in Python 2.1, where the modules are
autodetected using Python's distutils. It's less easy to supply your own
location, but it will hardly be necessary, in most cases. In Python 2.1, a
single 'make' after 'configure' will build everything for you :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jmg at ecs.soton.ac.uk  Mon Feb  5 11:50:29 2001
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: 05 Feb 2001 10:50:29 +0000
Subject: Command line editing in python shell
References: <g0elxgub86.fsf@scumbag.ecs.soton.ac.uk> <mailman.981144904.3081.python-list@python.org>
Message-ID: <g0y9vlo13e.fsf@scumbag.ecs.soton.ac.uk>

Thomas Wouters <thomas at xs4all.net> writes:

> On Fri, Feb 02, 2001 at 07:38:17PM +0000, Jacek Generowicz wrote:
> > I have compiled python 2.0 on a Mandrake 7.2 installation. Command
> > line editing is impossible: The up-arrow prints ^[[A and so on.
> > 
> > I guess that some libraries are not being found, as I have previously
> > compiled python 2.0 on a RedHat 6.2 installation and everything worked
> > fine.
> > Any ideas how I can fix this one ?
> 
> You have to compile in the 'readline' module. In Python 2.0 that means
> editing 'Modules/Setup' and uncommenting the line that mentions readline.

This solved the problem. (After obtaining and compiling the library)

> If that doesn't work, you might lack the headerfiles for readline, or they
> might be in an odd location. If you don't have them installed, you have to
> find an RPM that provides them (probably readline-devel.<something>.rpm.)

I used readline-4.1-9mdk-src.rpm which I obtained from rpmfind.net. I
applied the makefile patch that comes in the package before compiling.

Thanks,

Jacek


From geoff at homegain.com  Fri Feb  2 20:38:52 2001
From: geoff at homegain.com (Geoffrey Gerrietts)
Date: Fri, 2 Feb 2001 11:38:52 -0800 
Subject: fetching class by name
Message-ID: <393D961EA4FCD311B03F00D0B746D65802625F22@HQSXCH01>

I think I know the answer to this question before I ask it, but I
feel like I need to ask anyway.

Is there an easier way to retrieve a class object by name than
doing a manual lookup through the __dict__ of every module in
sys.modules?

I think maybe I'm just spoiled by the way objective c gives you a
cute little utility function to do this for you. Still, looking
it up manually seems extraordinarily costly.

Thanks,
--G.

---
Geoff Gerrietts <geoff at homegain.com>
Software Engineer, HomeGain.com
510-655-0800 x4320
 



From loewis at informatik.hu-berlin.de  Fri Feb  2 23:31:10 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 23:31:10 +0100
Subject: fetching class by name
References: <mailman.981142923.2378.python-list@python.org>
Message-ID: <p6q66isogy9.fsf@informatik.hu-berlin.de>

Geoffrey Gerrietts <geoff at homegain.com> writes:

> Is there an easier way to retrieve a class object by name than
> doing a manual lookup through the __dict__ of every module in
> sys.modules?

Depends on what you know about the class. If you *only* know the name,
there is probably no other way. Hint: if all you need is a class with
that name, how about

>>> import new
>>> new.classobj("thename",(),{})
<class __main__.thename at 0x811dbcc>

> I think maybe I'm just spoiled by the way objective c gives you a
> cute little utility function to do this for you. Still, looking
> it up manually seems extraordinarily costly.

More explicitly: there might be more than one class with that name,
the class might be known under a different name, or it might not be
bound to a variable altogether:

>>> class Foo:pass
...
>>> foo=Foo()
>>> del Foo
>>> foo.__class__
<class __main__.Foo at 0x811c72c>

Classes are objects like any other: some are referred-to by global
variables, others are not.

Regards,
Martin




From marchign at science.unitn.it  Wed Feb  7 13:57:40 2001
From: marchign at science.unitn.it (Davide Marchignoli)
Date: Wed, 07 Feb 2001 13:57:40 +0100
Subject: fetching class by name
In-Reply-To: <mailman.981142923.2378.python-list@python.org>
References: <mailman.981142923.2378.python-list@python.org>
Message-ID: <E14QUAG-0005oT-00@dott14.science.unitn.it>

In comp.lang.python, you wrote:
>
>I think I know the answer to this question before I ask it, but I
>feel like I need to ask anyway.
>
>Is there an easier way to retrieve a class object by name than
>doing a manual lookup through the __dict__ of every module in
>sys.modules?
>
>I think maybe I'm just spoiled by the way objective c gives you a
>cute little utility function to do this for you. Still, looking
>it up manually seems extraordinarily costly.
>
>Thanks,
>--G.
>
maybe eval suits your needs:

>>> class C : pass
... 
>>> x = eval('C')
>>> x
<class __main__.C at 80708b0>

Bye,

--
				Davide



From greg at cosc.canterbury.ac.nz  Wed Feb 14 01:18:10 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 14 Feb 2001 13:18:10 +1300
Subject: fetching class by name
References: <mailman.981142923.2378.python-list@python.org> <mailman.981550720.29353.python-list@python.org>
Message-ID: <3A89CEC2.DB75C7CE@cosc.canterbury.ac.nz>

Someone whose attribution has been lost said:

> >I think maybe I'm just spoiled by the way objective c gives you a
> >cute little utility function to do this for you. Still, looking
> >it up manually seems extraordinarily costly.

All the class names live in a single namespace in Objective C.
In Python the situation is more complicated. If you have no idea
which module the class is defined in, then you've no choice but
to go searching.

On the other hand, if you know both the module name and the 
class name, then you can go straight to it:

  import sys
  c = getattr(sys.modules[module_name], class_name)

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From geoff at homegain.com  Sat Feb  3 00:48:01 2001
From: geoff at homegain.com (Geoffrey Gerrietts)
Date: Fri, 2 Feb 2001 15:48:01 -0800 
Subject: fetching class by name
Message-ID: <393D961EA4FCD311B03F00D0B746D65802625F25@HQSXCH01>

 

Martin von Loewis wrote:
> 
> More explicitly: there might be more than one class with that name,
> the class might be known under a different name, or it might not be
> bound to a variable altogether:

All excellent points, the last of which is most surprising of the three
(I wouldn't have thought of that case if you hadn't pointed it out), 
and most alarming. Definitely the point that politely suggests an 
alternate implementation for the case I'm looking at.

Sometimes hard to chase away idioms of the past....

Thanks very much,
--G.

---
Geoff Gerrietts <geoff at homegain.com>
Software Engineer, HomeGain.com
510-655-0800 x4320 



From fluxent at yahoo.com  Fri Feb  2 21:26:28 2001
From: fluxent at yahoo.com (Bill Seitz)
Date: Fri, 02 Feb 2001 20:26:28 GMT
Subject: ADO/ODBC call via Python/IIS
Message-ID: <95f558$jgl$1@nnrp1.deja.com>

Architecture: ASP page using <%@ LANGUAGE = Python%>, calling Python
class which calls ADO to pass SQL command via ODBC driver to MsSQL db.
(Everything running on Win2K.) MsSQL configured to authenticate against
ActiveDirectory only. Sysadmin created userID 'pyasp'. I created MsSQL
login and assigned user rights in the appropriate db. I created an ODBC
source named 'pyasp', assigned the current db as the default.

Here is the class that does the work (in file 'aAdoBiz.py')
class aConnect:
    def __init__(self):
        import win32com.client
        adoConn = win32com.client.Dispatch('ADODB.Connection')
        adoConn.Open('pyasp')
        (adoRS, success) = adoConn.Execute('Select * from RateUnits')
        adoRS.MoveFirst()
        adoRS.MoveNext()
        self.fieldName = adoRS.Fields("Name").Value

When I call this code interactively, I get the right result:
>>> import aAdoBiz
>>> z=aAdoBiz.aConnect()
>>> z.fieldName.encode()
'Days           '

But then I have the same idea implemented in an ASP page:
<%
import aAdoBiz
x = aAdoBiz.aConnect()
out = x.fieldName.encode()
%>
<%= out %>

(A very similar ASP page, calling a simple stub class with no ODBC
activity, works fine.)

When I call this ASP page, I get an error to the browser...
HTTP 500.100 - Internal Server Error - ASP error
Error Type:
Python ActiveX Scripting Engine (0x80020009)
Traceback (innermost last): File "<Script Block >", line 3, in ? x =
aAdoBiz.aConnect() File "aAdoBiz.py", line 7, in __init__ adoConn.Open
('pyasp') File "c:\python20\win32com\gen_py\00000206-0000-0010-8000-
00AA006D2EA4x0x2x6.py", line 1833, in Open return self._ApplyTypes_
(0xa, 1, (24, 32), ((8, 49), (8, 49), (8, 49), (3, 49)), 'Open',
None,ConnectionString, UserID, Password, Options) File "c:\python20
\win32com\client\__init__.py", line 334, in _ApplyTypes_ return
self._get_good_object_(apply(self._oleobj_.InvokeTypes, (dispid, 0,
wFlags, retType, argTypes) + args), user, resultCLSID) COM Error:
[Microsoft][ODBC Driver Manager] Data source name not found and no
default driver specified

What is happening here? Why the error when called from IIS, but not
when run interactively?

thx


Sent via Deja.com
http://www.deja.com/


From sholden at holdenweb.com  Fri Feb  2 22:49:32 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 2 Feb 2001 16:49:32 -0500
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com>
Message-ID: <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com>

"Bill Seitz" <fluxent at yahoo.com> wrote in message
news:95f558$jgl$1 at nnrp1.deja.com...
> Architecture: ASP page using <%@ LANGUAGE = Python%>, calling Python
> class which calls ADO to pass SQL command via ODBC driver to MsSQL db.
> (Everything running on Win2K.) MsSQL configured to authenticate against
> ActiveDirectory only. Sysadmin created userID 'pyasp'. I created MsSQL
> login and assigned user rights in the appropriate db. I created an ODBC
> source named 'pyasp', assigned the current db as the default.
>
> Here is the class that does the work (in file 'aAdoBiz.py')
> class aConnect:
>     def __init__(self):
>         import win32com.client
>         adoConn = win32com.client.Dispatch('ADODB.Connection')
>         adoConn.Open('pyasp')
>         (adoRS, success) = adoConn.Execute('Select * from RateUnits')
>         adoRS.MoveFirst()
>         adoRS.MoveNext()
>         self.fieldName = adoRS.Fields("Name").Value
>
> When I call this code interactively, I get the right result:
> >>> import aAdoBiz
> >>> z=aAdoBiz.aConnect()
> >>> z.fieldName.encode()
> 'Days           '
>
> But then I have the same idea implemented in an ASP page:
> <%
> import aAdoBiz
> x = aAdoBiz.aConnect()
> out = x.fieldName.encode()
> %>
> <%= out %>
>
> (A very similar ASP page, calling a simple stub class with no ODBC
> activity, works fine.)
>
> When I call this ASP page, I get an error to the browser...
> HTTP 500.100 - Internal Server Error - ASP error
> Error Type:
> Python ActiveX Scripting Engine (0x80020009)
> Traceback (innermost last): File "<Script Block >", line 3, in ? x =
> aAdoBiz.aConnect() File "aAdoBiz.py", line 7, in __init__ adoConn.Open
> ('pyasp') File "c:\python20\win32com\gen_py\00000206-0000-0010-8000-
> 00AA006D2EA4x0x2x6.py", line 1833, in Open return self._ApplyTypes_
> (0xa, 1, (24, 32), ((8, 49), (8, 49), (8, 49), (3, 49)), 'Open',
> None,ConnectionString, UserID, Password, Options) File "c:\python20
> \win32com\client\__init__.py", line 334, in _ApplyTypes_ return
> self._get_good_object_(apply(self._oleobj_.InvokeTypes, (dispid, 0,
> wFlags, retType, argTypes) + args), user, resultCLSID) COM Error:
> [Microsoft][ODBC Driver Manager] Data source name not found and no
> default driver specified
>
> What is happening here? Why the error when called from IIS, but not
> when run interactively?
>
Is it possobile pyasp is a user DSN, when ASP expects systems DSNs?

regards
 Steve




From dsavitsk at e-coli.net  Fri Feb  2 23:38:50 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 2 Feb 2001 16:38:50 -0600
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com>
Message-ID: <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net>

> >         adoConn.Open('pyasp')

> Is it possobile pyasp is a user DSN, when ASP expects systems DSNs?

if this is the issue, asp and python work well (faster in fact on a web
server) with dsn-less connection strings.

ds




From fluxent at yahoo.com  Wed Feb  7 23:40:14 2001
From: fluxent at yahoo.com (Bill Seitz)
Date: Wed, 07 Feb 2001 22:40:14 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net>
Message-ID: <95sisd$ahc$1@nnrp1.deja.com>

Well, it's a User DSN. Do you know that ASP expects a system DSN?

And, I'm not clear on what a DSN-less connection is. Does this mean
that I don't set up an ODBC source at all, but just pass all the
connection details along in a code call? If so, do I need to use a
different COM object (see my original posting for code), or just call
the same object with a bigger set of params? (Sorry, I'm new at this,
probably need some additional books of examples...)


In article <%FGe6.31314$Ch.7369562 at newsrump.sjc.telocity.net>,
  "dsavitsk" <dsavitsk at e-coli.net> wrote:
> > >         adoConn.Open('pyasp')
>
> > Is it possobile pyasp is a user DSN, when ASP expects systems DSNs?
>
> if this is the issue, asp and python work well (faster in fact on a
web
> server) with dsn-less connection strings.
>
> ds
>
>


Sent via Deja.com
http://www.deja.com/


From bseitz at my-deja.com  Wed Feb  7 23:53:45 2001
From: bseitz at my-deja.com (Bill Seitz)
Date: Wed, 07 Feb 2001 22:53:45 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com>
Message-ID: <95sjll$bbs$1@nnrp1.deja.com>

Did some more research. Changed a line of code to look like
adoConn = win32com.client.Dispatch('ADODB.Connection')
adoConn.Open('SQL Server;server=sql0;uid=xxx;pwd=yyy;database=zzz')

Get similar error message:
Error Type:
Python ActiveX Scripting Engine (0x80020009)
Traceback (innermost last): File "<Script Block >", line 3, in ? x =
aAdoDsn.aConnect() File "c:\python20\ipath\aAdoDsn.py", line 8, in
__init__ adoConn.Open('SQL
Server;server=sql0;uid=xxx;pwd=yyy;database=zzz') File "c:\python20
\win32com\gen_py\00000206-0000-0010-8000-00AA006D2EA4x0x2x6.py", line
1833, in Open return self._ApplyTypes_(0xa, 1, (24, 32), ((8, 49), (8,
49), (8, 49), (3, 49)), 'Open', None,ConnectionString, UserID,
Password, Options) File "c:\python20\win32com\client\__init__.py", line
334, in _ApplyTypes_ return self._get_good_object_(apply
(self._oleobj_.InvokeTypes, (dispid, 0, wFlags, retType, argTypes) +
args), user, resultCLSID) COM Error: [Microsoft][ODBC Driver Manager]
Data source name not found and no default driver specified

What am I doing wrong?


In article <95sisd$ahc$1 at nnrp1.deja.com>,
  Bill Seitz <fluxent at yahoo.com> wrote:
> Well, it's a User DSN. Do you know that ASP expects a system DSN?
>
> And, I'm not clear on what a DSN-less connection is. Does this mean
> that I don't set up an ODBC source at all, but just pass all the
> connection details along in a code call? If so, do I need to use a
> different COM object (see my original posting for code), or just call
> the same object with a bigger set of params? (Sorry, I'm new at this,
> probably need some additional books of examples...)
>
> In article <%FGe6.31314$Ch.7369562 at newsrump.sjc.telocity.net>,
>   "dsavitsk" <dsavitsk at e-coli.net> wrote:
> > > >         adoConn.Open('pyasp')
> >
> > > Is it possobile pyasp is a user DSN, when ASP expects systems
DSNs?
> >
> > if this is the issue, asp and python work well (faster in fact on a
> web
> > server) with dsn-less connection strings.
> >
> > ds
> >
> >


Sent via Deja.com
http://www.deja.com/


From fluxent at yahoo.com  Thu Feb  8 00:06:17 2001
From: fluxent at yahoo.com (Bill Seitz)
Date: Wed, 07 Feb 2001 23:06:17 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com> <95sjll$bbs$1@nnrp1.deja.com>
Message-ID: <95skd2$c13$1@nnrp1.deja.com>

D'oh, had a typo. Now I get a different error.
adoConn.Open('driver=SQL
Server;server=sql0;uid=xxx;pwd=yyy;database=zzz')

....COM Error: [Microsoft][ODBC SQL Server Driver][SQL Server]Login
failed for user 'xxx'. Reason: Not associated with a trusted SQL Server
connection.

OK, this is more familiar. I'm guessing either (a) it's not recognizing
the server name, or (b) it's trying to do a "SQL Server
authentication", when our SQL Server is set up to only do Active
Directory authentication. I suspect it's (b).


In article <95sjll$bbs$1 at nnrp1.deja.com>,
  Bill Seitz <bseitz at my-deja.com> wrote:
> Did some more research. Changed a line of code to look like
> adoConn = win32com.client.Dispatch('ADODB.Connection')
> adoConn.Open('SQL Server;server=sql0;uid=xxx;pwd=yyy;database=zzz')
>
> Get similar error message:
> Error Type:
> Python ActiveX Scripting Engine (0x80020009)
> Traceback (innermost last): File "<Script Block >", line 3, in ? x =
> aAdoDsn.aConnect() File "c:\python20\ipath\aAdoDsn.py", line 8, in
> __init__ adoConn.Open('SQL
> Server;server=sql0;uid=xxx;pwd=yyy;database=zzz') File "c:\python20
> \win32com\gen_py\00000206-0000-0010-8000-00AA006D2EA4x0x2x6.py", line
> 1833, in Open return self._ApplyTypes_(0xa, 1, (24, 32), ((8, 49), (8,
> 49), (8, 49), (3, 49)), 'Open', None,ConnectionString, UserID,
> Password, Options) File "c:\python20\win32com\client\__init__.py",
line
> 334, in _ApplyTypes_ return self._get_good_object_(apply
> (self._oleobj_.InvokeTypes, (dispid, 0, wFlags, retType, argTypes) +
> args), user, resultCLSID) COM Error: [Microsoft][ODBC Driver Manager]
> Data source name not found and no default driver specified
>
> What am I doing wrong?
>
> In article <95sisd$ahc$1 at nnrp1.deja.com>,
>   Bill Seitz <fluxent at yahoo.com> wrote:
> > Well, it's a User DSN. Do you know that ASP expects a system DSN?
> >
> > And, I'm not clear on what a DSN-less connection is. Does this mean
> > that I don't set up an ODBC source at all, but just pass all the
> > connection details along in a code call? If so, do I need to use a
> > different COM object (see my original posting for code), or just
call
> > the same object with a bigger set of params? (Sorry, I'm new at
this,
> > probably need some additional books of examples...)
> >
> > In article <%FGe6.31314$Ch.7369562 at newsrump.sjc.telocity.net>,
> >   "dsavitsk" <dsavitsk at e-coli.net> wrote:
> > > > >         adoConn.Open('pyasp')
> > >
> > > > Is it possobile pyasp is a user DSN, when ASP expects systems
> DSNs?
> > >
> > > if this is the issue, asp and python work well (faster in fact on
a
> > web
> > > server) with dsn-less connection strings.
> > >
> > > ds
> > >
> > >
>
> Sent via Deja.com
> http://www.deja.com/
>


Sent via Deja.com
http://www.deja.com/


From dsavitsk at e-coli.net  Thu Feb  8 00:57:47 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 7 Feb 2001 17:57:47 -0600
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com> <95sjll$bbs$1@nnrp1.deja.com> <95skd2$c13$1@nnrp1.deja.com>
Message-ID: <1ilg6.61167$Ch.11522747@newsrump.sjc.telocity.net>

maybe not the error here, but you need a ";" at the end of the string ...
see below.

> adoConn.Open('driver=SQL
> Server;server=sql0;uid=xxx;pwd=yyy;database=zzz;')





From fuess at att.net  Thu Feb  8 02:50:55 2001
From: fuess at att.net (David Fuess)
Date: Thu, 08 Feb 2001 01:50:55 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com> <95sjll$bbs$1@nnrp1.deja.com> <95skd2$c13$1@nnrp1.deja.com> <1ilg6.61167$Ch.11522747@newsrump.sjc.telocity.net>
Message-ID: <lku38t0dl2jb5210e7ler5kv3c7qcfiadk@4ax.com>

Here is the beginning of my SQL Server class:

class Database:
   def __init__(self,Name,Server="Default",Arg=""):
      self.adoConn = None
      self.adoRS = None
      self.RecordCount = 0
      self.eof = 1
      self.ConnectString = "database=%s;driver={SQL
Server};QueryTimeout=300;server=%s;uid=sa%s"%(Name,Server,Arg)
      self.Schema = {}
      if AutoOpenOnInit:
         self.Open(Name)

   def Open(self,Name):
      if self.adoConn is not None:
         self.Close()
      self.adoConn = win32com.client.Dispatch('ADODB.Connection')
      self.adoConn.CursorLocation = 3
      self.adoConn.CommandTimeout = 900
      self.adoConn.Open(self.ConnectString)
      if AutoloadSchema:
         self.LoadSchema()

And so on ...

I use it regularly to connect to SQL Server Databases. It is DSNless
in that it attaches directly to SQL Server without any specific DSN
assignments. It has nothing to do with IIS, only ADO.

Dave
On Wed, 7 Feb 2001 17:57:47 -0600, "dsavitsk" <dsavitsk at e-coli.net>
wrote:

>maybe not the error here, but you need a ";" at the end of the string ...
>see below.
>
>> adoConn.Open('driver=SQL
>> Server;server=sql0;uid=xxx;pwd=yyy;database=zzz;')
>
>



From fluxent at yahoo.com  Fri Feb  9 22:52:51 2001
From: fluxent at yahoo.com (Bill Seitz)
Date: Fri, 09 Feb 2001 21:52:51 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com> <95sjll$bbs$1@nnrp1.deja.com> <95skd2$c13$1@nnrp1.deja.com> <1ilg6.61167$Ch.11522747@newsrump.sjc.telocity.net> <lku38t0dl2jb5210e7ler5kv3c7qcfiadk@4ax.com>
Message-ID: <961orf$nsk$1@nnrp1.deja.com>

Well, my simpler DSN-less code seems to be doing the right thing. But
get that lovely "Not associated with a trusted SQL Server connection."
error message. I've tried with and without a domain prefix on the uid.

Could this be related to running SQL in "active directory only" mode?

In article <lku38t0dl2jb5210e7ler5kv3c7qcfiadk at 4ax.com>,
  David Fuess <fuess at att.net> wrote:
>
> Here is the beginning of my SQL Server class:
>
> class Database:
>    def __init__(self,Name,Server="Default",Arg=""):
>       self.adoConn = None
>       self.adoRS = None
>       self.RecordCount = 0
>       self.eof = 1
>       self.ConnectString = "database=%s;driver={SQL
> Server};QueryTimeout=300;server=%s;uid=sa%s"%(Name,Server,Arg)
>       self.Schema = {}
>       if AutoOpenOnInit:
>          self.Open(Name)
>
>    def Open(self,Name):
>       if self.adoConn is not None:
>          self.Close()
>       self.adoConn = win32com.client.Dispatch('ADODB.Connection')
>       self.adoConn.CursorLocation = 3
>       self.adoConn.CommandTimeout = 900
>       self.adoConn.Open(self.ConnectString)
>       if AutoloadSchema:
>          self.LoadSchema()
>
> And so on ...
>
> I use it regularly to connect to SQL Server Databases. It is DSNless
> in that it attaches directly to SQL Server without any specific DSN
> assignments. It has nothing to do with IIS, only ADO.
>
> Dave
> On Wed, 7 Feb 2001 17:57:47 -0600, "dsavitsk" <dsavitsk at e-coli.net>
> wrote:
>
> >maybe not the error here, but you need a ";" at the end of the
string ...
> >see below.
> >
> >> adoConn.Open('driver=SQL
> >> Server;server=sql0;uid=xxx;pwd=yyy;database=zzz;')
> >
> >
>
>


Sent via Deja.com
http://www.deja.com/


From MarkH at ActiveState.com  Thu Feb  8 01:34:37 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 08 Feb 2001 00:34:37 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com>
Message-ID: <3A81E8B4.2000402@ActiveState.com>

Bill Seitz wrote:

> Well, it's a User DSN. Do you know that ASP expects a system DSN?

ASP does not execute as your user, but as its own system user.  Thus, a user DSN created for your user is simply not visible by ASP.  It must be system DSN, or a user DSN for the ASP specific user

> And, I'm not clear on what a DSN-less connection is. Does this mean

Neither do I - hopefully someone else will ;-)

Mark.



From fuess at att.net  Thu Feb  8 02:56:54 2001
From: fuess at att.net (David Fuess)
Date: Thu, 08 Feb 2001 01:56:54 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com> <3A81E8B4.2000402@ActiveState.com>
Message-ID: <ssu38t8nqkai2fmv14hl3rcp8d3hpgf691@4ax.com>

See my response to dsavitsk for connecting to a database using ADO. As
for DSNless, I take it to mean that no DSN need be created for the
database to be visible. For instance, SQL Server databases that you
connect using the ADODB.Connection mechanism do not have DSN's (either
user or system). If you want to use ODBC mechanisms to connect then
you do have to have a system DSN assigned for it to be visible. This
is what one does to get to an Access database using the Jet engine.

Dave

On Thu, 08 Feb 2001 00:34:37 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>Bill Seitz wrote:
>
>> Well, it's a User DSN. Do you know that ASP expects a system DSN?
>
>ASP does not execute as your user, but as its own system user.  Thus, a user DSN created for your user is simply not visible by ASP.  It must be system DSN, or a user DSN for the ASP specific user
>
>> And, I'm not clear on what a DSN-less connection is. Does this mean
>
>Neither do I - hopefully someone else will ;-)
>
>Mark.



From dsavitsk at e-coli.net  Thu Feb  8 05:23:44 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 7 Feb 2001 22:23:44 -0600
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com> <3A81E8B4.2000402@ActiveState.com> <ssu38t8nqkai2fmv14hl3rcp8d3hpgf691@4ax.com>
Message-ID: <mbpg6.62548$Ch.11694590@newsrump.sjc.telocity.net>

> >> And, I'm not clear on what a DSN-less connection is.

this isn't going to be technical ...

a DSN-less connection is placing the entire DSN file (a text file) in a
string and executing it there.  There are 3 benefits.  One is that if your
web site runs on a win32 server, you don't need admin access to set up
database access (making hosting simpler and less expensive). The second is
that since there is a little less communication with the OS (no searching
for or calling DSN files and what not), DSN-Less runs a little faster (I
think this is only true for 'small' uses (and i don't know what that means
exactly either)). The third is that one doesn't need to create a system/user
DSN when installing software.

i am not an expert ... or even good at ADO (i'm in law school and computer
time is limited), but i have a little Python/ADO/MS Access tutorial (using
DSN-Less connection strings) @ www.e-coli.net/pyado.html

doug







From fluxent at yahoo.com  Fri Feb  9 18:36:01 2001
From: fluxent at yahoo.com (Bill Seitz)
Date: Fri, 09 Feb 2001 17:36:01 GMT
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com> <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net> <95sisd$ahc$1@nnrp1.deja.com> <3A81E8B4.2000402@ActiveState.com>
Message-ID: <9619q0$96e$1@nnrp1.deja.com>

OK, I tried going back to the DSN approach. I dumped my user DSN and
created a system DSN. It looks to me like:

* you have to turn off IIS anonymous access, else you get an error
message of "COM Error: [Microsoft][ODBC SQL Server Driver][SQL Server]
Login failed for user 'NT AUTHORITY\ANONYMOUS LOGON'."

* you have to turn on Basic Authentication for IIS, supporting only the
(more secure) Windows-integrated or digest authentication doesn't do it!

* since the web-page-user's ID is what's being passed along, you have
to set up SQL server accounts for all potential users, right?

Am I missing something, or is this horribly lame?

In article <3A81E8B4.2000402 at ActiveState.com>,
  Mark Hammond <MarkH at ActiveState.com> wrote:
> Bill Seitz wrote:
>
> > Well, it's a User DSN. Do you know that ASP expects a system DSN?
>
> ASP does not execute as your user, but as its own system user.  Thus,
a user DSN created for your user is simply not visible by ASP.  It must
be system DSN, or a user DSN for the ASP specific user
>
> > And, I'm not clear on what a DSN-less connection is. Does this mean
>
> Neither do I - hopefully someone else will ;-)
>
> Mark.
>
>


Sent via Deja.com
http://www.deja.com/


From dnedrow at usa.net  Fri Feb  2 22:30:47 2001
From: dnedrow at usa.net (dnedrow at usa.net)
Date: Fri, 02 Feb 2001 21:30:47 GMT
Subject: Substring functions
Message-ID: <95f8u7$n1f$1@nnrp1.deja.com>

I'm trying to fixup someone elses work and am a complete Python neophyte.

The are two functions (is that even the term?) as follows:

def shellProtect(file):
  """Protects a filename against shell interpretation.
  This is done by putting the name in single quotes, and by
  escaping single quotes in the filename.
  """
  return "'%s'" % string.replace(file, "'", "'\\''")


def mv(self, src, dest):
  """Shorthand for the pattern trace('mv...);os.system("mv...)."""
  self.system("mv %s %s" % (shellProtect(src), shellProtect(dest)))

Basically, mv() is called with source and destination parameters. These
are quoted via shellProtext() and then passed as input parameters for the
Unix mv command.

E.g. if src=/usr/tmp/1/* and dest=/usr/tmp/2, the mv command issued by
calling mv() above would be:

  mv '/usr/tmp/1/*' '/usr/tmp/2'

Unfortunately, a wildcard doesn't work when quoted as above. How do I
check for an * at the end of the string and quote the input as
  '/usr/tmp/1/'*
rather than
  '/usr/tmp/1/*'

Any pointers would be appreciated.

-David


Sent via Deja.com
http://www.deja.com/


From loewis at informatik.hu-berlin.de  Fri Feb  2 23:45:47 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 02 Feb 2001 23:45:47 +0100
Subject: Substring functions
References: <95f8u7$n1f$1@nnrp1.deja.com>
Message-ID: <p6q3ddwog9w.fsf@informatik.hu-berlin.de>

dnedrow at usa.net writes:

> The are two functions (is that even the term?)

Sure. What you got are functions. 

> def shellProtect(file):
>   """Protects a filename against shell interpretation.
>   This is done by putting the name in single quotes, and by
>   escaping single quotes in the filename.
>   """
>   return "'%s'" % string.replace(file, "'", "'\\''")
> 
> 
> def mv(self, src, dest):
>   """Shorthand for the pattern trace('mv...);os.system("mv...)."""
>   self.system("mv %s %s" % (shellProtect(src), shellProtect(dest)))

As a side note: Isn't os.rename good enough? On Unix, os.rename won't
work across filesystem boundaries, but some versions of mv will; if
you can exclude that case, os.rename is faster and more reliable than
invoking a shell.

> Basically, mv() is called with source and destination parameters. These
> are quoted via shellProtext() and then passed as input parameters for the
> Unix mv command.
> 
> E.g. if src=/usr/tmp/1/* and dest=/usr/tmp/2, the mv command issued by
> calling mv() above would be:
> 
>   mv '/usr/tmp/1/*' '/usr/tmp/2'
> 
> Unfortunately, a wildcard doesn't work when quoted as above. How do I
> check for an * at the end of the string and quote the input as
>   '/usr/tmp/1/'*
> rather than
>   '/usr/tmp/1/*'
> 
> Any pointers would be appreciated.

Again, I suggest to operate differently. glob.glob will expand
wildcards, and just leave the name as-is if it is a non-pattern valid
file name (actually, it returns a list of file names in either case).

It turns out that os.rename wants the target filename; it doesn't
automatically generate a file name if the target is a directory. That
is where glob.glob1 helps.

So to do the equivalent of your move, I'd do

import os,glob
srcdir = "/usr/tmp/1"
dstdir = "/usr/tmp/2"
for srcfile in glob.glob1(srcdir,"*"):
  src = os.path.join(srcdir,srcfile)
  dst = os.paht.join(dstdir,srcfile)
  os.rename(src, dst)

In addition, if the only pattern you care about is "*" (i.e. all
files), then

for srcfile in os.listdir(srcdir):
  ...

is even better: it is faster, clearer to read, and it will catch
dotfiles (.foo) in /usr/tmp/1 whereas * won't.

I don't know whether you have to support a single move-all-files
operation or whether this is just a special case of something more
general, so I can't advise a "best" solution.

Regards,
Martin


From jeremy at alum.mit.edu  Sat Feb  3 00:39:51 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 2 Feb 2001 18:39:51 -0500 (EST)
Subject: Python 2.1 alpha 2 released
Message-ID: <14971.17735.263154.15769@w221.z064000254.bwi-md.dsl.cnc.net>

While Guido was working the press circuit at the LinuxWorld Expo in New
York City, the Python developers, including the many volunteers and
the folks from PythonLabs, were busy finishing the second alpha
release of Python 2.1.

The release is currently available from SourceForge and will also be
available from python.org later today.  You can find the source
release at:

    http://sourceforge.net/project/showfiles.php?group_id=5470

The Windows installer will be ready shortly.

Fred Drake announced the documentation release earlier today.  You can
browse the new docs online at

    http://python.sourceforge.net/devel-docs/

or download them from

    ftp://ftp.python.org/pub/python/doc/2.1a2/

Please give it a good try!  The only way Python 2.1 can become a
rock-solid product is if people test the alpha releases.  If you are
using Python for demanding applications or on extreme platforms, we
are particularly interested in hearing your feedback.  Are you
embedding Python or using threads?  Please test your application using
Python 2.1a2!  Please submit all bug reports through SourceForge:

  http://sourceforge.net/bugs/?group_id=5470

Here's the NEWS file:

What's New in Python 2.1 alpha 2?
=================================

Core language, builtins, and interpreter

- Scopes nest.  If a name is used in a function or class, but is not
  local, the definition in the nearest enclosing function scope will
  be used.  One consequence of this change is that lambda statements
  could reference variables in the namespaces where the lambda is
  defined.  In some unusual cases, this change will break code.

  In all previous version of Python, names were resolved in exactly
  three namespaces -- the local namespace, the global namespace, and
  the builtin namespace.  According to this old definition, if a
  function A is defined within a function B, the names bound in B are
  not visible in A.  The new rules make names bound in B visible in A,
  unless A contains a name binding that hides the binding in B.

  Section 4.1 of the reference manual describes the new scoping rules
  in detail.  The test script in Lib/test/test_scope.py demonstrates
  some of the effects of the change.

  The new rules will cause existing code to break if it defines nested
  functions where an outer function has local variables with the same
  name as globals or builtins used by the inner function.  Example:

    def munge(str):
        def helper(x):
            return str(x)
        if type(str) != type(''):
            str = helper(str)
        return str.strip()

  Under the old rules, the name str in helper() is bound to the
  builtin function str().  Under the new rules, it will be bound to
  the argument named str and an error will occur when helper() is
  called.

- The compiler will report a SyntaxError if "from ... import *" occurs
  in a function or class scope.  The language reference has documented
  that this case is illegal, but the compiler never checked for it.
  The recent introduction of nested scope makes the meaning of this
  form of name binding ambiguous.  In a future release, the compiler
  may allow this form when there is no possibility of ambiguity.

- repr(string) is easier to read, now using hex escapes instead of octal,
  and using \t, \n and \r instead of \011, \012 and \015 (respectively):

  >>> "\texample \r\n" + chr(0) + chr(255)
  '\texample \r\n\x00\xff'         # in 2.1
  '\011example \015\012\000\377'   # in 2.0

- Functions are now compared and hashed by identity, not by value, since
  the func_code attribute is writable.

- Weak references (PEP 205) have been added.  This involves a few
  changes in the core, an extension module (_weakref), and a Python
  module (weakref).  The weakref module is the public interface.  It
  includes support for "explicit" weak references, proxy objects, and
  mappings with weakly held values.

- A 'continue' statement can now appear in a try block within the body
  of a loop.  It is still not possible to use continue in a finally
  clause.

Standard library

- mailbox.py now has a new class, PortableUnixMailbox which is
  identical to UnixMailbox but uses a more portable scheme for
  determining From_ separators.  Also, the constructors for all the
  classes in this module have a new optional `factory' argument, which
  is a callable used when new message classes must be instantiated by
  the next() method.

- random.py is now self-contained, and offers all the functionality of
  the now-deprecated whrandom.py.  See the docs for details.  random.py
  also supports new functions getstate() and setstate(), for saving
  and restoring the internal state of the generator; and jumpahead(n),
  for quickly forcing the internal state to be the same as if n calls to
  random() had been made.  The latter is particularly useful for multi-
  threaded programs, creating one instance of the random.Random() class for
  each thread, then using .jumpahead() to force each instance to use a
  non-overlapping segment of the full period.

- random.py's seed() function is new.  For bit-for-bit compatibility with
  prior releases, use the whseed function instead.  The new seed function
  addresses two problems:  (1) The old function couldn't produce more than
  about 2**24 distinct internal states; the new one about 2**45 (the best
  that can be done in the Wichmann-Hill generator).  (2) The old function
  sometimes produced identical internal states when passed distinct
  integers, and there was no simple way to predict when that would happen;
  the new one guarantees to produce distinct internal states for all
  arguments in [0, 27814431486576L).

- The socket module now supports raw packets on Linux.  The socket
  family is AF_PACKET.

- test_capi.py is a start at running tests of the Python C API.  The tests
  are implemented by the new Modules/_testmodule.c.

- A new extension module, _symtable, provides provisional access to the
  internal symbol table used by the Python compiler.  A higher-level
  interface will be added on top of _symtable in a future release.

Windows changes

- Build procedure:  the zlib project is built in a different way that
  ensures the zlib header files used can no longer get out of synch with
  the zlib binary used.  See PCbuild\readme.txt for details.  Your old
  zlib-related directories can be deleted; you'll need to download fresh
  source for zlib and unpack it into a new directory.

- Build:  New subproject _test for the benefit of test_capi.py (see above).

- Build:  subproject ucnhash is gone, since the code was folded into the
  unicodedata subproject.

What's New in Python 2.1 alpha 1?
=================================

Core language, builtins, and interpreter

- There is a new Unicode companion to the PyObject_Str() API
  called PyObject_Unicode(). It behaves in the same way as the
  former, but assures that the returned value is an Unicode object
  (applying the usual coercion if necessary).

- The comparison operators support "rich comparison overloading" (PEP
  207).  C extension types can provide a rich comparison function in
  the new tp_richcompare slot in the type object.  The cmp() function
  and the C function PyObject_Compare() first try the new rich
  comparison operators before trying the old 3-way comparison.  There
  is also a new C API PyObject_RichCompare() (which also falls back on
  the old 3-way comparison, but does not constrain the outcome of the
  rich comparison to a Boolean result).

  The rich comparison function takes two objects (at least one of
  which is guaranteed to have the type that provided the function) and
  an integer indicating the opcode, which can be Py_LT, Py_LE, Py_EQ,
  Py_NE, Py_GT, Py_GE (for <, <=, ==, !=, >, >=), and returns a Python
  object, which may be NotImplemented (in which case the tp_compare
  slot function is used as a fallback, if defined).

  Classes can overload individual comparison operators by defining one
  or more of the methods__lt__, __le__, __eq__, __ne__, __gt__,
  __ge__.  There are no explicit "reflected argument" versions of
  these; instead, __lt__ and __gt__ are each other's reflection,
  likewise for__le__ and __ge__; __eq__ and __ne__ are their own
  reflection (similar at the C level).  No other implications are
  made; in particular, Python does not assume that == is the Boolean
  inverse of !=, or that < is the Boolean inverse of >=.  This makes
  it possible to define types with partial orderings.

  Classes or types that want to implement (in)equality tests but not
  the ordering operators (i.e. unordered types) should implement ==
  and !=, and raise an error for the ordering operators.

  It is possible to define types whose rich comparison results are not
  Boolean; e.g. a matrix type might want to return a matrix of bits
  for A < B, giving elementwise comparisons.  Such types should ensure
  that any interpretation of their value in a Boolean context raises
  an exception, e.g. by defining __nonzero__ (or the tp_nonzero slot
  at the C level) to always raise an exception.

- Complex numbers use rich comparisons to define == and != but raise
  an exception for <, <=, > and >=.  Unfortunately, this also means
  that cmp() of two complex numbers raises an exception when the two
  numbers differ.  Since it is not mathematically meaningful to compare
  complex numbers except for equality, I hope that this doesn't break
  too much code.

- Functions and methods now support getting and setting arbitrarily
  named attributes (PEP 232).  Functions have a new __dict__
  (a.k.a. func_dict) which hold the function attributes.  Methods get
  and set attributes on their underlying im_func.  It is a TypeError
  to set an attribute on a bound method.

- The xrange() object implementation has been improved so that
  xrange(sys.maxint) can be used on 64-bit platforms.  There's still a
  limitation that in this case len(xrange(sys.maxint)) can't be
  calculated, but the common idiom "for i in xrange(sys.maxint)" will
  work fine as long as the index i doesn't actually reach 2**31.
  (Python uses regular ints for sequence and string indices; fixing
  that is much more work.)

- Two changes to from...import:

  1) "from M import X" now works even if M is not a real module; it's
     basically a getattr() operation with AttributeError exceptions
     changed into ImportError.

  2) "from M import *" now looks for M.__all__ to decide which names to
     import; if M.__all__ doesn't exist, it uses M.__dict__.keys() but
     filters out names starting with '_' as before.  Whether or not
     __all__ exists, there's no restriction on the type of M.

- File objects have a new method, xreadlines().  This is the fastest
  way to iterate over all lines in a file:

  for line in file.xreadlines():
      ...do something to line...

  See the xreadlines module (mentioned below) for how to do this for
  other file-like objects.

- Even if you don't use file.xreadlines(), you may expect a speedup on
  line-by-line input.  The file.readline() method has been optimized
  quite a bit in platform-specific ways:  on systems (like Linux) that
  support flockfile(), getc_unlocked(), and funlockfile(), those are
  used by default.  On systems (like Windows) without getc_unlocked(),
  a complicated (but still thread-safe) method using fgets() is used by
  default.

  You can force use of the fgets() method by #define'ing
  USE_FGETS_IN_GETLINE at build time (it may be faster than
  getc_unlocked()).

  You can force fgets() not to be used by #define'ing
  DONT_USE_FGETS_IN_GETLINE (this is the first thing to try if std test
  test_bufio.py fails -- and let us know if it does!).

- In addition, the fileinput module, while still slower than the other
  methods on most platforms, has been sped up too, by using
  file.readlines(sizehint).

- Support for run-time warnings has been added, including a new
  command line option (-W) to specify the disposition of warnings.
  See the description of the warnings module below.

- Extensive changes have been made to the coercion code.  This mostly
  affects extension modules (which can now implement mixed-type
  numerical operators without having to use coercion), but
  occasionally, in boundary cases the coercion semantics have changed
  subtly.  Since this was a terrible gray area of the language, this
  is considered an improvement.  Also note that __rcmp__ is no longer
  supported -- instead of calling __rcmp__, __cmp__ is called with
  reflected arguments.

- In connection with the coercion changes, a new built-in singleton
  object, NotImplemented is defined.  This can be returned for
  operations that wish to indicate they are not implemented for a
  particular combination of arguments.  From C, this is
  Py_NotImplemented.

- The interpreter accepts now bytecode files on the command line even
  if they do not have a .pyc or .pyo extension. On Linux, after executing

  echo ':pyc:M::\x87\xc6\x0d\x0a::/usr/local/bin/python:' > /proc/sys/fs/binfmt_misc/register

  any byte code file can be used as an executable (i.e. as an argument
  to execve(2)).

- %[xXo] formats of negative Python longs now produce a sign
  character.  In 1.6 and earlier, they never produced a sign,
  and raised an error if the value of the long was too large
  to fit in a Python int.  In 2.0, they produced a sign if and
  only if too large to fit in an int.  This was inconsistent
  across platforms (because the size of an int varies across
  platforms), and inconsistent with hex() and oct().  Example:

  >>> "%x" % -0x42L
  '-42'      # in 2.1
  'ffffffbe' # in 2.0 and before, on 32-bit machines
  >>> hex(-0x42L)
  '-0x42L'   # in all versions of Python

  The behavior of %d formats for negative Python longs remains
  the same as in 2.0 (although in 1.6 and before, they raised
  an error if the long didn't fit in a Python int).

  %u formats don't make sense for Python longs, but are allowed
  and treated the same as %d in 2.1.  In 2.0, a negative long
  formatted via %u produced a sign if and only if too large to
  fit in an int.  In 1.6 and earlier, a negative long formatted
  via %u raised an error if it was too big to fit in an int.

- Dictionary objects have an odd new method, popitem().  This removes
  an arbitrary item from the dictionary and returns it (in the form of
  a (key, value) pair).  This can be useful for algorithms that use a
  dictionary as a bag of "to do" items and repeatedly need to pick one
  item.  Such algorithms normally end up running in quadratic time;
  using popitem() they can usually be made to run in linear time.

Standard library

- In the time module, the time argument to the functions strftime,
  localtime, gmtime, asctime and ctime is now optional, defaulting to
  the current time (in the local timezone).

- The ftplib module now defaults to passive mode, which is deemed a
  more useful default given that clients are often inside firewalls
  these days.  Note that this could break if ftplib is used to connect
  to a *server* that is inside a firewall, from outside; this is
  expected to be a very rare situation.  To fix that, you can call
  ftp.set_pasv(0).

- The module site now treats .pth files not only for path configuration,
  but also supports extensions to the initialization code: Lines starting
  with import are executed.

- There's a new module, warnings, which implements a mechanism for
  issuing and filtering warnings.  There are some new built-in
  exceptions that serve as warning categories, and a new command line
  option, -W, to control warnings (e.g. -Wi ignores all warnings, -We
  turns warnings into errors).  warnings.warn(message[, category])
  issues a warning message; this can also be called from C as
  PyErr_Warn(category, message).

- A new module xreadlines was added.  This exports a single factory
  function, xreadlines().  The intention is that this code is the
  absolutely fastest way to iterate over all lines in an open
  file(-like) object:

  import xreadlines
  for line in xreadlines.xreadlines(file):
      ...do something to line...

  This is equivalent to the previous the speed record holder using
  file.readlines(sizehint).  Note that if file is a real file object
  (as opposed to a file-like object), this is equivalent:

  for line in file.xreadlines():
      ...do something to line...

- The bisect module has new functions bisect_left, insort_left,
  bisect_right and insort_right.  The old names bisect and insort
  are now aliases for bisect_right and insort_right.  XXX_right
  and XXX_left methods differ in what happens when the new element
  compares equal to one or more elements already in the list:  the
  XXX_left methods insert to the left, the XXX_right methods to the
  right.  Code that doesn't care where equal elements end up should
  continue to use the old, short names ("bisect" and "insort").

- The new curses.panel module wraps the panel library that forms part
  of SYSV curses and ncurses.  Contributed by Thomas Gellekum.

- The SocketServer module now sets the allow_reuse_address flag by
  default in the TCPServer class.

- A new function, sys._getframe(), returns the stack frame pointer of
  the caller.  This is intended only as a building block for
  higher-level mechanisms such as string interpolation.

Build issues

- For Unix (and Unix-compatible) builds, configuration and building of
  extension modules is now greatly automated.  Rather than having to
  edit the Modules/Setup file to indicate which modules should be
  built and where their include files and libraries are, a
  distutils-based setup.py script now takes care of building most
  extension modules.  All extension modules built this way are built
  as shared libraries.  Only a few modules that must be linked
  statically are still listed in the Setup file; you won't need to
  edit their configuration.

- Python should now build out of the box on Cygwin.  If it doesn't,
  mail to Jason Tishler (jlt63 at users.sourceforge.net).

- Python now always uses its own (renamed) implementation of getopt()
  -- there's too much variation among C library getopt()
  implementations.

- C++ compilers are better supported; the CXX macro is always set to a
  C++ compiler if one is found.

Windows changes

- select module:  By default under Windows, a select() call
  can specify no more than 64 sockets.  Python now boosts
  this Microsoft default to 512.  If you need even more than
  that, see the MS docs (you'll need to #define FD_SETSIZE
  and recompile Python from source).

- Support for Windows 3.1, DOS and OS/2 is gone.  The Lib/dos-8x3
  subdirectory is no more!

-- Jeremy Hylton <http://www.python.org/~jeremy/>



From jafo at tummy.com  Sun Feb  4 05:50:15 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Sat, 3 Feb 2001 21:50:15 -0700
Subject: Python 2.1 alpha 2 released
In-Reply-To: <14971.17735.263154.15769@w221.z064000254.bwi-md.dsl.cnc.net>; from jeremy@alum.mit.edu on Fri, Feb 02, 2001 at 06:39:51PM -0500
References: <14971.17735.263154.15769@w221.z064000254.bwi-md.dsl.cnc.net>
Message-ID: <20010203215015.B29866@tummy.com>

On Fri, Feb 02, 2001 at 06:39:51PM -0500, Jeremy Hylton wrote:
>The release is currently available from SourceForge and will also be

My SRPM is available at:

	ftp://ftp.tummy.com/pub/tummy/RPMS/SRPMS/

To turn it into a binary RPM for your rpm-based system, run
"rpm --rebuild python-2.1a2-1tummy.src.rpm", get a cup of coffee, and
then install the resulting binary RPMs (probably under
"/usr/src/redhat/RPMS/i386").

Enjoy,
Sean
-- 
 What no spouse of a programmer can ever understand is that a programmer is
 working when he's staring out the window.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From baasad at qualitynet.net  Sat Feb  3 00:55:00 2001
From: baasad at qualitynet.net (bashar A. Asad)
Date: Sat, 03 Feb 2001 02:55:00 +0300
Subject: help with process monitoring using python
Message-ID: <95g62e$ne31@news.qualitynet.net>

hello ;
since am new to python I would like to start my experience with a small
project that will my life easier.  I would like to connect to a remote linux
machines and check for a certain process "may be useing a system command is
their something like that in python??" eg. httpd if its up or not if it is
up the program will produce an activity file or whatever... I have no idea
how to start working on that in python..any hints guys :)
thanx

bashar


From jafo at tummy.com  Sat Feb  3 17:41:13 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Sat, 3 Feb 2001 09:41:13 -0700
Subject: help with process monitoring using python
In-Reply-To: <95g62e$ne31@news.qualitynet.net>; from baasad@qualitynet.net on Sat, Feb 03, 2001 at 02:55:00AM +0300
References: <95g62e$ne31@news.qualitynet.net>
Message-ID: <20010203094113.C4938@tummy.com>

On Sat, Feb 03, 2001 at 02:55:00AM +0300, bashar A. Asad wrote:
>project that will my life easier.  I would like to connect to a remote linux
>machines and check for a certain process "may be useing a system command is
>their something like that in python??" eg. httpd if its up or not if it is
>up the program will produce an activity file or whatever... I have no idea
>how to start working on that in python..any hints guys :)

Check out:

	ftp://www.tummy.com/pub/pythonExamples/checkpid.py

which includes a sample of doing that.  The pid file you want may be
"/var/run/httpd.pid".

I'm actually working on the ultimate process restarter right now.  The
code is done, but I don't have any docs for it yet.  It'll probably
be under ftp.tummy.com:/pub/tummy/ called something like "trestarter"
when it's finally released (hopefully in the next day or so).

Sean
-- 
 YOU ARE WITNESSING A FRONT THREE-QUARTER VIEW OF TWO ADULTS SHARING A
 TENDER MOMENT.  -- Gordon Cole, _Twin_Peaks_
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From claird at starbase.neosoft.com  Mon Feb  5 15:06:23 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Feb 2001 08:06:23 -0600
Subject: help with process monitoring using python
References: <95g62e$ne31@news.qualitynet.net> <mailman.981218524.27316.python-list@python.org>
Message-ID: <3ADDADEF1AF041A6.5188C6AE0188D1C8.788BE319FE924227@lp.airnews.net>

In article <mailman.981218524.27316.python-list at python.org>,
Sean Reifschneider  <jafo at tummy.com> wrote:
			.
			.
			.
>I'm actually working on the ultimate process restarter right now.  The
>code is done, but I don't have any docs for it yet.  It'll probably
>be under ftp.tummy.com:/pub/tummy/ called something like "trestarter"
>when it's finally released (hopefully in the next day or so).
			.
			.
			.
What's the relation between trestarter and init(8)?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From jafo at tummy.com  Thu Feb  8 07:49:25 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 7 Feb 2001 23:49:25 -0700
Subject: help with process monitoring using python
In-Reply-To: <3ADDADEF1AF041A6.5188C6AE0188D1C8.788BE319FE924227@lp.airnews.net>; from claird@starbase.neosoft.com on Mon, Feb 05, 2001 at 08:06:23AM -0600
References: <95g62e$ne31@news.qualitynet.net> <mailman.981218524.27316.python-list@python.org> <3ADDADEF1AF041A6.5188C6AE0188D1C8.788BE319FE924227@lp.airnews.net>
Message-ID: <20010207234925.H505@tummy.com>

On Mon, Feb 05, 2001 at 08:06:26AM -0600, Cameron Laird wrote:
>What's the relation between trestarter and init(8)?

trestarter works with the existing "init.d" start and stop scripts.
It doesn't require disabing the "standard" RedHat method of starting
services and plunking them into xinit.d or inittab.  It can be configured
to not restart services when the admin has done a "service named stop",
for example (or not).  It sends reports of restarting services (including
the output of scripts) to an e-mail address (or not).

In addition to simply checking for the process existing, it can also run
an arbitrary script which, for example, does a sample transaction against
the service and does a restart if it fails.

trestarter can be run from init and/or from cron.  Init being the preferred
way since it can then watch cron.  trestarter also has rather serious logging
of what it's doing and why, if you enable it.

The target is that it will be included in KRUD (our RedHat-based
distribution) and therefore needs to provide functionality without
confusing people.  Disabling the init.d scripts and putting things
into inittab just isn't an option...

Sean
--
 "All I'm saying is that when I'm around you I find myself showing off,
 which is the idiots version of being interesting."  -- _LA_Story_
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python




From rturpin at my-deja.com  Sat Feb  3 01:13:28 2001
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Sat, 03 Feb 2001 00:13:28 GMT
Subject: Will nested scoping break restricted execution ??
Message-ID: <95fif3$vm6$1@nnrp1.deja.com>

I thought that the restricted execution module functioned
partly by controlling the builtin, local, and global
namespaces. With nested scoping, the namespace model
becomes significantly more complex. Will the restricted
execution module be badly affected by nested scoping? If
so that is a bad trade. A really bad trade.

Russell


Sent via Deja.com
http://www.deja.com/


From tim.one at home.com  Sat Feb  3 09:30:21 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 03:30:21 -0500
Subject: Will nested scoping break restricted execution ??
In-Reply-To: <95fif3$vm6$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDKINAA.tim.one@home.com>

[rturpin at my-deja.com]
> I thought that the restricted execution module functioned
> partly by controlling the builtin, local, and global
> namespaces.

Right.

> With nested scoping, the namespace model becomes significantly
> more complex. Will the restricted execution module be badly
> affected by nested scoping?

Jeremy Hylton implemented nested scopes, and is also Guido's "security
geek".  I doubt he would tolerate compromise here.  But he'll have to speak
for himself, because I can only channel Guido <wink>.

My offhand *opinion* is that it's darned hard to see any way in which it
could hurt at all:  there's nothing you can do with nested functions that
you couldn't do before with classes and instance data.  Reaching outside the
module still requires getting at the builtins or module globals, and those
are still under control.  You can create an unbounded number of new local
namespaces within a single function now, but you can't bind the new local
names to anything that isn't visible in their scope, which means crawling up
the nested chain of scopes and again topping out at the globals or builtins.

> If so that is a bad trade. A really bad trade.

And if not so, it's not a trade at all <wink>.




From loewis at informatik.hu-berlin.de  Sat Feb  3 12:34:16 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 03 Feb 2001 12:34:16 +0100
Subject: Will nested scoping break restricted execution ??
References: <mailman.981189063.17430.python-list@python.org>
Message-ID: <p6qelxgroef.fsf@informatik.hu-berlin.de>

"Tim Peters" <tim.one at home.com> writes:

> [rturpin at my-deja.com]
> > I thought that the restricted execution module functioned
> > partly by controlling the builtin, local, and global
> > namespaces.
> 
> Right.

Actually wrong, I think. It does not control local and global
namespaces. Instead, it controls what names you can get from
elsewhere.

I.e. you can't prohibit a program doing

  socket = 1

but you can prevent that it does

  import socket

although either code creates a socket variable, from a local and
global namespace point of view. Likewise, you can prevent the code
from doing

  import socket as string

even if importing "string" is allowed, as long import "socket" is not.

> My offhand *opinion* is that it's darned hard to see any way in
> which it could hurt at all: there's nothing you can do with nested
> functions that you couldn't do before with classes and instance
> data.

Exactly. Furthermore, with the current code, rexec operates on a
per-module basis: r_exec, r_execfile etc always take an entire module
code, and execute it restrictedly (which sometimes means that they
create a new __main__ to do so). So there never is a need to prevent
restricted code from accessing its global variables.

I believe there isn't a way to prevent restricted code from accessing
a specific __builtin__, either - instead, you have to provide
alternative or missing versions of the builtins in advance.

Regards,
Martin


From tim.one at home.com  Sun Feb  4 05:20:02 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 3 Feb 2001 23:20:02 -0500
Subject: Will nested scoping break restricted execution ??
In-Reply-To: <p6qelxgroef.fsf@informatik.hu-berlin.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEFNINAA.tim.one@home.com>

[Martin von Loewis]
> ...
> I believe there isn't a way to prevent restricted code from accessing
> a specific __builtin__, either - instead, you have to provide
> alternative or missing versions of the builtins in advance.

RExec.nok_builtin_names is a tuple of the builtin names to be rendered
invisible.  A subclass can define its own tuple of forbidden builtins.
Attempts to reference one of these names in restricted code just raises a
NameError, because it's plain missing:

>>> from rexec import RExec
>>> class NoIntRExec(RExec):
...     nok_builtin_names = RExec.nok_builtin_names + ("int",)
...
>>> def int_abuser():
...     print int(44.4)
...
>>> RExec().r_exec(int_abuser.func_code)  # fine by regular RExec
44
>>> NoIntRExec().r_exec(int_abuser.func_code) # but not mine
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  ... [lotsa rexec stuff] ...
  File "c:\code\python\dist\src\lib\ihooks.py", line 326, in load_module
    exec code in m.__dict__
  File "c:\code\python\dist\src\lib\string.py", line 189, in ?
    _int = int
NameError: name 'int' is not defined
>>>

simple-and-effective-is-a-delightful-combo-ly y'rs  - tim




From baasad at qualitynet.net  Sat Feb  3 01:39:56 2001
From: baasad at qualitynet.net (bashar A. Asad)
Date: Sat, 03 Feb 2001 03:39:56 +0300
Subject: help with python
Message-ID: <95g8mv$ne21@news.qualitynet.net>

hello ;
since am new to python I would like to start my experience with a small
project that will my life easier.  I would like to connect to a remote linux
machines and check for a certain process "may be useing a system command is
their something like that in python??" eg. httpd if its up or not if it is
up the program will produce an activity file or whatever... I have no idea
how to start working on that in python..any hints guys :)
thanx

bashar



From kwasi007uk at yahoo.co.uk  Sat Feb  3 01:39:57 2001
From: kwasi007uk at yahoo.co.uk (kwasi007uk at yahoo.co.uk)
Date: Sat, 03 Feb 2001 00:39:57 -0000
Subject: Wish there was an error message
Message-ID: <95fk0t+uu67@eGroups.com>

I was using poplib's example as in the documentation. Unfortunately I 
typed in the following at the end of the script:

M.quit instead of M.quit()!

I forgot the parenthesis. The effect was, that my pop-server produced 
really strange things, doubled the number of messages each time I 
queried it with poplib.

But M.quit does not exist. So it should give me an error message? Why 
doesn't it? Can I turn it on somehow?




From bbollenbach at homenospam.com  Sat Feb  3 03:53:49 2001
From: bbollenbach at homenospam.com (Brad Bollenbach)
Date: Sat, 03 Feb 2001 02:53:49 GMT
Subject: Wish there was an error message
References: <mailman.981165805.9847.python-list@python.org>
Message-ID: <1pKe6.8850$wI1.683172@news1.rdc1.mb.home.com>

Actually, if you meant to access M.quit() I'll assume quit() is a valid
method of object M. That said, M.quit is a _reference_ to M.quit(), and in
fact does exist.

For examples of using references to functions look at the documentation on
custom sorting a list, and you'll see that if you define a function called
funcname() you also have reference to that function with funcname.

<kwasi007uk at yahoo.co.uk> wrote in message
news:mailman.981165805.9847.python-list at python.org...
> I was using poplib's example as in the documentation. Unfortunately I
> typed in the following at the end of the script:
>
> M.quit instead of M.quit()!
>
> I forgot the parenthesis. The effect was, that my pop-server produced
> really strange things, doubled the number of messages each time I
> queried it with poplib.
>
> But M.quit does not exist. So it should give me an error message? Why
> doesn't it? Can I turn it on somehow?
>
>




From alain at onesite.org  Sun Feb 18 22:55:22 2001
From: alain at onesite.org (Alain TESIO)
Date: Sun, 18 Feb 2001 22:55:22 +0100
Subject: Wish there was an error message
References: <mailman.981165805.9847.python-list@python.org>
Message-ID: <mvg09tse3j87b1tkk3u74fttbek7kabu5h@4ax.com>

On Sat, 03 Feb 2001 00:39:57 -0000, kwasi007uk at yahoo.co.uk wrote:

>I was using poplib's example as in the documentation. Unfortunately I 
>typed in the following at the end of the script:
>
>M.quit instead of M.quit()!
>
>I forgot the parenthesis. The effect was, that my pop-server produced 
>really strange things, doubled the number of messages each time I 
>queried it with poplib.
>
>But M.quit does not exist. So it should give me an error message? Why 
>doesn't it? Can I turn it on somehow?
>

You get the function object, like for this :

x=1
x

This syntax is useful for interactive use if you want to know what is in an object, but
for non-interactive scripts I don't see what is the interest, maybe an option to turn
raise an error would help.

Alain



From thomas at xs4all.net  Sun Feb  4 03:05:04 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 4 Feb 2001 03:05:04 +0100
Subject: Wish there was an error message
In-Reply-To: <95fk0t+uu67@eGroups.com>; from kwasi007uk@yahoo.co.uk on Sat, Feb 03, 2001 at 12:39:57AM -0000
References: <95fk0t+uu67@eGroups.com>
Message-ID: <20010204030503.B962@xs4all.nl>

On Sat, Feb 03, 2001 at 12:39:57AM -0000, kwasi007uk at yahoo.co.uk wrote:

> But M.quit does not exist. So it should give me an error message? Why 
> doesn't it? Can I turn it on somehow?

M.quit does exist. It's a function or method object. M.quit is simply an
attribute access, which yields the object that is 'stored' in the 'quit'
attribute of 'M'. If you then 'call' that object, using ()'s, you execute
the code. This would do exactly the same:

  q = M.quit
  q()

As would

  getattr(M, "quit")()

There is nothing special about functions or methods, they are just objects
like everything else. In fact, you can create your own class that behaves
just like a function, by supplying a __call__ method.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From trace at reinventnow.com  Sat Feb  3 06:25:40 2001
From: trace at reinventnow.com (trace s. ruggles)
Date: Fri, 02 Feb 2001 23:25:40 -0600
Subject: MySQLdb on a VPS...
Message-ID: <B6A0F273.81D%trace@reinventnow.com>

I'm at a total loss here and desperately need help!!! I'm trying to install
the MySQLdb module but this is what I get... (running FreeBSD4.2 on a VPS,
note my "/usr/home/axiomfir/" prefix to all folder listings)

..

axiomfire% ./python Lib/site-packages/MySQLdb/setup.py build
running build
running build_py
warning: build_py: file MySQLdb.py (for module MySQLdb) not found
warning: build_py: file CompatMysqldb.py (for module CompatMysqldb) not
found
warning: build_py: file MySQLdb.py (for module MySQLdb) not found
warning: build_py: file CompatMysqldb.py (for module CompatMysqldb) not
found
running build_ext
Traceback (most recent call last):
  File "Lib/site-packages/MySQLdb/setup.py", line 90, in ?
    extra_objects=extra_objects,
  File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/core.py", line
138, in setup
  File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/dist.py", line
829, in run_commands
  File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/dist.py", line
849, in run_command
  File 
"/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/command/build.py",
line 106, in run
  File "/usr/home/axiomfir/usr/local/lib/python2.0/Lib/cmd.py", line 328, in
run_command
    print "\n"
  File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/dist.py", line
849, in run_command
  File 
"/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/command/build_ext.py",
line 200, in run
  File "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/sysconfig.py",
line 107, in customize_compiler
  File "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/sysconfig.py",
line 369, in get_config_vars
  File "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/sysconfig.py",
line 281, in _init_posix
distutils.errors.DistutilsPlatformError: invalid Python installation: unable
to open /usr/home/axiomfir/usr/local/lib/python2.0/config/Makefile (No such
file or directory)



From chris at voodooland.net  Sat Feb  3 06:49:48 2001
From: chris at voodooland.net (Chris Watson)
Date: Fri, 2 Feb 2001 23:49:48 -0600 (CST)
Subject: MySQLdb on a VPS...
In-Reply-To: <B6A0F273.81D%trace@reinventnow.com>
Message-ID: <Pine.BSF.4.21.0102022347380.18732-100000@open-systems.net>

> I'm at a total loss here and desperately need help!!! I'm trying to install
> the MySQLdb module but this is what I get... (running FreeBSD4.2 on a VPS,
> note my "/usr/home/axiomfir/" prefix to all folder listings)

Trace,

	If your on FreeBSD why dont you just use the py-MySQLdb port?

cd /usr/ports/databases/py-MySQLdb
make install
make clean

	And thats it. It's already been ported. Is there a reason the port
wont work in your scenario?

--
=============================================================================
-Chris Watson         (316) 326-3862 | FreeBSD Consultant, FreeBSD Geek 
Work:              scanner at jurai.net | Open Systems Inc., Wellington, Kansas
Home:  scanner at deceptively.shady.org | http://open-systems.net
=============================================================================
WINDOWS: "Where do you want to go today?"
LINUX: "Where do you want to go tomorrow?"
BSD: "Are you guys coming or what?"
=============================================================================
GNU = Gnu's Non-portable & Unstable
irc.openprojects.net #FreeBSD -Join the revolution!
ICQ: 20016186




From gerhard.nospam at bigfoot.de  Sat Feb  3 07:36:43 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 03 Feb 2001 07:36:43 +0100
Subject: MySQLdb on a VPS...
References: <B6A0F273.81D%trace@reinventnow.com>
Message-ID: <3A7BA6FB.46F33B56@bigfoot.de>

"trace s. ruggles" wrote:
> 
> I'm at a total loss here and desperately need help!!! I'm trying to install
> the MySQLdb module but this is what I get... (running FreeBSD4.2 on a VPS,
> note my "/usr/home/axiomfir/" prefix to all folder listings)
> 
> ..
> 
> axiomfire% ./python Lib/site-packages/MySQLdb/setup.py build
> running build
> running build_py
> warning: build_py: file MySQLdb.py (for module MySQLdb) not found
> warning: build_py: file CompatMysqldb.py (for module CompatMysqldb) not
> found
> warning: build_py: file MySQLdb.py (for module MySQLdb) not found
> warning: build_py: file CompatMysqldb.py (for module CompatMysqldb) not
> found
> running build_ext
> Traceback (most recent call last):
>   File "Lib/site-packages/MySQLdb/setup.py", line 90, in ?
>     extra_objects=extra_objects,
>   File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/core.py", line
> 138, in setup
>   File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/dist.py", line
> 829, in run_commands
>   File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/dist.py", line
> 849, in run_command
>   File
> "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/command/build.py",
> line 106, in run
>   File "/usr/home/axiomfir/usr/local/lib/python2.0/Lib/cmd.py", line 328, in
> run_command
>     print "\n"
>   File "/usr/home/axiomfir/usr/local/python2/Lib/distutils/dist.py", line
> 849, in run_command
>   File
> "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/command/build_ext.py",
> line 200, in run
>   File "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/sysconfig.py",
> line 107, in customize_compiler
>   File "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/sysconfig.py",
> line 369, in get_config_vars
>   File "/usr/home/axiomfir/usr/local/python2.0/Lib/distutils/sysconfig.py",
> line 281, in _init_posix
> distutils.errors.DistutilsPlatformError: invalid Python installation: unable
> to open /usr/home/axiomfir/usr/local/lib/python2.0/config/Makefile (No such
> file or directory)

This sounds as if your Python installation were screwed up in the first
place. Does basic Python work? You should check first if simple things
work, like importing modules and such:

import socket, sys, os
print socket.gethostname(), 'running', sys.platform
print os.system('uname -a')

If Python is ok, the MySQLdb module is usually built by just untarring
it somewhere (*not* in the lib/site-packages of your Python
installation). Somewhere like /tmp. Then check if the paths to MySQL are
ok in the setup.py file. Then issuing python setup.py build. Then python
setup.py install.

Gerhard

PS: Why is the path like this:
/usr/home/axiomfir/usr/local/python2.0/Lib/...

Why is the Lib with an uppercase L? This is very strange.
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From thelazydogsback at my-deja.com  Sat Feb  3 07:30:05 2001
From: thelazydogsback at my-deja.com (thelazydogsback at my-deja.com)
Date: Sat, 03 Feb 2001 06:30:05 GMT
Subject: Newest release issues (was: Nested Scopes)
Message-ID: <95g8h9$ghg$1@nnrp1.deja.com>

i'm sure I'm not understanding things fully and also don't want to
start a religous war... (ok, what the heck...)

I've seen Py described as both a "scripting language" and "programming
language" - I think one could equally read the universe or nothing into
these adjectives, but one qualification that comes to mind that may
grant a language status in the latter camp is run-time efficiency. In
evaluating Py I was surprised that apparently the compiler creates code
such that in "f(x)" the binding for "f" is looked up in a dictionary
every time "f" is executed. (E.g, even in "for...for...f(x)") This
seems very slow (as was shown in an article showing how to optimize
Python) and allows the (dubious, IHMO) flexibility of allowing "f" to
be re-bound at will, even given the "static" nature of the syntax. I
would think that "f(x)" (calling a function that you assume to be bound
to a single fn object for the duration) is such a common metaphor that
the defualt would be to do all one could to optimize for this case. If
the user really wants the flexibily of "f" to be re-bound, (s)he should
be forced to lookup the reference and call apply() (or similar)
explicitly - this usage is also far more descriptive of the actual
effect. Anyway... I was wondering if the new "nested scope" (2 pass?)
compiler did anything differently as far as late vs. earlier binding of
the fn obj. If I'm misunderstanding the way things work, sorry...

On a related note, on "weak references" w.r.t. gc(), I thought cycles
have been collected correctly for a few versions now?

thanks,
mike





Sent via Deja.com
http://www.deja.com/


From loewis at informatik.hu-berlin.de  Sat Feb  3 12:54:50 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 03 Feb 2001 12:54:50 +0100
Subject: Newest release issues (was: Nested Scopes)
References: <95g8h9$ghg$1@nnrp1.deja.com>
Message-ID: <p6qbsskrng5.fsf@informatik.hu-berlin.de>

thelazydogsback at my-deja.com writes:

> I've seen Py described as both a "scripting language" and
> "programming language" - I think one could equally read the universe
> or nothing into these adjectives, but one qualification that comes
> to mind that may grant a language status in the latter camp is
> run-time efficiency.

Yes, certainly. However, run-time efficiency does not mean "as fast as
possible".

> In evaluating Py I was surprised that apparently the compiler
> creates code such that in "f(x)" the binding for "f" is looked up in
> a dictionary every time "f" is executed. (E.g, even in
> "for...for...f(x)") This seems very slow

That conclusion is incorrect - dictionary lookups are very fast in
Python. By the same token, you might argue that virtual functions in
C++ and Java are very slow, because an access to the virtual function
table is necessary.

> and allows the (dubious, IHMO) flexibility of allowing "f" to be
> re-bound at will, even given the "static" nature of the syntax. I
> would think that "f(x)" (calling a function that you assume to be
> bound to a single fn object for the duration) is such a common
> metaphor that the defualt would be to do all one could to optimize
> for this case.

Be assured - all that can be done has been :-)

> If the user really wants the flexibily of "f" to be re-bound, (s)he
> should be forced to lookup the reference and call apply() (or
> similar) explicitly - this usage is also far more descriptive of the
> actual effect.

That would be a change in the language. Another aspect to qualify as a
programming language is that programs that use it should not break
with new releases - atleast the majority should not. Your proposed
change likely breaks more programs than acceptable.

> Anyway... I was wondering if the new "nested scope" (2 pass?)
> compiler did anything differently as far as late vs. earlier binding
> of the fn obj.

No, it still works the same way: accessing a global means a dictionary
lookup.

> On a related note, on "weak references" w.r.t. gc(), I thought
> cycles have been collected correctly for a few versions now?

Sure, since 2.0. However, there are cases where objects are still kept
alive by the application when they really should go away. The typical
case is a cache: to speed-up access to some object, you cache it in a
dictionary. That makes the cached object life, even if the cache holds
the last reference. With weak references, the object will go away if
the only reference to it is from the cache.

In addition, if you produce cycles, the objects participating in the
cycle will go away only when the collector runs. If you want to clear
them as soon as the last "external" reference to the structure is
released, you should use weak references. E.g. in a tree, you could
make parent link weak. That way, the tree will go away once the
reference to the root is released, as the tree does not contain cycles
anymore.

Regards,
Martin


From thelazydogsback at my-deja.com  Tue Feb  6 05:56:15 2001
From: thelazydogsback at my-deja.com (thelazydogsback at my-deja.com)
Date: Tue, 06 Feb 2001 04:56:15 GMT
Subject: Newest release issues (was: Nested Scopes)
References: <95g8h9$ghg$1@nnrp1.deja.com> <p6qbsskrng5.fsf@informatik.hu-berlin.de>
Message-ID: <95o05d$9oo$1@nnrp1.deja.com>

In article <p6qbsskrng5.fsf at informatik.hu-berlin.de>,
  Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
> thelazydogsback at my-deja.com writes:
>
> > I've seen Py described as both a "scripting language" and
> > "programming language" - I think one could equally read the universe
> > or nothing into these adjectives, but one qualification that comes
> > to mind that may grant a language status in the latter camp is
> > run-time efficiency.
>
> Yes, certainly. However, run-time efficiency does not mean "as fast as
> possible".

Agreed. (Notice I haven't asked about producing native code...)

>
> > In evaluating Py I was surprised that apparently the compiler
> > creates code such that in "f(x)" the binding for "f" is looked up in
> > a dictionary every time "f" is executed. (E.g, even in
> > "for...for...f(x)") This seems very slow
>
> That conclusion is incorrect - dictionary lookups are very fast in
> Python. By the same token, you might argue that virtual functions in
> C++ and Java are very slow, because an access to the virtual function
> table is necessary.

Actually, I wouldn't argue that. :) As I  understand it, a C++ compiler
analyses namespaces and the class hierarchy at compile-time, so a
virtual method call is a simple indirection with a known offset into
the vtable. (I.e, virtual foo(x,y) is always ((obj->vtable)[36])(x,y) )
A dynamic dictionary lookup (string hash + dealing w/collisions, or the
dict algol of your choice) will be a lot slower. (Again, if I'm wrong
here, I'd like to know.) (I guess if you're using an interned symbol
rather than a plain string, the hash fn would not be necessary, but
you'd still have to do the lookup.)

> > and allows the (dubious, IHMO) flexibility of allowing "f" to be
> > re-bound at will, even given the "static" nature of the syntax. I
> > would think that "f(x)" (calling a function that you assume to be
> > bound to a single fn object for the duration) is such a common
> > metaphor that the defualt would be to do all one could to optimize
> > for this case.
>
> Be assured - all that can be done has been :-)

I don't necessarily beleive that - but then again, I don't expect it
either. Certainly a lot of rocket science (at compile-time and run-
time) could produce global and local optimizations that have not been
applied yet - but this stuff is really hard to do and may not be worth
it.

> > If the user really wants the flexibily of "f" to be re-bound, (s)he
> > should be forced to lookup the reference and call apply() (or
> > similar) explicitly - this usage is also far more descriptive of the
> > actual effect.
>
> That would be a change in the language. Another aspect to qualify as a
> programming language is that programs that use it should not break
> with new releases - atleast the majority should not. Your proposed
> change likely breaks more programs than acceptable.

Really? I'd expect that most programs (and programmers) expect names
representing functions in a particular scope to remain bound to the
same fn obj for the duration of the scope's activation. Is there a
prevailing Python coding paradigm that goes against this princpal?

> > Anyway... I was wondering if the new "nested scope" (2 pass?)
> > compiler did anything differently as far as late vs. earlier binding
> > of the fn obj.
>
> No, it still works the same way: accessing a global means a dictionary
> lookup.

Hmm... Maybe that's a good reason to use functions defined in a nested
scope when possible - if they are treated as (faster) locals. (Then
again, I think that the new scoping w/o decls is potentially confusing -
 I'd
love to see optional decl's in Python. (Whether they're typed is
another question.) The rule could be as simple as
(?): if you want simple module/fn level scoping, do what you do now,
and if you want to use nested lexical scoping, use decls.)

> > On a related note, on "weak references" w.r.t. gc(), I thought
> > cycles have been collected correctly for a few versions now?
>
> Sure, since 2.0. However, there are cases where objects are still kept
> alive by the application when they really should go away. The typical
> case is a cache: to speed-up access to some object, you cache it in a
> dictionary. That makes the cached object life, even if the cache holds
> the last reference. With weak references, the object will go away if
> the only reference to it is from the cache.
>
> In addition, if you produce cycles, the objects participating in the
> cycle will go away only when the collector runs. If you want to clear
> them as soon as the last "external" reference to the structure is
> released, you should use weak references. E.g. in a tree, you could
> make parent link weak. That way, the tree will go away once the
> reference to the root is released, as the tree does not contain cycles
> anymore.

Thanks - good explanation. I figured most of that out, but the article
I read mislead me into beleiving that cycles would *never* get
collected w/o the weak refs.

thanks for the reply,
mike

ps -
One interesting idea (?) would be for a dev tool to observe scoping
rules
during coding and add "shadow decls" in the form of comments at the
level appropiate, so the use of a var that creates new scope would
create/update the comments in real-time.

> Regards,
> Martin
>


Sent via Deja.com
http://www.deja.com/


From fredrik at effbot.org  Sat Feb  3 16:45:54 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 15:45:54 GMT
Subject: ANN: prebuilt Tkinter driver for PIL 1.1.1 / Python2.0
Message-ID: <SIVe6.11290$AH6.1826366@newsc.telia.net>

I just posted a prebuilt Tkinter driver (_imagingtk.dll) for Python
2.0 and Tk 8.3.2 on Windows.  Get your copy from here:

    http://effbot.org/pil

This driver adds PIL PhotoImage support to stock Tkinter builds,
including the BeOpen 2.0 Windows distribution.

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From joconnor at cybermesa.com  Sat Feb  3 17:48:11 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 03 Feb 2001 16:48:11 +0000
Subject: Python and GTK
Message-ID: <3A7C364B.695D274A@cybermesa.com>

I tried using Python with GTK and ran into some problems. It seems I
didn't have GTK support installed. So I downloaded it and attempted to
compile it, but I was missing a lot of the GTK dev header files needed
to compile. So I got them all installed, I think. The Python/GTK stuff
at least compiles.

My first script has the following two lines

from gtk import *
from GTK import *

and on the first line I get...

                 
[jay at altaica gtk_test]$ ./gtktest.py 
Traceback (most recent call last):
  File "./gtktest.py", line 3, in ?
    from gtk import *
File "/usr/local/lib/python2.0/site-packages/gtk.py", line 21, in ?
    import _gtk
ImportError: /usr/local/lib/python2.0/site-packages/_gtkmodule.so:
undefined symbol: GTK_TYPE_GDK_COLOR
[jay at altaica gtk_test]$ 

Anyone else run into anything like this or have success or failure
running GTK with Python?

Thanks

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From dsh8290 at rit.edu  Sat Feb  3 18:34:46 2001
From: dsh8290 at rit.edu (D-Man)
Date: Sat, 3 Feb 2001 12:34:46 -0500
Subject: Python and GTK
In-Reply-To: <3A7C364B.695D274A@cybermesa.com>; from joconnor@cybermesa.com on Sat, Feb 03, 2001 at 04:48:11PM +0000
References: <3A7C364B.695D274A@cybermesa.com>
Message-ID: <20010203123446.A18900@harmony.cs.rit.edu>

On Sat, Feb 03, 2001 at 04:48:11PM +0000, Jay O'Connor wrote:
[snip]
| ImportError: /usr/local/lib/python2.0/site-packages/_gtkmodule.so:
| undefined symbol: GTK_TYPE_GDK_COLOR
| [jay at altaica gtk_test]$ 
| 
| Anyone else run into anything like this or have success or failure
| running GTK with Python?

I haven't had problems on my system.  It looks like _gtkmodule.so (the
C shared library part of the GTK bindins) makes a reference to the
symbol GTK_TYPE_GDK_COLOR.  Do you have GTK installed on the system?

I think this is a problem with the C libraries and ld's config on your
system.

HTH,
-D




From joconnor at cybermesa.com  Sat Feb  3 18:57:47 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 03 Feb 2001 17:57:47 +0000
Subject: Python and GTK
References: <3A7C364B.695D274A@cybermesa.com> <mailman.981221704.28538.python-list@python.org>
Message-ID: <3A7C469B.11E7F1C5@cybermesa.com>

D-Man wrote:
> Do you have GTK installed on the system?

Yes, I was able to compile, link and execute a simple C program using
GTK


-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From joconnor at cybermesa.com  Sat Feb  3 20:09:14 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 03 Feb 2001 19:09:14 +0000
Subject: Python and GTK
References: <3A7C364B.695D274A@cybermesa.com> <mailman.981221704.28538.python-list@python.org>
Message-ID: <3A7C575A.7ADEF12D@cybermesa.com>

D-Man wrote:

> 
> I haven't had problems on my system.  It looks like _gtkmodule.so (the
> C shared library part of the GTK bindins) makes a reference to the
> symbol GTK_TYPE_GDK_COLOR.  Do you have GTK installed on the system?
> 
> I think this is a problem with the C libraries and ld's config on your
> system.


I figured out the problem.  I had tried to run configure and make and
make install before all the base gtk stuff was installed and something
go built improperly that didn't get rebuilt properly later.

I just deleted the whole pygtk installed directory and untarred it again
and re rean the install and now it works fine

Thanks

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From morten at esol.no  Sat Feb  3 19:19:48 2001
From: morten at esol.no (Morten W. Petersen)
Date: Sat, 3 Feb 2001 19:19:48 +0100
Subject: Creating a safe counter with Python for Zope
Message-ID: <Pine.LNX.4.21.0102031858450.7792-100000@bcryachts.atsat.com>

Hi guys,

I've been contemplating to create a safe counter for Zope
using Python.  I've tried using the FSCounter product
(available from zope.org), but that occassionally
barfs up an Exception.

Now, the problem is that I don't really know too much
about threads and semaphores (which I'm assuming is
the best/easiest way to do it), much less how to make
this work safely with Zope (the ZODB).

This counter is going to be used as a system wide
unique identifier, and it therefore needs to be
completely safe and also persistent.

I've built a simple counter without persistence that
seems to work:

"""
import thread
import threading
import time

class test_counter:

        def __init__(self, start=0):

                self.counter = start
                self.lock = thread.allocate_lock()

        def __call__(self):

                self.lock.acquire()

                counter = self.counter = self.counter + 1

                self.lock.release()

                return counter

get_unique_id = test_counter()
id_list = []

def crack_it():

        while 1:

                id_list.append(get_unique_id())
                time.sleep(0.01)


threading._start_new_thread(crack_it, ())
threading._start_new_thread(crack_it, ())
threading._start_new_thread(crack_it, ())
threading._start_new_thread(crack_it, ())
"""

And I've run tests on this that append around 100000
integers to the list, all unique.

Now, with this working, can I be sure that if it
subclasses the persistence machinery that it will
always be unique?  And that multiple threads or
whatever can't modify it at the same time and
screw up the uniqueness of the value?..

Thank you for your time.

-Morten



From keisari_ at hotmail.com  Sat Feb  3 20:00:32 2001
From: keisari_ at hotmail.com (Joonas)
Date: Sat, 03 Feb 2001 21:00:32 +0200
Subject: How can I redirect stderr into stdout.
Message-ID: <3A7C554F.60D5556D@hotmail.com>

How can I redirect stderr to stdout.

Joonas


From chris.gonnerman at usa.net  Sun Feb  4 01:51:47 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 3 Feb 2001 18:51:47 -0600
Subject: How can I redirect stderr into stdout.
References: <3A7C554F.60D5556D@hotmail.com>
Message-ID: <003501c08e44$abc35740$a100000a@local>

----- Original Message ----- 
From: "Joonas" <keisari_ at hotmail.com>
Subject: How can I redirect stderr into stdout.


> How can I redirect stderr to stdout.

Question is too general.  Python?

    import sys
    sys.stderr = sys.stdout

Unix/Linux?

    myscript.py 2>&1

Windows 9x?  You need a stderr redirector program.  Visit:

    http://www.teaser.fr/~amajorel/stderr/

Windows NT/2000?  Unix method may work.







From chris at cogdon.org  Sat Feb  3 20:08:15 2001
From: chris at cogdon.org (Chris Cogdon)
Date: Sat, 3 Feb 2001 11:08:15 -0800 (PST)
Subject: Help detecting partial upload
Message-ID: <Pine.LNX.4.30.0102031105470.15856-100000@onca.catsden.net>

Hi folks. Hope someone can help me out with this (quick?) question.

I need to be able to detect when a file upload through a web form fails
due to a timeout, or other partial transfer.

I've hunted through the cgi.py module, and I cant see any facility to
represent a incomplete transfer. I know exceptions certainly aren't being
raised

If people want to send private replies, I'll post a summary back to the
list. Thanks in advance.


   ("`-/")_.-'"``-._        Chris Cogdon <chris at cogdon.org>
    . . `; -._    )-;-,_`)
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'
((,.-'   ((,/   fL




From pearu at cens.ioc.ee  Sat Feb  3 20:31:01 2001
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sat, 3 Feb 2001 21:31:01 +0200
Subject: NameError in cross imports of modules
Message-ID: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee>

Hi!

I have two classes defined in separate files and both classes use each
other:

File A.py:
  from B import *
  class A:
      def asB(self): return B()

File B.py:
  from A import *
  class B:
      def asA(self): return A()

In Python session I get NameError:

>>> from A import *
>>> from B import *
>>> B().asA()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "B.py", line 5, in asA
    return A()
NameError: global name 'A' is not defined

but

>>> A().asB()
<B.B instance at 0x810ea4c>


Note: I am not interested in converting the 'from-import' statements to
'import' statements, rather I have collected the classes A and B into the
same file but it's really inconvenient to maintain a large package in one
file :(


This behavior is well-known and has, roughly speaking, the following
explanation:
>>> from A import *    --- imports also module B but only "partly": module
                           A is not imported to module B.
>>> from B import *    --- nothing is done because module B is already in
                           sys.modules.

>>> reload(sys.modules['B'])  --- is needed in order to fix the
                           NameError: module B will be imported
                           "completely".

My question is: 
	could this reload be done automatically? 

Say, when module B is "partly" imported during `from A import *', then a
flag is set. This flag is used when one imports module B: instead of
doing nothing when it is already in sys.modules, module B will be reloaded
and the flag will be unset.

What do you think? Is is possible to get rid of this asymmetric behavior
of cross imports in future Python releases?

Pearu





From fredrik at effbot.org  Sat Feb  3 21:14:21 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 20:14:21 GMT
Subject: NameError in cross imports of modules
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee>
Message-ID: <xEZe6.11365$AH6.1836591@newsc.telia.net>

Pearu Peterson wrote:
> Note: I am not interested in converting the 'from-import' statements to
> 'import' statements,

> This behavior is well-known and has, roughly speaking, the following
> explanation:
> >>> from A import *    --- imports also module B but only "partly": module
>                            A is not imported to module B.
> >>> from B import *    --- nothing is done because module B is already in
>                            sys.modules.
> >>> reload(sys.modules['B'])  --- is needed in order to fix the
>                            NameError: module B will be imported
>                            "completely".

if you don't understand how things work today, are you
really the right person to suggest changes?

Cheers /F




From pearu at cens.ioc.ee  Sat Feb  3 22:14:49 2001
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sat, 3 Feb 2001 23:14:49 +0200
Subject: NameError in cross imports of modules
In-Reply-To: <xEZe6.11365$AH6.1836591@newsc.telia.net>
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee> <xEZe6.11365$AH6.1836591@newsc.telia.net>
Message-ID: <Pine.LNX.4.21.0102032251480.9712-100000@kev.ioc.ee>


On Sat, 3 Feb 2001, Fredrik Lundh wrote:

> Pearu Peterson wrote:
> > Note: I am not interested in converting the 'from-import' statements to
> > 'import' statements,
> 
> > This behavior is well-known and has, roughly speaking, the following
> > explanation:
> > >>> from A import *    --- imports also module B but only "partly": module
> >                            A is not imported to module B.
> > >>> from B import *    --- nothing is done because module B is already in
> >                            sys.modules.
> > >>> reload(sys.modules['B'])  --- is needed in order to fix the
> >                            NameError: module B will be imported
> >                            "completely".
> 
> if you don't understand how things work today, are you
> really the right person to suggest changes?

Probably not. I don't even understand what you are referring to when
asking such a question. Shouldn't I use 'from-import' statements? Is my
**rough** explanation completely wrong and misunderstood? (Sure, I didn't
bother to  check out that I was using proper terminology, my apologies for
that) Or what?

a-python-user-who-tends-to-abuse-it-because-of-lack-of-style'ly
Pearu



From thomas at xs4all.net  Sun Feb  4 03:31:11 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 4 Feb 2001 03:31:11 +0100
Subject: NameError in cross imports of modules
In-Reply-To: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee>; from pearu@cens.ioc.ee on Sat, Feb 03, 2001 at 09:31:01PM +0200
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee>
Message-ID: <20010204033111.C962@xs4all.nl>

On Sat, Feb 03, 2001 at 09:31:01PM +0200, Pearu Peterson wrote:

> I have two classes defined in separate files and both classes use each
> other:

> File A.py:
>   from B import *
>   class A:
>       def asB(self): return B()

> File B.py:
>   from A import *
>   class B:
>       def asA(self): return A()
> 
> In Python session I get NameError:

> >>> from A import *
> >>> from B import *
> >>> B().asA()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "B.py", line 5, in asA
>     return A()
> NameError: global name 'A' is not defined

> but

> >>> A().asB()
> <B.B instance at 0x810ea4c>

> Note: I am not interested in converting the 'from-import' statements to
> 'import' statements, rather I have collected the classes A and B into the
> same file but it's really inconvenient to maintain a large package in one
> file :(

You really should be interested in not using from-import; it solves your
problem. It works like this:

You execute 'from A import *':
  'A' has not been loaded yet; A.py is found and loaded:
  A new module object is created, and stored in sys.modules['A']
  A.py is executed, in the namespace of the newly created module:
    On the first line, the interpreter finds 'from B import *'.
    'B' has not been loaded yet; B.py is found and loaded:
      A new module object is created and stored in sys.modules['B']
      B.py is executed, in the namespace of the newly created module.
      On the first line, the interpreter finds 'From A import *'.
      'A' has already been loaded (it's present in sys.modules)
      'A' is retrieved from sys.modules, and scanned for names: none found
      B executes the rest of the code, defining class B.
    A retrieves module 'B' from sys.modules, and scans it for names: B found
    A executes the rest of the code, defining class A.

(I know from your posting you know at least part of this, but it's also to
educate others who might not :)

By the time 'B' loads 'A' (too early), there is no class 'A' to from-import,
and because you use 'from A import *', it isn't added later. Where we to
'automatically' reload the 'right' module (magically finding out the right
time and module to reload) we could very easily start looping. In fact, this
entire behaviour is intended to avoid looping. 

> >>> reload(sys.modules['B'])  --- is needed in order to fix the
>                            NameError: module B will be imported
>                            "completely".

This is not true, and implies that you didn't really understand what was
happening. 'B' isn't the one incompletely imported, 'A' was! And because you
use 'from-import-*', which makes *copies* of the variables (which are
themselves references) in a module, reloading 'A' won't help, because 'B'
won't see the new variables.

You have two real ways to fix this: don't use 'from-import *'. Honestly, it is
the best solution. 'from-import *' was not intended for this kind of
hackery, and as far as I'm concerned there are only two viable uses: the way
os.py uses it (import any and all symbols from a platform-dependant C
module) and the way Mailman uses it (import all symbols from a template
config file into the user-editable config file, so it only needs to contain
overriding values.) I still owe Fred a "From module import * considered
harmful" tutorial-section ;-P If prefixing all module variables with the
module name is too painful because the module name is too long or unclear,
consider using 'import my_long_module as _a' or some such.

Alternatively, move the 'from-import *' to the bottom of the file, instead
of the top. That way, 'A' will have defined all the variables before it
starts loading B, and B won't have any problem finding them. Definately the
inferior solution, though ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From joconnor at cybermesa.com  Sun Feb  4 08:39:21 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 07:39:21 +0000
Subject: NameError in cross imports of modules
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee> <mailman.981253934.10580.python-list@python.org>
Message-ID: <3A7D0729.C7BF5EF1@cybermesa.com>

Thomas Wouters wrote:

> and as far as I'm concerned there are only two viable uses: the way
> os.py uses it (import any and all symbols from a platform-dependant C
> module) and the way Mailman uses it (import all symbols from a template
> config file into the user-editable config file, so it only needs to contain
> overriding values.) I still owe Fred a "From module import * considered
> harmful" tutorial-section ;-P If prefixing all module variables with the
> module name is too painful because the module name is too long or unclear,
> consider using 'import my_long_module as _a' or some such.


I've followed the convention (maybe I made it up for myself :)  That if
I'm importing classes, I use "from package import *" such as "from gtk
import *", but if I'm importing a module of functions, I just use
"import package name" such as "import string" and use the module name
prefix for all function calles.

Don't know if it's good or not...


Take care,
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From thomas at xs4all.net  Sun Feb  4 16:54:37 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 4 Feb 2001 16:54:37 +0100
Subject: NameError in cross imports of modules
In-Reply-To: <3A7D0729.C7BF5EF1@cybermesa.com>; from joconnor@cybermesa.com on Sun, Feb 04, 2001 at 07:39:21AM +0000
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee> <mailman.981253934.10580.python-list@python.org> <3A7D0729.C7BF5EF1@cybermesa.com>
Message-ID: <20010204165437.F962@xs4all.nl>

On Sun, Feb 04, 2001 at 07:39:21AM +0000, Jay O'Connor wrote:
> Thomas Wouters wrote:

> > and as far as I'm concerned there are only two viable uses: the way
> > os.py uses it (import any and all symbols from a platform-dependant C
> > module) and the way Mailman uses it (import all symbols from a template
> > config file into the user-editable config file, so it only needs to contain
> > overriding values.) I still owe Fred a "From module import * considered
> > harmful" tutorial-section ;-P If prefixing all module variables with the
> > module name is too painful because the module name is too long or unclear,
> > consider using 'import my_long_module as _a' or some such.

> I've followed the convention (maybe I made it up for myself :)  That if
> I'm importing classes, I use "from package import *" such as "from gtk
> import *", but if I'm importing a module of functions, I just use
> "import package name" such as "import string" and use the module name
> prefix for all function calles.

> Don't know if it's good or not...

Well, I already made clear here that the answer is 'no' :-) I'm not sure
what kind of situation you are in when you say you are 'importing classes'
(I've never done that, myself, and it's only possible since Python 2.1 in
any case :), but 'from module import *' imports a lot more than classes, and
can do all kinds of nasty things, like shadowing builtins and globals, and
clobbering your own variables. If you want to import a couple of class names
from a module to your own namespace, you best name them explicitly, so you
are sure those are the only names imported. And more importantly, so that
you, and everyone else, can always find out where a name comes from. For
instance:

from spamsong import *

<X lines of code>

from Eric import *

<several hundred lines of code>

class ExtendedViking(Viking):
    <lots of code relying on what Viking is>

Where did the 'Viking' come from ? Everyone knows both the spamsong and Eric
the Viking feature Vikings. Maybe spamsong named its viking differently when
you first wrote the module, or you never noticed spamsong had a viking
because it was always overwritten by the one in Eric. Or maybe it's the
other way 'round, you intended to use spamsong's Viking, but didn't know
Eric had one, too. Or maybe Eric grew its viking later. Or even worse, both
spamsong and Eric had vikings, but the on in Eric wasn't supposed to be
exported (like, it should have been named _Viking and someone forgot) and
they removed it in a later release -- but you were relying on that
undocumented misfeature. Suddenly your extended viking behaves differently,
'for no apparent reason'.

Now if you had written 
 
from spamsong import Viking
from Eric import Viking

It would have been obvious from the start that you were overwriting the
first with the second. You wouldn't have run into all those ambiguities
because if either Viking hadn't existed when you first wrote it, you would
have seen the import errors. And if either Viking disappeared in a later
release, this would also cause a nice clean error.

The only real, justifiable (in my eyes) use for 'from ... import *' is when
you have two modules, and you want one to act as a 'front' for the other.
This is what os.py does: it grabs a platform-dependant module and all the
names it defines. (posix on POSIXish systems, nt on windont, mac on
macintoshes, etc. And more can be added without breaking or confusing code.)
os.py and the platform-specifc module are not meant to be used seperately,
and not meant to be of different versions, either. 

The same goes for the Mailman mm_cfg.py and Defaults.py. mm_cfg.py is a
config file (in handy Python format) that is created if it doesn't exist,
but is never overwritten by vesion updates. Instead, Defaults.py is
provided, which contains default values for all configuration directives.
This file should not be edited, because it *is* overwritten by Mailman
version updates. mm_cfg.py does not provide any functionality itself, and
the user is supposed to be able to override values from Defaults.py. The
user should not invent new values or add new code to mm_cfg.py, because that
might mask new configuration directives in future versions, causing strange
breakage. It's purely a 'override the defaults' opportunity.

These two uses are so specific, and the techniques so infrequently used, I
really wish there *was* no 'from-import *'. The above could have been done
in a more elaborate, verbose and less abusable way :-) I doubt 'from-import
*' will be removed even in Py3K, but I can still wish, can't I ? :-)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From joconnor at cybermesa.com  Sun Feb  4 20:55:28 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 19:55:28 +0000
Subject: NameError in cross imports of modules
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee> <mailman.981253934.10580.python-list@python.org> <3A7D0729.C7BF5EF1@cybermesa.com> <mailman.981302118.31502.python-list@python.org>
Message-ID: <3A7DB3B0.9425BF64@cybermesa.com>

Thomas Wouters wrote:

> 
> Well, I already made clear here that the answer is 'no' :-) I'm not sure
> what kind of situation you are in when you say you are 'importing classes'
> (I've never done that, myself, and it's only possible since Python 2.1 in
> any case :),

I was careless in my terminology, I meant importing a module for the
sake of access to the classes the module contains.

> but 'from module import *' imports a lot more than classes, and
> can do all kinds of nasty things, like shadowing builtins and globals, and
> clobbering your own variables. If you want to import a couple of class names
> from a module to your own namespace, you best name them explicitly, so you
> are sure those are the only names imported. And more importantly, so that
> you, and everyone else, can always find out where a name comes from. For
> instance:
> 
> from spamsong import *
> 
> <X lines of code>
> 
> from Eric import *
> 
> <several hundred lines of code>
> 
> class ExtendedViking(Viking):
>     <lots of code relying on what Viking is>
> 
> Where did the 'Viking' come from ? Everyone knows both the spamsong and Eric
> the Viking feature Vikings. Maybe spamsong named its viking differently when
> you first wrote the module, or you never noticed spamsong had a viking
> because it was always overwritten by the one in Eric. Or maybe it's the
> other way 'round, you intended to use spamsong's Viking, but didn't know
> Eric had one, too. Or maybe Eric grew its viking later. Or even worse, both
> spamsong and Eric had vikings, but the on in Eric wasn't supposed to be
> exported (like, it should have been named _Viking and someone forgot) and
> they removed it in a later release -- but you were relying on that
> undocumented misfeature. Suddenly your extended viking behaves differently,
> 'for no apparent reason'.

I haven't run into this problem (yet!) because I tend to prefix my class
names with an identifier that indicates what project, etc...the class is
from.  That is a habit from using pre-namespace Smalltalk where the was
no scoping tool for namespaces so you had to create Class names likely
not to be clobbered

Of course the return argument is that using "moduleName.ClassName" is
the correct scoping tool.

I see your points; I will rethink my approach

Thank you

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From thomas at xs4all.net  Mon Feb  5 00:25:24 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 5 Feb 2001 00:25:24 +0100
Subject: NameError in cross imports of modules
In-Reply-To: <3A7DB3B0.9425BF64@cybermesa.com>; from joconnor@cybermesa.com on Sun, Feb 04, 2001 at 07:55:28PM +0000
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee> <mailman.981253934.10580.python-list@python.org> <3A7D0729.C7BF5EF1@cybermesa.com> <mailman.981302118.31502.python-list@python.org> <3A7DB3B0.9425BF64@cybermesa.com>
Message-ID: <20010205002524.I962@xs4all.nl>

On Sun, Feb 04, 2001 at 07:55:28PM +0000, Jay O'Connor wrote:

> Of course the return argument is that using "moduleName.ClassName" is
> the correct scoping tool.

> I see your points; I will rethink my approach

Note that the whole thing was just my humble opinion. I'm not trying to
force people to follow my guidelines or my ideas, just trying to explain
them in the most elaborate way I can think of, and hoping it will open your
eyes to the foolishness of your ways, like it did with Jay <wink>.

I also have a particular weakness in this area that not everyone shares: I
type too fast for the average keyboard to keep up(*). So I *really* don't
care about the extra typing to get 'Tkinter.' in front of each Tkinter
variable. The reading part is much, much more important to me :)

I-bet-even-Guido-uses-from-import-*-more-often-than-me-:-)-ly y'rs 

*) This is particularly true for cheapo keyboards and laptop-keyboards. They
use a limited number of wires to connect the keys with the keyboard
processor, and some combinations introduce an extra 'ghost' keystroke. For
instance, my own laptop insists 'pyt' is followed by a semi-colon, and 'bsd'
is followed by a 'u', regardless of the fact I follow them with 'hon' and
'i', respectively. If you ever see me write 'pyt;hon', you know I was typing
on my keyboard :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From dboeren at gw.total-web.net  Sat Feb  3 20:48:00 2001
From: dboeren at gw.total-web.net (David Boeren)
Date: Sat, 03 Feb 2001 19:48:00 GMT
Subject: [NEWBIE] Priority Queue in Python
Message-ID: <3a7c5ada.67417497@news.gw.total-web.net>

I'm trying to implement something like a priority queue in Python, but
it's quite slow compared to a similar implementation in Java.  I tried
running the List speed test I found on this web page
	http://www.twistedmatrix.com/~glyph/rant/python-vs-java.html

and the Java version was about 7.7 * faster, although the page author
reported only 2.4 * faster.

Anyway, I've coded the Python version in several different ways and
measured them all.  Here's the code:

def benchmark(func, times=1):
    t0=clock()
    for i in xrange(times): func()
    t1=clock()
    print "Time =", (t1-t0)

def shuffle(list):
    "Randomly shuffles a list"
    if type(list) is IntType:
        list=range(list)
    length=len(list)
    for i in xrange(length):
        pos=randrange(i, length)
        (list[i], list[pos]) = (list[pos], list[i])
    return list

def insert_priority(list, tuple, key_index):
    i=0
    key = tuple[key_index]
    try:
        while key < list[i][key_index]: i=i+1
        list.insert(i, tuple)
    except IndexError:
        list.append(tuple)

def insert_priority2(list, tuple, key_index):
    key = tuple[key_index]
    for i in list:
        if i[key_index] < key:
            list.insert(list.index(i), tuple)
            break
    else:
        list.append(tuple)

def insert_priority3(list, tuple, key_index):
    key = tuple[key_index]
    n=0
    for i in list:
        if i[key_index] < key:
            list.insert(n, tuple)
            break
        n=n+1
    else:
        list.append(tuple)

def insert_priority4(list, item):
    i=0
    try:
        while item < list[i]: i=i+1
        list.insert(i, item)
    except IndexError:
        list.append(item)

def test0():
    "benchmark(test0, 100) =  2.4259 test"
    l=[]
    for i in shuffle(1000):
        l.append((0,i))

def test1():
    "benchmark(test1, 100) = 26.6688 sec"
    l=[]
    for i in shuffle(1000):
        insert_priority(l, (0, i), 1)

def test2():
    "benchmark(test2, 100) = 31.7182 sec"
    l=[]
    for i in shuffle(1000):
        insert_priority2(l, (0, i), 1)

def test3():
    "benchmark(test3, 100) = 30.8639 sec"
    l=[]
    for i in shuffle(1000):
        insert_priority3(l, (0, i), 1)

def test4():
    "benchmark(test4, 100) = 21.6703 sec"
    l=[]
    for i in shuffle(1000):
        insert_priority4(l, i)


As you can see, I've coded the insert_priority() several different
ways.  1-3 each take a tuple or list, and sort based on the indicated
item in the sequence.  In case the extra tuple creation and lookup was
costing too much, I also coded a 4th version to which I just pass an
int to sort on.  Turns out it *IS* significant, but it's not the
biggest cost.  I also timed the creation of my random list to be
inserted (test0), and it doesn't take that much time either.

Out of curiosity, I coded a similar function in Java to test the
speed.  The code is rather ugly, I was just trying to get it working
quickly (still took a lot longer to write than Python, and this is
some of my first Python code!)  The java version (which is
functionally equivalent to test4 in that it uses ints and not tuples
or other more advanced structures) runs in about 1.87 seconds,
which makes it about 11.6 times slower than the Python version.

I'm using Python 2.0 and Java 1.3, both on an Athlon Thunderbird 900
running under Win98SE.

I'm wondering if I'm just missing something here, or should there
really be this big of a difference between the speed of Java and
Python?  I expected maybe around a 4X difference, but 12X is starting
to get uncomfortable.  Am I implementing my algorithm badly?  I freely
admit to being a total beginner in Python, so that may be the case.  I
do like the way Python takes less time and fewer lines to write, and
the code is more general.  I've got a 16-line version of
insert_priority that I'm working on that handles keys by either direct
use, sequence-position, attribute-name, or a compare-function!

Also, one specific question.  Maybe it's because I don't know the
idioms yet, but I keep wanting to have access to a loop counter
variable in my for loops.  For example:

for i in list:
    print i, "is the", list.index(i), "element"

The index() call takes too long, I wish there was like a hidden
variable so I could say it like this.  Is there?

for i in list:
    print i, "is the", __loopindex__, "element"

Or is there just a better way to do this that I don't know?

Well, thanks for putting up with the ravings of a mad Python newbie.
I love the language so far, and I'd like to make it my main language
for coding at home, I just have some doubts about the speed that are
making me a little hesitant so far.  Oh, and if anyone knows of a good
tutorial for wxWindows, let me know, ok?  :)

import java.util.*;

class tools {
    Vector v;
    int range[];

    void shuffle(int n) {
        Random randy=new Random();
        int pos, tmp;
        v=new Vector(n);

        range = new int[n];
        for (int i=0; i < n; i++) range[i]=i;
        for (int i=0; i < (n-1); i++) {
            pos=i+randy.nextInt(n-i);
            tmp=range[i];
            range[i]=range[pos];
            range[pos]=tmp;
        }
    }

    void insert_priority(int n) {
        for (int pos=0; pos < v.size(); pos++) {
            if (n > ((Integer)v.elementAt(pos)).intValue()) {
                v.add(pos, new Integer(n));
                return;
            }
        }
        v.add(new Integer(n));
    }

    void test(int n) {
        shuffle(n);
        for (int i=0; i < n; i++) insert_priority(range[i]);
    }

    public static void main(String[] args) {
        tools t=new tools();

        long t0=System.currentTimeMillis();
        for (int i=0; i<100; i++) t.test(1000);
        long t1=System.currentTimeMillis();
        System.out.println("\nTime = " + (t1-t0));
    }
}



From fredrik at effbot.org  Sat Feb  3 21:35:01 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 20:35:01 GMT
Subject: [NEWBIE] Priority Queue in Python
References: <3a7c5ada.67417497@news.gw.total-web.net>
Message-ID: <VXZe6.11367$AH6.1837105@newsc.telia.net>

David Boeren wrote:
> Also, one specific question.  Maybe it's because I don't know the
> idioms yet, but I keep wanting to have access to a loop counter
> variable in my for loops.  For example:
>
> for i in list:
>     print i, "is the", list.index(i), "element"

Ouch.

> The index() call takes too long, I wish there was like a hidden
> variable so I could say it like this.  Is there?
>
> for i in list:
>     print i, "is the", __loopindex__, "element"
>
> Or is there just a better way to do this that I don't know?

There's is an index counter deep inside Python, but you
cannot access it from the "outside" [1].  But there are
more efficient ways to do what you want:

    for i in range(len(list)):
        item = list[i]
        print item, "is the", i, "element"

or

    i = 0
    for item in list:
        print item, "is the", i, "element"
        i += 1

Cheers /F

1) the counter is passed to the list's __getitem__ method.
you can wrap the list in an adapter object whose __getitem__
method returns both the element and the index as a tuple
(but don't do that if you care about performance).




From tjreedy at udel.edu  Sun Feb  4 01:36:36 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 3 Feb 2001 19:36:36 -0500
Subject: [NEWBIE] Priority Queue in Python
References: <3a7c5ada.67417497@news.gw.total-web.net>
Message-ID: <95i80p$f90$1@news.udel.edu>

> for i in list:
>     print i, "is the", list.index(i), "element"
>
> The index() call takes too long, I wish there was like a hidden
> variable so I could say it like this.  Is there?

Yes, tidbit is O(len(list))**2

>
> for i in list:
>     print i, "is the", __loopindex__, "element"

Since list() is a builtin function, best not to use as a variable name.
Common idiom is as follows:

for i in range(len(listx)):
    print listx[i], 'is the', i, 'th element

Exposing the internal loopindex has been discused, but never agreement
reached.




From tim.one at home.com  Sun Feb  4 07:01:57 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 4 Feb 2001 01:01:57 -0500
Subject: [NEWBIE] Priority Queue in Python
In-Reply-To: <3a7c5ada.67417497@news.gw.total-web.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEGAINAA.tim.one@home.com>

[David Boeren, timing priority queues]

Some notes:

Three of your Python priority queue guys have the extra burden (compared to
your Python insert_priority4 and your Java insert_priority) of mucking
around with an extra key_index argument.  Stick to the same thing
throughout.

Effort devoted to timing naive algorithms is better spent dreaming up better
algorithms.  "The usual" naive priority queue in Python uses the std bisect
module directly:

import bisect
bisect.insort(list, item)

Then finding the place to insert is an O(log(len(list))) time operation,
although doing the actual insert remains O(len(list)) time, but where the
latter is at full C speed.  While faster then doing a linear search, if your
priority queues are going to get large, this is a disastrous approach
regardless of language.

Note that Python compares tuples lexicographically (strings and lists too),
so even if the item is a tuple this works fine provided item[0] is the
primary key.

If your benchmarks are typical of your expected access pattern (i.e.,
nothing but inserts -- no lookups), or an approximation to your expected
access pattern (perhaps lots of inserts followed by lots of lookups), it
would be much faster to do the insert via

list.append(item)

and do

list.sort()

later before your first access.  Python's list.sort() realizes if a list is
"sorted at the start, and then has 'some small amount of' disordered gunk at
the end", and in that case does a fast sequence of binary inserts under
covers to clean up the gunk.  Yet this is still a disaster if priority
queues get very large.

A much better way to do priority queues if they can get large is via a heap.
Then insertion, deletion and lookup never take worse than O(log(n)) time.

Finally, note that the std module random has a random.shuffle function.

luckily-the-python-libraries-are-much-smaller-than-java's-so-it's-
    possible-to-learn-all-of-them-in-less-than-a-decade<wink>-ly y'rs
    - tim




From me at nospam.net  Sun Feb 11 10:38:04 2001
From: me at nospam.net (Scottie)
Date: Sun, 11 Feb 2001 01:38:04 -0800
Subject: [NEWBIE] Priority Queue in Python
References: <mailman.981266535.14704.python-list@python.org>
Message-ID: <t8cnbsrbguh77d@corp.supernews.com>

You could look at the pre-release code in PriQueue.sourceforge.net for an
example of the heap version.  I have a few changes yet in mind.  "skim"
rather than "cream" for the faster equivalent of:
    def ?(self,x): self.push(x); return self.pop()
This I only decided on as I started implementing double-ended queues and
determined that skim has dredge as a reasonable "other-ended" verb.  The
last interface decision is in exposure of the contents.  I am leaning
towards a sequence-like interface (so index will work to find elements),
allowing delete of single internal elements (after which everything moves
appropriately), and re-prioritizing by replacing an entry by one with a new
priority (which also has the slightly unsettling effect of moving things
around on you, so that after:
    p,v = pq[N]
    pq[N] = (p+1,v)
you typically have pq[N] with some completely unrelated value, and (p+1,v)
is at some completely different index, M.  Of course very few ranges make
sense.  pq[:] works, as does pq[1:], but I can think of very few others.

I have the code to expand/contract the array at less frequent times, but
have left it out until I'm happy with the interface itself.  Anybody have
any comments on the interface as is?

-Scott David Daniels
Scott.Daniels at Acm.Org




From michael.ellwood at virgin.net  Sat Feb  3 21:30:28 2001
From: michael.ellwood at virgin.net (Michael Ellwood)
Date: Sat, 3 Feb 2001 20:30:28 -0000
Subject: Basic problem - Syntax error
Message-ID: <nOZe6.13296$LQ2.213892@news2-win.server.ntlworld.com>

Im writing a programme to decrypt a casaer cipher. Im using a while loop
with a counter, and when the counter exceeds a certain value , the loop
stops and the alphabet generated is printed eg:

plaintextalphabet="abcdefghijklmnopqrstuvwxyz"
ciphertext="mhily lza zbhl xbpzxbl mvyabuhl hwwpbz jshbkpbz jhljbz kpjabt
hyjhubt lza ulbayvu"

counter=0
alphabetpos=0

while counter < 26:
    cipheralphabet.append(plaintextalphabet[counter:]
**    counter = counter + 1

print cipheralphabet



Python tells me there is a syntax error on the starred line - can some one
please explain why!! Im so annoyed - this is such a basic problem!!

-=+Stephen+=-






From joconnor at cybermesa.com  Sat Feb  3 21:30:01 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sat, 03 Feb 2001 20:30:01 +0000
Subject: Basic problem - Syntax error
References: <nOZe6.13296$LQ2.213892@news2-win.server.ntlworld.com>
Message-ID: <3A7C6A49.A77161BF@cybermesa.com>

> while counter < 26:
>     cipheralphabet.append(plaintextalphabet[counter:]
> **    counter = counter + 1
> 
> print cipheralphabet
> 
> Python tells me there is a syntax error on the starred line - can some one
> please explain why!! Im so annoyed - this is such a basic problem!!


Umm..if this is an accuracte piece of code....yoo forgot to close the
parenthesis on the line before.

It should read

	... [counter:] )   <-close paren

Take care,
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From chris at onca.catsden.net  Sat Feb  3 21:33:34 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Sat, 3 Feb 2001 12:33:34 -0800 (PST)
Subject: Basic problem - Syntax error
In-Reply-To: <nOZe6.13296$LQ2.213892@news2-win.server.ntlworld.com>
Message-ID: <Pine.LNX.4.30.0102031231110.15856-100000@onca.catsden.net>

On Sat, 3 Feb 2001, Michael Ellwood wrote:

> Im writing a programme to decrypt a casaer cipher. Im using a while loop
> with a counter, and when the counter exceeds a certain value , the loop
> stops and the alphabet generated is printed eg:
>
> plaintextalphabet="abcdefghijklmnopqrstuvwxyz"
> ciphertext="mhily lza zbhl xbpzxbl mvyabuhl hwwpbz jshbkpbz jhljbz kpjabt
> hyjhubt lza ulbayvu"
>
> counter=0
> alphabetpos=0
>
> while counter < 26:
>     cipheralphabet.append(plaintextalphabet[counter:]
> **    counter = counter + 1
>
> print cipheralphabet

You're missing a ) on the line beforehand.


   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From fredrik at effbot.org  Sat Feb  3 22:11:05 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 21:11:05 GMT
Subject: Basic problem - Syntax error
References: <nOZe6.13296$LQ2.213892@news2-win.server.ntlworld.com>
Message-ID: <Jt_e6.11373$AH6.1838692@newsc.telia.net>

Michael Ellwood wrote:
> counter=0
> alphabetpos=0
>
> while counter < 26:
>     cipheralphabet.append(plaintextalphabet[counter:]
> **    counter = counter + 1
>
> print cipheralphabet
>
> Python tells me there is a syntax error on the starred line - can some one
> please explain why!! Im so annoyed - this is such a basic problem!!

if you cannot figure out what's wrong with the line Python
complains about, look at the line just above it.

(hint: something's missing at the end of the append line

Cheers /F




From randy at spoke.net  Sat Feb  3 22:02:56 2001
From: randy at spoke.net (Randall Kern)
Date: Sat, 03 Feb 2001 21:02:56 GMT
Subject: linuxnewbie: clean handling of SIGTERM
Message-ID: <4m_e6.678$GV6.260515@dfiatx1-snr1.gtei.net>

This is hopefully a simple "everybody has done that" kind of question, but
my Linux (unix) experience is sort of lacking.

I have a daemon that runs a loop, waiting for work.  When a request arrives,
it processes the request, and goes back into the loop.

Now I want to check for shutdown on each iteration of the loop.  I thought I
could register a handler for SIGTERM, and simply set a variable to tell my
loop to terminate:

_shutdown = 0
def sigterm(sig, frame):
    global _shutdown
    _shutdown = 1

signal.signal(signal.SIGTERM, sigterm)
while not _shutdown:
    request = get_request()    # returns None if _shutdown gets set (select
based loop)
    if request is not None:
        process_request(request)

The problem is that I get various exceptions when the signal is raised
(EINTR, but in various guises depending on the code that's running).  It
seems to me I could fix this if I could use reliable (BSD) signals from
python, but I don't see a way to do that.

Otherwise, I'll have to trap each and every EINTR derived exception, which
will mean copying a lot of library code (generally, a big hairy mess).

Is there a better way?

Thanks,
-Randy

P.S. My system does recover from a forced shutdown, but I would like to
complete any running requests before shutting down.




From kwasi007uk at yahoo.co.uk  Sat Feb  3 23:33:33 2001
From: kwasi007uk at yahoo.co.uk (kwasi007uk at yahoo.co.uk)
Date: Sat, 03 Feb 2001 22:33:33 -0000
Subject: Syntax Error
Message-ID: <95i0vt+kju6@eGroups.com>

Using e.g. poplib, m=poplib.POP3 ...
I typed in error m.quit.
It should have been m.quit().
Why doesn't Python give me an error message?
It definitely should since q.quit does not exist!




From chris at onca.catsden.net  Sat Feb  3 23:34:26 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Sat, 3 Feb 2001 14:34:26 -0800 (PST)
Subject: Syntax Error
In-Reply-To: <95i0vt+kju6@eGroups.com>
Message-ID: <Pine.LNX.4.30.0102031429230.15856-100000@onca.catsden.net>

On Sat, 3 Feb 2001 kwasi007uk at yahoo.co.uk wrote:

> Using e.g. poplib, m=poplib.POP3 ...
> I typed in error m.quit.
> It should have been m.quit().
> Why doesn't Python give me an error message?
> It definitely should since q.quit does not exist!

m.quit returns a callable function object, and () calls it.

If m.quit were to return some sort of error, then you would not be able to
assign it to something else. Eg: x = m.quit

However, I'm sure there are a lot of inadvertant errors that could have
been picked up if Python issued a warning if return results are not used
unless a special construct is employed. I'm sure this has all been gone
over before, but its fun... lets do it again:

Eg, say a function f  returned a value, and had a side effect (yes, this
is bad practise, but this is just an example, okay? :) Now say I wanted to
call this function f /just/ for the side effect:

   f()

With the scheme above, this would issue a warning. However, if the
language had a discard construct:

  discard f()

This would solve kwasi's problem. 'm.quit' would return a warning since
the value is not used anywhere. He (she?) would them either change that to
m.quit() as he intended, or add the 'discard' keyword, telling the
interpretor to suppress the warning.

Personally, I dont think its worth the extra work. I've learned not to
screw up so often ;)


   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From tjreedy at udel.edu  Sun Feb  4 01:27:45 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 3 Feb 2001 19:27:45 -0500
Subject: Syntax Error
References: <mailman.981239644.2167.python-list@python.org>
Message-ID: <95i7g3$f65$1@news.udel.edu>

<kwasi007uk at yahoo.co.uk> wrote in message
news:mailman.981239644.2167.python-list at python.org...
> Using e.g. poplib, m=poplib.POP3 ...
> I typed in error m.quit.
> It should have been m.quit().
> Why doesn't Python give me an error message?
> It definitely should since q.quit does not exist!

(You meant m.quit on last line?)

If m.quit() is legal (evaluating to the result of calling m.quit), then
m.quit is also
(evaluating to the function itself).  So there should be no error message




From fredrik at effbot.org  Sun Feb  4 09:00:19 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 04 Feb 2001 08:00:19 GMT
Subject: Syntax Error
References: <mailman.981239644.2167.python-list@python.org>
Message-ID: <n_7f6.11671$AH6.1852702@newsc.telia.net>

kwasi007uk at yahoo.co.uk wrote:
> Using e.g. poplib, m=poplib.POP3 ...
> I typed in error m.quit.
> It should have been m.quit().
> Why doesn't Python give me an error message?
> It definitely should since q.quit does not exist!

Python hasn't changed much since you last asked this
question...

>>> import poplib
>>> mailbox = poplib.POP3("localhost")
>>> mailbox.user("spam")
'+OK User name accepted, password please'
>>> mailbox.pass_("egg")
'+OK Mailbox open, 191 messages'

>>> mailbox.exit
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: 'POP3' instance has no attribute 'exit'

exit didn't exist.

>>> mailbox.quit
<method POP3.quit of POP3 instance at 008AC4EC>

quit sure exists -- it's a method (a callable object).

if you want, you can store that method in a variable
for later use, or pass it to a function, etc.

>>> callback = mailbox.quit
>>> callback
<method POP3.quit of POP3 instance at 008AC4EC>

>>> callback()
'+OK Sayonara'

Cheers /F




From bbollenbach at homenospam.com  Tue Feb  6 00:51:17 2001
From: bbollenbach at homenospam.com (Brad Bollenbach)
Date: Mon, 05 Feb 2001 23:51:17 GMT
Subject: Syntax Error
References: <mailman.981239644.2167.python-list@python.org>
Message-ID: <V%Gf6.11692$wI1.864939@news1.rdc1.mb.home.com>

You've already asked this question once, and recieved two accurate and
helpful replies. Was there a reason for asking it again (the next day)?

<kwasi007uk at yahoo.co.uk> wrote in message
news:mailman.981239644.2167.python-list at python.org...
> Using e.g. poplib, m=poplib.POP3 ...
> I typed in error m.quit.
> It should have been m.quit().
> Why doesn't Python give me an error message?
> It definitely should since q.quit does not exist!
>
>




From dan at eevolved.com  Sun Feb  4 02:13:03 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 04 Feb 2001 01:13:03 GMT
Subject: Function Attribute funkiness...
Message-ID: <z02f6.44367$Pm2.910555@news20.bellglobal.com>

I was reading up on function attributes. Incredible new feature... You can 
now add member functions to classes and class instances without having to 
inherit from them. Can be done at runtime!

Look what you can do with them


------picklesocket.py------
from cPickle import loads, dumps, UnpickleError

def unpickle(maxbytes):
        data = self.sock.recv(maxbytes)
        return dumps(data)

def pickle(anything):
        """anything can be a list, dictionary, tuple, string.. you name it
        """
        data = loads(anything)
        self.sock.send(data)

def patchSocket(instance):
        pickle.sock = unpickle.sock = instance
        instance.pickle = pickle
        instance.unpickle = unpickle
        

sock = socket(...)

patchSocket(sock) # adds new member methods : pickle and unpickle



From chris at onca.catsden.net  Sun Feb  4 02:29:42 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Sat, 3 Feb 2001 17:29:42 -0800 (PST)
Subject: Function Attribute funkiness...
In-Reply-To: <z02f6.44367$Pm2.910555@news20.bellglobal.com>
Message-ID: <Pine.LNX.4.30.0102031727080.15856-100000@onca.catsden.net>

On Sun, 4 Feb 2001, Dan Parisien wrote:

> I was reading up on function attributes. Incredible new feature... You can
> now add member functions to classes and class instances without having to
> inherit from them. Can be done at runtime!

Err... not to rain on your parade, but we've had that particular feature
for some time, now.




   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From jmarshal at mathworks.com  Sun Feb  4 03:36:55 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 4 Feb 2001 02:36:55 GMT
Subject: Stringbuffer
Message-ID: <95if87$juo$1@news.mathworks.com>

I'm plucking characters off a stream and concatenating them.
Currently, I'm doing something like:

   buf = []
   while stream.more():
      buf = buf + stream.read()
   string.join(buf, '')

Is there a faster way?  I'm thinking of writing a Stringbuffer
class in C.  Or does a quick, reliable one already exist?


From chris at onca.catsden.net  Sun Feb  4 03:52:09 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Sat, 3 Feb 2001 18:52:09 -0800 (PST)
Subject: Stringbuffer
In-Reply-To: <95if87$juo$1@news.mathworks.com>
Message-ID: <Pine.LNX.4.30.0102031851070.15856-100000@onca.catsden.net>

On 4 Feb 2001, Joshua Marshall wrote:

> I'm plucking characters off a stream and concatenating them.
> Currently, I'm doing something like:
>
>    buf = []
>    while stream.more():
>       buf = buf + stream.read()
>    string.join(buf, '')
>
> Is there a faster way?  I'm thinking of writing a Stringbuffer
> class in C.  Or does a quick, reliable one already exist?

You could try buf.append(stream.read())

And... I dont know what you're trying to achive with string.join(buf,'').
Are you just adding a space?? :)


   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From jmarshal at mathworks.com  Sun Feb  4 04:37:35 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 4 Feb 2001 03:37:35 GMT
Subject: Stringbuffer
References: <mailman.981255616.11015.python-list@python.org>
Message-ID: <95iipv$lro$1@news.mathworks.com>

chris at onca.catsden.net wrote:
> On 4 Feb 2001, Joshua Marshall wrote:

>> I'm plucking characters off a stream and concatenating them.
>> Currently, I'm doing something like:
>>
>>    buf = []
>>    while stream.more():
>>       buf = buf + stream.read()
>>    string.join(buf, '')
>>
>> Is there a faster way?  I'm thinking of writing a Stringbuffer
>> class in C.  Or does a quick, reliable one already exist?

> You could try buf.append(stream.read())

> And... I dont know what you're trying to achive with string.join(buf,'').
> Are you just adding a space?? :)

I'm using string.join to convert a list of strings into a single string.
My use of a list is for efficiency.  Intuitively, I'm trying to do:

   ret = ''
   while stream.more():
      ret = ret + stream.read()

but this is too slow.  Appending to a list is better, but I'm looking
for more speed.


From jmarshal at mathworks.com  Sun Feb  4 04:44:21 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 4 Feb 2001 03:44:21 GMT
Subject: Stringbuffer
References: <mailman.981255616.11015.python-list@python.org> <95iipv$lro$1@news.mathworks.com>
Message-ID: <95ij6l$mc3$1@news.mathworks.com>

Joshua Marshall <jmarshal at mathworks.com> wrote:
> chris at onca.catsden.net wrote:
>> On 4 Feb 2001, Joshua Marshall wrote:

>>> I'm plucking characters off a stream and concatenating them.
>>> Currently, I'm doing something like:
>>>
>>>    buf = []
>>>    while stream.more():
>>>       buf = buf + stream.read()
>>>    string.join(buf, '')
>>>
>>> Is there a faster way?  I'm thinking of writing a Stringbuffer
>>> class in C.  Or does a quick, reliable one already exist?

>> You could try buf.append(stream.read())

>> And... I dont know what you're trying to achive with string.join(buf,'').
>> Are you just adding a space?? :)

> I'm using string.join to convert a list of strings into a single string.
> My use of a list is for efficiency.  Intuitively, I'm trying to do:

>    ret = ''
>    while stream.more():
>       ret = ret + stream.read()

> but this is too slow.  Appending to a list is better, but I'm looking
> for more speed.

Oh - and my mistake, I was typing the code from memory.  My first example
errors if stream.read() returns a string.  But you see my intent.


From dalke at acm.org  Sun Feb  4 06:50:30 2001
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 3 Feb 2001 22:50:30 -0700
Subject: Stringbuffer
References: <95if87$juo$1@news.mathworks.com>
Message-ID: <95iqk3$msa$1@slb1.atl.mindspring.net>

Joshua Marshall:
>   buf = []
>   while stream.more():
>      buf = buf + stream.read()
>   string.join(buf, '')
>
>Is there a faster way?  I'm thinking of writing a Stringbuffer
>class in C.  Or does a quick, reliable one already exist?

Might want to look at cStringIO

import cStringIO
buf = cStringIO.StringIO()
while stream.more():
  buf.write(stream.read())
s = buf.getvalue()

Or you can use an array

import array
buf = array.array("c", "")
while stream.more():
  buf.fromstring(stream.read())
s = buf.tostring()


Must admit that I would like the method names to be more
similar between these two.

                    Andrew
                    dalke at acm.org





From jmarshal at mathworks.com  Sun Feb  4 05:07:19 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 4 Feb 2001 04:07:19 GMT
Subject: Gratuitous Change
References: <mailman.981258255.11795.python-list@python.org>
Message-ID: <95ikhn$n13$1@news.mathworks.com>

Tim Peters <tim.one at home.com> wrote:
...

> 2.1 is registerless instead.  We were getting too many complaints that
> making Python faster was disturbing carefully tuned Python delay loops
> <wink>.

Complaints that Python was too fast?
What constructs were being used in the delay loops that were sensitive
to the efficiency changes?


From peter at engcorp.com  Sun Feb  4 05:40:47 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 03 Feb 2001 23:40:47 -0500
Subject: Gratuitous Change
References: <mailman.981258255.11795.python-list@python.org> <95ikhn$n13$1@news.mathworks.com>
Message-ID: <3A7CDD4F.93A1C7B@engcorp.com>

Joshua Marshall wrote:
> 
> Tim Peters <tim.one at home.com> wrote:
> > 2.1 is registerless instead.  We were getting too many complaints that
> > making Python faster was disturbing carefully tuned Python delay loops
> > <wink>.
> 
> Complaints that Python was too fast?
> What constructs were being used in the delay loops that were sensitive
> to the efficiency changes?

Apparently it takes about three months for newcomers to catch on,
but I've finally graduated so:  notice the "<wink>"?  :-)


From dnew at san.rr.com  Sun Feb  4 21:43:02 2001
From: dnew at san.rr.com (Darren New)
Date: Sun, 04 Feb 2001 20:43:02 GMT
Subject: Gratuitous Change
References: <mailman.981258255.11795.python-list@python.org> <95ikhn$n13$1@news.mathworks.com> <3A7CDD4F.93A1C7B@engcorp.com>
Message-ID: <3A7DBED7.79251515@san.rr.com>

Peter Hansen wrote:
> Apparently it takes about three months for newcomers to catch on,
> but I've finally graduated so:  notice the "<wink>"?  :-)

Smiley Captioned, for the Humor Impaired.

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
                 Ignorance can be cured. Naivety cures itself.


From gloth at my-deja.com  Sun Feb  4 06:34:21 2001
From: gloth at my-deja.com (gloth at my-deja.com)
Date: Sun, 04 Feb 2001 05:34:21 GMT
Subject: Q: name of current function?
Message-ID: <95ipkt$ab0$1@nnrp1.deja.com>

I try to implement python wrappers around functions in c++. At compile
time, I can't determine which functions will be wrappered, so I need a
mechanism that allows me to create wrappers dynamically.

My idea was to create a single entry-point for all functions that I
create in python, and do the dispatching there by querying the name
under which the function was called. Basically this...

    PyThreadState *state = PyThreadState_Get();
    PyFrameObject *frame = state->frame;
    PyCodeObject *code = frame->f_code;
    PyObject *name = code->co_name;
    const char *str = PyString_AsString(name);

Unfortunately, I always get "?" as the function name. Basically, the
whole code object is not very useful. The sample code was based on a
solution to the same problem (obtaining the name of the current
function), only in Pythin itself. It seems that this mechanism does not
work for builtin functions.

Is there a way to get this information, perhaps the whole PyMethodDef
structure that I used to register the new function?

Please also reply to mail email address: tobias (at) gloth.net

Any hints or suggestions are greatly appreciated!

-Tobi


Sent via Deja.com
http://www.deja.com/


From nospam at mega-nerd.com  Sun Feb  4 06:39:26 2001
From: nospam at mega-nerd.com (Erik de Castro Lopo)
Date: Sun, 04 Feb 2001 05:39:26 GMT
Subject: Attention Travis Oliphant
Message-ID: <3A7CE79E.455935D0@mega-nerd.com>

Sorry for posting this to the group but email to Travis does
not get answered.

Travis,

I have been trying for two weeks now to get a copy of two Python
extensions you wrote and listed on Vaults of Parnassus.

Both are advertised as being available at 

     http://oliphant.netpedia.net/

which requires a password. 

Is this software still available? If so can you either fix your
web server so that it can be accessed without a password or
move it to as server that doesn't require one.

Erik
-- 
+----------------------------------------------------------+
  Erik de Castro Lopo  nospam at mega-nerd.com (Yes its valid)
+----------------------------------------------------------+
Linux: Because rebooting is for adding new hardware


From jdgorman at nospam.ieee.org  Wed Feb 14 05:39:51 2001
From: jdgorman at nospam.ieee.org (John D. Gorman)
Date: Wed, 14 Feb 2001 04:39:51 GMT
Subject: Attention Travis Oliphant
References: <3A7CE79E.455935D0@mega-nerd.com>
Message-ID: <3A8A0C15.D2F13D84@nospam.ieee.org>

Eric,
    I had the same question a few weeks ago.  Those files are now being
maintained at  Travis Oliphant's Python Pages which is at
http://pylab.sourceforge.net/ .

Cheers,
John Gorman


Erik de Castro Lopo wrote:

> Sorry for posting this to the group but email to Travis does
> not get answered.
>
> Travis,
>
> I have been trying for two weeks now to get a copy of two Python
> extensions you wrote and listed on Vaults of Parnassus.
>
> Both are advertised as being available at
>
>      http://oliphant.netpedia.net/
>
> which requires a password.
>
> Is this software still available? If so can you either fix your
> web server so that it can be accessed without a password or
> move it to as server that doesn't require one.
>
> Erik
> --
> +----------------------------------------------------------+
>   Erik de Castro Lopo  nospam at mega-nerd.com (Yes its valid)
> +----------------------------------------------------------+
> Linux: Because rebooting is for adding new hardware
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010214/de62c928/attachment.htm>

From nospam at mega-nerd.com  Wed Feb 14 08:49:08 2001
From: nospam at mega-nerd.com (Erik de Castro Lopo)
Date: Wed, 14 Feb 2001 07:49:08 GMT
Subject: Attention Travis Oliphant
References: <3A7CE79E.455935D0@mega-nerd.com> <3A8A0C15.D2F13D84@nospam.ieee.org>
Message-ID: <3A8A364C.2F124DA4@mega-nerd.com>

John D. Gorman wrote:
> 
> Eric,
>     I had the same question a few weeks ago.  Those files are now
> being maintained at  Travis Oliphant's Python Pages which is at
> http://pylab.sourceforge.net/ .
> 
> Cheers,
> John Gorman

Thank you, thank you thank you!!!!!

Erik
-- 
+----------------------------------------------------------+
  Erik de Castro Lopo  nospam at mega-nerd.com (Yes its valid)
+----------------------------------------------------------+
"Using Java as a general purpose application development language 
is like going big game hunting armed with Nerf weapons." 
-- Author Unknown


From linux at thewebsons.com  Sun Feb  4 08:48:08 2001
From: linux at thewebsons.com (Ben Ocean)
Date: Sat, 03 Feb 2001 23:48:08 -0800
Subject: HTTPD Writable
Message-ID: <5.0.2.1.0.20010203234754.00a4d0d0@thewebsons.com>

Hi;
I'm trying to install a program that requires I create an *httpd writable* 
file. How do I do this? What does it mean to have httpd write to a file? Do 
I need to assign httpd a passwd? Do I need to group it with root? Please 
advise.
BenO 




From thomas at xs4all.net  Sun Feb  4 16:18:22 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 4 Feb 2001 16:18:22 +0100
Subject: HTTPD Writable
In-Reply-To: <5.0.2.1.0.20010203234754.00a4d0d0@thewebsons.com>; from linux@thewebsons.com on Sat, Feb 03, 2001 at 11:48:08PM -0800
References: <5.0.2.1.0.20010203234754.00a4d0d0@thewebsons.com>
Message-ID: <20010204161822.E962@xs4all.nl>

On Sat, Feb 03, 2001 at 11:48:08PM -0800, Ben Ocean wrote:

> I'm trying to install a program that requires I create an *httpd writable* 
> file. How do I do this? What does it mean to have httpd write to a file? Do 
> I need to assign httpd a passwd? Do I need to group it with root? Please 
> advise.

Unclear why you are posting it here; it's terribly off-topic, even if the
program you are installing is a Python program. Nevertheless, what you are
probably supposed to do is make a file writable for the httpd process. It
depends on your installation (and indeed, your 'brand' of httpd) what userid
and permissions the httpd process is running as, but as a wild guess, you
can use 'chmod a+w <file>' (dangerous, but sure to work) or 'chown nobody
<file>' (sometimes inconvenient, but slightly less dangerous.) If you happen
to know that httpd is running as a particular group, and noone else is in
that group, chgrp the file to that group and make it only writable for the
group, not the rest of the world: chmod g+w <file>.

For further questions I suggest you ask the author or supplier of the
program you are installing, they're sure to know exactly what they mean :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From zope at thewebsons.com  Mon Feb  5 00:45:46 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Sun, 04 Feb 2001 15:45:46 -0800
Subject: HTTPD Writable
In-Reply-To: <20010204161822.E962@xs4all.nl>
References: <5.0.2.1.0.20010203234754.00a4d0d0@thewebsons.com>
 <5.0.2.1.0.20010203234754.00a4d0d0@thewebsons.com>
Message-ID: <5.0.2.1.0.20010204154508.00a12960@thewebsons.com>

My apologies! Did not mean to post here!
BenO

At 04:18 PM 2/4/2001 +0100, you wrote:
>On Sat, Feb 03, 2001 at 11:48:08PM -0800, Ben Ocean wrote:
>
> > I'm trying to install a program that requires I create an *httpd writable*
> > file. How do I do this? What does it mean to have httpd write to a 
> file? Do
> > I need to assign httpd a passwd? Do I need to group it with root? Please
> > advise.
>
>Unclear why you are posting it here; it's terribly off-topic, even if the
>program you are installing is a Python program. Nevertheless, what you are
>probably supposed to do is make a file writable for the httpd process. It
>depends on your installation (and indeed, your 'brand' of httpd) what userid
>and permissions the httpd process is running as, but as a wild guess, you
>can use 'chmod a+w <file>' (dangerous, but sure to work) or 'chown nobody
><file>' (sometimes inconvenient, but slightly less dangerous.) If you happen
>to know that httpd is running as a particular group, and noone else is in
>that group, chgrp the file to that group and make it only writable for the
>group, not the rest of the world: chmod g+w <file>.
>
>For further questions I suggest you ask the author or supplier of the
>program you are installing, they're sure to know exactly what they mean :)
>
>--
>Thomas Wouters <thomas at xs4all.net>
>
>Hi! I'm a .signature virus! copy me into your .signature file to help me 
>spread!
>
>--
>http://mail.python.org/mailman/listinfo/python-list




From arildh at stud.cs.uit.no  Sun Feb  4 13:15:45 2001
From: arildh at stud.cs.uit.no (Arild Hansen)
Date: Sun, 4 Feb 2001 13:15:45 +0100
Subject: IPv6 patch for python 2.0 anyuone?
Message-ID: <Pine.LNX.4.21.0102041314490.13943-100000@arildh.pasta.cs.uit.no>

Does anyone know if an IPv6 patch has been released for python 2.0 yet?

A. Hansen



From erno-news at erno.iki.fi  Mon Feb  5 01:15:55 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 05 Feb 2001 02:15:55 +0200
Subject: IPv6 patch for python 2.0 anyuone?
References: <Pine.LNX.4.21.0102041314490.13943-100000@arildh.pasta.cs.uit.no>
Message-ID: <kuvgqq2ddw.fsf@lasipalatsi.fi>

In article
<Pine.LNX.4.21.0102041314490.13943-100000 at arildh.pasta.cs.uit.no>,
Arild Hansen <arildh at stud.cs.uit.no> writes:

| Does anyone know if an IPv6 patch has been released for python 2.0 yet?

looks like there is one at <URL:http://sourceforge.net/patch/?
func=detailpatch&patch_id=101196&group_id=5470>.

  -- erno


From arildh at stud.cs.uit.no  Mon Feb  5 14:00:02 2001
From: arildh at stud.cs.uit.no (Arild Hansen)
Date: Mon, 5 Feb 2001 14:00:02 +0100
Subject: IPv6 patch for python 2.0 anyuone?
In-Reply-To: <kuvgqq2ddw.fsf@lasipalatsi.fi>
References: <Pine.LNX.4.21.0102041314490.13943-100000@arildh.pasta.cs.uit.no> <kuvgqq2ddw.fsf@lasipalatsi.fi>
Message-ID: <Pine.LNX.4.21.0102051359400.16039-100000@arildh.pasta.cs.uit.no>

Thx a lot !! I really wanted that patch,

A. Hansen





On 5 Feb 2001, Erno Kuusela wrote:

> In article
> <Pine.LNX.4.21.0102041314490.13943-100000 at arildh.pasta.cs.uit.no>,
> Arild Hansen <arildh at stud.cs.uit.no> writes:
> 
> | Does anyone know if an IPv6 patch has been released for python 2.0 yet?
> 
> looks like there is one at <URL:http://sourceforge.net/patch/?
> func=detailpatch&patch_id=101196&group_id=5470>.
> 
>   -- erno
> 



From gry at ll.mit.edu  Sun Feb  4 14:38:39 2001
From: gry at ll.mit.edu (John Gault)
Date: Sun, 04 Feb 2001 13:38:39 GMT
Subject: threading.RLock not subclassible?
Message-ID: <3A7D5B5F.AADAA673@ll.mit.edu>

[python2.0, i386-linux 2.2.16]
I would like to make a subclass of threading.RLock so I can
add a 'safe_acquire' method.  Safe_acquire would use non-
blocking acquires and sleeps over several iterations to try
to acquire the lock, but would eventually return so my thread
is not forever blocked.

But the RLock class seems to be written in a way discouraging
subclassing: in the threading class I see --

def RLock(*args, **kwargs):
    return apply(_RLock, args, kwargs)class _RLock(_Verbose):
    
    def __init__(self, verbose=None):
        _Verbose.__init__(self, verbose)
        self.__block = _allocate_lock()
        self.__owner = None
        self.__count = 0

It looks like the author(not identified in the code) is explictly trying
to avoid subclassing...  Does anyone see danger (or stupidity?) in my
doing:

class Safe_RLock(threading._RLock):
    '''reentrant 'RLock' with safe acquire method'''
    def __init__(self, timeout=2.0): #default cumulative timeout of 2.0
seconds 
        threading._RLock.__init__(self)
        self.timeout = timeout

    def safe_acquire(self, timeout=None):
        timeout = timeout or self.timeout
        cumulative_backoff = 0
        while not self.acquire(0):
            # pseudorandom 50ms<delay<200ms to retry lock acquire
            backoff = whrandom.uniform(0.05, 0.2)
            dbug('delaying %.3f seconds for blocked lock' % timeout)
            time.sleep(backoff)
            cumulative_backoff += backoff
            if cumulative_timeout > timeout: # give up after timeout
seconds
                dbug('failed to acquire lock after %s seconds' %
cumulative_backoff)
                return 0
        else:
            return 1

used later as:
        # try to get the lock, but don't block forever...
        if dblock.safe_acquire():
            db[ip] = hostname or '$hostnotfound'
            dblock.release()
        else:
            bug('inverse_lookup: failed to acquire db lock')
        result_queue.put((ip, hostname))

Thanks,
	George
------
George Young, gry at mediaone dot net


From aahz at panix.com  Sun Feb  4 18:14:57 2001
From: aahz at panix.com (Aahz Maruch)
Date: 4 Feb 2001 09:14:57 -0800
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu>
Message-ID: <95k2mh$fj0$1@panix3.panix.com>

In article <3A7D5B5F.AADAA673 at ll.mit.edu>, John Gault  <gry at ll.mit.edu> wrote:
>
>I would like to make a subclass of threading.RLock so I can
>add a 'safe_acquire' method.  Safe_acquire would use non-
>blocking acquires and sleeps over several iterations to try
>to acquire the lock, but would eventually return so my thread
>is not forever blocked.

I think that it would be better for you to just create a brand-new class
that contains an RLock object.  Is there some reason to avoid that?
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From tim.one at home.com  Mon Feb  5 02:22:23 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 05 Feb 2001 01:22:23 GMT
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu>
Message-ID: <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com>

[John Gault, or maybe George Young]
> I would like to make a subclass of threading.RLock so I can
> add a 'safe_acquire' method.  Safe_acquire would use non-
> blocking acquires and sleeps over several iterations to try
> to acquire the lock, but would eventually return so my thread
> is not forever blocked.
>
> But the RLock class seems to be written in a way discouraging
> subclassing:
> ...
> It looks like the author(not identified in the code) is explictly
> trying to avoid subclassing...

The author is Guido Himself, and he was indeed trying to avoid subclassing.
I didn't know why either, but had a chance ask him.  The deal is that some
platforms may have a much more efficient way to implement one or more of
these abstractions, in which case threading.py would be happy to export the
platform-specific way when running on that platform.

There are no instances of that yet, although threading.py gets kinda close
with its

Lock = _allocate_lock

You can't subclass from Lock because its *implementation*
(thread.allocate_lock) is not a class.  The same may well be true of future
platform-specific implementations of synch objects.  So threading.py exposes
factory functions instead of classes.

There's no immediate danger that you'll get in trouble by subclassing from
_RLock instead, but there's no guarantee that will remain safe (the leading
underscore means, as you know, "touch at your own risk").

BTW, an acquire with a timeout is a cool idea!  Feel encouraged to submit a
patch to add such a thing directly to threading.py:

    http://sourceforge.net/patch/?group_id=5470

I suggest a name like timed_acquire, though -- if it's in the core, the name
safe_acquire reads more like a snotty value judgment about the other
.acquire() <wink>.

safety-depends-more-on-how-it's-used-than-what-it-does-ly y'rs  - tim





From db3l at fitlinxx.com  Tue Feb  6 01:54:22 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 05 Feb 2001 19:54:22 -0500
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com>
Message-ID: <uk874zl4x.fsf@ctwd0143.fitlinxx.com>

"Tim Peters" <tim.one at home.com> writes:

> BTW, an acquire with a timeout is a cool idea!  Feel encouraged to submit a
> patch to add such a thing directly to threading.py:

If anyone does think along these lines I'd suggest investigating
adding a timeout to the code per-platform thread code as opposed to
something at a high level in threading.py.

Using a loop construct around a non-blocking lock acquisition runs the
risk of never getting the lock, even if you were the first to start
waiting, such as in the case where you are in the loop part of the
code when the lock releases.

What's really needed is a true blocking acquire with timeout, so that
you remain constantly in the queue for the lock at the OS level, and
are guaranteed to get it if it frees up within your timeout.

For systems where that isn't supported at the OS level, the
non-blocking loop could be implemented in the appropriate thread_* OS
module.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From rwklee at home.com  Tue Feb  6 03:07:52 2001
From: rwklee at home.com (Rick Lee)
Date: Tue, 06 Feb 2001 02:07:52 GMT
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A7F5C74.6C9BC395@home.com>

It seems to me that OS-specific behaviour of thread switching is _very_
different among platforms.  For example, if the parent thread starts up 100
different threads each with some long loop to run:

    for i in xrange(100):
        try:
            t = threading.Thread(target=runObj, args=())
            t.start()
        except:
            break
    snapshot = threading.activeCount()

If I run it in different platforms, snapshot is consistently different, anywhere
between very close to 100, or just 1.  You may argue it depends on how long
runObj takes to run; but I can change runObj to run for a very long time, say
seconds, and the results are consistently different across the platforms (NT,
Mac, Linux).

I have some even weirder differences, if I use say time.sleep(.001) inside a
thread, as the FAQ says.  And I have a few other examples I ran across.  (The
one that bugs me the most is that some blocked threads in the Mac just don't get
to run, when the blocking condition is removed.)

This makes me very nervous about implementing a multi-thread cross-platform
application.

Anyone has a list of what the differences are across the platforms?

- Rick Lee

David Bolen wrote:

> "Tim Peters" <tim.one at home.com> writes:
>
> > BTW, an acquire with a timeout is a cool idea!  Feel encouraged to submit a
> > patch to add such a thing directly to threading.py:
>
> If anyone does think along these lines I'd suggest investigating
> adding a timeout to the code per-platform thread code as opposed to
> something at a high level in threading.py.
>
> Using a loop construct around a non-blocking lock acquisition runs the
> risk of never getting the lock, even if you were the first to start
> waiting, such as in the case where you are in the loop part of the
> code when the lock releases.
>
> What's really needed is a true blocking acquire with timeout, so that
> you remain constantly in the queue for the lock at the OS level, and
> are guaranteed to get it if it frees up within your timeout.
>
> For systems where that isn't supported at the OS level, the
> non-blocking loop could be implemented in the appropriate thread_* OS
> module.
>
> --
> -- David
> --
> /-----------------------------------------------------------------------\
>  \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
>   |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
>  /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
> \-----------------------------------------------------------------------/



From db3l at fitlinxx.com  Tue Feb  6 20:42:31 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Feb 2001 14:42:31 -0500
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com> <3A7F5C74.6C9BC395@home.com>
Message-ID: <uvgqna994.fsf@ctwd0143.fitlinxx.com>

Rick Lee <rwklee at home.com> writes:

> It seems to me that OS-specific behaviour of thread switching is _very_
> different among platforms.  For example, if the parent thread starts up 100
> different threads each with some long loop to run:
> 
>     for i in xrange(100):
>         try:
>             t = threading.Thread(target=runObj, args=())
>             t.start()
>         except:
>             break
>     snapshot = threading.activeCount()
> 
> If I run it in different platforms, snapshot is consistently
> different, anywhere between very close to 100, or just 1.  You may
> argue it depends on how long runObj takes to run; but I can change
> runObj to run for a very long time, say seconds, and the results are
> consistently different across the platforms (NT, Mac, Linux).

I suppose the question is why would the difference in shapshot be a
problem to an application?  What you're seeing is that how threads are
dispatched at creation with respect to the currently executing thread
can differ across platforms (and presumably whether the thread support
is native or how deep in the OS it is supported).  But it's not really
a problem, per se, except that it's something that's part and parcel
of multi-threaded development.

By it's nature, when you attempt to start threads their execution with
respect to the starting thread is asynchronous and unrelated.  I'm
guessing the two major cases you're seeing are either (a) when each
new thread gets some execution time right at creation and (b) when the
main thread runs to completion before any of the created threads
really got started.  Thus having nearly a full count versus just a
count of your main thread.  Changing the runtime of runObj won't
likely change that since it's the the scheduling aspect of the
separate threads with respect to the main thread that is causing the
behavior.  You might find though that adding a delay to your main
thread in between (or just after) the creation of the other threads
gives them all a chance to start.

If it's important to you that all of the threads actually begin
executing before you continue the main thread, you'll need to use some
sort of event or other notification mechanism (the quickest way being
a slow loop that waits until the activeCount shows all threads).  You
can't assume that the execution timeframe of the main thread will have any
specific correlation (technically even on the same platform, much less
cross-platform) to the execution timeframe of the other threads.

> I have some even weirder differences, if I use say time.sleep(.001)
> inside a thread, as the FAQ says.  And I have a few other examples I
> ran across.  (The one that bugs me the most is that some blocked
> threads in the Mac just don't get to run, when the blocking
> condition is removed.)

Can you be more specific as to the other wierder differences - they
may just be natural behavior in a multi-threaded environment.  Adding
a time.sleep() within a thread will typically cause that thread to
yield it's control of the processor, and some other thread will get
dispatched, but that's about it.  Within Python, time.sleep() will
also release the global interpreter lock, thus permitting other Python
threads to execute.  That can be important if you have a thread that
otherwise would loop rapidly (without any natural blocking point),
since it relieves some of the unnecessary CPU load.  The best case is
having a thread block until some event occurs that needs to be
serviced.

For your Mac case, how are you blocking the threads?  If you have
multiple threads all using wait() with a timeout on the same event,
then the threading module implements the timeout wait as a loop around
a non-blocking wait, and (as discussed in this thread) that can result
in a thread never being woken up, since the looping construct is not
atomic.  But I believe that uncertainty is cross-platform :-)

But if you use non-timeout waits, then I believe it should eventually
wake up, although the order may be non-deterministic, although someone
else may know more about the Mac thread implementation to comment
differently.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From rwklee at home.com  Wed Feb  7 02:48:25 2001
From: rwklee at home.com (Rick Lee)
Date: Wed, 07 Feb 2001 01:48:25 GMT
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com> <3A7F5C74.6C9BC395@home.com> <uvgqna994.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A80A969.54B41DFF@home.com>

Thanks for your explanation about thread execution with respect to the main
thread.  I accept it, but only to a point, because in Python Essential Reference
on Threads, pg. 174, it is stated that:

"A multithreaded program executes by dividing its processing time between all
active threads.  For example, a program with 10 active threads of execution
would allow approximately 1/10 of its CPU time to each thread and cycle between
threads in rapid succession."

I certainly was not seeing this behaviour on some platforms, even for threads
that take a long time to perform computations.

You asked why I should care.  Actually, I had two questions I tried to find out
an answer for:

- what is the thread switching performance penalty, and how sensitive is the
penalty to the number of threads?

- is there an upper limit to the number of threads?

I asked this question here in comp.lang.python.  Getting no response, I tried to
write some small programs to help me see the effects.  So that was the bit of
history behind this.

Regarding what happens on the Mac:  I have a multi-threaded program with several
threads, each of which blocks without timeout on sockets.accept, sockets.recv,
and queues.get respectively.  This program runs perfectly on NT and Linux, but
the blocked threads on the Mac don't seem to execute when the blocking condition
is removed (I am 99% sure that's what's happening).  If anyone can shed light on
this, I will be very grateful.

Another weirdness includes a bit of code like this, which is a recursive call on
itself to create threads:

def run():
    me = len(runners)
    started.append (me)
    if len(runners) > 0:
        runners.pop()
        try:
            t = threading.Thread(target=run, args=())
            t.start()
        except:
            pass
        else:
            runObj()
            t.join()
        completedRunObj.append(me)

This scheme runs fine on NT, and produces sensible results; ie., the amount of
the time for the program to run to completion grows fairly linearly with the
number of runners.  On Linux, the time for the program to run to completion is
always constant: as if only between 1 and 2 runners had really executed runObj;
all the other threads seem to take no time to run.  This one also boggled my
mind.  (This was one of my several attempts at trying to characterize thread
switching performance penalty, and max. limit of threads.)


David Bolen wrote:

> Rick Lee <rwklee at home.com> writes:
>
> > It seems to me that OS-specific behaviour of thread switching is _very_
> > different among platforms.  For example, if the parent thread starts up 100
> > different threads each with some long loop to run:
> >
> >     for i in xrange(100):
> >         try:
> >             t = threading.Thread(target=runObj, args=())
> >             t.start()
> >         except:
> >             break
> >     snapshot = threading.activeCount()
> >
> > If I run it in different platforms, snapshot is consistently
> > different, anywhere between very close to 100, or just 1.  You may
> > argue it depends on how long runObj takes to run; but I can change
> > runObj to run for a very long time, say seconds, and the results are
> > consistently different across the platforms (NT, Mac, Linux).
>
> I suppose the question is why would the difference in shapshot be a
> problem to an application?  What you're seeing is that how threads are
> dispatched at creation with respect to the currently executing thread
> can differ across platforms (and presumably whether the thread support
> is native or how deep in the OS it is supported).  But it's not really
> a problem, per se, except that it's something that's part and parcel
> of multi-threaded development.
>
> By it's nature, when you attempt to start threads their execution with
> respect to the starting thread is asynchronous and unrelated.  I'm
> guessing the two major cases you're seeing are either (a) when each
> new thread gets some execution time right at creation and (b) when the
> main thread runs to completion before any of the created threads
> really got started.  Thus having nearly a full count versus just a
> count of your main thread.  Changing the runtime of runObj won't
> likely change that since it's the the scheduling aspect of the
> separate threads with respect to the main thread that is causing the
> behavior.  You might find though that adding a delay to your main
> thread in between (or just after) the creation of the other threads
> gives them all a chance to start.
>
> If it's important to you that all of the threads actually begin
> executing before you continue the main thread, you'll need to use some
> sort of event or other notification mechanism (the quickest way being
> a slow loop that waits until the activeCount shows all threads).  You
> can't assume that the execution timeframe of the main thread will have any
> specific correlation (technically even on the same platform, much less
> cross-platform) to the execution timeframe of the other threads.
>
> > I have some even weirder differences, if I use say time.sleep(.001)
> > inside a thread, as the FAQ says.  And I have a few other examples I
> > ran across.  (The one that bugs me the most is that some blocked
> > threads in the Mac just don't get to run, when the blocking
> > condition is removed.)
>
> Can you be more specific as to the other wierder differences - they
> may just be natural behavior in a multi-threaded environment.  Adding
> a time.sleep() within a thread will typically cause that thread to
> yield it's control of the processor, and some other thread will get
> dispatched, but that's about it.  Within Python, time.sleep() will
> also release the global interpreter lock, thus permitting other Python
> threads to execute.  That can be important if you have a thread that
> otherwise would loop rapidly (without any natural blocking point),
> since it relieves some of the unnecessary CPU load.  The best case is
> having a thread block until some event occurs that needs to be
> serviced.
>
> For your Mac case, how are you blocking the threads?  If you have
> multiple threads all using wait() with a timeout on the same event,
> then the threading module implements the timeout wait as a loop around
> a non-blocking wait, and (as discussed in this thread) that can result
> in a thread never being woken up, since the looping construct is not
> atomic.  But I believe that uncertainty is cross-platform :-)
>
> But if you use non-timeout waits, then I believe it should eventually
> wake up, although the order may be non-deterministic, although someone
> else may know more about the Mac thread implementation to comment
> differently.
>
> --
> -- David
> --
> /-----------------------------------------------------------------------\
>  \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
>   |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
>  /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
> \-----------------------------------------------------------------------/



From db3l at fitlinxx.com  Thu Feb  8 04:37:54 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 07 Feb 2001 22:37:54 -0500
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com> <3A7F5C74.6C9BC395@home.com> <uvgqna994.fsf@ctwd0143.fitlinxx.com> <3A80A969.54B41DFF@home.com>
Message-ID: <uvgqlvo8d.fsf@ctwd0143.fitlinxx.com>

Rick Lee <rwklee at home.com> writes:

> "A multithreaded program executes by dividing its processing time
> between all active threads.  For example, a program with 10 active
> threads of execution would allow approximately 1/10 of its CPU time
> to each thread and cycle between threads in rapid succession."
> 
> I certainly was not seeing this behaviour on some platforms, even
> for threads that take a long time to perform computations.

Well, that may be a different issue than your posted point about the
active thread count.  It is possible that depending on the local
thread implementation (at the process level, OS kernel level,
pre-emptive versus cooperative, etc...) that you may get better or
worse sharing.  Really the only bad case is if the threads are
cooperative (that is, a thread has to yield before the scheduler moves
to another thread).  I'm not sure which if any platforms fall into
that category (perhaps someone else can answer for the typical Unix
pthreads implementation).

How were you measuring the division of labor amongst the threads?

> - what is the thread switching performance penalty, and how sensitive is the
> penalty to the number of threads?

There's a hit to the Python interpreter to support threading (I think
I recall seeing something like 10%) in general, but actually switching
between the threads should be quite low overhead.  Certainly when
threads are native to the platform's OS (such as with Windows NT), the
system scheduler really only thinks in terms of threads anyway, so
thread context switches are just the normal mode of dispatching.

It should also be fairly insensitive to the number of threads,
providing overall system resources are not exceeded.

> - is there an upper limit to the number of threads?

That depends on the platform - you'll definitely run out of resources
eventually, but it's sort of like asking what's the upper limit on the
number of processes on the system, which can vary.

It also depends on what they are doing, since you can likely support
far more threads if the majority are blocked, then if you are trying
to share the CPU among them all as active threads.

Certainly I would expect that hundreds should not be out of the
question, and probably thousands.  My desktop NT machine just with a
normal mix of stuff active sits with 42 processes and about 250
threads, and I expect it could easily handle an order of magnitude
higher - as long as they weren't all trying to run simultaneously.

But I don't have a hard limit for you, nor do I personally know of
such values across all platforms.  It's a per-platform resource issue
and is related to the activity of the code as well, so I'm not sure
there's a fixed answer.  Certainly if you're thinking in terms of
thousands you may want to find a way to share them - but if you're
thinking 10s or hundreds, you're probably ok.

> Regarding what happens on the Mac: I have a multi-threaded program
> with several threads, each of which blocks without timeout on
> sockets.accept, sockets.recv, and queues.get respectively.  This
> program runs perfectly on NT and Linux, but the blocked threads on
> the Mac don't seem to execute when the blocking condition is removed
> (I am 99% sure that's what's happening).  If anyone can shed light
> on this, I will be very grateful.

By blocking condition removed, do you mean that data shows up on the
socket or queues?  I really can't think why if such data were to
arrive that the system calls wouldn't be satisfied.  I don't see the
Mac implementation in the CVS tree I've currently got checked out, so
I'm not sure how the Mac threading is implemented - perhaps someone
else could chime in with some info.

> Another weirdness includes a bit of code like this, which is a
> recursive call on itself to create threads:

Well, my first inclination is that the code itself appears dangerous -
and unpredictable - because you don't have any protection against
simultaneous access to the shared objects (e.g., runners).  But it's
hard to say without seeing more of the surrounding code.  But for
example, it would appear that by the time you get around to appending
"me" to your completedRunObj, it's probably been changed by a new
thread.

Personally, if you're trying to observe behavior of threads, I find
that a few judicial print statements can help out a lot.  Yes, if your
problems have a time component, the additional time to do the output
can interfere, but for basic scheduling (what thread ran when) it can
be enlightening to just print the names of threads and points of their
execution all in sequence.  If you keep your output as a single string
to the "print" statement, it won't get split amongst different threads
either, which can be a problem sometimes with this approach.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From rwklee at home.com  Fri Feb  9 02:02:43 2001
From: rwklee at home.com (Rick Lee)
Date: Fri, 09 Feb 2001 01:02:43 GMT
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com> <3A7F5C74.6C9BC395@home.com> <uvgqna994.fsf@ctwd0143.fitlinxx.com> <3A80A969.54B41DFF@home.com> <uvgqlvo8d.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A8341B1.6B7CB4C6@home.com>

Thanks for your explanations and tips.  I found the thread behaviour on the Mac
extremely perplexing.

Regarding your comment on what you think may be the unsafe and unpredictable
code: the code does access a couple of global lists, the only things done with
these lists are len() and append(), which I read somewhere are "atomic".

David Bolen wrote:

> Rick Lee <rwklee at home.com> writes:
>
> > "A multithreaded program executes by dividing its processing time
> > between all active threads.  For example, a program with 10 active
> > threads of execution would allow approximately 1/10 of its CPU time
> > to each thread and cycle between threads in rapid succession."
> >
> > I certainly was not seeing this behaviour on some platforms, even
> > for threads that take a long time to perform computations.
>
> Well, that may be a different issue than your posted point about the
> active thread count.  It is possible that depending on the local
> thread implementation (at the process level, OS kernel level,
> pre-emptive versus cooperative, etc...) that you may get better or
> worse sharing.  Really the only bad case is if the threads are
> cooperative (that is, a thread has to yield before the scheduler moves
> to another thread).  I'm not sure which if any platforms fall into
> that category (perhaps someone else can answer for the typical Unix
> pthreads implementation).
>
> How were you measuring the division of labor amongst the threads?
>
> > - what is the thread switching performance penalty, and how sensitive is the
> > penalty to the number of threads?
>
> There's a hit to the Python interpreter to support threading (I think
> I recall seeing something like 10%) in general, but actually switching
> between the threads should be quite low overhead.  Certainly when
> threads are native to the platform's OS (such as with Windows NT), the
> system scheduler really only thinks in terms of threads anyway, so
> thread context switches are just the normal mode of dispatching.
>
> It should also be fairly insensitive to the number of threads,
> providing overall system resources are not exceeded.
>
> > - is there an upper limit to the number of threads?
>
> That depends on the platform - you'll definitely run out of resources
> eventually, but it's sort of like asking what's the upper limit on the
> number of processes on the system, which can vary.
>
> It also depends on what they are doing, since you can likely support
> far more threads if the majority are blocked, then if you are trying
> to share the CPU among them all as active threads.
>
> Certainly I would expect that hundreds should not be out of the
> question, and probably thousands.  My desktop NT machine just with a
> normal mix of stuff active sits with 42 processes and about 250
> threads, and I expect it could easily handle an order of magnitude
> higher - as long as they weren't all trying to run simultaneously.
>
> But I don't have a hard limit for you, nor do I personally know of
> such values across all platforms.  It's a per-platform resource issue
> and is related to the activity of the code as well, so I'm not sure
> there's a fixed answer.  Certainly if you're thinking in terms of
> thousands you may want to find a way to share them - but if you're
> thinking 10s or hundreds, you're probably ok.
>
> > Regarding what happens on the Mac: I have a multi-threaded program
> > with several threads, each of which blocks without timeout on
> > sockets.accept, sockets.recv, and queues.get respectively.  This
> > program runs perfectly on NT and Linux, but the blocked threads on
> > the Mac don't seem to execute when the blocking condition is removed
> > (I am 99% sure that's what's happening).  If anyone can shed light
> > on this, I will be very grateful.
>
> By blocking condition removed, do you mean that data shows up on the
> socket or queues?  I really can't think why if such data were to
> arrive that the system calls wouldn't be satisfied.  I don't see the
> Mac implementation in the CVS tree I've currently got checked out, so
> I'm not sure how the Mac threading is implemented - perhaps someone
> else could chime in with some info.
>
> > Another weirdness includes a bit of code like this, which is a
> > recursive call on itself to create threads:
>
> Well, my first inclination is that the code itself appears dangerous -
> and unpredictable - because you don't have any protection against
> simultaneous access to the shared objects (e.g., runners).  But it's
> hard to say without seeing more of the surrounding code.  But for
> example, it would appear that by the time you get around to appending
> "me" to your completedRunObj, it's probably been changed by a new
> thread.
>
> Personally, if you're trying to observe behavior of threads, I find
> that a few judicial print statements can help out a lot.  Yes, if your
> problems have a time component, the additional time to do the output
> can interfere, but for basic scheduling (what thread ran when) it can
> be enlightening to just print the names of threads and points of their
> execution all in sequence.  If you keep your output as a single string
> to the "print" statement, it won't get split amongst different threads
> either, which can be a problem sometimes with this approach.
>
> --
> -- David
> --
> /-----------------------------------------------------------------------\
>  \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
>   |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
>  /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
> \-----------------------------------------------------------------------/



From db3l at fitlinxx.com  Fri Feb  9 19:54:19 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 09 Feb 2001 13:54:19 -0500
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com> <3A7F5C74.6C9BC395@home.com> <uvgqna994.fsf@ctwd0143.fitlinxx.com> <3A80A969.54B41DFF@home.com> <uvgqlvo8d.fsf@ctwd0143.fitlinxx.com> <3A8341B1.6B7CB4C6@home.com>
Message-ID: <uu263wuuc.fsf@ctwd0143.fitlinxx.com>

Rick Lee <rwklee at home.com> writes:

> Regarding your comment on what you think may be the unsafe and
> unpredictable code: the code does access a couple of global lists,
> the only things done with these lists are len() and append(), which
> I read somewhere are "atomic".

Perhaps, although given that any given method might behind the scenes
have been overridden with arbitrary code, I would probably suggest
never making such assumptions and instead always protecting common
data.  (And in particular when you are having uncertain behavior :-))

One point was that the code seemed to assume the object pointed to by
"me" is the same between when you append it to started and when you
later append it to completedRunObj (if the completed is supposed to
track the started), but since I think "me" may end as the same local
label accessed simultaneously by multiple threads, you could fine "me"
change in between (so perhaps you'd add "n" to the started list, but
then append "n-1" to the completed list when the new thread just
started re-evaluates the length of the runners and assigns it to "me").

Basically, with threading it's always better to play it safe when it
involves any potential for multi-thread access to the same data.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From aahz at panix.com  Tue Feb  6 07:12:53 2001
From: aahz at panix.com (Aahz Maruch)
Date: 5 Feb 2001 22:12:53 -0800
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com>
Message-ID: <95o4l5$oak$1@panix6.panix.com>

In article <uk874zl4x.fsf at ctwd0143.fitlinxx.com>,
David Bolen  <db3l at fitlinxx.com> wrote:
>"Tim Peters" <tim.one at home.com> writes:
>>
>> BTW, an acquire with a timeout is a cool idea!  Feel encouraged to
>> submit a patch to add such a thing directly to threading.py:
>
>If anyone does think along these lines I'd suggest investigating adding
>a timeout to the code per-platform thread code as opposed to something
>at a high level in threading.py.

Given how little of Python is currently platform-specific, what kinds of
arguments can you marshal in favor of this?
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Life as a contractor: when they tell you it's going to rain, sometimes
it drizzles and sometimes it pours.


From db3l at fitlinxx.com  Tue Feb  6 20:28:02 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Feb 2001 14:28:02 -0500
Subject: threading.RLock not subclassible?
References: <3A7D5B5F.AADAA673@ll.mit.edu> <jfnf6.66068$B6.17478061@news1.rdc1.md.home.com> <uk874zl4x.fsf@ctwd0143.fitlinxx.com> <95o4l5$oak$1@panix6.panix.com>
Message-ID: <uy9vja9x9.fsf@ctwd0143.fitlinxx.com>

aahz at panix.com (Aahz Maruch) writes:

> Given how little of Python is currently platform-specific, what kinds of
> arguments can you marshal in favor of this?

At the high level library layer there isn't much, but there's a decent
amount of platform code within the core itself.  An admirably small
amount to be sure, but thread support is one of those areas where it's
hard to avoid.

In particular, the low level thread support is already
platform-specific in the core (see all the thread_xxx.c modules),
including the lock acquisition we're talking about modifying, so I
don't think that having a platform-specific implementation of a new
timeout parameter is much of a difference at all implementation wise.

Note that the existence of the timeout parameter would be
platform-independent, and the existing core OS modules would just use
the native support if present or implement a non-blocking loop if not.
You could probably abstract that non-blocking loop as a single piece
of code for all thread modules that needed it.  For many cases it
would be a trivial change to the core code (e.g., under Windows just
passing on the timeout parameter rather than the current hardcoded
INFINITE if a blocking acquisition was used).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From joconnor at cybermesa.com  Sun Feb  4 15:19:04 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 14:19:04 +0000
Subject: While everyone is saying what they want in Python :)
Message-ID: <3A7D64D8.74E71826@cybermesa.com>

Smalltalk-style cascade operations would be *very* cool

win = GtkWindow();
	set_title("Hello World");
	set_name ("window");
	set_usize (400,200)

versus

win = GtkWindow()
win.set_title("Hello World")
win.set_name ("window")
win.set_usize (400,200)

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From peter at engcorp.com  Sun Feb  4 15:40:31 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 04 Feb 2001 09:40:31 -0500
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com>
Message-ID: <3A7D69DF.99352687@engcorp.com>

Jay O'Connor wrote:
> 
> Smalltalk-style cascade operations would be *very* cool
> 
> win = GtkWindow();
>         set_title("Hello World");
>         set_name ("window");
>         set_usize (400,200)
> 
> versus
> 
> win = GtkWindow()
> win.set_title("Hello World")
> win.set_name ("window")
> win.set_usize (400,200)

Why?  'Cause it saves a bit of typing?


From dan at eevolved.com  Mon Feb  5 01:50:41 2001
From: dan at eevolved.com (Dan Parisien)
Date: Mon, 05 Feb 2001 00:50:41 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D69DF.99352687@engcorp.com>
Message-ID: <BNmf6.53327$Pm2.1068813@news20.bellglobal.com>

Peter Hansen wrote:

> Jay O'Connor wrote:
> > 
> > Smalltalk-style cascade operations would be *very* cool
> > 
> > win = GtkWindow();
> >         set_title("Hello World");
> >         set_name ("window");
> >         set_usize (400,200)
> > 
> > versus
> > 
> > win = GtkWindow()
> > win.set_title("Hello World")
> > win.set_name ("window")
> > win.set_usize (400,200)
> 
> Why?  'Cause it saves a bit of typing?

YES!!!

That is not a good enough reason not to add a feature like this :)

Dan




From peter at engcorp.com  Mon Feb  5 06:10:39 2001
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 05 Feb 2001 00:10:39 -0500
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D69DF.99352687@engcorp.com> <BNmf6.53327$Pm2.1068813@news20.bellglobal.com>
Message-ID: <3A7E35CF.65F1C91F@engcorp.com>

Dan Parisien wrote:
> 
> Peter Hansen wrote:
> 
> > Jay O'Connor wrote:
> > >
> > > Smalltalk-style cascade operations would be *very* cool
> > >
[snip]
> >
> > Why?  'Cause it saves a bit of typing?
> 
> YES!!!
> 
> That is not a good enough reason not to add a feature like this :)

Absolutely not.  Look at http://www.pauahtun.org/TYPython/PythonWay.html
for some concise reasons why (scroll to the bottom).

As Erik Max wrote in another response, clarity of code is more 
important than quantity of keystrokes.


From nickm at mit.edu  Mon Feb  5 18:03:02 2001
From: nickm at mit.edu (Nick Mathewson)
Date: Mon, 05 Feb 2001 17:03:02 GMT
Subject: Cut-rate cascades [was Re: While everyone is saying what they want in Python :)]
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D69DF.99352687@engcorp.com> <BNmf6.53327$Pm2.1068813@news20.bellglobal.com> <3A7E35CF.65F1C91F@engcorp.com>
Message-ID: <a1Bf6.12085$fa2.1028297@typhoon.ne.mediaone.net>

On Mon, 05 Feb 2001 00:10:39 -0500, Peter Hansen <peter at engcorp.com> wrote:
>Dan Parisien wrote:
>> 
>> Peter Hansen wrote:
>> 
>> > Jay O'Connor wrote:
>> > >
>> > > Smalltalk-style cascade operations would be *very* cool
>> > >
>[snip]
>> >
>> > Why?  'Cause it saves a bit of typing?
>> 
>> YES!!!
>> 
>> That is not a good enough reason not to add a feature like this :)
>
>Absolutely not.  Look at http://www.pauahtun.org/TYPython/PythonWay.html
>for some concise reasons why (scroll to the bottom).
>
>As Erik Max wrote in another response, clarity of code is more 
>important than quantity of keystrokes.

OTOH, if you really want cascades, you can do something like this:

from types import StringType

def with1(obj, *commands):
    for command in commands:
	if type(command) == StringType:
	    attribute, arguments = command, ()
	else:
	    attribute, arguments = command[0], command[1:]
	getattr(obj, attribute).(*arguments)
    
    return obj

win = with1(GtkWindow(),
	    ('connect', "destroy", mainquit),
	    ('set_name', "window"),
	    ('set_title', "My Window"),
	    ('set_location', 100, 100),
	    ('add', with1(GtkHBox(),
			  ('add', with1(GtkLabel("Hi There"),
					 ("show"))),
			  ('add', with1(GtkButton("Push Me"),
					("connect","clicked",buttonClicked))),
			  ('show'))),
	    ('show'))

===================================

Or, if you absolutely _must_ evaluate all the withs in order:

class lazy:
    pass

class lazywith(lazy):
    def __init__(self, initializer, *commands):
	self.initializer = initializer
	self.commands = commands

    def __call__(self):
	if type(self.initializer) == ClassType:
	    obj = self.initializer()
	else:
	    obj = self.initializer[0](initializer[1:])

	for command in commands:
	    if type(command) == StringType:
		attribute, arguments = command, ()
	    else:
		attribute, arguments = command[0], command[1:]
	
	    arguments = [ isinstance(arg, lazy) and arg() or arg 
		          for arg in arguments ]
			  
	    getattr(obj, attribute).(*arguments)

	return obj
	

win = lazywith(GtkWindow,
	       ('connect', "destroy", mainquit),
	       ('set_name', "window"),
	       ('set_title', "My Window"),
	       ('set_location', 100, 100),
	       ('add', lazywith(GtkHBox,
				('add', lazywith( (GtkLabel, "Hi There"),
						  ("show"))),
				('add', lazywith( (GtkButton, "Push Me"),
					("connect","clicked",buttonClicked))),
			  ('show'))),
	       ('show')) ()

===================================

Or, if you think a pretty syntax is important:


class _LazyCall(lazy):
    def __init__(self,func, *args, **kwargs):
	self.func = func
	self.args = args
	self.kwargs = kwarfs

    def __call__(self, target):
	func = getattr(target, self.func)
	args = []
	for arg in self.args:
	    if isinstance(arg, lazy):
		args.append(arg())
	    else:
		args.append(arg)
	for key, val in self.kwargs.items():
	    if isinstance(val, lazy):
		self.kwargs[key] = val()

	return func(*args, **self.kwargs)

class WithFunction
    def __init__(self,func):
	self.func = func
    def __call__(self,*args,**kwargs):
	return _LazyCall(self.func, *args, **kwargs)

class WithWrapper:
    def __getattr__(self, attr):
	return WithFunction(attr)
    
class with3(lazy):
    def __init__(self, initializer, *commands):
	self.initializer = initializer
	self.commands = commands

    def __call__(self):
	if isinstance(self.initializer, lazy):
	    obj = self.initializer()
	else:
	    obj = self.initializer

	for command in commands:
	    assert isinstance(command, _LazyCall)
	    command(obj)
	
	return obj

W = WithWrapper()
  
win = with3(W.GtkWindow(),
	    W.connect("destroy", mainquit),
	    W.set_name("window"),
	    W.set_title("My Window"),
	    W.set_location(100,100),
	    W.add( with3( W.GtkHBox(),
			  W.add( with3( W.GtkLabel("Hi There"),
					W.show())),
			  W.add( with3( W.GtkButton("Push Me"),
					W.connect("clicked", buttonClicked),
					W.show())),
			  W.show())),
	    W.show()) ()


===================================

Rewrite-with-nested-scopes-for-extra-credit-ly Y'rs,

-- 
Nick Mathewson     <nickm at alum dot mit dot edu> 


From danielk at aracnet.com  Sun Feb  4 17:20:24 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 04 Feb 2001 08:20:24 -0800
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com>
Message-ID: <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com>

On Sun, 04 Feb 2001 14:19:04 +0000, Jay O'Connor <joconnor at cybermesa.com>
wrote:

>Smalltalk-style cascade operations would be *very* cool
>
>win = GtkWindow();
>	set_title("Hello World");
>	set_name ("window");
>	set_usize (400,200)
>
>versus
>
>win = GtkWindow()
>win.set_title("Hello World")
>win.set_name ("window")
>win.set_usize (400,200)


That's cos Smalltalk returns 'self' by default when a there is no explicit
return value. This can be done in Python if your 'set' methods return 'self'
instead of 'None'. For example:

>>> class A:
	def m1(self, t1):
		self.t1 = t1
		return self
	def m2(self, t2):
		self.t2 = t2
		return self
	def m3(self, h, w):
		self.h = h
		self.w = w
		return self
	
>>> a = A().m1("Hello World").m2("window").m3(400,200)
>>> a.h
400

pythonic-ly yr's,

Daniel Klein
Portland OR USA



From aahz at panix.com  Sun Feb  4 18:16:36 2001
From: aahz at panix.com (Aahz Maruch)
Date: 4 Feb 2001 09:16:36 -0800
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com>
Message-ID: <95k2pk$fn9$1@panix3.panix.com>

In article <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv at 4ax.com>,
Daniel Klein  <danielk at aracnet.com> wrote:
>
>>>> a = A().m1("Hello World").m2("window").m3(400,200)

Oh, that is s[l]ick.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From warlock at eskimo.com  Sun Feb  4 19:44:07 2001
From: warlock at eskimo.com (Jim Richardson)
Date: Sun, 4 Feb 2001 10:44:07 -0800
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com>  <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com>
Message-ID: <slrn97r8nn.gp.warlock@gargoyle.myth>

On Sun, 04 Feb 2001 08:20:24 -0800, 
 Daniel Klein, in the persona of <danielk at aracnet.com>,
 brought forth the following words...:

>On Sun, 04 Feb 2001 14:19:04 +0000, Jay O'Connor <joconnor at cybermesa.com>
>
>That's cos Smalltalk returns 'self' by default when a there is no explicit
>return value. This can be done in Python if your 'set' methods return 'self'
>instead of 'None'. For example:

can someone explain to me what exactly "self" is for? I just don't get it.
(I am trying to learn this language, but this puzzles me.) Every time I think I
get it, I find proof otherwise :)


-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From mwh21 at cam.ac.uk  Sun Feb  4 19:26:51 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Feb 2001 18:26:51 +0000
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com> <slrn97r8nn.gp.warlock@gargoyle.myth>
Message-ID: <m3lmrm4844.fsf@atrus.jesus.cam.ac.uk>

warlock at eskimo.com (Jim Richardson) writes:

> On Sun, 04 Feb 2001 08:20:24 -0800, 
>  Daniel Klein, in the persona of <danielk at aracnet.com>,
>  brought forth the following words...:
> 
> >On Sun, 04 Feb 2001 14:19:04 +0000, Jay O'Connor <joconnor at cybermesa.com>
> >
> >That's cos Smalltalk returns 'self' by default when a there is no explicit
> >return value. This can be done in Python if your 'set' methods return 'self'
> >instead of 'None'. For example:
> 
> can someone explain to me what exactly "self" is for? I just don't
> get it.  (I am trying to learn this language, but this puzzles me.)
> Every time I think I get it, I find proof otherwise :)

Well, it's so methods can refer to the object they are invoked on.  If
they didn't have a way of doing that, there wouldn't be much point in
them being methods would there?

Consider:

class C:
    def __init__(i):
        m_j = i
        if m_j > 0:
           m_i = m_j
        else:
           m_i = -m_j
        
is "m_j" an instance or a local variable?

The point is that Python needs some way to distinguish instance &
local variables.  You could have some special attribute like:

class C:
    __attrs__ = ['m_a']
    def __init__(i):
        m_a = i

but this isn't very Pythonic (declarations aren't, in general).

It's just a difference from C++ and Java.  Get used to it.

If I haven't covered your reason for "not getting it", please feel
free to be more specific.

Cheers,
M.

-- 
  Now this is what I don't get.  Nobody said absolutely anything 
  bad about anything.  Yet it is always possible to just pull 
  random flames out of ones ass.
         -- http://www.advogato.org/person/vicious/diary.html?start=60


From aahz at panix.com  Sun Feb  4 19:31:11 2001
From: aahz at panix.com (Aahz Maruch)
Date: 4 Feb 2001 10:31:11 -0800
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com> <slrn97r8nn.gp.warlock@gargoyle.myth>
Message-ID: <95k75f$oqb$1@panix3.panix.com>

In article <slrn97r8nn.gp.warlock at gargoyle.myth>,
Jim Richardson <warlock at eskimo.com> wrote:
>
>can someone explain to me what exactly "self" is for? I just don't get
>it.  (I am trying to learn this language, but this puzzles me.) Every
>time I think I get it, I find proof otherwise :)

Let's start by considering a class:

class foo:
    pass

To create an instance of the class we call it like a function:

a = foo()
b = foo()

We now have two instances of the class.  Normally classes have functions
associated with them (called methods):

class foo:
    def __init__(self, bar):
        self.bar = bar
    def print(self):
        x = type(self.bar)
        print self.bar, "is type", x

In order to create instances of this new class, we do this:

a = foo(5)
b = foo('Hi there!')

In order to call the print() method of foo, we do this:

a.print()
b.print()

Now, inside the print() method, how is Python supposed to distinguish
between a variable that belongs to the instance versus one that is
strictly local to the method (as x is)?
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From m.faassen at vet.uu.nl  Wed Feb  7 00:34:36 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 6 Feb 2001 23:34:36 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com>  <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com> <slrn97r8nn.gp.warlock@gargoyle.myth>
Message-ID: <95q1mc$hji$3@newshost.accu.uu.nl>

Jim Richardson <warlock at eskimo.com> wrote:
> On Sun, 04 Feb 2001 08:20:24 -0800, 
>  Daniel Klein, in the persona of <danielk at aracnet.com>,
>  brought forth the following words...:

>>On Sun, 04 Feb 2001 14:19:04 +0000, Jay O'Connor <joconnor at cybermesa.com>
>>
>>That's cos Smalltalk returns 'self' by default when a there is no explicit
>>return value. This can be done in Python if your 'set' methods return 'self'
>>instead of 'None'. For example:

> can someone explain to me what exactly "self" is for? I just don't get it.
> (I am trying to learn this language, but this puzzles me.) Every time I think I
> get it, I find proof otherwise :)

I'll add another answer to the list of answers given.

Let's first consider the case not using self at all. We have a class
that's just empty (you can see it as a dynamic structure or as a
type of dictionary; it is in fact a fancy dictionary):

class Record:
    pass

# let's make an initialization function
def record_init(record, name, street, city, country):
    record.name = name
    record.street = street
    record.city = city
    record.country = country

# let's make a function that prints the record
def record_print(record):
    print record.name, record.street, record.city, record.country

# and a function that can compare two records
def record_is_same(record, other_record):
    if (record.name == other_record.name and
        record.street == other_record.street and
        record.city == other_record.city and
        record.country = other_record.country):
        return 1
    else:
        return 0

You can imagine a whole host of functions that may do stuff with your
records.

Let's use them:

record = Record() # create a record
record_init(record, "Mr. Foo", "FooStreet 1", "Fooville", "Foogistan")
record_print(record)
# if record equal to itself?
if record_is_same(record, record):
    print "It's the same!"

Now let's write down the same, but using 'self' and methods instead.
Note that it's very much the same!

class Record:
    def __init__(self, name, street, city, country):
        self.name = name
        self.street = street
        self.city = city
        self.country = country

    def myprint(self):
        print self.name, self.street, self.city, self.country

    def is_same(self, other):
        if (self.name == other.name and
            self.street == other.street and
            self.city == other.city and
            self.country = other.country):
            return 1
        else:
            return 0

Let's use this:

record = Record("Bar", "44th Bar Street", "Bartown", "Baronia")
record.myprint()
record.is_same(record)

As you can see, the differences are minimal. We're using 'self' instead of
'record' (this is just a convention). We're using methods to the class
instead of functions. This makes it very easy make a slightly different
type of record later, which has a different print function, for instance,
but can still be used by the same code as nothing else is different.
That's a very powerful OO concept, called 'polymorphism'. Classes also
can inherit from each other, making polymorphism easier to accomplish
(as the methods inherited are automatically shared unless you define them
in some other way). These kind of things are somewhat harder to accomplish
by functions alone.

Another difference is that we use __init__ as a special method name; Python
knows to call this when we're constructing an instance of the class
(with Record(...)).

The 'self' is explicit as opposed to what's usual in C++ and Java because
it's clearer this way to see what's going on; it's always clear when
you're calling a method on a class instead of calling some function or
using a local variable.

I hope this helps.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From joconnor at cybermesa.com  Sun Feb  4 20:59:03 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 19:59:03 +0000
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com>
Message-ID: <3A7DB487.20A27050@cybermesa.com>

Daniel Klein wrote:

> That's cos Smalltalk returns 'self' by default when a there is no explicit
> return value. 

That is true, but also not how cascading works in Smalltalk.  When using
a cascaded message send, Smalltalk will send all messages to the single
receivng object regardless of the return value of the message.  The
classic example is #add:  which returns the new object being added; but
which is a very popular use of cascading.

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From max at alcyone.com  Sun Feb  4 19:14:53 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 04 Feb 2001 10:14:53 -0800
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com>
Message-ID: <3A7D9C1D.781F8E5@alcyone.com>

Jay O'Connor wrote:
> 
> Smalltalk-style cascade operations would be *very* cool
> 
> win = GtkWindow();
>         set_title("Hello World");
>         set_name ("window");
>         set_usize (400,200)
> 
> versus
> 
> win = GtkWindow()
> win.set_title("Hello World")
> win.set_name ("window")
> win.set_usize (400,200)

This general type of functionality (compare to the `with' keyword in
Pascal/Delphi) is generally a lot more trouble than it works.  It saves
a bit of typing, but sacrifices it for the sake of obscurity and
occasionally ambiguity when there are name collisions.

The amount of typing required should never be the primary concern in
desiging a language; clarity of what you're typing is much more
important than how many keys you physically have to hit.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I never think of the future.  It comes soon enough.
\__/ Albert Einstein
    Official Buh rules / http://www.alcyone.com/max/projects/cards/buh/
 The official rules to the betting card game, Buh.


From joconnor at cybermesa.com  Sun Feb  4 20:45:24 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 19:45:24 +0000
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com>
Message-ID: <3A7DB154.87E04245@cybermesa.com>

Erik Max Francis wrote:

> This general type of functionality (compare to the `with' keyword in
> Pascal/Delphi) 

I'm not familiar with the 'with' keyword in Delphi.

> The amount of typing required should never be the primary concern in
> desiging a language; clarity of what you're typing is much more
> important than how many keys you physically have to hit.

Actually, this construct in Smalltalk adds clarity because you don't
have the redundancy of the same object name repeated so the whitespace
helps clarify which objects are receiving which messages.

Plus it allows building complex structures without building needing to
create temporary variables to hold them.  Such as... (this is quasi
Python with  what I have in mind based on Smalltalk)

	win = GtKWindow();
		connect ("destroy", mainquit);
		set_name("window");
		set_title("My Window");
		set_location (100,100);
		add (GtkHBox();
			add (GtkLabel ("Hi There");
				show());
			add (GtkButton ("Push Me");
				connect ("clicked", buttonClicked);
				show());
			show());
		show()

Here I've created four widgets without needing extra temp variables. 
More importantly, the whitespace indicates the exact structural
relationship of the widgets.  I find this much easier to read than the
normal way of doing it.

	win = GtkWindow()
	
	win.connect("destroy", mainquit)
	win.set_name ("window")	
	win.set_title ("My Window")
	win.show()
	box = GtkHBox ()
	label = GtkLabel("Hi There")
	button = GtkButton ("Hi there")
	box.add (label)
	box.add (button)
	win.add (box)
	box.show()
	label.show()
	button.show()
	button.connect ("clicked", buttonClicked)
	win.show()


Take care,

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From joconnor at cybermesa.com  Sun Feb  4 21:18:43 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 20:18:43 +0000
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com>
Message-ID: <3A7DB923.AD330F7E@cybermesa.com>

Jay O'Connor wrote:

Actually, it's not *that* big a deal.  There seemed to be a lot of
traffic about what people like or dislike about 2.x, so I thought I'd
throw out an idea of something I liked about Smalltalk that I hadn't
seen in other languages.

I can live without it :) (do I have a choice?)

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From cjc26 at nospam.cornell.edu  Tue Feb  6 02:34:47 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Tue, 06 Feb 2001 01:34:47 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <3A7DB923.AD330F7E@cybermesa.com>
Message-ID: <XwIf6.63152$o91.6456979@typhoon.nyroc.rr.com>

* Jay O'Connor <joconnor at cybermesa.com> menulis:
| Jay O'Connor wrote:
| 
| Actually, it's not *that* big a deal.  There seemed to be a lot of
| traffic about what people like or dislike about 2.x, so I thought I'd
| throw out an idea of something I liked about Smalltalk that I hadn't
| seen in other languages.
| 
| I can live without it :) (do I have a choice?)

You could try writing a PEP for it.  I would mention the Tkinter example
in it, that gives it a slight chance of being considered at least, since
the inconvenience of using Tkinter was the impetus for adding a previous
change to the language (namely, keyword arguments).


-- 
Cliff Crawford               http://www.people.cornell.edu/pages/cjc26/
                             print "Just another Python hacker"


From tuttledon at hotmail.com  Tue Feb  6 02:35:02 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Tue, 06 Feb 2001 01:35:02 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com>
Message-ID: <axIf6.35318$p8.7719400@typhoon.southeast.rr.com>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:3A7DB154.87E04245 at cybermesa.com...
> I'm not familiar with the 'with' keyword in Delphi.

Here's a vbscript example.
With MyLabel
   .Height = 2000
   .Width = 2000
   .Caption = "This is MyLabel"
End With
Personally, I'd love to see Python include the 'with' keyword! It improves
readability and should execute a bit faster as well.
Don






From ullrich at math.okstate.edu  Tue Feb  6 16:17:05 2001
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Tue, 06 Feb 2001 15:17:05 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com>
Message-ID: <3a8012f5.7625600@nntp.sprynet.com>

On Tue, 06 Feb 2001 01:35:02 GMT, "Don Tuttle" <tuttledon at hotmail.com>
wrote:

>
>"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
>news:3A7DB154.87E04245 at cybermesa.com...
>> I'm not familiar with the 'with' keyword in Delphi.
>
>Here's a vbscript example.
>With MyLabel
>   .Height = 2000
>   .Width = 2000
>   .Caption = "This is MyLabel"
>End With
>Personally, I'd love to see Python include the 'with' keyword! It improves
>readability and should execute a bit faster as well.

Delphi's "with" is more or less the same except without the
dots at the start of the field names. For a long time I've
seen Delphi people say that with is bad, the reason
being it leads to bugs

with AnObject do
  begin
    OneThing
    AnotherThing
  end

when AnObject does not have a OneThing method
but there _is_ a OneThing in the surrounding scope.

For years I thought this was just stupid - you should
know what sort of object you're dealing with. The 
Python slogans about how explicit is better than
implicit just about have me convinced that 
people are right, with is bad.

Right or not, it certainly seems contrary to the
"explciit is better than implicit" philosophy.

>Don
>
>
>
>



From joconnor at cybermesa.com  Tue Feb  6 16:44:06 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 06 Feb 2001 15:44:06 +0000
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com>
Message-ID: <3A801BC6.DF764D5E@cybermesa.com>

"David C. Ullrich" wrote:
> 
> On Tue, 06 Feb 2001 01:35:02 GMT, "Don Tuttle" <tuttledon at hotmail.com>
> wrote:
> 
> >
> >"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
> >news:3A7DB154.87E04245 at cybermesa.com...
> >> I'm not familiar with the 'with' keyword in Delphi.
> >
> >Here's a vbscript example.
> >With MyLabel
> >   .Height = 2000
> >   .Width = 2000
> >   .Caption = "This is MyLabel"
> >End With
> >Personally, I'd love to see Python include the 'with' keyword! It improves
> >readability and should execute a bit faster as well.
> 
> Delphi's "with" is more or less the same except without the
> dots at the start of the field names. For a long time I've
> seen Delphi people say that with is bad, the reason
> being it leads to bugs
> 
> with AnObject do
>   begin
>     OneThing
>     AnotherThing
>   end
> 


Eeewwww....that's ugly :)

Smalltalk cascading really isn't the same.  "with" seems to be
eastablishing a sort of local scope f reference.  Smalltalk cascading
just says "send this message to the same object you sent the last
message."

One point that may have gotten missed.  Smalltalk doesn't use ";" as a
delimiter between statements, it uses "." as a delimiter and a ";" to
indicate a cascade.  So the code

	anObject
		messageOne;
		messageTwo.

It's syntactically obvious that messageTwo is sent to anObject; there
can be no other confusion.

> when AnObject does not have a OneThing method
> but there _is_ a OneThing in the surrounding scope.

Oh, yes...I could see that being a big issue.  Smalltalk doesn't have
that problem because a) the syntax is explicit that you are cascading a
message send and b)  Smalltalk doesn't have the concept of functions not
bound to objects so you don't run into a problem of "this method doesn't
apply in this scope, let me try an outer scope".

> For years I thought this was just stupid - you should
> know what sort of object you're dealing with. The
> Python slogans about how explicit is better than
> implicit just about have me convinced that
> people are right, with is bad.

I agree.  Explicit is better than implicit (I still have Perl
nightmares...), and the Pascal 'with' statement seems wrought with
potential disaster.  Fortunately, the simularity between 'with' and
Smalltalk cascading is only superficial, at best.

Take care,

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discusssion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython


From tuttledon at hotmail.com  Tue Feb  6 16:57:20 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Tue, 06 Feb 2001 15:57:20 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com>
Message-ID: <A9Vf6.38067$p8.8472953@typhoon.southeast.rr.com>

"David C. Ullrich" <ullrich at math.okstate.edu> wrote in message
news:3a8012f5.7625600 at nntp.sprynet.com...
> Delphi's "with" is more or less the same < as vbscript> except without the
> dots at the start of the field names. For a long time I've
> seen Delphi people say that with is bad, the reason
> being it leads to bugs
>
> with AnObject do
>   begin
>     OneThing
>     AnotherThing
>   end
>
> when AnObject does not have a OneThing method
> but there _is_ a OneThing in the surrounding scope.

I'm not familar with Delphi, but I suspect this is why vbscript uses the
leading 'dot' notation.

> Right or not, it certainly seems contrary to the
> "explciit is better than implicit" philosophy.

Perhaps I'm missing something but using a keyword to define a single block
of code seems very explicit (ie Select Case). Again, maybe the leading 'dot'
makes all the difference.

Don

> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

> >Here's a vbscript example.
>>
> >With MyLabel
> >     .Height = 2000
> >     .Width = 2000
> >     .Caption = "This is MyLabel"
> >End With
>>
> >Personally, I'd love to see Python include the 'with' keyword! It
improves
> >readability and should execute a bit faster as well.






From ullrich at math.okstate.edu  Tue Feb  6 22:51:19 2001
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Tue, 06 Feb 2001 21:51:19 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com> <A9Vf6.38067$p8.8472953@typhoon.southeast.rr.com>
Message-ID: <95prki$tt4$1@nnrp1.deja.com>

In article <A9Vf6.38067$p8.8472953 at typhoon.southeast.rr.com>,
  "Don Tuttle" <tuttledon at hotmail.com> wrote:
> "David C. Ullrich" <ullrich at math.okstate.edu> wrote in message
> news:3a8012f5.7625600 at nntp.sprynet.com...
> > Delphi's "with" is more or less the same < as vbscript> except
without the
> > dots at the start of the field names. For a long time I've
> > seen Delphi people say that with is bad, the reason
> > being it leads to bugs
> >
> > with AnObject do
> >   begin
> >     OneThing
> >     AnotherThing
> >   end
> >
> > when AnObject does not have a OneThing method
> > but there _is_ a OneThing in the surrounding scope.
>
> I'm not familar with Delphi, but I suspect this is why vbscript uses
the
> leading 'dot' notation.

No doubt.

> > Right or not, it certainly seems contrary to the
> > "explciit is better than implicit" philosophy.
>
> Perhaps I'm missing something but using a keyword to define a single
block
> of code seems very explicit (ie Select Case). Again, maybe the leading
'dot'
> makes all the difference.

I didn't mean to get into off-topic heresy, but I don't know what
you mean by calling this "explicit". The problem is that the methods
could be methods of AnObject or they could have nothing to do with
AnObject. As opposed to an explicit AnObject,OneThing.

> Don
>
> > >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>
> > >Here's a vbscript example.
> >>
> > >With MyLabel
> > >     .Height = 2000
> > >     .Width = 2000
> > >     .Caption = "This is MyLabel"
> > >End With
> >>
> > >Personally, I'd love to see Python include the 'with' keyword! It
> improves
> > >readability and should execute a bit faster as well.
>
>

--
Oh, dejanews lets you add a sig - that's useful...


Sent via Deja.com
http://www.deja.com/


From daniel.dittmar at sap.com  Tue Feb  6 17:46:15 2001
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Tue, 6 Feb 2001 17:46:15 +0100
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com>
Message-ID: <95p9om$h23$1@news1.wdf.sap-ag.de>

In VB, although WITH can be nested, only the innermost is valid. Together
with the .dot notation makes your intentions pretty explicit. In Pascal, you
have
- nested with
- multiple expressions allowed in the with clause
- no dots
which leads to very messy code.

Although cascading in Smalltalk and with in VB are very different, in most
cases, they lead to similar code (Pythonified):

sys.stdout
    .write (...);
    .write (...);
    .write (...)

with sys.stdout:
    .write (...);
    .write (...);
    .write (...)

The main exception is object creation, where in Smalltalk, you can create an
object and send it a bunch of cascaded initialization messages.

One problem with cascading is that I don't know yet of a syntax using
indentation. And using ';' as a delimiter? There are already far too many
people writing 'if <cond>: thenstmt', all in one line.

Daniel

--
Daniel Dittmar
daniel.dittmar at sap.com
SAP DB, SAP Labs Berlin
http://www.sapdb.org/





From Daniel.Kinnaer at Advalvas.be  Tue Feb  6 22:48:19 2001
From: Daniel.Kinnaer at Advalvas.be (Daniel)
Date: Tue, 06 Feb 2001 21:48:19 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com>
Message-ID: <3a806c08.26417386@news.skynet.be>

On Tue, 06 Feb 2001 15:17:05 GMT, ullrich at math.okstate.edu (David C.
Ullrich) wrote:

>
>Delphi's "with" is more or less the same except without the
>dots at the start of the field names. For a long time I've
>seen Delphi people say that with is bad, the reason
>being it leads to bugs

I've been programming in Delphi for quite some years now (only
"discovered" Python a few months ago), and  I'd like to react on this,
though.

The part where I find the 'with' command interesting is for
readability.  Consider : 

MySpecialButton := TSpecialButton;
MySpecialButton.caption := 'Busy';
MySpecialButton.method2;
MySpecialButton.caption := "Ok";

or

MySpecialButton := TSpecialButton;
with MySpecialButton do
begin
   caption := 'Busy';
   method2;
   caption := 'Done';
end;

Which code would you prefer?

Perhaps that there is a parent or child of this  class which also
knows method2 (and perhaps is overrridden).  If I don't specify, the
Delphi-compiler will either complain or weird things will start to
happen :)
E.g.  table1.close (to close a table) and Form1.Close (to close the
program). If I would just use 'Close' there is a big chance the
program just terminates (instead of closing the table :)

Just my 2cents...

groetjes,

Daniel



From fredrik at effbot.org  Tue Feb  6 23:02:44 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 06 Feb 2001 22:02:44 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com> <3a806c08.26417386@news.skynet.be>
Message-ID: <8w_f6.13141$AH6.1969901@newsc.telia.net>

Daniel Kinnaer wrote:
> The part where I find the 'with' command interesting is for
> readability.  Consider :
>
> MySpecialButton := TSpecialButton;
> MySpecialButton.caption := 'Busy';
> MySpecialButton.method2;
> MySpecialButton.caption := "Ok";
>
> or
>
> MySpecialButton := TSpecialButton;
> with MySpecialButton do
> begin
>    caption := 'Busy';
>    method2;
>    caption := 'Done';
> end;
>
> Which code would you prefer?

b = mybutton = TSpecialButton()
b.caption = "Busy"
b.method2()
b.caption = "Ok"

(variables are just names in Python -- you can have any
number of names pointing to the same thing...)

Cheers /F




From ullrich at math.okstate.edu  Wed Feb  7 16:43:10 2001
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Wed, 07 Feb 2001 15:43:10 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com> <3a806c08.26417386@news.skynet.be> <8w_f6.13141$AH6.1969901@newsc.telia.net>
Message-ID: <95rqec$hot$1@nnrp1.deja.com>

In article <8w_f6.13141$AH6.1969901 at newsc.telia.net>,
  "Fredrik Lundh" <fredrik at effbot.org> wrote:
> Daniel Kinnaer wrote:
> > The part where I find the 'with' command interesting is for
> > readability.  Consider :
> >
> > MySpecialButton := TSpecialButton;
> > MySpecialButton.caption := 'Busy';
> > MySpecialButton.method2;
> > MySpecialButton.caption := "Ok";
> >
> > or
> >
> > MySpecialButton := TSpecialButton;
> > with MySpecialButton do
> > begin
> >    caption := 'Busy';
> >    method2;
> >    caption := 'Done';
> > end;
> >
> > Which code would you prefer?
>
> b = mybutton = TSpecialButton()
> b.caption = "Busy"
> b.method2()
> b.caption = "Ok"
>
> (variables are just names in Python -- you can have any
> number of names pointing to the same thing...)

You can do more or less the same thing in Pascal. Doesn't
have the same effect of reducing the amount of typing while
keeping things explicit, because (at least in Delphi) the
syntax would be

NewNameFortheButtonThatIInventedToSaveKeystrokes:= mybutton;
NewNameFortheButtonThatIInventedToSaveKeystrokes.caption:= 'Busy';

> Cheers /F
>
>

--
Oh, dejanews lets you add a sig - that's useful...


Sent via Deja.com
http://www.deja.com/


From sdm7g at virginia.edu  Wed Feb  7 00:48:40 2001
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 6 Feb 2001 18:48:40 -0500 (EST)
Subject: While everyone is saying what they want in Python :)
In-Reply-To: <3a806c08.26417386@news.skynet.be>
Message-ID: <Pine.NXT.4.21.0102061842480.306-100000@localhost.virginia.edu>


On Tue, 6 Feb 2001, Daniel wrote:

> The part where I find the 'with' command interesting is for
> readability.  Consider : 
> 
> MySpecialButton := TSpecialButton;
> MySpecialButton.caption := 'Busy';
> MySpecialButton.method2;
> MySpecialButton.caption := "Ok";
> 
> or
> 
> MySpecialButton := TSpecialButton;
> with MySpecialButton do
> begin
>    caption := 'Busy';
>    method2;
>    caption := 'Done';
> end;
> 
> Which code would you prefer?

Having had to maintain Pascal code that used the 'with' construct,
I would always choose the former. 'With' is less readable because
it reduces locality of the code. You have to look at the data
type definition of TSpecialButton to understand what's going on
in that code. Try reading and maintaining code that has nested
withs 3 or more deep, with 20-60 fields for each struct, and 
I think you'll change your mind about it. 

It saves some typing but it reduces context, locality and readability. 

-- Steve Majewski

"Ontology recapitulates philology" -- W.V.Quine





From gzeljko at sezampro.yu  Wed Feb  7 02:31:50 2001
From: gzeljko at sezampro.yu (gzeljko)
Date: Wed, 7 Feb 2001 02:31:50 +0100
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com> <3a806c08.26417386@news.skynet.be>
Message-ID: <95q8l6$htb$1@neptun.beotel.net>

I agree,
but this construct can't go in Python
(what about assigment in this context ?)

Zeljko

Daniel <Daniel.Kinnaer at Advalvas.be> wrote in message
news:3a806c08.26417386 at news.skynet.be...
>
> The part where I find the 'with' command interesting is for
> readability.  Consider :
>
> MySpecialButton := TSpecialButton;
> MySpecialButton.caption := 'Busy';
> MySpecialButton.method2;
> MySpecialButton.caption := "Ok";
>
> or
>
> MySpecialButton := TSpecialButton;
> with MySpecialButton do
> begin
>    caption := 'Busy';
>    method2;
>    caption := 'Done';
> end;
>
> Which code would you prefer?
>





From joconnor at cybermesa.com  Wed Feb  7 05:30:54 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Wed, 07 Feb 2001 04:30:54 GMT
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com> <3a806c08.26417386@news.skynet.be> <95q8l6$htb$1@neptun.beotel.net>
Message-ID: <3a80ca91.21486373@news.cybermesa.com>

On Wed, 7 Feb 2001 02:31:50 +0100, "gzeljko" <gzeljko at sezampro.yu>
wrote:

>
>I agree,
>but this construct can't go in Python
>(what about assigment in this context ?)
>
>Zeljko

Well, I'm not sure about Pascal's 'with' clause (actually, I'm not too
interested in Pascal's 'with' clause.  I was thinking of Smalltalk's
cascading, which is a different concept)

In Smalltalk. the value of a cascaded set of message sends is whatever
the return was of the last messages (intermediate returns get
discarded)

thus in

win := ApplicationWindow new
	name: 'My Window';
	location: (100 @ 100) 

win will be equal to th ereturn value of #location:  Since #location:
is mostly likely an accessor with no explicit return, the default
return of 'self' will be return and win will equal whatever the
receiving object is (in this case, the newly created ApplicationWindow
object)

This gives rise to the rather strange use of #yourself, being a simple
method that returns self.  This is most often used for creating a new
collection with a cascaded approach to populating the collection.

collection := OrderedCollection new
	add: "Terry Gillian";
	add: "John Cleeese";
	add: "Kevin Kilne";
	yourself.

In Smalltalk, the collection method #add: returns the newly added
object.  So in the above case would make collection equal to the
String "Kevin Kline" and not to the newly created collection.
Therefore, #yourself is used as the last cascaded message which simply
returns 'self' and makes sure that collection is indeed set to the
newly created collection 


Take care,


Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum - http://pub1.ezboard.com/fobjectorienteddevelopmentpython

"God himself plays on the bass strings first, when he tunes the soul"


From erno-news at erno.iki.fi  Tue Feb  6 18:05:56 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 06 Feb 2001 19:05:56 +0200
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com>
Message-ID: <kuy9vj213f.fsf@lasipalatsi.fi>

In article <axIf6.35318$p8.7719400 at typhoon.southeast.rr.com>, "Don
Tuttle" <tuttledon at hotmail.com> writes:

| "Jay O'Connor" <joconnor at cybermesa.com> wrote in message
| news:3A7DB154.87E04245 at cybermesa.com...
|| I'm not familiar with the 'with' keyword in Delphi.

| Here's a vbscript example.
| With MyLabel
|    .Height = 2000
|    .Width = 2000
|    .Caption = "This is MyLabel"
| End With

>>> class foo:
...     pass
... 
>>> def with(obj, code):
...     exec code in obj.__dict__
...     del obj.__dict__['__builtins__'] # this gets added by exec
... 
>>> my_foo = foo()
>>> with(my_foo,  '''
... height = 2000
... width = 2000
... caption = "this is mylabel"
... ''')
>>> 
>>> print my_foo.caption
this is mylabel

;)

   -- erno


From jurgen.defurne at philips.com  Wed Feb  7 08:17:03 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Wed, 7 Feb 2001 08:17:03 +0100
Subject: While everyone is saying what they want in Python :)
Message-ID: <0056900015716907000002L072*@MHS>

In fact, 'with' is not an artefact of Delphi, but from the plain, original Pascal!

Jurgen




ullrich at math.okstate.edu@SMTP at python.org on 06/02/2001 16:42:59
Please respond to ullrich at math.okstate.edu@SMTP 
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	Re: While everyone is saying what they want in Python :)
Classification:	

On Tue, 06 Feb 2001 01:35:02 GMT, "Don Tuttle" <tuttledon at hotmail.com>
wrote:

>
>"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
>news:3A7DB154.87E04245 at cybermesa.com...
>> I'm not familiar with the 'with' keyword in Delphi.
>
>Here's a vbscript example.
>With MyLabel
>   .Height = 2000
>   .Width = 2000
>   .Caption = "This is MyLabel"
>End With
>Personally, I'd love to see Python include the 'with' keyword! It improves
>readability and should execute a bit faster as well.

Delphi's "with" is more or less the same except without the
dots at the start of the field names. For a long time I've
seen Delphi people say that with is bad, the reason
being it leads to bugs

with AnObject do
  begin
    OneThing
    AnotherThing
  end

when AnObject does not have a OneThing method
but there _is_ a OneThing in the surrounding scope.

For years I thought this was just stupid - you should
know what sort of object you're dealing with. The
Python slogans about how explicit is better than
implicit just about have me convinced that
people are right, with is bad.

Right or not, it certainly seems contrary to the
"explciit is better than implicit" philosophy.

>Don
>
>
>
>

--
http://mail.python.org/mailman/listinfo/python-list





From brucepm at pacbell.net  Sun Feb  4 15:25:43 2001
From: brucepm at pacbell.net (brucepm at pacbell.net)
Date: Sun, 04 Feb 2001 14:25:43 GMT
Subject: I WILL MAKE YOU RICH BEYOND YOUR DREAMS.  4144
Message-ID: <GDdf6.13581$GV6.674589@dfiatx1-snr1.gtei.net>

CONTACT ME FOR DETAILS ON GETTING FILTHY RICH

BRUCE P. MEYERS
10517 VALPARAISO STREET 
LOS ANGELES, CA 90034
PHONE:(310)202.9005
CELL:(310)213.1148
E-Mail:brucepm at pacbell.net

hmlwdqgcvnokqdxzslpmfpkhmdhqpfwqhmnmpcgeorfdkdjplwjqeiszhqbrmjvgchlwzbxjcqtlrseggjjlnzjyrpoqxdjlghyt



From phil at river-bank.demon.co.uk  Sun Feb  4 15:45:56 2001
From: phil at river-bank.demon.co.uk (phil at river-bank.demon.co.uk)
Date: Sun, 04 Feb 2001 14:45:56 -0000
Subject: ANNOUNCE: SIP v2.3 (Python Bindings Generator for C++ Class Libraries)
Message-ID: <95jpv4+dfnk@eGroups.com>

SIP v2.3 has been released and is available from
http://www.thekompany.com/projects/pykde/.

SIP does a similar job to SWIG but is specifically aimed at generating
Python bindings for C++ class libraries. SIP is used to generated PyQt
(the Python bindings for the Qt GUI toolkit).

Phil




From phil at river-bank.demon.co.uk  Sun Feb  4 15:50:51 2001
From: phil at river-bank.demon.co.uk (phil at river-bank.demon.co.uk)
Date: Sun, 04 Feb 2001 14:50:51 -0000
Subject: ANNOUNCE: PyQt v2.3 (Python Bindings for Qt GUI Toolkit)
Message-ID: <95jq8b+3tt4@eGroups.com>

PyQt v2.3 has been released and is available from
http://www.thekompany.com/projects/pykde/.

PyQt is a complete set of Python bindings for Trolltech's Qt GUI
toolkit. It implements over 200 classes and includes pyuic, a utility
for generating Python code for user interfaces created with Qt
Designer.

Phil




From lwss at china.com  Sun Feb  4 17:21:52 2001
From: lwss at china.com (oboy)
Date: Mon, 05 Feb 2001 00:21:52 +0800
Subject: Ask for help of VB
Message-ID: <3A7D81A0.9B6C2B4B@china.com>

I have read many book but i still don't know how to create a help file
Anyone know how to create a "Help file"(.hlp)?
thx



From stephen_purcell at yahoo.com  Sun Feb  4 17:23:56 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sun, 4 Feb 2001 17:23:56 +0100
Subject: [Announce] PyUnit 1.3 released
Message-ID: <20010204172355.C16150@freedom.puma-ag.com>

Version 1.3 of the PyUnit unit testing framework is out, for those who are
interested.

Quick summary for those not familiar with PyUnit:-

 - PyUnit is a Python port of Java's popular JUnit testing framework
 - It provides a framework for easily writing unit and functional tests
   for your Python code, for aggregating the tests to run en-masse, and
   for running tests in text or GUI mode
 - PyUnit has been around since 1999 and is used in many projects,
   including Zope.

Here's what's new in this release:-

 - Clearer and more verbose text output format
 - JPython and Python 2 compatibility
 - New FunctionTestCase class provides support for wrapping legacy test
   functions into PyUnit test case instances
 - Support for short descriptions of tests, taken from __doc__ strings
   by default
 - Updated and expanded documentation
 - Convenient new unittest.main() function for use by all test modules
 - Tests run in text mode can now be interrupted using ctrl-c

For further information and downloadable packages, visit:-

 - http://pyunit.sourceforge.net/

Happy testing!

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From stephen_purcell at yahoo.com  Sun Feb  4 17:23:56 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sun, 4 Feb 2001 17:23:56 +0100
Subject: [Announce] PyUnit 1.3 released
Message-ID: <mailman.981362542.20434.clpa-moderators@python.org>

Version 1.3 of the PyUnit unit testing framework is out, for those who are
interested.

Quick summary for those not familiar with PyUnit:-

 - PyUnit is a Python port of Java's popular JUnit testing framework
 - It provides a framework for easily writing unit and functional tests
   for your Python code, for aggregating the tests to run en-masse, and
   for running tests in text or GUI mode
 - PyUnit has been around since 1999 and is used in many projects,
   including Zope.

Here's what's new in this release:-

 - Clearer and more verbose text output format
 - JPython and Python 2 compatibility
 - New FunctionTestCase class provides support for wrapping legacy test
   functions into PyUnit test case instances
 - Support for short descriptions of tests, taken from __doc__ strings
   by default
 - Updated and expanded documentation
 - Convenient new unittest.main() function for use by all test modules
 - Tests run in text mode can now be interrupted using ctrl-c

For further information and downloadable packages, visit:-

 - http://pyunit.sourceforge.net/

Happy testing!

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright




From micheljonas at web.de  Sun Feb  4 17:42:14 2001
From: micheljonas at web.de (Michael Jonas)
Date: Sun, 4 Feb 2001 17:42:14 +0100
Subject: Win Development
Message-ID: <NDBBICIKFJMPFCCMBLNIEEINCDAA.micheljonas@web.de>

I'm going to start a bigger project which is primary based on a Win32
Environment. Cause I used python only on IRIX and Linux the last years, I'm
not used to the Win32 environment and I think I need some tips how to
organize the whole stuff.

What I did till here:
I installed the python20 binary dist to C:\PYTHON20 and put the source dist
to C:\PYTHON20SRC.
I compiled the Release and the Debug version. Now there are some conceptual
questions...

Should I copy the self_build (debug and release?) libs, dll and exe files to
the binary dist and do my Development there?!?
Or do I leave the stuff in the source dist to code my C Modules?!? And build
python Modules in the binary dist!?
Or should I develop my C-Modules in the source dist and put the result to
the binary Dist?!?
Or should I put the source into the root of my binary dist ?!?

I've allready seen that VC debugger (and the look at Python's source) is one
of my best friends and I want to make it find everything as required. ( I
used vi the last years...) So maybe there are some additional benefits using
VC. By the way... is there a way to do Python development in VC's
interface?!?
I could go on and try every solution myself, but there's a lack of
expirience with the Win32 Concept. On Linux I was used to build my C-Modules
in the source tree and get them installed using make. I was a little bit
confused that there's no installer which put's everything in place from the
source dist, or at least I did not find it.

Thanks in advance...

Michael

-----------------------------------------
Michael Jonas
Das Werk Digitale Bildverarbeitungs GmbH
mail:  mj at muc.das-werk.de
phone: +49(89)368148-700
fax:   +49(89)368148-111
-----------------------------------------




From MarkH at ActiveState.com  Sun Feb  4 23:05:20 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 04 Feb 2001 22:05:20 GMT
Subject: Win Development
References: <mailman.981305180.32424.python-list@python.org>
Message-ID: <3A7DD132.7050309@ActiveState.com>

Michael Jonas wrote:

> Should I copy the self_build (debug and release?) libs, dll and exe files to
> the binary dist and do my Development there?!?
> Or do I leave the stuff in the source dist to code my C Modules?!? And build
> python Modules in the binary dist!?
> Or should I develop my C-Modules in the source dist and put the result to
> the binary Dist?!?
> Or should I put the source into the root of my binary dist ?!?

What I do is to build everything I can from sources, and run them directly.  Thus, Python itself is always run from the CVS source directory, with the executables directly in the "PCBuild" directory.

If you run Python in this way, it doesnt need any PYTHONPATH setup at all for its core library.

This just leaves you with your extension modules.  Again, you should be able to setup to run directly from the source tree.  Hit the registry editor, and manually add the paths to your extensions to the PYTHONPATH.


> I've allready seen that VC debugger (and the look at Python's source) is one
> of my best friends and I want to make it find everything as required. ( I
> used vi the last years...) 

Doing what I descrribed keeps VC happy, and it never needs to ask where source-code is.


> VC. By the way... is there a way to do Python development in VC's
> interface?!?

Not yet, but ActiveState are working on "Visual Python" for the VC7 environment.

Mark.



From aleaxit at yahoo.com  Mon Feb  5 14:14:52 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 5 Feb 2001 14:14:52 +0100
Subject: Win Development
References: <mailman.981305180.32424.python-list@python.org>
Message-ID: <95m90f01do3@news1.newsguy.com>

"Michael Jonas" <micheljonas at web.de> wrote in message
news:mailman.981305180.32424.python-list at python.org...
    [snip]
> I've allready seen that VC debugger (and the look at Python's source) is
one
> of my best friends and I want to make it find everything as required. ( I
> used vi the last years...) So maybe there are some additional benefits
using

I see Hammond already gave precious advice, and just want to add two
cents -- VIM (at www.vim.org) is just great (on Windows as well as on
Unixy platforms) if you like VI, and, among many other abilities, it
can hook itself up as the editor to use within Visual Studio *AND* it
can be built to use Python as its editor scripting language.  Try it,
you'll love it (the gui-enabled version, GVIM, is what you'll need to
use for Visual Studio integration -- but you also get a console based
version, VIM, for those occasional editing needs right from a console
commandline...).


Alex





From Jan.Kybic at epfl.ch  Sun Feb  4 17:52:41 2001
From: Jan.Kybic at epfl.ch (Jan Kybic)
Date: 04 Feb 2001 17:52:41 +0100
Subject: Q: shelve ovewrites for different users
Message-ID: <lxk1ytev19i.fsf@ioasun3.epfl.ch>

Hello,
	I am using 'shelve' with Berkeley DB to store Numerical Python
arrays using a code like:

f=shelve.open('file.dat')
f['coefs']=zeros((10,20))
f.close()

for writing and

f=shelve.open('file.dat')
c=array(f['coefs'])
f.close()

for reading. I am sharing the shelve file between two users on Unix. My
problem is that when one writes it and the second one reads it, the
shelve file shrinks and does not contain anything any longer. Anybody
has seen this behaviour? Any hints? Thanks.

Jan

-- 
-------------------------------------------------------------------------
Jan Kybic <Jan.Kybic at epfl.ch>      BIG IOA DMT EPFL Lausanne, Switzerland
       or <kybic at ieee.org>         tel. work +41 (21) 693 5142
For PGP key see my WWW page.       http://bigwww.epfl.ch/kybic


From erno-news at erno.iki.fi  Sun Feb  4 19:46:37 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 04 Feb 2001 20:46:37 +0200
Subject: Q: shelve ovewrites for different users
References: <lxk1ytev19i.fsf@ioasun3.epfl.ch>
Message-ID: <kuzog22smq.fsf@lasipalatsi.fi>

In article <lxk1ytev19i.fsf at ioasun3.epfl.ch>, Jan Kybic
<Jan.Kybic at epfl.ch> writes:

| I am sharing the shelve file between two users on Unix.

are your two different users reading and writing it at the
same time? if so, problems are to be expected. you
need to either

1) set up locking manually, there are some functions
   in the fcntl module to do that, or

2) use gdbm for the shelf object (gdbm does locking itself): copy the
   DbfilenameShelf class from shelve.py and change "anydbm" to "gdbm".

  -- erno


From chris.gonnerman at usa.net  Sun Feb  4 17:57:31 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: 4 Feb 2001 10:57:31 CST
Subject: ANNOUNCE:  WConio
Message-ID: <20010204165731.23240.qmail@www0h.netaddress.usa.net>

WConio -- Windows CONsole I/O module for Python

Created 2/3/2001 by Chris Gonnerman chris.gonnerman at usa.net
based on code by Daniel Guerrero Miralles
This source code is public domain.

This module is based heavily on the TCCONIO package; the notes from the top of
the source are as follows:

    Old Turbo-C CONIO.H compatibility library for LCC-Win32
    and GCC/EGCS Mingw32 compilers.
    Version 1.0 (September 1999).
    Created by Daniel Guerrero Miralles (daguer at geocities.com).

Visit
    http://newcenturycomputers.net/projects/wconio

for more information or to download.



____________________________________________________________________
Get free email and a permanent address at http://www.netaddress.com/?N=1



From danielk at aracnet.com  Sun Feb  4 18:53:33 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 04 Feb 2001 09:53:33 -0800
Subject: Preventing direct access to a class
Message-ID: <dc4r7t41dchrbrdabst1ung9mop3u4dnoj@4ax.com>

I have a situation where I need to prevent direct access to a class. IOW, the
only way I wish a class to be instantiated is via another class. For example

>>> class A:
	pass

>>> class B:
	def foo(self):
		return A()
	
>>> b = B()
>>> a = b.foo()

...and there might be other classes that 'return A()'.

I do not want the user to be allowed to do this...

>>>a = A()

It doesn't look like classes can be made 'private to the module' as I tried to
define class A as 'class __A'. The user could still access with a=__A() (no
mangling needed).

Other than placing some convoluted logic in each class, is there a way to
accomplish this pythonic-ly?

Thanks,
Daniel Klein


From mwh21 at cam.ac.uk  Sun Feb  4 19:16:25 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Feb 2001 18:16:25 +0000
Subject: Preventing direct access to a class
References: <dc4r7t41dchrbrdabst1ung9mop3u4dnoj@4ax.com>
Message-ID: <m3ofwi48li.fsf@atrus.jesus.cam.ac.uk>

Daniel Klein <danielk at aracnet.com> writes:

> I have a situation where I need to prevent direct access to a class. IOW, the
> only way I wish a class to be instantiated is via another class. For example
> 
> >>> class A:
> 	pass
> 
> >>> class B:
> 	def foo(self):
> 		return A()
> 	
> >>> b = B()
> >>> a = b.foo()
> 
> ...and there might be other classes that 'return A()'.
> 
> I do not want the user to be allowed to do this...
> 
> >>>a = A()
> 
> It doesn't look like classes can be made 'private to the module' as
> I tried to define class A as 'class __A'. The user could still
> access with a=__A() (no mangling needed).
> 
> Other than placing some convoluted logic in each class, is there a way to
> accomplish this pythonic-ly?

You won't be able to do this completely.  You can make it sufficiently
hard that people won't do it by mistake, which is good enough for most
folk (it is for me).

Here's one way:

-------------------------------
class A:
 pass

class B:
 def foo(self,A=A):
  return A()

del A
-------------------------------

the class A is still there:

>>> B.foo.im_func.func_defaults 
(<class __main__.A at 0x8195024>,)

but I think you can rely on people not doing

>>> B.foo.im_func.func_defaults[0]()

accidentally <wink>.

Cheers,
M.

-- 
  I'm okay with intellegent buildings, I'm okay with non-sentient
  buildings. I have serious reservations about stupid buildings.
     -- Dan Sheppard, ucam.chat (from Owen Dunn's summary of the year)


From danielk at aracnet.com  Sun Feb  4 22:56:07 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 04 Feb 2001 13:56:07 -0800
Subject: Preventing direct access to a class
References: <dc4r7t41dchrbrdabst1ung9mop3u4dnoj@4ax.com> <m3ofwi48li.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <9mjr7tc4r1q2h5ju7792hk41jks4a79f2m@4ax.com>

On 04 Feb 2001 18:16:25 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:

>You won't be able to do this completely.  You can make it sufficiently
>hard that people won't do it by mistake, which is good enough for most
>folk (it is for me).
>
>Here's one way:
>
>-------------------------------
>class A:
> pass
>
>class B:
> def foo(self,A=A):
>  return A()
>
>del A
>-------------------------------
>
>the class A is still there:
>
>>>> B.foo.im_func.func_defaults 
>(<class __main__.A at 0x8195024>,)
>
>but I think you can rely on people not doing
>
>>>> B.foo.im_func.func_defaults[0]()
>
>accidentally <wink>.

Sneaky!

I actually resolved my own problem by shifting some of the responsibility of
B() back over to A() (where it belongs), so I don't have the problem anymore.
But thanks for the idea, I'll keep it on tap.

Dan


From joconnor at cybermesa.com  Sun Feb  4 21:15:26 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 20:15:26 +0000
Subject: Preventing direct access to a class
References: <dc4r7t41dchrbrdabst1ung9mop3u4dnoj@4ax.com>
Message-ID: <3A7DB85E.786352C4@cybermesa.com>

Daniel Klein wrote:
> 
> I have a situation where I need to prevent direct access to a class. IOW, the
> only way I wish a class to be instantiated is via another class.

Any reason you wish to enforce this particular restriction?  Usually
such notions are really means of communicating with other developers the
proper use and intent of the code you've written and I've found that
commenting with something akin to  "This is a private class, do not
instantiate directly" (but with more information) is sufficient to
communicate to other developers that they shouldn't use the class like
that.


So if you have a class that shoudn't be instantiated except from another
class, it should be sufficient to comment it as such.  And if someone
does it anyway, they get to deal with the ramifications :)

I think code is as much communicating intent to other developers as much
as communicating instructions to the computer and I like that Python
doesn't have a lot of syntax (or mechanisms) to enforce decisions that
are really communications issues that can be handled without extra
language features

Take care,

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From aahz at panix.com  Sun Feb  4 21:58:32 2001
From: aahz at panix.com (Aahz Maruch)
Date: 4 Feb 2001 12:58:32 -0800
Subject: Preventing direct access to a class
References: <dc4r7t41dchrbrdabst1ung9mop3u4dnoj@4ax.com>
Message-ID: <95kfpo$e9s$1@panix2.panix.com>

In article <dc4r7t41dchrbrdabst1ung9mop3u4dnoj at 4ax.com>,
Daniel Klein  <danielk at aracnet.com> wrote:
>
>I have a situation where I need to prevent direct access to a class. 

BTW, the standard Python idiom for declaring this is to precede the
class name with a single underscore (see threading.py for a real
example):

class _foo:
    pass
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"People sometimes focus too much on the fact of communication rather than
the substance of communication."  --Dave Morton


From danielk at aracnet.com  Sun Feb  4 22:50:57 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 04 Feb 2001 13:50:57 -0800
Subject: Preventing direct access to a class
References: <dc4r7t41dchrbrdabst1ung9mop3u4dnoj@4ax.com> <95kfpo$e9s$1@panix2.panix.com>
Message-ID: <dijr7t8fc2cilu9n9bl4fmbmb1bm4ssfsc@4ax.com>

On 4 Feb 2001 12:58:32 -0800, aahz at panix.com (Aahz Maruch) wrote:

>In article <dc4r7t41dchrbrdabst1ung9mop3u4dnoj at 4ax.com>,
>Daniel Klein  <danielk at aracnet.com> wrote:
>>
>>I have a situation where I need to prevent direct access to a class. 
>
>BTW, the standard Python idiom for declaring this is to precede the
>class name with a single underscore (see threading.py for a real
>example):
>
>class _foo:
>    pass

Yes, I see. But this only prevents access to class A if the user does
'from/import'.

Thanks,
Dan


From aahz at panix.com  Mon Feb  5 02:16:55 2001
From: aahz at panix.com (Aahz Maruch)
Date: 4 Feb 2001 17:16:55 -0800
Subject: Preventing direct access to a class
References: <dc4r7t41dchrbrdabst1ung9mop3u4dnoj@4ax.com> <95kfpo$e9s$1@panix2.panix.com> <dijr7t8fc2cilu9n9bl4fmbmb1bm4ssfsc@4ax.com>
Message-ID: <95kuu7$im$1@panix2.panix.com>

In article <dijr7t8fc2cilu9n9bl4fmbmb1bm4ssfsc at 4ax.com>,
Daniel Klein  <danielk at aracnet.com> wrote:
>On 4 Feb 2001 12:58:32 -0800, aahz at panix.com (Aahz Maruch) wrote:
>>In article <dc4r7t41dchrbrdabst1ung9mop3u4dnoj at 4ax.com>,
>>Daniel Klein  <danielk at aracnet.com> wrote:
>>>
>>>I have a situation where I need to prevent direct access to a class. 
>>
>>BTW, the standard Python idiom for declaring this is to precede the
>>class name with a single underscore (see threading.py for a real
>>example):
>>
>>class _foo:
>>    pass
>
>Yes, I see. But this only prevents access to class A if the user does
>'from/import'.

Thing is, because of the way Python works, there is no way to absolutely
prevent someone from doing anything.  I mean, if nothing else, they've
got access to your source code, right?  ;-)  So what you want to do is
be clear about your intentions and trust that anyone using your code is
going to Do The Right Thing.  Python has a number of mechanisms for
making your intentions clear; using a leading underscore is one of the
simplest and most powerful.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Let's go home and turn on MTV.  I want to watch some Pop-Up Videos."
"Pop-Up Videos is not on MTV, it's on VH-1."
"'MTV' is a generic."


From reborn at neozone.net  Sun Feb  4 20:05:31 2001
From: reborn at neozone.net (syKim)
Date: Mon, 5 Feb 2001 04:05:31 +0900
Subject: About applet under jython in Web browser.
Message-ID: <95k94s$t1l$1@news1.kornet.net>

I made a simple jython code as you know that just print "hello world!!"

It works so well under appletviewer.

but, It just show rectangle box In MS explorer.

I didn't know why the result like that..

Today, I suddenly think up reload..

I showed "hello world!!" in explorer at last.. :)

but, It's nessary reloading.

When applet first loaded in explorer and bring up the mouse cursor on
applet,

the error message is "

load: HelloApplet can't be instantiated"

What can I do something for it? (for show applet when it's first loaded)





From phlip_cpp at my-deja.com  Sun Feb  4 20:59:52 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 04 Feb 2001 19:59:52 GMT
Subject: Please translate this easy snip of C++ to Python
Message-ID: <95kcbo$k8s@dispatch.concentric.net>

Pythonographers:

Ogle this awesome snip of C++:

        #define TRACE_(x) cout << #x ": " << x << endl

When you use that like this:

        TRACE_(y + z);
        TRACE_(strAnimals);
        TRACE_(__LINE__);

the program emits this:

        y + z: 12
        strAnimals: Lemmings
        __LINE__:  69

Mighty healthy and easy to write debug statements with, huh? Saves a lot of 
keystrokes, huh?

How do I do that, just as easy to call, in Python?

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  In the beginning, God distinguished heaven and earth  --


From chris at onca.catsden.net  Sun Feb  4 21:07:20 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Sun, 4 Feb 2001 12:07:20 -0800 (PST)
Subject: Please translate this easy snip of C++ to Python
In-Reply-To: <95kcbo$k8s@dispatch.concentric.net>
Message-ID: <Pine.LNX.4.30.0102041205280.24132-100000@onca.catsden.net>

On 4 Feb 2001, Phlip wrote:

> Pythonographers:
>
> Ogle this awesome snip of C++:
>
>         #define TRACE_(x) cout << #x ": " << x << endl
>
> When you use that like this:
>
>         TRACE_(y + z);
>         TRACE_(strAnimals);
>         TRACE_(__LINE__);
>
> the program emits this:
>
>         y + z: 12
>         strAnimals: Lemmings
>         __LINE__:  69
>
> Mighty healthy and easy to write debug statements with, huh? Saves a lot of
> keystrokes, huh?
>
> How do I do that, just as easy to call, in Python?

Try this:

def TRACE ( str ):
	print "%s: %s" % ( str, eval(str) )

TRACE('y+z')
TRACE('strAnimals')

Sorry... you /do/ need to put the parameters in quotes, so Python doesnt
try to evaluate them before calling TRACE

And... not sure if there's a equivalent of __line__


   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From moshez at zadka.site.co.il  Sun Feb  4 21:21:32 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun,  4 Feb 2001 22:21:32 +0200 (IST)
Subject: Please translate this easy snip of C++ to Python
In-Reply-To: <95kcbo$k8s@dispatch.concentric.net>
References: <95kcbo$k8s@dispatch.concentric.net>
Message-ID: <20010204202132.0B13BA840@darjeeling.zadka.site.co.il>

On 04 Feb 2001 19:59:52 GMT, Phlip <phlip_cpp at my-deja.com> wrote:

> Pythonographers:
> 
> Ogle this awesome snip of C++:
> 
>         #define TRACE_(x) cout << #x ": " << x << endl

You can do it in Python the other way around:

>>> def trace_(x):
...         print x+":", eval(x)
...
>>> y = z = 6; animal='lemming'
>>> trace_('y+z')
y+z: 12
>>> trace_('animal')
animal: lemming

HIH!
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From fredrik at effbot.org  Sun Feb  4 21:24:01 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 04 Feb 2001 20:24:01 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net>
Message-ID: <BTif6.12144$AH6.1880632@newsc.telia.net>

"Phlip" <phlip_cpp at my-deja.com> wrote:
> Ogle this awesome snip of C++:
>
>         #define TRACE_(x) cout << #x ": " << x << endl

That's a preprocessor trick, not a C++ trick.  Replace <<
with printf, and it works just as fine under C.

(iirc, bjarne thinks that real C++ programmers shouldn't
use the preprocessor at all, but what does he know...)

> How do I do that, just as easy to call, in Python?

Run your Python code through C's preprocessor?

Or use something like this:

    def trace(expr):
        # evaluate expression in callers namespace
        import sys
        try:
            raise None
        except:
            frame = sys.exc_info()[2].tb_frame.f_back
        print expr, "=", eval(expr, frame.f_globals, frame.f_locals)

    x = 10
    y = 20
    trace("x + y")

    ## prints:
    ## x + y = 30

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From max at alcyone.com  Sun Feb  4 21:44:12 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 04 Feb 2001 12:44:12 -0800
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net>
Message-ID: <3A7DBF1C.B03B8200@alcyone.com>

Fredrik Lundh wrote:

> (iirc, bjarne thinks that real C++ programmers shouldn't
> use the preprocessor at all, but what does he know...)

C++ features such as namespaces, templates, and const variables acting
as true constants generally render it unnecessary to use the
preprocessor to accomplish most tasks it was traditionally used for. 
It's generally considered bad style to make excessive use of the
preprocessor in Standard C++.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ When one is in love, a cliff becomes a meadow.
\__/ (an Ethiopian saying)
    Official Buh rules / http://www.alcyone.com/max/projects/cards/buh/
 The official rules to the betting card game, Buh.


From bowman at montana.com  Sun Feb  4 21:58:17 2001
From: bowman at montana.com (bowman)
Date: Sun, 4 Feb 2001 13:58:17 -0700
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net>
Message-ID: <yljf6.293$K77.3010@newsfeed.slurp.net>

Fredrik Lundh <fredrik at effbot.org> wrote in message news:BTif6.12144
>
> (iirc, bjarne thinks that real C++ programmers shouldn't
> use the preprocessor at all, but what does he know...)

"The C preprocessor is a powerful but blunt tool, however, and macros are a
dangerous way to program because they change the lexical structure of the
program underfoot. Let the language proper do the work."

Kernighan & Pike, _The Practice of Programming_.


iirc, either K or R have mentioned they would have left macros out of the
language if they could have forseen the clever and wondrous uses that would
be found for them.





From thomas at xs4all.net  Mon Feb  5 00:00:09 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 5 Feb 2001 00:00:09 +0100
Subject: Please translate this easy snip of C++ to Python
In-Reply-To: <yljf6.293$K77.3010@newsfeed.slurp.net>; from bowman@montana.com on Sun, Feb 04, 2001 at 01:58:17PM -0700
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <yljf6.293$K77.3010@newsfeed.slurp.net>
Message-ID: <20010205000008.G962@xs4all.nl>

On Sun, Feb 04, 2001 at 01:58:17PM -0700, bowman wrote:

> Fredrik Lundh <fredrik at effbot.org> wrote in message news:BTif6.12144
> >
> > (iirc, bjarne thinks that real C++ programmers shouldn't
> > use the preprocessor at all, but what does he know...)

> "The C preprocessor is a powerful but blunt tool, however, and macros are a
> dangerous way to program because they change the lexical structure of the
> program underfoot. Let the language proper do the work."

> Kernighan & Pike, _The Practice of Programming_.

> iirc, either K or R have mentioned they would have left macros out of the
> language if they could have forseen the clever and wondrous uses that would
> be found for them.

Note how Python lacks a preprocessor and macros :) People occasionally rally
for a macro preprocessor, but usually not of the C type. Being fairly
ignorant in languages featuring macros other than the C style, I can't
really say what they *are* rallying for, but apparently there are sane macro
preprocessors in existance, somewhere :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From moshez at zadka.site.co.il  Mon Feb  5 12:16:50 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Mon,  5 Feb 2001 13:16:50 +0200 (IST)
Subject: Please translate this easy snip of C++ to Python
In-Reply-To: <20010205000008.G962@xs4all.nl>
References: <20010205000008.G962@xs4all.nl>, <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <yljf6.293$K77.3010@newsfeed.slurp.net>
Message-ID: <20010205111650.BE66BA840@darjeeling.zadka.site.co.il>

On Mon, 5 Feb 2001 00:00:09 +0100, Thomas Wouters <thomas at xs4all.net> wrote:
 
> Note how Python lacks a preprocessor and macros :) People occasionally rally
> for a macro preprocessor, but usually not of the C type. Being fairly
> ignorant in languages featuring macros other than the C style, I can't
> really say what they *are* rallying for, but apparently there are sane macro
> preprocessors in existance, somewhere :)

If I'm in a mood to annoy Guido, I propose hygienic(sp?) macros, like
in R5RS. These are things which are macros, but on the verge on being
functions: no accidental name collision, for one. Real macros work on
the parser level, not the tokenizer levels, so you can't do idiotic
stuff like 

#define BEGIN {
#define END ;}

Or other shoot-on-sight offense.
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From bowman at montana.com  Mon Feb  5 07:42:55 2001
From: bowman at montana.com (bowman)
Date: Sun, 4 Feb 2001 23:42:55 -0700
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <yljf6.293$K77.3010@newsfeed.slurp.net> <mailman.981327676.7786.python-list@python.org>
Message-ID: <zVrf6.15$RM.332@newsfeed.slurp.net>

Thomas Wouters <thomas at xs4all.net> wrote in message
news:mailman.981327676.7786.python-list at python.org...
>
> Note how Python lacks a preprocessor and macros :) People occasionally
rally
> for a macro preprocessor, but usually not of the C type. Being fairly
> ignorant in languages featuring macros other than the C style, I can't
> really say what they *are* rallying for, but apparently there are sane
macro
> preprocessors in existance, somewhere :)

Other than C, my exposure has been to various macro assemblers. With some
creativity, it is possible to write your own language to the extent the
underlying asm never surfaces. In fact, it isn't uncommon in embedded work
for people to use macros to write code for a PIC or AVR device almost
completely in '51 syntax.

I personally am uncomfortable when a language allows one to create a private
dialect; FORTH and many of the functional languages come to mind. Factoring
a problem into functions is one thing, but I prefer to see the bare metal of
the language and well known library calls rather than essentially having to
learn yet another language someone has lovingly handcrafted (and generally
not documented).






From csrabak at uol.com.br  Tue Feb  6 23:53:28 2001
From: csrabak at uol.com.br (Cesar Rabak)
Date: Tue, 06 Feb 2001 20:53:28 -0200
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <yljf6.293$K77.3010@newsfeed.slurp.net> <mailman.981327676.7786.python-list@python.org>
Message-ID: <3A808068.D8D38007@uol.com.br>


Thomas Wouters wrote:
> 
> Note how Python lacks a preprocessor and macros :) People occasionally rally
> for a macro preprocessor, but usually not of the C type. Being fairly
> ignorant in languages featuring macros other than the C style, I can't
> really say what they *are* rallying for, but apparently there are sane macro
> preprocessors in existance, somewhere :)

Yep! If I felt need of one, I would craft macros in M4. But them to have
your scripts you?ll need to explicitly run some 'wrapper' (shell) script
or a Makefile. After all this, would it be still worthwhile?


From peter at engcorp.com  Mon Feb  5 06:37:08 2001
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 05 Feb 2001 00:37:08 -0500
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net>
Message-ID: <3A7E3C04.9EB02932@engcorp.com>

Fredrik Lundh wrote:
> 
> "Phlip" <phlip_cpp at my-deja.com> wrote:
> > Ogle this awesome snip of C++:
> >
> >         #define TRACE_(x) cout << #x ": " << x << endl
> 
> > How do I do that, just as easy to call, in Python?
> 
> Or use something like this:
> 
[code snippet snipped]
> 
> Cheers /F

If I may humbly offer a small change which even handles 
something like the original __LINE__ request, I believe:

def trace(expr = None):
    # evaluate expression in callers namespace
    import sys
    try:
        raise None
    except:
        frame = sys.exc_info()[2].tb_frame.f_back
    if expr != None:
        print expr, ":", eval(expr, frame.f_globals, frame.f_locals)
    else:
        print '__LINE__ :', frame.f_lineno

x = 10
y = 20
strAnimals = "Lemmings"

trace("x + y")
trace("strAnimals")
trace()

## prints:
## x + y : 30
## strAnimals : Lemmings
## __LINE__ : 19


From root at rainerdeyke.com  Mon Feb  5 16:56:33 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Mon, 05 Feb 2001 15:56:33 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com>
Message-ID: <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3A7E3C04.9EB02932 at engcorp.com...
> def trace(expr = None):
>     # evaluate expression in callers namespace
>     import sys
>     try:
>         raise None
>     except:

Better make that 'except None'.  Unqualified 'except' clauses are generally
a bad idea.  In this case, you really don't want to catch 'MemoryError',
'KeyboardInterrupt', or 'ComputerExplodingError'.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From joconnor at cybermesa.com  Mon Feb  5 17:02:51 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 05 Feb 2001 16:02:51 +0000
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com>
Message-ID: <3A7ECEAB.DB6F8648@cybermesa.com>

Rainer Deyke wrote:

> Better make that 'except None'.  Unqualified 'except' clauses are generally
> a bad idea.  In this case, you really don't want to catch...'ComputerExplodingError'.

I hadn't seen that exception raised yet....is it new?  What do you
usually do to handle it?

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From max at alcyone.com  Mon Feb  5 18:09:48 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 05 Feb 2001 09:09:48 -0800
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com>
Message-ID: <3A7EDE5C.5D3F7050@alcyone.com>

Jay O'Connor wrote:

> Rainer Deyke wrote:
> 
> > Better make that 'except None'.  Unqualified 'except' clauses are
> > generally
> > a bad idea.  In this case, you really don't want to
> > catch...'ComputerExplodingError'.
> 
> I hadn't seen that exception raised yet....is it new?  What do you
> usually do to handle it?

Right past you.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    7 sisters productions / http://www.7sisters.com/
 Web design for the future.


From joconnor at cybermesa.com  Mon Feb  5 18:45:59 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 05 Feb 2001 17:45:59 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com> <3A7EDE5C.5D3F7050@alcyone.com>
Message-ID: <3a7ee6b3.673800@news.cybermesa.com>

On Mon, 05 Feb 2001 09:09:48 -0800, Erik Max Francis <max at alcyone.com>
wrote:

>Jay O'Connor wrote:
>
>> Rainer Deyke wrote:

>> > catch...'ComputerExplodingError'.
>> 
>> I hadn't seen that exception raised yet....is it new?  What do you
>> usually do to handle it?
>
>Right past you.


Nope, just being silly :)

Take care,
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays on the bass strings first, when he tunes the soul"


From root at rainerdeyke.com  Mon Feb  5 18:44:06 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Mon, 05 Feb 2001 17:44:06 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com>
Message-ID: <GDBf6.175479$ge4.60994110@news2.rdc2.tx.home.com>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:3A7ECEAB.DB6F8648 at cybermesa.com...
> Rainer Deyke wrote:
>
> > Better make that 'except None'.  Unqualified 'except' clauses are
generally
> > a bad idea.  In this case, you really don't want to
catch...'ComputerExplodingError'.
>
> I hadn't seen that exception raised yet....is it new?  What do you
> usually do to handle it?

ComputerExplodingError is one of the "hidden" exceptions of the Underground
of Pythonic Secrecy (UPS), created in response to the much cooler names of
rival groups Secret Pythonic Underground (SPU) amd Pythonic Secret
Underground (SPU).  It is raised automatically when a computer "goes
postal".  The proper response is to panic and scream a lot.  This is done
automatically by the interpreter when stack unwinding is complete.
Obviously this requires that any 'except' clause that intercepts the
exception re-throws it.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From dalke at acm.org  Mon Feb  5 19:36:52 2001
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 5 Feb 2001 11:36:52 -0700
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com> <GDBf6.175479$ge4.60994110@news2.rdc2.tx.home.com>
Message-ID: <95mru3$2ar$1@slb2.atl.mindspring.net>

Rainer Deyke:
> ComputerExplodingError is one of the "hidden" exception
> of the Underground of Pythonic Secrecy (UPS), created in
> response to the much cooler names of rival groups Secret
> Pythonic Underground (SPU) amd Pythonic Secret Underground
> (SPU).  It is raised automatically when a computer "goes
> postal".  The proper response is to panic and scream a
> lot.  This is done automatically by the interpreter when
> stack unwinding is complete.

And if you have sound support then the computer says
 "I'm sorry %{USER}s, I can't do that right now" % os.environ
if os.environ.has_key("USER"), else it plays
 "Daisy, Daisy, give me your answer true"

But I don't actually know this since I'm not a member of





From quinn at ngwee.ugcs.caltech.edu  Tue Feb  6 04:43:22 2001
From: quinn at ngwee.ugcs.caltech.edu (Quinn Dunkan)
Date: 6 Feb 2001 03:43:22 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com> <GDBf6.175479$ge4.60994110@news2.rdc2.tx.home.com>
Message-ID: <slrn97usmq.b5b.quinn@ngwee.ugcs.caltech.edu>

On Mon, 05 Feb 2001 17:44:06 GMT, Rainer Deyke <root at rainerdeyke.com> wrote:
>ComputerExplodingError is one of the "hidden" exceptions of the Underground
>of Pythonic Secrecy (UPS), created in response to the much cooler names of
>rival groups Secret Pythonic Underground (SPU) amd Pythonic Secret
>Underground (SPU).  It is raised automatically when a computer "goes
>postal".

So long as it goes via normal routes, there shouldn't be a problem.  But know
that the Tristero network takse a dim view of transporting equipment for other
nonexistent organizations.

Don't ever antagonize the horn.


From m.faassen at vet.uu.nl  Wed Feb  7 00:47:00 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 6 Feb 2001 23:47:00 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com> <GDBf6.175479$ge4.60994110@news2.rdc2.tx.home.com>
Message-ID: <95q2dk$hji$4@newshost.accu.uu.nl>

Rainer Deyke <root at rainerdeyke.com> wrote:
> "Jay O'Connor" <joconnor at cybermesa.com> wrote in message
> news:3A7ECEAB.DB6F8648 at cybermesa.com...
>> Rainer Deyke wrote:
>>
>> > Better make that 'except None'.  Unqualified 'except' clauses are
> generally
>> > a bad idea.  In this case, you really don't want to
> catch...'ComputerExplodingError'.
>>
>> I hadn't seen that exception raised yet....is it new?  What do you
>> usually do to handle it?

> ComputerExplodingError is one of the "hidden" exceptions of the Underground
> of Pythonic

Something must be wrong with some Usenet feed somewhere. Your post got
cut off in the middle. I suspect that this was due to


From sholden at holdenweb.com  Wed Feb  7 16:57:50 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 7 Feb 2001 10:57:50 -0500
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com> <GDBf6.175479$ge4.60994110@news2.rdc2.tx.home.com> <95q2dk$hji$4@newshost.accu.uu.nl>
Message-ID: <Yjeg6.38501$o5.480868@e420r-atl1.usenetserver.com>

"Martijn Faassen" <m.faassen at vet.uu.nl> wrote in message
news:95q2dk$hji$4 at newshost.accu.uu.nl...
> Rainer Deyke <root at rainerdeyke.com> wrote:
> > "Jay O'Connor" <joconnor at cybermesa.com> wrote in message
> > news:3A7ECEAB.DB6F8648 at cybermesa.com...
> >> Rainer Deyke wrote:
> >>
> >> > Better make that 'except None'.  Unqualified 'except' clauses are
> > generally
> >> > a bad idea.  In this case, you really don't want to
> > catch...'ComputerExplodingError'.
> >>
> >> I hadn't seen that exception raised yet....is it new?  What do you
> >> usually do to handle it?
>
> > ComputerExplodingError is one of the "hidden" exceptions of the
Underground
> > of Pythonic
>
> Something must be wrong with some Usenet feed somewhere. Your post got
> cut off in the middle. I suspect that this was due to
>
Look, cut this out. I'm channelling Tim again. For the last time, there IS
no



From phlip_cpp at my-deja.com  Mon Feb  5 06:59:47 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 05 Feb 2001 05:59:47 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net>
Message-ID: <95lfgj$kbg@dispatch.concentric.net>

Fredrik Lundh wrote:

> That's a preprocessor trick, not a C++ trick.  Replace <<
> with printf, and it works just as fine under C.

Could'a mentioned that's where the trick started with me. Before the # 
stringerizer operator got invented, too...

> (iirc, bjarne thinks that real C++ programmers shouldn't
> use the preprocessor at all, but what does he know...)

The exact rule is "Don't use a low-tech feature when a high-tech one will 
do". You'l notice this rule explains why I'm deep in the coils of Python 
these days...

One should not use the C preprocessor to perform simple administration, 
such as declaring a constant. But only the preprocessor has the 
stringerizer operator #; and it's a quite recent addition, based on bugs 
exploited in earlier preprocessors to stringerize without it, so it 
qualifies as high-tech.

> Run your Python code through C's preprocessor?

We are already deep in a stack of script, Perforce Jam, Python, SWIG and 
C++ code; we don't need another operation on our source!

Thomas Wouters wrote:

> Note how Python lacks a preprocessor and macros :)

No, reaching down into the interpret's own pants and grabbing its nads with 
lines like these will do nicely:

        frame = sys.exc_info()[2].tb_frame.f_back

chris at onca.catsden.net wrote:

> And... not sure if there's a equivalent of __line__

I just threw that one in worrying the first question might turn out too 
easy...

Thanks to all the replies; I'l try the suggestions (and learn what 
"trace()" does) as soon as I'm on the clock ;-)

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  Who needs a degree when we have Web e-search engines?  --


From gangli at msn.com  Thu Feb  8 16:34:47 2001
From: gangli at msn.com (gangli at msn.com)
Date: Thu, 08 Feb 2001 15:34:47 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net>
Message-ID: <95uean$qsf$1@nnrp1.deja.com>

In article <BTif6.12144$AH6.1880632 at newsc.telia.net>,
  "Fredrik Lundh" <fredrik at effbot.org> wrote:

How do you make trace("a = 1") work ?


> "Phlip" <phlip_cpp at my-deja.com> wrote:
> Or use something like this:
>
>     def trace(expr):
>         # evaluate expression in callers namespace
>         import sys
>         try:
>             raise None
>         except:
>             frame = sys.exc_info()[2].tb_frame.f_back
>         print expr, "=", eval(expr, frame.f_globals, frame.f_locals)
>
>     x = 10
>     y = 20
>     trace("x + y")
>
>     ## prints:
>     ## x + y = 30
>
> Cheers /F
>
> <!-- (the eff-bot guide to) the standard python library:
> http://www.pythonware.com/people/fredrik/librarybook.htm
> -->
>
>


Sent via Deja.com
http://www.deja.com/


From joconnor at cybermesa.com  Sun Feb  4 21:28:57 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 20:28:57 +0000
Subject: Please translate this easy snip of C++ to Python
References: <mailman.981317716.4248.python-list@python.org>
Message-ID: <3A7DBB89.A79A769F@cybermesa.com>

chris at onca.catsden.net wrote:

> And... not sure if there's a equivalent of __line__

Of more interest would be to be able to get a hold of execution stack
information, such as what method called this method (and maybe the
parameter values, such as 'self' of *that* method)

If it's called 'trace'?  it'd be nice to trace back  :)

This may be possible, I don't know how the Traceback works in Python; I
haven't explored it yet.

Take care,

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From clarence at netlojix.com  Sun Feb  4 22:11:57 2001
From: clarence at netlojix.com (Clarence Gardner)
Date: Sun, 4 Feb 2001 13:11:57 -0800
Subject: Please translate this easy snip of C++ to Python
References: <mailman.981317716.4248.python-list@python.org>
Message-ID: <01020413173100.10388@liberty.sba2.netlojix.net>

On Sun, 04 Feb 2001, chris at onca.catsden.net wrote:
>On 4 Feb 2001, Phlip wrote:
[C debugging macro snipped]

>Try this:
>
>def TRACE ( str ):
>	print "%s: %s" % ( str, eval(str) )
>
>TRACE('y+z')
>TRACE('strAnimals')
>
>Sorry... you /do/ need to put the parameters in quotes, so Python doesnt
>try to evaluate them before calling TRACE

You also need to provide the TRACE function with the caller's local
namespace, or TRACE will, in general, get NameErrors.

def TRACE(str, locals):
    print "%s: %s" % (str, eval(str, globals(), locals))

and call it like
    TRACE('y+z', locals())

(I know I'm stomping on a predefined function name in TRACE, but I
have no problem with it in this case.)
--
Clarence Gardner
Software Engineer
NetLojix Communications
clarence at netlojix.com



From clarence at netlojix.com  Sun Feb  4 22:11:57 2001
From: clarence at netlojix.com (Clarence Gardner)
Date: Sun, 4 Feb 2001 13:11:57 -0800
Subject: Please translate this easy snip of C++ to Python
References: <mailman.981317716.4248.python-list@python.org>
Message-ID: <981321451.86433377@news.silcom.com>

On Sun, 04 Feb 2001, chris at onca.catsden.net wrote:
>On 4 Feb 2001, Phlip wrote:
[C debugging macro snipped]

>Try this:
>
>def TRACE ( str ):
>	print "%s: %s" % ( str, eval(str) )
>
>TRACE('y+z')
>TRACE('strAnimals')
>
>Sorry... you /do/ need to put the parameters in quotes, so Python doesnt
>try to evaluate them before calling TRACE

You also need to provide the TRACE function with the caller's local
namespace, or TRACE will, in general, get NameErrors.

def TRACE(str, locals):
    print "%s: %s" % (str, eval(str, globals(), locals))

and call it like
    TRACE('y+z', locals())

(I know I'm stomping on a predefined function name in TRACE, but I
have no problem with it in this case.)
--
Clarence Gardner
Software Engineer
NetLojix Communications
clarence at netlojix.com



From jick78 at my-deja.com  Sun Feb  4 21:40:34 2001
From: jick78 at my-deja.com (jick78 at my-deja.com)
Date: Sun, 04 Feb 2001 20:40:34 GMT
Subject: ANN: mod_snake-0.5.0
Message-ID: <95keo2$csn$1@nnrp1.deja.com>

                        Mod Snake 0.5.0
                        ---------------

An Apache module which allows plugins to be written in Python.

Mod Snake is an Apache module written for the purpose to give Python
developers the same power that C module writers have.  It currently
runs in both Apache 1.3 and Apache 2.0, providing access to new
functionality such as writing protocol modules and filtering.  It
includes modules for Python CGI acceleration, embedded Python in HTML,
and other example modules.


         URL:  http://modsnake.sourceforge.net

  Catagories:  Internet


This release includes a huge number of enhancements and bug fixes.
For the definitive list, see the ChangeLog on the mod_snake web site.

        - Created a RedHat RPM
        - Setup a testsuite framework
        - Numerous improvements to the build system
        - Fixed dependencies on GNU make
        - Added support for Python 2.0
        - Sped up HTML with embedded Python processing
(jepler at inetnebr.com)
        - Fixed an issue in the CGI acceleration module
        - etc. ;-)

Jon Travis (jtravis at p00p.org)
http://modsnake.sourceforge.net


Sent via Deja.com
http://www.deja.com/


From chris.gonnerman at usa.net  Sun Feb  4 22:03:46 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sun, 4 Feb 2001 15:03:46 -0600
Subject: Paging Richard Jones...
Message-ID: <009501c08eee$1489a6e0$a100000a@local>

Richard, if you are reading this list, I've been trying to reach you
regarding the Python
GD module you say you aren't maintaining anymore.  I've updated it
dramatically and
am willing to take it over from you.  If you have changed your email address
you need
to update your website.

To all:  If you know how to contact Richard, please let me know (or forward
this to him).





From jick78 at my-deja.com  Sun Feb  4 22:15:56 2001
From: jick78 at my-deja.com (jick78 at my-deja.com)
Date: Sun, 04 Feb 2001 21:15:56 GMT
Subject: ANN: Auth DBAPI 0.10
Message-ID: <95kgq9$ect$1@nnrp1.deja.com>


                        Auth DBAPI 0.10
                        ---------------

A mod_snake plugin for Apache authentication and authorization via
DBAPI.

DBAPI is a specification created by the Python Database SIG which
defines a common interface for accessing database modules.  Auth DBAPI
is a module which allows mod_snake users to use DBAPI 1.0 and 2.0
database modules to authenticate and authorize users
(e.g. via Gadfly && MySQLdb).

         URL:  http://modsnake.sourceforge.net/msmnet/auth_dbapi

  Catagories:  Internet

This is the initial public announcement.

Jon Travis (jtravis at p00p.org)
http://modsnake.sourceforge.net


Sent via Deja.com
http://www.deja.com/


From syver at NOSPAMcyberwatcher.com  Sun Feb  4 22:37:04 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sun, 4 Feb 2001 22:37:04 +0100
Subject: Jython incompatibilities
Message-ID: <95ki6h$gst$1@troll.powertech.no>

I've tried to reach the Jython list but it wouldn't listen to me so....

I'm checking out Jython while waiting for .NET python.

One thing that was a bit weird was that the max function behaves differently
on Jython than Cpython.

Ex under python:
min([])

throws a ValueError exception

Ex under jython:
min([])

throws a TypeError exception.

This is not so great as it can break an application, it did to one of mine.





From mlh at idi.ntnu.no  Mon Feb  5 09:08:37 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Mon, 5 Feb 2001 09:08:37 +0100
Subject: Jython incompatibilities
References: <95ki6h$gst$1@troll.powertech.no>
Message-ID: <95ln7m$jq6$1@tyfon.itea.ntnu.no>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote in message
news:95ki6h$gst$1 at troll.powertech.no...
> I've tried to reach the Jython list but it wouldn't listen to me so....
>
> I'm checking out Jython while waiting for .NET python.
>
> One thing that was a bit weird was that the max function behaves
differently
> on Jython than Cpython.
>
[snip]

Hm. They don't even have the same error message... How odd.

>
> This is not so great as it can break an application, it did to one of
mine.

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From shredwheat at mediaone.net  Sun Feb  4 22:42:30 2001
From: shredwheat at mediaone.net (Pete Shinners)
Date: Sun, 04 Feb 2001 21:42:30 GMT
Subject: new command for getting traceback?
Message-ID: <a1kf6.168776$y9.29864178@typhoon.we.rr.com>

one trick in python i've always thought was bizarre was getting
the current stack trace by raising a false exception.

        import sys
        try:
            raise None
        except:
            return sys.exc_info()[2].tb_frame

obviously grabbing the current stack is usually a sign of some
sort of "magic ahead", this is one of those things that is a
bit obscure. even harder from inside my C extension modules. It
would make things a bit nicer and more understandable if there
was some sort of traceback() builtin. (accessable from C too!)





From tim.one at home.com  Sun Feb  4 23:45:50 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 4 Feb 2001 17:45:50 -0500
Subject: new command for getting traceback?
In-Reply-To: <a1kf6.168776$y9.29864178@typhoon.we.rr.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEHIINAA.tim.one@home.com>

[Pete Shinners]
> one trick in python i've always thought was bizarre was getting
> the current stack trace by raising a false exception.
>
>         import sys
>         try:
>             raise None
>         except:
>             return sys.exc_info()[2].tb_frame
>
> obviously grabbing the current stack is usually a sign of some
> sort of "magic ahead", this is one of those things that is a
> bit obscure. even harder from inside my C extension modules. It
> would make things a bit nicer and more understandable if there
> was some sort of traceback() builtin. (accessable from C too!)

The details of frame manipulation are internal, and not guaranteed to work
the same way across releases.  So it's *good* that deep magic looks like
deep magic -- you're playing with fire.

Nevertheless,

Python 2.1a2 (#10, Feb  3 2001, 22:01:45) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import sys
>>> print sys._getframe.__doc__
_getframe([depth]) -> frameobject

Return a frame object from the call stack.  If optional integer depth is
given, return the frame object that many calls below the top of the stack.
If that is deeper than the call stack, ValueError is raised.  The default
for depth is zero, returning the frame at the top of the call stack.

This function should be used for internal and specialized
purposes only.
>>>

That's new for 2.1, and can replace current uses of the try/except trick, if
you're willing to live with that frame interals are subject to change
without notice.  Note that there have long been traceback-manipulating
functions in the std traceback module, if you want to work at a more
abstract (and safer) level.

If you want more than that, you'll have to talk someone into submitting a
patch and making a case for it.




From fredrik at effbot.org  Sun Feb  4 23:52:22 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sun, 04 Feb 2001 22:52:22 GMT
Subject: new command for getting traceback?
References: <a1kf6.168776$y9.29864178@typhoon.we.rr.com>
Message-ID: <G2lf6.12189$AH6.1887988@newsc.telia.net>

Pete wrote:
> one trick in python i've always thought was bizarre was getting
> the current stack trace by raising a false exception.
>
>         import sys
>         try:
>             raise None
>         except:
>             return sys.exc_info()[2].tb_frame

in 2.1, you can use sys._getframe():

    Python 2.1a1
    >>> import sys
    >>> print sys._getframe.__doc__
    _getframe([depth]) -> frameobject

    Return a frame object from the call stack.  If optional integer depth is
    given, return the frame object that many calls below the top of the stack.
    If that is deeper than the call stack, ValueError is raised.  The default
    for depth is zero, returning the frame at the top of the call stack.

    This function should be used for internal and specialized
    purposes only.

for more traceback/frame manipulation stuff, see the traceback
module.

Cheers /F




From bashanguy at hotmail.com  Sun Feb  4 23:29:41 2001
From: bashanguy at hotmail.com (guy bashan)
Date: Mon, 5 Feb 2001 00:29:41 +0200
Subject: TKInter installation problems
Message-ID: <3a7dd337@news.bezeqint.net>

Hello,

I have just installed the ActiveState Komodo. I thought that the tkinter
package should be already included in the ActivePython, but when I tried
running some simple samlpe demo using the tkinter, it wrote me that I'm
missing: tcl83.dll.  After a little search I downloaded the PythonWare
package, which should have the tkinter in it. The PythonWare looks pretty
much like the ActivePython, but it really did include the missing DLL. after
copying the requested DLL to the folder of the ActivePython (under the DLL
folder), I tried running the sample again, but encountered some weird long
sequence of errors, and at the end got a message that the TCL is probably
not installed well . . .     What have I done wrong ???

thanks in advance,
Guy Bashan.
bashan at newmail.net




From fuess at att.net  Mon Feb  5 04:49:32 2001
From: fuess at att.net (David Fuess)
Date: Mon, 05 Feb 2001 03:49:32 GMT
Subject: TKInter installation problems
References: <3a7dd337@news.bezeqint.net>
Message-ID: <4i8s7to9bm1csairuf9toe3or7fm47v3ar@4ax.com>

I found that on some installs tcl83.dll is not places in a searched
path so the system can't find it even though it's really there. My
solutions was to put the Tcl library path in my PATH. You can also
copy tcl83.dll to a system directory (like \windows\system32).

Dave

On Mon, 5 Feb 2001 00:29:41 +0200, "guy bashan"
<bashanguy at hotmail.com> wrote:

>Hello,
>
>I have just installed the ActiveState Komodo. I thought that the tkinter
>package should be already included in the ActivePython, but when I tried
>running some simple samlpe demo using the tkinter, it wrote me that I'm
>missing: tcl83.dll.  After a little search I downloaded the PythonWare
>package, which should have the tkinter in it. The PythonWare looks pretty
>much like the ActivePython, but it really did include the missing DLL. after
>copying the requested DLL to the folder of the ActivePython (under the DLL
>folder), I tried running the sample again, but encountered some weird long
>sequence of errors, and at the end got a message that the TCL is probably
>not installed well . . .     What have I done wrong ???
>
>thanks in advance,
>Guy Bashan.
>bashan at newmail.net
>



From fredrik at effbot.org  Mon Feb  5 18:56:40 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 05 Feb 2001 17:56:40 GMT
Subject: TKInter installation problems
References: <3a7dd337@news.bezeqint.net>
Message-ID: <sPBf6.12356$AH6.1910515@newsc.telia.net>

guy bashan wrote:
> missing: tcl83.dll.  After a little search I downloaded the PythonWare
> package, which should have the tkinter in it. The PythonWare looks pretty
> much like the ActivePython, but it really did include the missing DLL. after
> copying the requested DLL to the folder of the ActivePython (under the DLL
> folder), I tried running the sample again, but encountered some weird long
> sequence of errors, and at the end got a message that the TCL is probably
> not installed well . . .     What have I done wrong ???

you need tcl83.dll, tk83.dll, and also the tcl/tk library files (they're
in the tcl subdirectory in the PY20 distribution).  see our install faq
for some clues on where to put them:

    http://www.pythonware.com/products/tkinter/faq.htm

or you could get scriptics tcl/tk installer, and pray that activestate's
distro will find the dlls if they're installed in the default location:

    http://dev.scriptics.com/

hope this helps!

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From dustin at blacksky.org  Mon Feb  5 01:02:46 2001
From: dustin at blacksky.org (J Dustin Flesher)
Date: Sun, 4 Feb 2001 17:02:46 -0700
Subject: calldll with Python 2.0
Message-ID: <E4mf6.2531$mO6.467240@news.uswest.net>

Does anyone have a compiled version of calldll for ActiveState ActivePython
2.0? Has anyone got calldll working with ActiveState Python 2.0? Thank you.




From chris.gonnerman at usa.net  Mon Feb  5 03:13:25 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sun, 4 Feb 2001 20:13:25 -0600
Subject: calldll with Python 2.0
References: <E4mf6.2531$mO6.467240@news.uswest.net>
Message-ID: <005401c08f19$3e415fc0$a100000a@local>

----- Original Message -----
From: "J Dustin Flesher" <dustin at blacksky.org>
Subject: calldll with Python 2.0


> Does anyone have a compiled version of calldll for ActiveState
ActivePython
> 2.0? Has anyone got calldll working with ActiveState Python 2.0? Thank
you.

IMHO the standard calldll for Python 2.0 should work equally with the BeOpen
and
ActiveState releases; get it from http://www.nightmare.com/software.html.





From samschul at pacbell.net  Mon Feb  5 08:59:37 2001
From: samschul at pacbell.net (Sam Schulenburg)
Date: Mon, 05 Feb 2001 07:59:37 GMT
Subject: calldll with Python 2.0
References: <E4mf6.2531$mO6.467240@news.uswest.net>
Message-ID: <95lmh7$7iq$1@nnrp1.deja.com>

In article <E4mf6.2531$mO6.467240 at news.uswest.net>,
  "J Dustin Flesher" <dustin at blacksky.org> wrote:
> Does anyone have a compiled version of calldll for ActiveState
ActivePython
> 2.0? Has anyone got calldll working with ActiveState Python 2.0?
Thank you.
>
>
see following web site:

http://www.nightmare.com/software.html


Sent via Deja.com
http://www.deja.com/


From binary3418 at my-deja.com  Mon Feb  5 03:51:04 2001
From: binary3418 at my-deja.com (Deja User)
Date: Mon, 05 Feb 2001 02:51:04 GMT
Subject: Install PythonWin on top of Python 2.0
Message-ID: <95l4en$rbs$1@nnrp1.deja.com>

Hi, there:

  This is a question from a new user. I recently installed the BeOpen
Python 2.0 on my Win98 PC. Now I'd like to install the Win32 extension
so I download "win32all-125.exe". Unfortunately, I cannot install it
because this build insists on the presence of Python 1.5.x. Now, do I
have to install 1.5 so as to install PythonWin? Would the author of
PythonWin please do us a favor to relax the version check in PythonWin?

  Thanks,

Bin


Sent via Deja.com
http://www.deja.com/


From chris.gonnerman at usa.net  Mon Feb  5 04:39:38 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sun, 4 Feb 2001 21:39:38 -0600
Subject: Install PythonWin on top of Python 2.0
References: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <008601c08f25$48e987c0$a100000a@local>

----- Original Message -----
From: "Deja User" <binary3418 at my-deja.com>
Subject: Install PythonWin on top of Python 2.0


> Hi, there:
>
>   This is a question from a new user. I recently installed the BeOpen
> Python 2.0 on my Win98 PC. Now I'd like to install the Win32 extension
> so I download "win32all-125.exe". Unfortunately, I cannot install it
> because this build insists on the presence of Python 1.5.x. Now, do I
> have to install 1.5 so as to install PythonWin? Would the author of
> PythonWin please do us a favor to relax the version check in PythonWin?

You have the wrong version.  Sadly the python.org site is a bit confusing;
ActiveState
distributes this software, and you need the 2.0 compatible version.  Visit:

    http://www.activestate.com/Products/ActivePython/win32all.html

for downloads and info.






From fuess at att.net  Mon Feb  5 04:46:32 2001
From: fuess at att.net (David Fuess)
Date: Mon, 05 Feb 2001 03:46:32 GMT
Subject: Install PythonWin on top of Python 2.0
References: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <je8s7tsh9sj8mdki71q9kf3b54og3f0nca@4ax.com>

You need version 135 of Win32 all. You can get it from Active State
at:
http://www.activestate.com/download/ActivePython/windows/win32all/win32all-135.exe

Dave

On Mon, 05 Feb 2001 02:51:04 GMT, Deja User <binary3418 at my-deja.com>
wrote:

>Hi, there:
>
>  This is a question from a new user. I recently installed the BeOpen
>Python 2.0 on my Win98 PC. Now I'd like to install the Win32 extension
>so I download "win32all-125.exe". Unfortunately, I cannot install it
>because this build insists on the presence of Python 1.5.x. Now, do I
>have to install 1.5 so as to install PythonWin? Would the author of
>PythonWin please do us a favor to relax the version check in PythonWin?
>
>  Thanks,
>
>Bin
>
>
>Sent via Deja.com
>http://www.deja.com/



From neilh at scintilla.org  Mon Feb  5 04:47:13 2001
From: neilh at scintilla.org (Neil Hodgson)
Date: Mon, 05 Feb 2001 03:47:13 GMT
Subject: Install PythonWin on top of Python 2.0
References: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <5npf6.276$tT3.2649@news-server.bigpond.net.au>

    Hi Bin,

>   This is a question from a new user. I recently installed the BeOpen
> Python 2.0 on my Win98 PC. Now I'd like to install the Win32 extension
> so I download "win32all-125.exe". Unfortunately, I cannot install it
> because this build insists on the presence of Python 1.5.x. Now, do I
> have to install 1.5 so as to install PythonWin? Would the author of
> PythonWin please do us a favor to relax the version check in PythonWin?

   You should install a version of PythonWin designed to work with Python
2.0. Version 135 is a good one and can be downloaded from
http://www.activestate.com/Products/ActivePython/win32all.html

   Neil




From tim.one at home.com  Mon Feb  5 04:52:03 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 4 Feb 2001 22:52:03 -0500
Subject: Install PythonWin on top of Python 2.0
In-Reply-To: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEIHINAA.tim.one@home.com>

[Deja User]
>   This is a question from a new user. I recently installed the
> BeOpen Python 2.0 on my Win98 PC. Now I'd like to install the Win32
> xtension so I download "win32all-125.exe". Unfortunately, I cannot
> install it because this build insists on the presence of Python 1.5.x.

That's because 125 only knew about 1.5.x -- unfortunately, it wasn't able to
see into the future so that it could work correctly with a release of Python
that didn't yet exist <wink>.

> Now, do I have to install 1.5 so as to install PythonWin? Would the
> author of PythonWin please do us a favor to relax the version check
> in PythonWin?

Heh.  Some favor!  That would probably send your machine crashing in flames.

You simply need to get a newer version of win32all.  Look here:

http://www.activestate.com/Products/ActivePython/win32all.html

The version for use with 2.0 is clearly marked.

Or you can uninstall BeOpen 2.0 and get ActiveState's 2.0 package (which
includes the Win32 extensions) instead:

http://www.activestate.com/Products/ActivePython/Download.html

consumer-choice-is-a-wonderful-thing-although-it's-just-confusing-
    at-first-ly y'rs  - tim




From samschul at pacbell.net  Mon Feb  5 09:07:41 2001
From: samschul at pacbell.net (Sam Schulenburg)
Date: Mon, 05 Feb 2001 08:07:41 GMT
Subject: Install PythonWin on top of Python 2.0
References: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <95ln0b$7vb$1@nnrp1.deja.com>

In article <95l4en$rbs$1 at nnrp1.deja.com>,
  Deja User <binary3418 at my-deja.com> wrote:

You need to get the 2.0 build I think it is build 135

> Hi, there:
>
>   This is a question from a new user. I recently installed the BeOpen
> Python 2.0 on my Win98 PC. Now I'd like to install the Win32 extension
> so I download "win32all-125.exe". Unfortunately, I cannot install it
> because this build insists on the presence of Python 1.5.x. Now, do I
> have to install 1.5 so as to install PythonWin? Would the author of
> PythonWin please do us a favor to relax the version check in
PythonWin?
>
>   Thanks,
>
> Bin
>
> Sent via Deja.com
> http://www.deja.com/
>


Sent via Deja.com
http://www.deja.com/


From binary3418 at my-deja.com  Mon Feb  5 11:16:49 2001
From: binary3418 at my-deja.com (Deja User)
Date: Mon, 05 Feb 2001 10:16:49 GMT
Subject: Install PythonWin on top of Python 2.0
References: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <95luie$e1g$1@nnrp1.deja.com>

Hi,

  Thanks to all the replies. I've downloaded the right build and the
installation is smooth.

  Regards,

Bin


Sent via Deja.com
http://www.deja.com/


From parz at home.com  Mon Feb  5 04:05:20 2001
From: parz at home.com (Parzival Herzog)
Date: Mon, 05 Feb 2001 03:05:20 GMT
Subject: How to detect client closed socket in server?
Message-ID: <QLof6.10526$wI1.817063@news1.rdc1.mb.home.com>

I have written a client-server application using socket.py. The model is one
where the
server loops accepting connections, starting worker threads for each
connection.
I am having two problems:

1) the client is able to detect when the server disconnects (it gets an
exception),
    but the server is unable to detect when the client disconnects using
socket.close()
   (no exception arises in the server). An exception will aris in the server
if  the
   client process terminates.

2) Although the maximum number of connections is limited to 16, in practise
my
   server accepts an unlimited number of connections.


An earlier newsgroup thread mentioned a similar, with some responders
asserted that
the client side "should" receive an exception, but this seemed to be an
opinion, not
a hard fact. I most definitely want an exception when the client closes, and
can't seem to
get one.

I am running Python 2.0 on Win 2000. Can anyone tell if this is a bug (in
socket.py),
or if I am doing something incorrect?

-- Parzival.

----------------------------------------------------------------------------
---------------------
# extract of the server code:

ServerPort = 10800
MaxConnections = 16         # Maximum number of connections to serve

class Server(threading.Thread):

    def __init__(self,name,ServerClass):
        self.ServerClass = ServerClass
        threading.Thread.__init__(self, name=name)

    def run(self):
        try:
            sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
            sock.bind(('', ServerPort))
            sock.listen(MaxConnections)
        except Exception, E:
            SocketDiagnostic(E, "Unable to start server.")
        else:
            try:
                try:
                    while 1:
                        self.ServerClass(sock.accept()).start()
                except Exception, E:
                    SocketDiagnostic(E, "Unable to accept new connections.")
            finally:
                sock.close()

class EmsUpdate(threading.Thread):

    def __init__(self,conn):
        self.client, self.addr = conn
        threading.Thread.__init__(self, name="EmsUpdate")

    def run(self):
        print "Connected to", self.addr
        try:
            try:
                MaxReqSize = 9600
                self.client.setsockopt(socket.SOL_SOCKET,
socket.SO_RCVTIMEO, struct.pack("ll",5000,0))
                while self.KeepRunning:
                    req = self.client.recv(MaxReqSize)
                    if not req:
                        print "Null request received from", self.addr
                        break
                    self.ProcessRequest(req)
            except Exception, E:
                # I am expecting an exception (socket.error) here if the
client closes its
                # socket, but none arises.
                SocketDiagnostic(E,"Unable to provide EmsUpdate services for
%s." % str(self.addr))
        finally:
            self.client.close()
            print "Disconnected from", self.addr

    def ProcessRequest(self,req):
        try:
            # service the request (specific request service code is not
shown)
            func, args = eval(req, {})
            res = apply(self.Requests[func],args)
            self.client.send(repr(res))
        except Exception, E:
            SocketDiagnostic(E,"Unable to complete '%s' request from %s." %
(func, str(self.addr)))

# Start a server using the EmsUpdate class to serve
Server("EmsUpdateServer", EmsUpdate).start()
----------------------------------------------------------------------------
----------------------

Extract of the client code.


    def run(self):
        while self.keeprunning:
            self.Log("Trying to connect to %s port %d" % (ServerHost,
ServerPort))
            sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
            try:
                try:
                    sock.connect((ServerHost, ServerPort))
                    sock.setsockopt(socket.SOL_SOCKET, socket.SO_RCVTIMEO,
struct.pack("ll",ServerTimeoutMs,0))
                    self.Log("Connected to " + str(sock.getpeername()))
                    # Compaose a request for the server
                    rq = repr(("PAGE", ((), self.Page,
self.DataEvents.keytexts())))
                    sock.send(rq)
                    rcv = sock.recv(32)
                    assert int(rcv) == len(self.DataEvents)

                    while self.keeprunning:
                        # Wait for an update of the page
                        rq=repr(("GETWAIT", (self.Page,)))
                        sock.send(rq)
                        rcv = sock.recv(UpdateBufferSize)
                        assert rcv
                        ...  do soemthing with rcv ...

                    # Politely end the connection
                    rq=repr(("CLOSE", ()))
                    sock.send(rq)
                    rcv = sock.recv(32)
                    assert rcv == "1"

                except Exception, E:
                    msg = "Exception: " +
SocketErrCodes.SpecificErrMsg(E)[1]
                    tb = sys.exc_info()[2]
                    msg = msg + " at line " + str(traceback.tb_lineno(tb))
                    del tb
                    self.Log(msg)
            finally:
                # ********************* This code has no effect on the
server ********************
                self.Log("Shutting down connection")
                sock.close()
                del sock # just to make sure!

            for i in range(ReConnectSecs):
                if self.keeprunning:
                    time.sleep(1.0)





From craigf at ilid.com.au  Mon Feb  5 06:52:31 2001
From: craigf at ilid.com.au (Craig Findlay)
Date: Mon, 05 Feb 2001 16:52:31 +1100
Subject: Socket woes and signals
Message-ID: <3dfs7tkv4g6ulmeng4a1pnf94b1dfak2rd@4ax.com>

I am writing a simple socket server on an OpenBSD machine, and I would
like to program the following functionality:

If I send the server a SIGTERM signal, I want it to break out of the
loop that it is in, and close gracefully. 

The loop is basically the standard:

while 1:
        conn, addr = s.accept() 
        data = conn.recv(bufsize)
        if not data: break
        do something with data
conn.close()

Q. This is fine if the client closes the connection, but how do I
break out of the look in response to say a signal?

If I force the loop to end ungracefully, the socket is left hanging
and I have to reboot the machine to free it up. But the code normally
stops at the accept line until a connection is accepted, so how do I
close the socket properly in that state?

Craig


From max at alcyone.com  Mon Feb  5 18:12:12 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 05 Feb 2001 09:12:12 -0800
Subject: Socket woes and signals
References: <3dfs7tkv4g6ulmeng4a1pnf94b1dfak2rd@4ax.com>
Message-ID: <3A7EDEEC.3F1D395B@alcyone.com>

Craig Findlay wrote:

> The loop is basically the standard:
> 
> while 1:
>         conn, addr = s.accept()
>         data = conn.recv(bufsize)
>         if not data: break
>         do something with data
> conn.close()
> 
> Q. This is fine if the client closes the connection, but how do I
> break out of the look in response to say a signal?

You can change the signal handler to raise a custom exception which you
can then catch and act on.

> If I force the loop to end ungracefully, the socket is left hanging
> and I have to reboot the machine to free it up.

It takes a few minutes for it to clear, but you won't need to actually
reboot to free it up.

> But the code normally
> stops at the accept line until a connection is accepted, so how do I
> close the socket properly in that state?

If that's all you need, there's an easier way.  AFter you create the
socket, call:

    s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)

Now you won't have to wait to rebind to the address.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    7 sisters productions / http://www.7sisters.com/
 Web design for the future.


From donn at u.washington.edu  Mon Feb  5 18:36:50 2001
From: donn at u.washington.edu (Donn Cave)
Date: 5 Feb 2001 17:36:50 GMT
Subject: Socket woes and signals
References: <3dfs7tkv4g6ulmeng4a1pnf94b1dfak2rd@4ax.com>
Message-ID: <95mobi$q44$1@nntp6.u.washington.edu>

Quoth Craig Findlay <craigf at ilid.com.au>:
| I am writing a simple socket server on an OpenBSD machine, and I would
| like to program the following functionality:
|
| If I send the server a SIGTERM signal, I want it to break out of the
| loop that it is in, and close gracefully. 
|
| The loop is basically the standard:
|
| while 1:
|         conn, addr = s.accept() 
|         data = conn.recv(bufsize)
|         if not data: break
|         do something with data
| conn.close()
|
| Q. This is fine if the client closes the connection, but how do I
| break out of the look in response to say a signal?
|
| If I force the loop to end ungracefully, the socket is left hanging
| and I have to reboot the machine to free it up. But the code normally
| stops at the accept line until a connection is accepted, so how do I
| close the socket properly in that state?

I think you're in a worse pickle than you realize.  It's easy to
interrupt accept() with a signal, you just need a signal handler
(see documentation for the signal module) and an exception handler
(after the signal, accept() will raise an EINTR, at least it does
on NetBSD 1.5.)

But in any event, when the process that bound the socket exits, with
any luck the bound port should close and you won't need to reboot
the computer.  When it gets like that, you probably aren't going to 
solve everything with a SIGTERM anyway.  It couldn't hurt to turn
on the SOL_SOCKET attribute SO_REUSEADDR (see setsockopt.)

	Donn Cave, donn at u.washington.edu


From craigf at ilid.com.au  Wed Feb  7 00:46:21 2001
From: craigf at ilid.com.au (Craig Findlay)
Date: Wed, 07 Feb 2001 10:46:21 +1100
Subject: Socket woes and signals
References: <3dfs7tkv4g6ulmeng4a1pnf94b1dfak2rd@4ax.com> <95mobi$q44$1@nntp6.u.washington.edu>
Message-ID: <b1318t8p409au8htbg9rci0igvqd6pmijb@4ax.com>

Donn Cave <donn at u.washington.edu> wrote:

>But in any event, when the process that bound the socket exits, with
>any luck the bound port should close and you won't need to reboot
>the computer.  When it gets like that, you probably aren't going to 
>solve everything with a SIGTERM anyway.  It couldn't hurt to turn
>on the SOL_SOCKET attribute SO_REUSEADDR (see setsockopt.)

Thanks Don for your help. I'll do some reading and some more playing
:)

Cheers, Craig




From p.g at figu.no  Mon Feb  5 09:01:00 2001
From: p.g at figu.no (Per Gummedal)
Date: Mon, 5 Feb 2001 09:01:00 +0100
Subject: ANNOUNCE: Python for AS/400.
Message-ID: <1921465137616.20010205090100@figu.no>

Here is a port of Python2.0 to AS/400.
http://home.no.net/pgummeda

It includes a File400 class.(recordlevel access to DB2 for AS/400)


Per





From p.g at figu.no  Thu Feb  8 00:31:24 2001
From: p.g at figu.no (Per Gummedal)
Date: Thu, 8 Feb 2001 00:31:24 +0100
Subject: ANNOUNCE: Python for AS/400, New update available
In-Reply-To: <1921465137616.20010205090100@figu.no>
References: <1921465137616.20010205090100@figu.no>
Message-ID: <1501693700732.20010208003124@figu.no>

A new update is now available at http://home.no.net/pgummeda

Changes since last update:
 - Codepage/ccsid problems fixed. (I hope)
 - Some bug-fixes on File400 module.


Per





From vonehr at ira.uka.de  Mon Feb  5 11:33:26 2001
From: vonehr at ira.uka.de (Markus von Ehr)
Date: Mon, 05 Feb 2001 11:33:26 +0100
Subject: COM objects in threads
Message-ID: <3A7E8176.2B2DBB@ira.uka.de>

Hallo,

I still didn't find out the solution of accessing a COM object from
a thread.
I play a mp3 File with MediaPlayer using COM like this:
(Globals. are my global variables)

  Globals.ie =
win32com.client.dynamic.Dispatch("InternetExplorer.Application")
  Globals.ie.Navigate(r'about:blank')
  Globals.doc = Globals.ie.Document
  Globals.doc.body.insertAdjacentHTML("afterBegin",
  """
  <OBJECT ID="MediaPlayer1"
  CLASSID="CLSID:22D6f312-B0F6-11D0-94AB-0080C74C7E95">
  </OBJECT>
  """)
    
  Globals.mp = Globals.doc.body.children("MediaPlayer1").object

  Globals.mp.FileName = "MySoundFile.mp3"

In another thread I want to set the New SoundFile:

  Globals.mp.FileName = "null.wav"

but I have a Pythoncommerror.

Anybody can help me accessing the mp object in the thread? How
can I set the new FileName?

Thanks for any comments,

Markus


From MarkH at ActiveState.com  Mon Feb  5 13:30:08 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Mon, 05 Feb 2001 12:30:08 GMT
Subject: COM objects in threads
References: <3A7E8176.2B2DBB@ira.uka.de>
Message-ID: <3A7E9BE2.1040203@ActiveState.com>

Markus von Ehr wrote:

 
> Anybody can help me accessing the mp object in the thread? How
> can I set the new FileName?

I have already replied to this - there is no easy answer.

Either, you must ensure _all_ threads, including the main thread, are in the free-threading apartment.  This will involve setting sys.coinit_flags=0 before importing pythoncom for the first time, and ensuring you call CoInitializeEx(0) on each thread you create.

The only other alternative is to use pythoncom.CoMarshalInterThreadInterfaceInStream and pythoncom.CoGetInterfaceAndReleaseStream to pass a COM object between 2 apartment-model threads.  The default is apartment-model threading.  COM says it is illegal to share objects between apartments, and as you have seen, often enforces it!

You are going to have to find a reasonable reference on COM threading models, or see the "threading" appendix in "Python Programming on Win32" - that appendix was hell to write, and I am not going to attempt it again on Usenet ;-)  Deja threads (no pun intended;) on the subject may also help.

Mark.



From aleaxit at yahoo.com  Mon Feb  5 14:02:02 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 5 Feb 2001 14:02:02 +0100
Subject: COM objects in threads
References: <3A7E8176.2B2DBB@ira.uka.de> <3A7E9BE2.1040203@ActiveState.com>
Message-ID: <95m88c01cph@news1.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A7E9BE2.1040203 at ActiveState.com...
    [snip]
> You are going to have to find a reasonable reference on COM threading
models, or see the "threading" appendix in "Python Programming on Win32" -
that appendix was hell to write, and I am not going to attempt it again on
Usenet ;-)  Deja threads (no pun intended;) on the subject may also help.

Don Box, "Essential COM", Addison-Wesley; a MUST if you're doing
anything advanced with COM -- and communicating COM objects
between multiple threads qualifies as advanced.  Also very
helpful reading, AFTER "Essential COM", is "Effective COM", by
Box et al, also Addison-Wesley.


Alex





From mbel44 at dial.pipex.net  Mon Feb  5 15:52:18 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Mon, 05 Feb 2001 14:52:18 +0000
Subject: COM objects in threads
References: <3A7E8176.2B2DBB@ira.uka.de> <3A7E9BE2.1040203@ActiveState.com>
Message-ID: <hdft7to8kqeg9m7l4dlk1eojg1t52adh9l@4ax.com>

Mark Hammond <MarkH at ActiveState.com> wrote:

>The only other alternative is to use pythoncom.CoMarshalInterThreadInterfaceInStream and pythoncom.CoGetInterfaceAndReleaseStream to pass a COM object between 2 apartment-model threads.  The default is apartment-model threading.  COM says it is illegal to share objects between apartments, and as you have seen, often enforces it!

Also, you will need to run a message loop in *both* threads.

>You are going to have to find a reasonable reference on COM threading models, or see the "threading" appendix in "Python Programming on Win32" - that appendix was hell to write, and I am not going to attempt it again on Usenet ;-)  Deja threads (no pun intended;) on the subject may also help.

I can recommend this; its one of the best summaries around.

Toby Dickenson
tdickenson at geminidataloggers.com


From vonehr at ira.uka.de  Wed Feb  7 11:52:47 2001
From: vonehr at ira.uka.de (Markus von Ehr)
Date: Wed, 07 Feb 2001 11:52:47 +0100
Subject: COM objects in threads
References: <3A7E8176.2B2DBB@ira.uka.de>
Message-ID: <3A8128FF.268F5EDC@ira.uka.de>

Thanx for all answers,

finally I could solve the problem using the 
  pythoncom.CoMarshalInterThreadInterfaceInStream
without the book (just kiddin Mark ;-) ), after having two beers
and taking a sharp look at the testMarshal.py program. The last 
problem was a repeated COM call in the Thread. It is only possible,
if CoUninitialize is omitted.

    try:
	pythoncom.CoInitialize()
	if first_ThreadCall == 1:
	    interp =
pythoncom.CoGetInterfaceAndReleaseStream(Globals.interpStream,
		pythoncom.IID_IDispatch)
	    interp = win32com.client.Dispatch(interp)
	    doc = interp.Document
	    mplayer = doc.body.children("MediaPlayer1").object
	    first_ThreadCall = 0
	        
	mplayer.FileName = "C:/Programme/Python/lightcontrol/null.wav"
	  
    except pythoncom.com_error, details:
        print "Pythoncomerror, ", details

The first time I thought the COM interface is the biggest shit in
the world but know I think it is quite a simple interface, when it
works. I'm just implementing my next COM interface using win32com.
But it's quite slow.

Markus.


From info at pythonware.com  Mon Feb  5 11:55:49 2001
From: info at pythonware.com (PythonWare)
Date: Mon, 5 Feb 2001 11:55:49 +0100
Subject: ANN: tkinter 3000 wck, release 1.0 alpha 1
Message-ID: <010c01c08f62$34446b40$0900a8c0@SPIFF>

the labs strikes again:
 
    The Tkinter 3000 project is working on a couple of enhancements to
    Python's popular Tkinter toolkit.  First out is the Widget Construction
    Kit, which allows you to write new Tkinter widgets in pure Python.

    The first public alpha is now available from:

        http://www.pythonware.com/products/tkinter/tkinter3000.htm

    This distribution includes library source code, a couple of simple
    demos, and prebuilt Windows binaries for Python 1.5.2 and 2.0.
    See the site for documentation and information about other Tkinter
    3000 3000 components.

enjoy,
the pythonware team

    "Secret Labs -- makers of fine pythonware since 1997."




From chris.gonnerman at usa.net  Mon Feb  5 15:19:36 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Mon, 5 Feb 2001 08:19:36 -0600
Subject: SIP
Message-ID: <000901c08f7e$b317b320$a100000a@local>

Does anyone know of a tutorial on SIP?  I am a C (not C++) programmer, but
there
is a specific C++ library I'd like to 'wrap' and I am not finding docs
anywhere.

Sorry if I've missed an obvious link.





From aleaxit at yahoo.com  Mon Feb  5 18:09:41 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 5 Feb 2001 18:09:41 +0100
Subject: SIP
References: <mailman.981382527.27449.python-list@python.org>
Message-ID: <95mmok020jv@news1.newsguy.com>

"Chris Gonnerman" <chris.gonnerman at usa.net> wrote in message
news:mailman.981382527.27449.python-list at python.org...
> Does anyone know of a tutorial on SIP?  I am a C (not C++) programmer, but
> there
> is a specific C++ library I'd like to 'wrap' and I am not finding docs
> anywhere.

Can't help you on SIP, but, have a look at http://www.boost.org -- the
Boost Python docs are a decent start (although I'm afraid they DO take
it for granted that the reader is fluent in C++ _and_ Python).


Alex





From phil at river-bank.demon.co.uk  Mon Feb  5 18:37:14 2001
From: phil at river-bank.demon.co.uk (phil at river-bank.demon.co.uk)
Date: Mon, 05 Feb 2001 17:37:14 -0000
Subject: SIP
In-Reply-To: <000901c08f7e$b317b320$a100000a@local>
Message-ID: <95moca+k29p@eGroups.com>

--- In python-list at y..., "Chris Gonnerman" <chris.gonnerman at u...>
wrote:
> Does anyone know of a tutorial on SIP?  I am a C (not C++)
programmer, but
> there
> is a specific C++ library I'd like to 'wrap' and I am not finding
docs
> anywhere.
> 
> Sorry if I've missed an obvious link.

Sorry - documentation is the weakest aspect of SIP. The best approach
is to use PyQt as a working example. Feel free to ask SIP questions on
the PyQt/PyKDE mailing list.

Phil




From chris.gonnerman at usa.net  Tue Feb  6 05:10:34 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Mon, 5 Feb 2001 22:10:34 -0600
Subject: SIP
References: <95moca+k29p@eGroups.com>
Message-ID: <00ba01c08ff2$c9b34300$a100000a@local>

----- Original Message -----
From: <phil at river-bank.demon.co.uk>
Subject: Re: SIP


> Sorry - documentation is the weakest aspect of SIP. The best approach
> is to use PyQt as a working example. Feel free to ask SIP questions on
> the PyQt/PyKDE mailing list.

Except I know squat about QT, and as I am a GNOME user I have little
interest.

What I am wanting to do is connect the Xbase library to Python.  Xbase is a
dBase
compatible C++ library.

I am probably not qualified to do this anyway.  I have many thousands of
lines of
C in my history but zero C++.






From phil at river-bank.demon.co.uk  Tue Feb  6 11:39:22 2001
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Tue, 06 Feb 2001 10:39:22 +0000
Subject: SIP
References: <95moca+k29p@eGroups.com> <00ba01c08ff2$c9b34300$a100000a@local>
Message-ID: <3A7FD45A.F5B581ED@river-bank.demon.co.uk>

Chris Gonnerman wrote:
> 
> ----- Original Message -----
> From: <phil at river-bank.demon.co.uk>
> Subject: Re: SIP
> 
> > Sorry - documentation is the weakest aspect of SIP. The best approach
> > is to use PyQt as a working example. Feel free to ask SIP questions on
> > the PyQt/PyKDE mailing list.
> 
> Except I know squat about QT, and as I am a GNOME user I have little
> interest.

I just meant to look at it as a good example of SIP specification files
for a variety of C++ classes. I wasn't suggesting polluting you system
by compiling it. :)

> What I am wanting to do is connect the Xbase library to Python.  Xbase is a
> dBase
> compatible C++ library.
> 
> I am probably not qualified to do this anyway.  I have many thousands of
> lines of
> C in my history but zero C++.

Phil



From g_will at cyberus.ca  Mon Feb  5 16:03:04 2001
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 5 Feb 2001 10:03:04 -0500
Subject: Installer problem with NT/ KERNEN32.dll error
Message-ID: <95mfak$2snv$1@news2.ottawa.cyberus.ca>

Hello All,

Using:
python 2.0
wxPython 2.2.2
W95/ NT 4 sp6a
Gordon's Installer 3i

I have a problem with a program that I  have made into a standalone exe
using Gordon's Installer.  I'm not sure where the problem lies.  It works
under W95 but not under NT.  When I try and start the program I get the
following message box on the screen:

_GSocketInternal_Window_Name: Scopal06B.exe - Entry Point Not Found
The procedure entry point GetHandleContext could not be lockated in the
dynamic link library KERNEN32.dll


The traceback from the dos window is:


 File "e:\program files\python\lib\imputil.py", line 170, in
_import_top_module

    module = item.import_top(name)
  File "e:\program files\python\lib\imputil.py", line 197, in import_top
    return self._import_one(None, name, name)
  File "e:\program files\python\lib\imputil.py", line 252, in _import_one
    module = self._process_result(result, fqname)
  File "e:\program files\python\lib\imputil.py", line 281, in
_process_result
    exec code in module.__dict__
  File "e:\program files\python\junk_files\serial.py", line 3, in ?
    from win32file import * # The base COM port and file IO functions.
  File "e:\program files\python\lib\imputil.py", line 88, in _import_hook
    top_module = self._import_top_module(parts[0])
  File "e:\program files\python\lib\imputil.py", line 168, in
_import_top_module

    module = self.fs_imp.import_from_dir(item, name)
  File "e:\program files\python\lib\imputil.py", line 512, in
import_from_dir
    result = self._import_pathname(_os_path_join(dir, fqname), fqname)
  File "e:\program files\python\lib\imputil.py", line 545, in
_import_pathname
    return importFunc(filename, finfo, fqname)
  File "e:\program files\python\lib\imputil.py", line 578, in import_file
    module = imp.load_module(fqname, fp, filename, self.desc)
ImportError: DLL load failed: The specified procedure could not be found.


I'm using win32file for some serial IO communications.

Any idea what is causing this and what the work around is?  I thought that
if it worked on W95 it would work on NT, but this is not the case:(

Regards,

Gordon Williams





From senux at my-deja.com  Mon Feb  5 16:14:22 2001
From: senux at my-deja.com (senux at my-deja.com)
Date: Mon, 05 Feb 2001 15:14:22 GMT
Subject: [mod_python] Getting data from FORM tag
Message-ID: <95mg04$sa6$1@nnrp1.deja.com>

Hi,

I'm new at mod_python and want to use the
fuction of embedded Python code in HTML file by
mod_snake.

There are 2 files. One form.html and the other
is login.epy. I want to get what the web visitor
inputed in <FORM>..</FORM> tag. How can I code
them?

Thanks in any postings. :)

 - Brian,.


Sent via Deja.com
http://www.deja.com/


From mrq at for.mat.bham.ac.uk  Mon Feb  5 16:35:27 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Mon, 5 Feb 2001 15:35:27 +0000
Subject: Beginning to learn python...help needed!
Message-ID: <Pine.GSO.4.21.0102051527590.21294-100000@babbage>

I'm just beginning to learn Python.  I've done a little programming before
(so I'm quite happy with manipulating lists and using for/while loops,
etc.), but I'm having trouble understanding the "Object Oriented" part of
the programming.  Can anyone give some advice when is the right time to
use a class, as opposed to just using the def command?

I'm also getting nowhere with Tkinter.  I tried to look at the Tkinter
tutorial which can be downloaded from python.org, but found that it went
far to fast without nearly enough explanation.  I've seen that wxPython is
an alternative, but someone said that there was less documentation
available for that.  Given what I've found with the documentation for
Tkinter, this comment puts me off slightly.  Can anyone point me in the
direction of detailed info on an appropriate GUI?  (Am I right in thinking
that the reason I'm having trouble with Tkinter is that the Object
Orientedness is more apparent early on?)

Apologies for writing a rather long and rambly message!

Thanks,

Martyn

P.S. I cannot afford to buy a book on the subject - I'm a rather junior
academic in the UK and don't get paid very highly!!!

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick



From bobhicks at adelphia.net  Mon Feb  5 17:51:13 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Mon, 05 Feb 2001 16:51:13 GMT
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage>
Message-ID: <B6A44420.A042%bobhicks@adelphia.net>

"Python an Tkinter Programming" by John Grayson is an excellent work and has
great info on Tkinter as well as PMW.

I think Mark Lutz is coming out with an update to his tome around March or
so.

Hmmm...I would use a class when I have a collection of objects with
identical characteristics.

Example:

class Message:
    def __init__(self, aString):
        self.text = aString
    def printIt(self):
        print self.text

Having defined a class you can now create instances of the Message class and
manipulate them:

m1 = Message("Hello world!")
m2 = Message("So long and thanks for all the fish!")

note = [m1, m2) # put the objects in a list
for msg in note:
    msg.printIt() # print each message in turn

Hope this helps. The example was taken from "Learning to Program using
Python" by A. Gauld. He also has a web site at:

http://www.crosswinds.net/~agauld/

I have found both to be extremely useful!

- Bob



> From: Martyn Quick <mrq at for.mat.bham.ac.uk>
> Organization: The University of Birmingham news server
> Newsgroups: comp.lang.python
> Date: Mon, 5 Feb 2001 15:35:27 +0000
> Subject: Beginning to learn python...help needed!
> 
> I'm just beginning to learn Python.  I've done a little programming before
> (so I'm quite happy with manipulating lists and using for/while loops,
> etc.), but I'm having trouble understanding the "Object Oriented" part of
> the programming.  Can anyone give some advice when is the right time to
> use a class, as opposed to just using the def command?
> 
> I'm also getting nowhere with Tkinter.  I tried to look at the Tkinter
> tutorial which can be downloaded from python.org, but found that it went
> far to fast without nearly enough explanation.  I've seen that wxPython is
> an alternative, but someone said that there was less documentation
> available for that.  Given what I've found with the documentation for
> Tkinter, this comment puts me off slightly.  Can anyone point me in the
> direction of detailed info on an appropriate GUI?  (Am I right in thinking
> that the reason I'm having trouble with Tkinter is that the Object
> Orientedness is more apparent early on?)
> 
> Apologies for writing a rather long and rambly message!
> 
> Thanks,
> 
> Martyn
> 
> P.S. I cannot afford to buy a book on the subject - I'm a rather junior
> academic in the UK and don't get paid very highly!!!
> 
> --------------------------------------------------------
> Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
> University of Birmingham, Edgbaston, Birmingham, UK.
> http://www.mat.bham.ac.uk/M.R.Quick
> 



From mrq at for.mat.bham.ac.uk  Tue Feb  6 13:00:27 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Tue, 6 Feb 2001 12:00:27 +0000
Subject: Beginning to learn python...help needed!
In-Reply-To: <B6A44420.A042%bobhicks@adelphia.net>
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage> <B6A44420.A042%bobhicks@adelphia.net>
Message-ID: <Pine.GSO.4.21.0102061155150.26792-100000@babbage>

On Mon, 5 Feb 2001, Robert L Hicks wrote:

> "Python an Tkinter Programming" by John Grayson is an excellent work and has
> great info on Tkinter as well as PMW.

What's PMW?

We don't have this book in the library at the university where I
work.  (Indeed, I don't think we have any books on python in the
library!)  I looked on Amazon and saw that the book cost 31 pounds - which
I guess is fairly standard for computing books, but rather expensive for
my budget.  Are there any *very cheap* ways of learning about Tkinter?

> Hope this helps. The example was taken from "Learning to Program using
> Python" by A. Gauld. He also has a web site at:
> 
> http://www.crosswinds.net/~agauld/
> 
> I have found both to be extremely useful!

I've been looking at this website and found it very useful.

Martyn

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick





From bobhicks at adelphia.net  Tue Feb  6 14:53:17 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Tue, 06 Feb 2001 13:53:17 GMT
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage> <B6A44420.A042%bobhicks@adelphia.net> <Pine.GSO.4.21.0102061155150.26792-100000@babbage>
Message-ID: <B6A56BF9.A111%bobhicks@adelphia.net>

 
>> "Python an Tkinter Programming" by John Grayson is an excellent work and has
>> great info on Tkinter as well as PMW.
> 
> What's PMW?

PMW = Python Mega-widgets, they are sorta super widgets built using Python
and Tkinter.
 
> We don't have this book in the library at the university where I
> work.  (Indeed, I don't think we have any books on python in the
> library!)  I looked on Amazon and saw that the book cost 31 pounds - which
> I guess is fairly standard for computing books, but rather expensive for
> my budget.  Are there any *very cheap* ways of learning about Tkinter?
> 
>> Hope this helps. The example was taken from "Learning to Program using
>> Python" by A. Gauld. He also has a web site at:
>> 
>> http://www.crosswinds.net/~agauld/
>> 
>> I have found both to be extremely useful!
> 
> I've been looking at this website and found it very useful.
> 
> Martyn
> 
> --------------------------------------------------------
> Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
> University of Birmingham, Edgbaston, Birmingham, UK.
> http://www.mat.bham.ac.uk/M.R.Quick
> 
> 
> 



From llewin at oreilly.com  Tue Feb 13 04:52:20 2001
From: llewin at oreilly.com (llewin at oreilly.com)
Date: 13 Feb 2001 03:52:20 GMT
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage> <B6A44420.A042%bobhicks@adelphia.net> <Pine.GSO.4.21.0102061155150.26792-100000@babbage>
Message-ID: <96ab1k$c56$1@news.netmar.com>

Hi,
If you go to oreillynet.com, to the Python Dev Center, you should find a
recent article that lists free Python books online, and provides links.  
Laura





 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse at newsone.net


From aleaxit at yahoo.com  Mon Feb  5 18:06:58 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 5 Feb 2001 18:06:58 +0100
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage>
Message-ID: <95mmjs020fr@news1.newsguy.com>

"Martyn Quick" <mrq at for.mat.bham.ac.uk> wrote in message
news:Pine.GSO.4.21.0102051527590.21294-100000 at babbage...
    [snip]
> the programming.  Can anyone give some advice when is the right time to
> use a class, as opposed to just using the def command?

For a very simple tutorial on why one might use a class as a simple
container of data, http://www.deja.com/getdoc.xp?AN=708325609 --
I never wrote the followups to that about 'more uses for methods',
'utilitarian classes', inheritance, etc, but, at least, it IS
a small start:-).


Alex





From bryan at eevolved.com  Mon Feb  5 18:17:33 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Mon, 05 Feb 2001 17:17:33 GMT
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage>
Message-ID: <NeBf6.203727$f36.8126916@news20.bellglobal.com>

Hi Martyn,

I think its great you chose python to expand on your programming abilities. 
Hopefully you'll become one of those programmers that learnt the best way 
of doing things right off the bat.  Object Oriented (OO) programming is 
built around one fundamental concept:

Everything can (and should) be built with small, *independent* components.

Break a task down conceptually first into its components. Then create 
objects that are ignorant of everything else but their assigned task. Then 
build your complete program by plugging your objects together.

This also has the effect of helping you code in a style that will permit 
you to reuse those objects over and over again, in different projects. 
Other people will be able to use your objects and save time because they 
are self-contained "mini-programs" that accomplish a single task.

> I'm also getting nowhere with Tkinter.  I tried to look at the Tkinter
> tutorial which can be downloaded from python.org, but found that it went
> far to fast without nearly enough explanation.  I've seen that wxPython is
> an alternative, but someone said that there was less documentation
> available for that.  Given what I've found with the documentation for
> Tkinter, this comment puts me off slightly.  Can anyone point me in the
> direction of detailed info on an appropriate GUI?  (Am I right in thinking
> that the reason I'm having trouble with Tkinter is that the Object
> Orientedness is more apparent early on?)
> 

Ugh TKinter. Well, I can make a few recommendations but most (read: all) 
GUI toolkits are heavily OO.  My experience has also taught me that good 
documentation is more important that the features of the toolkit itself, 
due to the fact that if you can't figure out how to use it, it's useless to 
you.

Trolltech develops a premiere GUI toolkit called QT which has by far the 
best documentation of any I've seen. Check out:

http://doc.trolltech.com/

for proof. :)  The toolkit is written in C++ but python wrappings that make 
your job incredibly easy are available at:

http://www.thekompany.com/projects/pykde/

For anything more complicated than a toy program, I would consider QT.

> Apologies for writing a rather long and rambly message!

The ramblier, the better... 
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
www.eevolved.com
<==================================>

"Now he has departed from this strange world a little ahead of me. That 
means nothing. People like us, who believe in physics, know that the 
distinction between past, present, and future is only a stubbornly 
persistent illusion."-- Einstein



From wware at world.std.com  Tue Feb  6 00:41:54 2001
From: wware at world.std.com (Will Ware)
Date: Mon, 5 Feb 2001 23:41:54 GMT
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage>
Message-ID: <G8B5tv.121@world.std.com>

Martyn Quick (mrq at for.mat.bham.ac.uk) wrote:
> I'm having trouble understanding the "Object Oriented" part of
> the programming.  Can anyone give some advice when is the right time to
> use a class, as opposed to just using the def command?

Here is what I do. I usually start out writing some functions and some
global data structures. Then I notice that data structure A is only used
by functions f and g, and data structure B is only used by functions d
and e. It kinda makes sense to think of (A, f, g) as a sort of package
deal: a computational beast that knows how to perform actions f and g,
and has internal knowledge A, but for the most part, nobody interacting
with this beast needs to deal much directly with A. This beast is then
an object:

class Beast_Afg:
  def __init__(self, ...):
    self.A = <initial value for A>
  def f(self, ...):
    do stuff with self.A
    return something, maybe
  def g(self, ...):
    do stuff with self.A
    return something, maybe

mybeast = Beast_Afg(...)    # and here is how you might use it
mybeast.f(...)
mybeast.g(...)

The next question is, why is this advantageous over the global functions
and global data structure? A couple of reasons. You can hide the definitions
for Beast away in some file that you don't look at very often, to avoid
confusing yourself with too many details. Because each copy of Beast has
its own copy of A, you can have several beasts existing simultaneously
without messing up each others' data. And once you get used to the idea
of structuring stuff as objects, you'll find it quite pleasant and easy
on the poor frazzled neurons.

The downside to objects is that, in order to have useful ones, you need
to invest some time and energy into designing them well. That is, choosing
good choices for A, f, and g, and giving them clear descriptive names.
People sometimes think that object-oriented programming is magical and they
won't have to do any more thinking or working. It's a different (and often
better) way to organize your work but the work still needs to be done. In
particular, people often claim that OOP leads directly to easily-reused
code, much more so than conventional (e.g. C) code. It can, if you invest
the effort to design objects well, and you're prepared to take feedback
from whoever uses those objects (since you probably won't get it exactly
right the first time). There are also tricks for making non-OOP C code
reusable, if you're willing to work equally hard at it.

> I'm also getting nowhere with Tkinter.

I've heard good things about the John Grayson book mentioned in another
response, and intend to pick up a copy one of these days. I find that a
great thing to look at is the Demo/tkinter directory in the Unix source
code for Python. There are lots of examples, and you can try them to see
what they do. Pick one that is doing approximately what you want, and
start tinkering with it.

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From tuttledon at hotmail.com  Tue Feb  6 03:09:20 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Tue, 06 Feb 2001 02:09:20 GMT
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage> <G8B5tv.121@world.std.com>
Message-ID: <k1Jf6.35380$p8.7765406@typhoon.southeast.rr.com>

"Will Ware" <wware at world.std.com> wrote in message
news:G8B5tv.121 at world.std.com...

> import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
> lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will
Ware
> *(x-3))+3)(x),24),range(1,28, 2))),"\n") ################
wware at world.std.com

hmm?...An entry for the "First Anual Obfuscated Python Contest"?
The motto "Why should Perl programmers have all the headaches!"
<WINK!!!> (lest I be too subtle ;-)

Don




From wware at world.std.com  Tue Feb  6 05:25:45 2001
From: wware at world.std.com (Will Ware)
Date: Tue, 6 Feb 2001 04:25:45 GMT
Subject: Beginning to learn python...help needed!
References: <Pine.GSO.4.21.0102051527590.21294-100000@babbage> <G8B5tv.121@world.std.com> <k1Jf6.35380$p8.7765406@typhoon.southeast.rr.com>
Message-ID: <G8BIyx.9I8@world.std.com>

"Will Ware" <wware at world.std.com> wrote in message....
> [[[ dried-out Xmas tree losing its needles ]]]

Don Tuttle (tuttledon at hotmail.com) wrote:
> hmm?...An entry for the "First Anual Obfuscated Python Contest"?

I can't claim authorship, as much as I'd like to. The first time I
saw it was here:
http://www.deja.com/getdoc.xp?AN=707068397
so I attribute it to gunter <gb at a3design.de>. Some day I hope to
devote some mental energy to coming up with something similarly
convoluted but not so seasonal.

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From akuchlin at mems-exchange.org  Mon Feb  5 16:46:39 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: Mon, 05 Feb 2001 10:46:39 -0500
Subject: Alpha release of ZODB programming guide
Message-ID: <E14Pnqh-0006eL-00@ute.cnri.reston.va.us>

For what seems like half an eternity, I've been working on a ZODB
programming guide, written for Python programmers who want to use the
ZODB to store data.  The document is placed under the GNU Free
Documentation License.  It's still far from being finished, but while
on vacation I filled it out enough to make it alpha-worthy, though
note that there are still empty sections.  Please take a look at:

http://www.amk.ca/zodb/guide/

What I'd like right now:
     * Suggestions for topics that should be covered.  This would be
the most helpful.

     * Comments on the organization (ordering of chapters, topics, &c.)

     * Reporting errors and inaccuracies would also be helpful, but
are less important at the moment.  Getting the outline correct
is my primary concern at the moment.

--amk

Table of contents:
     * 1 Introduction
          + 1.1 What is ZODB?
          + 1.2 OODBs vs. Relational DBs
          + 1.3 What is ZEO?
          + 1.4 About this guide
     * 2 ZODB Programming
          + 2.1 Installing ZODB
               o 2.1.1 Requirements
               o 2.1.2 Installing the Packages
          + 2.2 How ZODB Works
          + 2.3 Opening a ZODB
          + 2.4 Writing a Persistent Class
          + 2.5 Rules for Writing Persistent Classes
               o 2.5.1 Modifying Mutable Objects
               o 2.5.2 Some Special Methods Don't Work
               o 2.5.3 Fixing isinstance and issubclass
               o 2.5.4 __getattr__, __delattr__, and __setattr__
          + 2.6 Writing Persistent Classes
               o 2.6.1 Changing Class Attributes
     * 3 ZEO
          + 3.1 How ZEO Works
          + 3.2 Installing ZEO
               o 3.2.1 Requirements
               o 3.2.2 Installation
          + 3.3 Configuring and Running a ZEO Server
          + 3.4 Testing the ZEO Installation
          + 3.5 Sample Application: chatter.py
     * 4 Transactions and Versioning
          + 4.1 Subtransactions
          + 4.2 Undoing Changes
          + 4.3 Versions
          + 4.4 Multithreaded ZODB Programs
     * 5 Storages
          + 5.1 FileStorage
          + 5.2 BerkeleyStorage
          + 5.3 OracleStorage
     * 6 Related Modules
          + 6.1 ZODB.PersistentMapping
          + 6.2 ZODB.PersistentList
          + 6.3 B-tree Modules
     * A. Resources
     * B. GNU Free Documentation License



From bobhicks at adelphia.net  Mon Feb  5 17:54:44 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Mon, 05 Feb 2001 16:54:44 GMT
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org>
Message-ID: <B6A444F3.A043%bobhicks@adelphia.net>

AMK -

Will there be a PDF of the *final* version?

- Bob

> From: Andrew Kuchling <akuchlin at mems-exchange.org>
> Organization: Baymountain
> Reply-To: akuchlin at mems-exchange.org
> Newsgroups: comp.lang.python
> Date: Mon, 05 Feb 2001 10:46:39 -0500
> To: python-list at python.org
> Cc: zope-dev at zope.org, zope-zeo at zope.org
> Subject: Alpha release of ZODB programming guide
> 
> For what seems like half an eternity, I've been working on a ZODB
> programming guide, written for Python programmers who want to use the
> ZODB to store data.  The document is placed under the GNU Free
> Documentation License.  It's still far from being finished, but while
> on vacation I filled it out enough to make it alpha-worthy, though
> note that there are still empty sections.  Please take a look at:
> 
> http://www.amk.ca/zodb/guide/
> 
> What I'd like right now:
> * Suggestions for topics that should be covered.  This would be
> the most helpful.
> 
> * Comments on the organization (ordering of chapters, topics, &c.)
> 
> * Reporting errors and inaccuracies would also be helpful, but
> are less important at the moment.  Getting the outline correct
> is my primary concern at the moment.
> 
> --amk
> 
> Table of contents:
> * 1 Introduction
> + 1.1 What is ZODB?
> + 1.2 OODBs vs. Relational DBs
> + 1.3 What is ZEO?
> + 1.4 About this guide
> * 2 ZODB Programming
> + 2.1 Installing ZODB
> o 2.1.1 Requirements
> o 2.1.2 Installing the Packages
> + 2.2 How ZODB Works
> + 2.3 Opening a ZODB
> + 2.4 Writing a Persistent Class
> + 2.5 Rules for Writing Persistent Classes
> o 2.5.1 Modifying Mutable Objects
> o 2.5.2 Some Special Methods Don't Work
> o 2.5.3 Fixing isinstance and issubclass
> o 2.5.4 __getattr__, __delattr__, and __setattr__
> + 2.6 Writing Persistent Classes
> o 2.6.1 Changing Class Attributes
> * 3 ZEO
> + 3.1 How ZEO Works
> + 3.2 Installing ZEO
> o 3.2.1 Requirements
> o 3.2.2 Installation
> + 3.3 Configuring and Running a ZEO Server
> + 3.4 Testing the ZEO Installation
> + 3.5 Sample Application: chatter.py
> * 4 Transactions and Versioning
> + 4.1 Subtransactions
> + 4.2 Undoing Changes
> + 4.3 Versions
> + 4.4 Multithreaded ZODB Programs
> * 5 Storages
> + 5.1 FileStorage
> + 5.2 BerkeleyStorage
> + 5.3 OracleStorage
> * 6 Related Modules
> + 6.1 ZODB.PersistentMapping
> + 6.2 ZODB.PersistentList
> + 6.3 B-tree Modules
> * A. Resources
> * B. GNU Free Documentation License
> 



From akuchlin at mems-exchange.org  Mon Feb  5 18:19:54 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 05 Feb 2001 12:19:54 -0500
Subject: Alpha release of ZODB programming guide (PDF)
References: <mailman.981388047.29255.python-list@python.org> <B6A444F3.A043%bobhicks@adelphia.net>
Message-ID: <3dk875yrlx.fsf_-_@ute.cnri.reston.va.us>

Robert L Hicks <bobhicks at adelphia.net> writes:
> Will there be a PDF of the *final* version?

After nudges from Tony Ibbs and you, I've made a PDF version, now at
http://www.amk.ca/files/zodb/zodb-guide.pdf .  If you're printing it,
you can skip the last 5 pages, which are the text of the GFDL.  (I'll
fix that waste of space in a future version.)

--amk


From rkiendl at gmx.net  Mon Feb  5 18:24:19 2001
From: rkiendl at gmx.net (Robert Kiendl)
Date: Mon, 05 Feb 2001 18:24:19 +0100
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <B6A444F3.A043%bobhicks@adelphia.net>
Message-ID: <3A7EE1C3.EA8EE642@gmx.net>

I'd also like a PDF for alpha version(s)

-robert

Robert L Hicks wrote:
> 
> AMK -
> 
> Will there be a PDF of the *final* version?
> 
> - Bob
> 
> > From: Andrew Kuchling <akuchlin at mems-exchange.org>
> > Organization: Baymountain
> > Reply-To: akuchlin at mems-exchange.org
> > Newsgroups: comp.lang.python
> > Date: Mon, 05 Feb 2001 10:46:39 -0500
> > To: python-list at python.org
> > Cc: zope-dev at zope.org, zope-zeo at zope.org
> > Subject: Alpha release of ZODB programming guide
> >
> > For what seems like half an eternity, I've been working on a ZODB
> > programming guide, written for Python programmers who want to use the
> > ZODB to store data.  The document is placed under the GNU Free
> > Documentation License.  It's still far from being finished, but while
> > on vacation I filled it out enough to make it alpha-worthy, though
> > note that there are still empty sections.  Please take a look at:
> >
> > http://www.amk.ca/zodb/guide/
> >
> > What I'd like right now:
> > * Suggestions for topics that should be covered.  This would be
> > the most helpful.
> >
> > * Comments on the organization (ordering of chapters, topics, &c.)
> >
> > * Reporting errors and inaccuracies would also be helpful, but
> > are less important at the moment.  Getting the outline correct
> > is my primary concern at the moment.
> >
> > --amk
> >
> > Table of contents:
> > * 1 Introduction
> > + 1.1 What is ZODB?
> > + 1.2 OODBs vs. Relational DBs
> > + 1.3 What is ZEO?
> > + 1.4 About this guide
> > * 2 ZODB Programming
> > + 2.1 Installing ZODB
> > o 2.1.1 Requirements
> > o 2.1.2 Installing the Packages
> > + 2.2 How ZODB Works
> > + 2.3 Opening a ZODB
> > + 2.4 Writing a Persistent Class
> > + 2.5 Rules for Writing Persistent Classes
> > o 2.5.1 Modifying Mutable Objects
> > o 2.5.2 Some Special Methods Don't Work
> > o 2.5.3 Fixing isinstance and issubclass
> > o 2.5.4 __getattr__, __delattr__, and __setattr__
> > + 2.6 Writing Persistent Classes
> > o 2.6.1 Changing Class Attributes
> > * 3 ZEO
> > + 3.1 How ZEO Works
> > + 3.2 Installing ZEO
> > o 3.2.1 Requirements
> > o 3.2.2 Installation
> > + 3.3 Configuring and Running a ZEO Server
> > + 3.4 Testing the ZEO Installation
> > + 3.5 Sample Application: chatter.py
> > * 4 Transactions and Versioning
> > + 4.1 Subtransactions
> > + 4.2 Undoing Changes
> > + 4.3 Versions
> > + 4.4 Multithreaded ZODB Programs
> > * 5 Storages
> > + 5.1 FileStorage
> > + 5.2 BerkeleyStorage
> > + 5.3 OracleStorage
> > * 6 Related Modules
> > + 6.1 ZODB.PersistentMapping
> > + 6.2 ZODB.PersistentList
> > + 6.3 B-tree Modules
> > * A. Resources
> > * B. GNU Free Documentation License
> >


From erno-news at erno.iki.fi  Mon Feb  5 18:44:23 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 05 Feb 2001 19:44:23 +0200
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org>
Message-ID: <kuitmp2few.fsf@lasipalatsi.fi>

In article <mailman.981388047.29255.python-list at python.org>, Andrew
Kuchling <akuchlin at mems-exchange.org> writes:

| What I'd like right now:
|      * Suggestions for topics that should be covered.  This would be
| the most helpful.

here's one: how does one implement searching? i would have expected the
btree objects to have "find key nearest to this" and "next" methods,
(like bsddb's set_location)...

  -- erno


From syver at NOSPAMcyberwatcher.com  Mon Feb  5 19:21:32 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Mon, 5 Feb 2001 19:21:32 +0100
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org>
Message-ID: <95mr3t$ad2$1@troll.powertech.no>

"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:mailman.981388047.29255.python-list at python.org...
> For what seems like half an eternity, I've been working on a ZODB
> programming guide, written for Python programmers who want to use the
> ZODB to store data.

This sounds very exciting, I had been looking forward to an object database
on python. I've tried ZEO under MS Win 2000 but it doesn't seem to work. Is
it supposed to?




From akuchlin at mems-exchange.org  Mon Feb  5 20:03:46 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 05 Feb 2001 14:03:46 -0500
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95mr3t$ad2$1@troll.powertech.no>
Message-ID: <3dg0htymst.fsf@ute.cnri.reston.va.us>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:
> This sounds very exciting, I had been looking forward to an object database
> on python. I've tried ZEO under MS Win 2000 but it doesn't seem to work. Is
> it supposed to?

I don't know, but suspect it should work -- for some value of "work"
-- on Windows.  But note that I can't make Windows binaries available,
so you'd need to have a C compiler in order to install it.  Try asking
on one of the Zope mailing lists: zope-zeo at zope.org.  (Oh, wait,
they're apparently down at the moment, so maybe this should be kept on
python-list for now.)

--amk




From syver at NOSPAMcyberwatcher.com  Wed Feb  7 01:31:45 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 7 Feb 2001 01:31:45 +0100
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95mr3t$ad2$1@troll.powertech.no> <3dg0htymst.fsf@ute.cnri.reston.va.us>
Message-ID: <95q562$fpu$1@troll.powertech.no>

"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3dg0htymst.fsf at ute.cnri.reston.va.us...
> "Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:
> > This sounds very exciting, I had been looking forward to an object
database
> > on python. I've tried ZEO under MS Win 2000 but it doesn't seem to work.
Is
> > it supposed to?
>
> I don't know, but suspect it should work -- for some value of "work"
> -- on Windows.  But note that I can't make Windows binaries available,
> so you'd need to have a C compiler in order to install it.

I have msdev so that's okay, I just tried to install it and it complained
about various signals and other unix stuff not being available. Maybe I did
something wrong when installing?




From tuttledon at hotmail.com  Wed Feb  7 06:38:36 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Wed, 07 Feb 2001 05:38:36 GMT
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95mr3t$ad2$1@troll.powertech.no> <3dg0htymst.fsf@ute.cnri.reston.va.us> <95q562$fpu$1@troll.powertech.no>
Message-ID: <wb5g6.39505$p8.9293946@typhoon.southeast.rr.com>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote in message
news:95q562$fpu$1 at troll.powertech.no...

> I have msdev so that's okay, I just tried to install it and it complained
> about various signals and other unix stuff not being available. Maybe I
did
> something wrong when installing?

I got the installer error that "basetsd.h" couldn't be found in
"\Python20\include".  Must be a platform specific file not supported on
Windows :(

Don




From mbel44 at dial.pipex.net  Tue Feb  6 18:04:31 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Tue, 06 Feb 2001 17:04:31 +0000
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95mr3t$ad2$1@troll.powertech.no>
Message-ID: <c6b08t4ts7s8hlig4jg7ont008r986mn7e@4ax.com>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote:

>This sounds very exciting, I had been looking forward to an object database
>on python. I've tried ZEO under MS Win 2000 but it doesn't seem to work. Is
>it supposed to?

Some older versions did. Later versions added some socketness that
win32 didnt like, although I believe that is fixed on the most recent
version.


Toby Dickenson
tdickenson at geminidataloggers.com


From ndev42 at yahoo.com  Thu Feb  8 09:37:17 2001
From: ndev42 at yahoo.com (ndev42 at yahoo.com)
Date: Thu, 08 Feb 2001 08:37:17 GMT
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org>
Message-ID: <95tlrt$7f2$1@nnrp1.deja.com>

In article <mailman.981388047.29255.python-list at python.org>,
  akuchlin at mems-exchange.org wrote:
> For what seems like half an eternity, I've been working on a ZODB
> programming guide
[...]

Seems like you are exploring the namespace of ZO* words,
so just to avoid a faux pas I'd point out that "zob" is a very
rude word in french. Avoid naming a piece of software with anything
close to that! :-)

No kidding.
--
Nicolas


Sent via Deja.com
http://www.deja.com/


From fredp at mygale.org.nospam  Thu Feb  8 18:19:17 2001
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 8 Feb 2001 17:19:17 GMT
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95tlrt$7f2$1@nnrp1.deja.com>
Message-ID: <Xns9042BA83D34A2PaCmAnRDLM@161.48.76.20>

ndev42 at yahoo.com said :

>Seems like you are exploring the namespace of ZO* words,
>so just to avoid a faux pas I'd point out that "zob" is a very
>rude word in french. Avoid naming a piece of software with anything
>close to that! :-)

Actually it is a (north-African) Arab word that has become very common in 
colloquial French, meaning more or less "dick" (pardon my French :-).

>No kidding.

Indeed. As it is, both "ZOPE" and "ZODB" are close enough, phonetically, 
when pronounced in French, as to be incomfortable in corporate settings. 
This may sound trivial, but starting with "Python", a lot of valuable open 
source software meets credibility problems over here that stem from 
unfortunate (French-wise) naming choices. Curiously, "Perl" seems to come 
across much more easily, which is very frustrating :-)

-- 
YAFAP : http://www.multimania.com/fredp/


From akuchlin at mems-exchange.org  Thu Feb  8 19:41:24 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 08 Feb 2001 13:41:24 -0500
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95tlrt$7f2$1@nnrp1.deja.com> <Xns9042BA83D34A2PaCmAnRDLM@161.48.76.20>
Message-ID: <3dsnlp586j.fsf@ute.cnri.reston.va.us>

fredp at mygale.org.nospam (Fred Pacquier) writes:
> Indeed. As it is, both "ZOPE" and "ZODB" are close enough, phonetically, 
> when pronounced in French, as to be incomfortable in corporate settings. 

Well, you could expand the acronym to Z Object Database.  Or maybe
"Base a Donnees Z", which would have the French acronym BDZ.  

--amk


From fredp at mygale.org.nospam  Fri Feb  9 09:59:05 2001
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 9 Feb 2001 08:59:05 GMT
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95tlrt$7f2$1@nnrp1.deja.com> <Xns9042BA83D34A2PaCmAnRDLM@161.48.76.20> <3dsnlp586j.fsf@ute.cnri.reston.va.us>
Message-ID: <Xns904365B588AC8PaCmAnRDLM@161.48.76.20>

Andrew Kuchling <akuchlin at mems-exchange.org> said :

>> Indeed. As it is, both "ZOPE" and "ZODB" are close enough,
>> phonetically, when pronounced in French, as to be incomfortable in
>> corporate settings. 
>
>Well, you could expand the acronym to Z Object Database.  Or maybe
>"Base a Donnees Z", which would have the French acronym BDZ.  

Okay. That just leaves us with the question of explaining the meaning of 
"Z" in all of these terms. Is there an official answer to this ? :-)

-- 
YAFAP : http://www.multimania.com/fredp/


From sampe99 at my-deja.com  Mon Feb  5 17:07:45 2001
From: sampe99 at my-deja.com (sampe99 at my-deja.com)
Date: Mon, 05 Feb 2001 16:07:45 GMT
Subject: Square root of a matrix?
Message-ID: <95mj49$vbs$1@nnrp1.deja.com>

I'm looking for a module or a function that can
preferably perform a n-root of a matrix or the x-
power of a matrix (where x is a float). The
second best alternative is a square root of a
matrix.

The matrix is not symmetric so a cholesky
factorization is not an option.

I'd be forever greatful for any help.


Sent via Deja.com
http://www.deja.com/


From hinsen at cnrs-orleans.fr  Mon Feb  5 18:18:39 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 05 Feb 2001 18:18:39 +0100
Subject: Square root of a matrix?
References: <95mj49$vbs$1@nnrp1.deja.com>
Message-ID: <m3g0ht3v68.fsf@chinon.cnrs-orleans.fr>

sampe99 at my-deja.com writes:

> I'm looking for a module or a function that can
> preferably perform a n-root of a matrix or the x-
> power of a matrix (where x is a float). The
> second best alternative is a square root of a
> matrix.
> 
> The matrix is not symmetric so a cholesky
> factorization is not an option.

This is not really a Python question...

The most general approach is via eigenvalues, but for numerical
applications you should try to profit from any specificities of
your problem in order to get the best accuracy and speed. So the
starting point should be a book on numerical linear algebra,
or a more general one such as Numerical Recipes.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From nickb at earth.ox.ac.uk  Mon Feb  5 19:06:59 2001
From: nickb at earth.ox.ac.uk (nick)
Date: Mon, 05 Feb 2001 18:06:59 +0000
Subject: Q:python control of gnome-canvas
Message-ID: <3A7EEBC3.30684AC5@earth.ox.ac.uk>

I would be very grateful for any assistance here.

Q/ How do I get a  CanvasItem position information to update when I move
it?
I can move things but their position info seems to remain unchanged!

I can't find the answer to this particular problem though there is a lot
of info out there....

Setting up some test routines on a gnome-canvas
when moving widgets using event drivers and the mouse.

For a rectangle I have ...

the original location-
    ('rect',x1=x1,y1=y1....etc
and can see the location info by -
    widget['x1'] etc

and using an event driver to do-
    widget.move(x,y)

but widget['x1'] etc are not updated and only hold the original widget
coordinates.

I can only get the coordinates to update using
    widget.set(x1=newx1,x2=newx2...etc

I presume I am missing something.
It gets worse if I try and group things since it is rather
troublesome to have to explicitly set  multiple grouped widgets.

So - how do I get the position information to update when a widget
moves?

thanks if you can spare the time
Nick/Oxford







From btang at pacific.jpl.nasa.gov  Mon Feb  5 19:18:04 2001
From: btang at pacific.jpl.nasa.gov (Benyang Tang)
Date: Mon, 05 Feb 2001 10:18:04 -0800
Subject: speed comparison of IDL, numPy, Matlab
Message-ID: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov>

Out of curiosity, I did a quick benchmark test of IDL, NumPy and Matlab on my
desktop machine. I know benchmarking is a complicated issue; don't take my
naive test too serious.

======================
Conditions of the test:
======================

* Machine: a dual Intel Xeon 550 MHz box with 1GB ram, running RedHat Linux
6.2. The machine was not doing any serious service, so the test code should
have had close to 100% of the resources.

* IDL, version 5.3

* Python version 1.5.2; NumPy release 15.3-1

* Matlab version 5.3

* For each code, I ran it several times so the timing became somewhat stable. I
just took the last reading; for the size of 600X600, timing fluctuation of
less than 2% was observed.

* All tests were done in a period of 20 minutes; within this period the
following data were collected. About 2 hours later, I redid the tests and
obtained similar results.


=================================================================
Timing (in seconds) of matrix multiplication of 2 m-by-m matrixes
=================================================================

m       100   200   300   400   500  600
-----------------------------------------
     single precision
IDL    0.02  0.12  0.54  1.83  4.06  6.90
NumPy  0.01  0.12  0.46  1.37  3.05  5.49
Matlab           (1)

     double precision
IDL    0.04  0.41  1.89  5.14 10.72 18.61
NumPy  0.02  0.16  0.97  2.50  5.05  8.75
Matlab 0.01  0.15  0.94  2.50  4.90  8.69
-----------------------------------------
m       100   200   300   400   500  600

(1) Matlab does not do single-precision calculation.

===================
What the tests tell:
===================
1) NumPy is about as fast as Matlab;
2) NumPy is about 25% faster (in single precision), or more than 100% faster
(in double precision) than IDL.

==================================================
Here are the codes (for the double precision test):
==================================================

IDL:
===
for m = 100,600,100 do begin
  a = double(randomn(0,m,m))
  b = double(randomn(0,m,m))

  time0 = systime(1)
  c = a##b
  dtime = systime(1)-time0

  print, string(m,m,dtime, format='("multiplication of " ,i3, "X", i3, " matrixes takes ", f5.2)')
endfor

NumPy:
=====
from Numeric import *
from RandomArray import *
import time

for m in [100,200,300,400,500,600]:
  a= uniform(-1,1,(m,m,))
  b= uniform(-1,1,(m,m,))
  a=a.astype(Float64)
  b=b.astype(Float64)

  time0 = time.time()
  c = matrixmultiply(a,b)
  dtime = time.time() - time0

  print 'multiplication of %dX%d matrixes takes %5.2f' %(m,m,dtime)

Matlab:
======
for m = [100 200 300 400 500 600];
  a = randn(m);
  b = randn(m);

  time0 = clock;
  c = a*b;
  dtime = etime(clock,time0);

  fprintf('multiplication of %dX%d matrixes takes %5.2f\n', m,m,dtime)
end

-- 
<> Benyang Tang                              
<> 300-323, JPL     
<> 4800 Oak Grove Drive      
<> Pasadena, CA 91109, USA


From pvandelst at ncep.noaa.gov  Mon Feb  5 21:08:54 2001
From: pvandelst at ncep.noaa.gov (Paul van Delst)
Date: Mon, 05 Feb 2001 15:08:54 -0500
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov>
Message-ID: <3A7F0856.3CFA7FA2@ncep.noaa.gov>

Benyang Tang wrote:
> 
> Out of curiosity, I did a quick benchmark test of IDL, NumPy and Matlab on my
> desktop machine. I know benchmarking is a complicated issue; don't take my
> naive test too serious.

O.k.  :o)

I've never used Python, and Matlab only once or twice, but what's with the Python syntax?
Seems a tad wordy. Looks like Nick Bower's IDL-like package in Python
(http://nickbower.com/computer/pydl) is worth another look though......

paulv

P.S. BTW, in Matlab, how do you simply multiply the corresponding matrix elements? (i.e.
not a matrix multiply).


-- 
Paul van Delst                 A little learning is a dangerous thing;
CIMSS @ NOAA/NCEP              Drink deep, or taste not the Pierian spring;
Ph:    (301) 763-8000 x7274    There shallow draughts intoxicate the brain,
Fax:   (301) 763-8545          And drinking largely sobers us again.
Email: pvandelst at ncep.noaa.gov                    Alexander Pope.


From obas at mathworks.com  Mon Feb  5 21:43:29 2001
From: obas at mathworks.com (Omur Bas)
Date: Mon, 5 Feb 2001 15:43:29 -0500
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov> <3A7F0856.3CFA7FA2@ncep.noaa.gov>
Message-ID: <95n39h$2e6$1@news.mathworks.com>

Try

>> A.*B

for element-by-element multiplication.

Omur
>
> P.S. BTW, in Matlab, how do you simply multiply the corresponding matrix
elements? (i.e.
> not a matrix multiply).
>
>
> --
> Paul van Delst                 A little learning is a dangerous thing;
> CIMSS @ NOAA/NCEP              Drink deep, or taste not the Pierian
spring;
> Ph:    (301) 763-8000 x7274    There shallow draughts intoxicate the
brain,
> Fax:   (301) 763-8545          And drinking largely sobers us again.
> Email: pvandelst at ncep.noaa.gov                    Alexander Pope.




From deblNoNONspammy at theworld.com  Tue Feb  6 06:44:14 2001
From: deblNoNONspammy at theworld.com (David Lees)
Date: Tue, 06 Feb 2001 05:44:14 GMT
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov> <3A7F0856.3CFA7FA2@ncep.noaa.gov>
Message-ID: <3A7F8F35.13FD8BB8@theworld.com>

Paul,

Thanks for the URL to pydl.  It looks quite interesting.

I am a Python newbie, but your comments Python syntax and being "wordy"
seem a bit misplaced.  The 'for' statement could have been written with
a 'range' so that it looks similar to the IDL for statement.  The
wordiness of Python relative to IDL or Matlab is just due to the built
in functions used, not something inherent in Python.  A function could
easily be defined for generating the uniform square matrices used in the
benchmark.  I really like the balance that Python achieves between
clarity and program length.  Certainly there have been languages like
APL that were far more compact than IDL or Matlab, but they were a bit
cryptic for my taste.

david lees


Paul van Delst wrote:
> 
> Benyang Tang wrote:
> >
> > Out of curiosity, I did a quick benchmark test of IDL, NumPy and Matlab on my
> > desktop machine. I know benchmarking is a complicated issue; don't take my
> > naive test too serious.
> 
> O.k.  :o)
> 
> I've never used Python, and Matlab only once or twice, but what's with the Python syntax?
> Seems a tad wordy. Looks like Nick Bower's IDL-like package in Python
> (http://nickbower.com/computer/pydl) is worth another look though......
> 
> paulv
> 
<snip>


From pvandelst at ncep.noaa.gov  Tue Feb  6 14:46:41 2001
From: pvandelst at ncep.noaa.gov (Paul van Delst)
Date: Tue, 06 Feb 2001 08:46:41 -0500
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov> <3A7F0856.3CFA7FA2@ncep.noaa.gov> <3A7F8F35.13FD8BB8@theworld.com>
Message-ID: <3A800041.3AA72161@ncep.noaa.gov>

David Lees wrote:
> 
> Paul,
> 
> Thanks for the URL to pydl.  It looks quite interesting.
> 
> I am a Python newbie, but your comments Python syntax and being "wordy"
> seem a bit misplaced.

I should've put a smiley after that statement. That comment was meant as an indictment on
(towards?) me (feeling old and stuff, :o( ), not Python.

paulv

-- 
Paul van Delst           A little learning is a dangerous thing;
CIMSS @ NOAA/NCEP        Drink deep, or taste not the Pierian spring;
Ph: (301)763-8000 x7274  There shallow draughts intoxicate the brain,
Fax:(301)763-8545        And drinking largely sobers us again.
pvandelst at ncep.noaa.gov                   Alexander Pope.


From n8gray at caltech.edu.is.my.email.address  Tue Feb  6 08:58:50 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Mon, 5 Feb 2001 23:58:50 -0800
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov> <3A7F0856.3CFA7FA2@ncep.noaa.gov>
Message-ID: <95ob6p$oba@gap.cco.caltech.edu>

Paul van Delst wrote:

> I've never used Python, and Matlab only once or twice, but what's with the
> Python syntax? Seems a tad wordy. Looks like Nick Bower's IDL-like package
> in Python (http://nickbower.com/computer/pydl) is worth another look
> though......

It's true that Numpy code can end up more verbose than Matlab, although 
this example doesn't really hilight that issue.  This is to be expected 
from a language that wasn't designed from the ground up as a numerical 
computing language.  There's a faint glimmer of hope that we may get a new 
Python operator for matrix multiplication, though.  (See PEP 211)

On the flipside, have you ever tried to write user interfaces in Matlab?  
Have you ever tried to implement a dictionary in Matlab?  Have you ever 
tried to understand when Matlab is making a copy of a giant array and when 
it's using a reference?  Have you ever had your Matlab license expire the 
day of a meeting, _before_ you've made your plots?  

I have.  <shudder>

Having used IDL and Matlab quite extensively, I've realized that in any 
moderately complex numerical program 75% of the code is *programming* and 
only 25% of the code is *numerical*.  I'd take a great general-purpose 
language with decent numerical extensions over a decent language with great 
numerical capabilities any day of the week, and twice on Sunday. :^)

> paulv
> 
> P.S. BTW, in Matlab, how do you simply multiply the corresponding matrix
> elements? (i.e. not a matrix multiply).

mat1 .* mat2;

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From n8gray at caltech.edu.is.my.email.address  Tue Feb  6 08:58:50 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Mon, 5 Feb 2001 23:58:50 -0800
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov> <3A7F0856.3CFA7FA2@ncep.noaa.gov>
Message-ID: <95obfh$ocp@gap.cco.caltech.edu>

Paul van Delst wrote:

> I've never used Python, and Matlab only once or twice, but what's with the
> Python syntax? Seems a tad wordy. Looks like Nick Bower's IDL-like package
> in Python (http://nickbower.com/computer/pydl) is worth another look
> though......

It's true that Numpy code can end up more verbose than Matlab, although 
this example doesn't really hilight that issue.  This is to be expected 
from a language that wasn't designed from the ground up as a numerical 
computing language.  There's a faint glimmer of hope that we may get a new 
Python operator for matrix multiplication, though.  (See PEP 211)

On the flipside, have you ever tried to write user interfaces in Matlab?  
Have you ever tried to implement a dictionary in Matlab?  Have you ever 
tried to understand when Matlab is making a copy of a giant array and when 
it's using a reference?  Have you ever had your Matlab license expire the 
day of a meeting, _before_ you've made your plots?  

I have.  <shudder>

Having used IDL and Matlab quite extensively, I've realized that in any 
moderately complex numerical program 75% of the code is *programming* and 
only 25% of the code is *numerical*.  I'd take a great general-purpose 
language with decent numerical extensions over a decent language with great 
numerical capabilities any day of the week, and twice on Sunday. :^)

> paulv
> 
> P.S. BTW, in Matlab, how do you simply multiply the corresponding matrix
> elements? (i.e. not a matrix multiply).

mat1 .* mat2;

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From m.hadfield at niwa.cri.nz  Wed Feb  7 00:10:55 2001
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Wed, 7 Feb 2001 12:10:55 +1300
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov>
Message-ID: <981501055.966870@clam-ext>

"Benyang Tang" <btang at pacific.jpl.nasa.gov> wrote in message
news:3A7EEE5C.6419B707 at pacific.jpl.nasa.gov...
>
> Out of curiosity, I did a quick benchmark test of IDL, NumPy and Matlab on
my
> desktop machine. I know benchmarking is a complicated issue; don't take my
> naive test too serious.

I don't take it seriously at all, because you've only exercised a tiny
fraction of the code in each product.

As another very small contribution to the IDL vs NumPy vs Matlab discussion,
may I offer my observation that, when you are forced to write (quelle
horreur!) a for loop, IDL will process that loop significantly faster than
Matlab and orders of magnitude faster than Python. (I'm afraid I can't offer
the numbers to back up that assertion right now--it's based on comparisons I
did some time ago.)

Speaking as a long-time IDL user who has dabbled with Matlab & Python/NumPy,
I'd love to switch to NumPy, and I'd do it tomorrow if the graphics
capabilities were there.

---
Mark Hadfield
m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield/
National Institute for Water and Atmospheric Research
PO Box 14-901, Wellington, New Zealand





From nospam at no.spam  Fri Feb  9 00:27:45 2001
From: nospam at no.spam (Carsten Geckeler)
Date: Fri, 9 Feb 2001 00:27:45 +0100 (MET)
Subject: speed comparison of IDL, numPy, Matlab
In-Reply-To: <981501055.966870@clam-ext>
Message-ID: <Pine.LNX.4.10.10102090025540.886-100000@erde.kosmos.all>

On Wed, 7 Feb 2001, Mark Hadfield wrote:
[snip]
> 
> Speaking as a long-time IDL user who has dabbled with Matlab & Python/NumPy,
> I'd love to switch to NumPy, and I'd do it tomorrow if the graphics
> capabilities were there.

Although I haven't tried it myself yet, there was mentioned a package
collection called PyDL just a few days ago on this list.  The URL is

http://nickbower.com/computer/pydl

Ciao, Carsten
-- 
Carsten Geckeler:  carsten dot geckeler at gmx dot de
To get proper email-address replace `dot' and `at' by the corresponding symbols.




From gavrie at my-deja.com  Mon Feb 12 13:19:48 2001
From: gavrie at my-deja.com (Gavrie)
Date: Mon, 12 Feb 2001 12:19:48 GMT
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov>
Message-ID: <968kd0$rml$1@nnrp1.deja.com>

In article <3A7EEE5C.6419B707 at pacific.jpl.nasa.gov>,
  Benyang Tang <btang at pacific.jpl.nasa.gov> wrote:

[...]

> * Machine: a dual Intel Xeon 550 MHz box with 1GB ram, running RedHat
Linux
> 6.2. The machine was not doing any serious service, so the test code
should
> have had close to 100% of the resources.

Well, I don't know how you got these results on a
dual Xeon 550 with 1G of RAM.
I ran the same benchmarks on a dual PIII-550 with
256MB of RAM, and got:

> python:
multiplication of 100X100 matrixes takes  0.02
multiplication of 200X200 matrixes takes  0.20
multiplication of 300X300 matrixes takes  1.07
multiplication of 400X400 matrixes takes  2.81
multiplication of 500X500 matrixes takes  5.49
multiplication of 600X600 matrixes takes  9.58

> matlab:

multiplication of 100X100 matrixes takes  0.01
multiplication of 200X200 matrixes takes  0.06
multiplication of 300X300 matrixes takes  0.15
multiplication of 400X400 matrixes takes  0.35
multiplication of 500X500 matrixes takes  0.68
multiplication of 600X600 matrixes takes  1.19

So, this difference seems a bit strange, doesn't
it?
I'm using MATLAB 6 (whichis supposed to be
*slower* than 5.3), and Python 1.5.2.

-- Gavrie Philipson.


Sent via Deja.com
http://www.deja.com/


From clee at gnwy100.wuh.wustl.edu  Mon Feb 12 16:14:04 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 12 Feb 2001 09:14:04 -0600
Subject: speed comparison of IDL, numPy, Matlab
References: <3A7EEE5C.6419B707@pacific.jpl.nasa.gov> <968kd0$rml$1@nnrp1.deja.com>
Message-ID: <7u66igaq83.fsf@gnwy100.wuh.wustl.edu>

>>>>> "Gavrie" == Gavrie  <gavrie at my-deja.com> writes:

    Gavrie> In article <3A7EEE5C.6419B707 at pacific.jpl.nasa.gov>, Benyang
    Gavrie> Tang <btang at pacific.jpl.nasa.gov> wrote:


    Gavrie> So, this difference seems a bit strange, doesn't it?  I'm using
    Gavrie> MATLAB 6 (whichis supposed to be *slower* than 5.3), and Python
    Gavrie> 1.5.2.

I believe there's an explaination for this:

While I've heard that Matlab 6's footprint is much larger, especially the
interface, I have heard that Matlab 6 does (finally) use LAPACK and BLAS
which are optimized for modern machines w/ hierarchical memory.  I would
therefore expect that it would do substantially better than the
out-of-the-box numpy build for matrix multiplication.  By default, numpy
uses a naive algorithm for matrix multiplication which is quite cache
unfriendly.  When numpy is linked to ATLAS's BLAS routines and LAPACK, it's
more cache-friendly---and much faster.

I did some benchmarks myself: For matrix inversion of a 1000x1000 matrix,
numpy-atlas is 7 times faster than matlab 5.3 (no lapack).  The difference
is greater if you have a dual processor machine because ATLAS now has
options for multi-threaded operation.  And for matrix multiplication you
can compare the naive numpy multiplication with the optmized-dgemm based
version on the same matrix:

#  1000 x 1000 matrix 
time elapsed for matrixmultiply (sec) 24.142714
time elapsed for dgemm-matrixmultiply (sec) 5.997188


-chris

[BTW: I'm working on putting together an optimized-numpy distribution that
uses ATLAS by default.]




From phrxy at csv.warwick.ac.uk  Tue Feb 13 08:15:09 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Tue, 13 Feb 2001 07:15:09 +0000
Subject: speed comparison of IDL, numPy, Matlab
Message-ID: <Pine.SOL.4.30.0102130701320.9863-100000@mimosa.csv.warwick.ac.uk>

On 2001-02-06 15:26:01 PST,  Mark Hadfield (m.hadfield at niwa.cri.nz) wrote:
[...]
> Speaking as a long-time IDL user who has dabbled with Matlab &
> Python/NumPy, I'd love to switch to NumPy, and I'd do it tomorrow if
> the graphics capabilities were there.

I don't know if this is the kind of thing you'd be looking for, but...

vtk?  I don't know if it does 2D stuff well, though.  OpenDX apparently
does -- 'User Quote' currently on the latter's page:

----

In a nutshell, I use DX extensively for many types of applications using
both large and small data sets...

I'm basically using it in a 2-d world for remote sensing data
applications. 99% of the people in this area use IDL. DX stands out better
in 3-d applications, but from my experience it is faster and more reliable
than IDL even in 2-d applications (very large data sets). I use DX for
many data types, especially unstructured and polyline and I also use DX as
a debugging tool for just about anything. If an algorithm isn't doing what
I think it should, just start a new DX program from scratch and in minutes
you can be analizing the problem.

                                                             Robert Neely
                                                NASA Langley Research Center


----

There's PIL too of course, and if you just want to plot graphs, DISLIN,
PGPLOT, Gnuplot (ugly but simple), BLT, Qwt (both GUI toolkits with graph
widgets).  DISLIN has two python interfaces: the standard one and an OO
one that wraps it.

For 3D, there seem to be lots of things with python interfaces -- no idea
what they do, though.


John



From bill.xu at newlix.com  Mon Feb  5 19:59:31 2001
From: bill.xu at newlix.com (Bill Xu)
Date: Mon, 05 Feb 2001 18:59:31 GMT
Subject: Why "undefined reference to `Py_Initialize' " error?
Message-ID: <nKCf6.85784$V22.16880816@news4.rdc1.on.home.com>

Hi, Buddies:

Please help me with Python embeded C++ problem:
I compile the following file with gcc -I /usr/local/include/python2.0/
basic1.c
It gives me " undefined reference to `Py_Initialize' " error,
Do I miss some setting in Python 2.0 or miss some package?
I can't find my Modules, Python ... directorys in my python 2.0 directory.
Please give me some hints! Thanks in advance.

Bill

The makefile with this sample is:
PY = /usr/include/python1.5

PLIBS = $(PY)/Modules/libModules.a \
        $(PY)/Python/libPython.a \
        $(PY)/Objects/libObjects.a \
        $(PY)/Parser/libParser.a

POBJS = $(PY)/Modules/config.o $(PY)/Modules/getpath.o

basic1: basic1.o
 cc -g basic1.o $(POBJS) $(PLIBS) -lm -o basic1

basic1.o: basic1.c
 cc basic1.c -c -g -I$(PY)/Include -I$(PY)/.



$ gcc -I /usr/local/include/python2.0/ basic1.c
/tmp/ccNyWd2r.o: In function `main':
/tmp/ccNyWd2r.o(.text+0x2d): undefined reference to `Py_Initialize'
collect2: ld returned 1 exit status


#include "Python.h"
char *getprogramname() { return "basic1"; }

main() {
    printf("%s\n", getprogramname());
    Py_Initialize();
    // PyRun_SimpleString("import string");
    // PyRun_SimpleString("print string.uppercase");
    // PyRun_SimpleString("x = string.uppercase");
    // PyRun_SimpleString("print string.lower(x)");
}









From bryan at eevolved.com  Mon Feb  5 20:37:24 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Mon, 05 Feb 2001 19:37:24 GMT
Subject: Why "undefined reference to `Py_Initialize' " error?
References: <nKCf6.85784$V22.16880816@news4.rdc1.on.home.com>
Message-ID: <UhDf6.203892$f36.8144342@news20.bellglobal.com>

Have you tried explicitly including "pythonrun.h"?
That file specifies Py_Initialize I think.

Bill Xu wrote:

> Hi, Buddies:
> 
> Please help me with Python embeded C++ problem:
> I compile the following file with gcc -I /usr/local/include/python2.0/
> basic1.c
> It gives me " undefined reference to `Py_Initialize' " error,
> Do I miss some setting in Python 2.0 or miss some package?
> I can't find my Modules, Python ... directorys in my python 2.0 directory.
> Please give me some hints! Thanks in advance.
> 
> Bill
> 
> The makefile with this sample is:
> PY = /usr/include/python1.5
> 
> PLIBS = $(PY)/Modules/libModules.a \
>         $(PY)/Python/libPython.a \
>         $(PY)/Objects/libObjects.a \
>         $(PY)/Parser/libParser.a
> 
> POBJS = $(PY)/Modules/config.o $(PY)/Modules/getpath.o
> 
> basic1: basic1.o
>  cc -g basic1.o $(POBJS) $(PLIBS) -lm -o basic1
> 
> basic1.o: basic1.c
>  cc basic1.c -c -g -I$(PY)/Include -I$(PY)/.
> 
> 
> 
> $ gcc -I /usr/local/include/python2.0/ basic1.c
> /tmp/ccNyWd2r.o: In function `main':
> /tmp/ccNyWd2r.o(.text+0x2d): undefined reference to `Py_Initialize'
> collect2: ld returned 1 exit status
> 
> 
> #include "Python.h"
> char *getprogramname() { return "basic1"; }
> 
> main() {
>     printf("%s\n", getprogramname());
>     Py_Initialize();
>     // PyRun_SimpleString("import string");
>     // PyRun_SimpleString("print string.uppercase");
>     // PyRun_SimpleString("x = string.uppercase");
>     // PyRun_SimpleString("print string.lower(x)");
> }
> 
> 
> 
> 
> 
> 
> 

-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
www.eevolved.com
<==================================>

"Now he has departed from this strange world a little ahead of me. That 
means nothing. People like us, who believe in physics, know that the 
distinction between past, present, and future is only a stubbornly 
persistent illusion."-- Einstein



From fredrik at effbot.org  Mon Feb  5 20:57:08 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 05 Feb 2001 19:57:08 GMT
Subject: Why "undefined reference to `Py_Initialize' " error?
References: <nKCf6.85784$V22.16880816@news4.rdc1.on.home.com>
Message-ID: <oADf6.11577$Qb7.1749259@newsb.telia.net>

Bill Xu wrote:
> Please help me with Python embeded C++ problem:
> I compile the following file with gcc -I /usr/local/include/python2.0/
> basic1.c
> It gives me " undefined reference to `Py_Initialize' " error,
> Do I miss some setting in Python 2.0 or miss some package?
> I can't find my Modules, Python ... directorys in my python 2.0 directory.

the library has been called libpython<version>.a for the
last four years, or so...

try changing PY to point to your 2.0 build directory, and
change PLIBS to point to $(PY)/libpython2.0.a

(if you don't have that file, you don't have a full installation.
look for a "development" package, or build Python from the
source tarball)

...and yes:

$ gcc -I /usr/local/include/python2.0/basic1.c

if you have a makefile, you better use "make" to run it,
not "gcc"...  (especially not gcc without the -c flag...)

Cheers /F




From jimmypop181 at my-deja.com  Mon Feb  5 20:13:39 2001
From: jimmypop181 at my-deja.com (jimmypop181 at my-deja.com)
Date: Mon, 05 Feb 2001 19:13:39 GMT
Subject: Can't Generate Random #
Message-ID: <95mu0p$aqc$1@nnrp1.deja.com>

Does somebody know how to generate a random number in
Python?  I read the docs and it says rand(x, y) should do it.  But
whenever I try that, It says that "there is no variable named rand".
This also happens when trying to get the Square Root.  Please
help.


Sent via Deja.com
http://www.deja.com/


From grey at despair.rpglink.com  Mon Feb  5 20:21:40 2001
From: grey at despair.rpglink.com (Steve Lamb)
Date: Mon, 05 Feb 2001 19:21:40 -0000
Subject: Can't Generate Random #
References: <95mu0p$aqc$1@nnrp1.deja.com>
Message-ID: <slrn97tva4.hpf.grey@teleute.dmiyu.org>

On Mon, 05 Feb 2001 19:13:39 GMT, jimmypop181 at my-deja.com
<jimmypop181 at my-deja.com> wrote:
>Does somebody know how to generate a random number in
>Python?  I read the docs and it says rand(x, y) should do it.  But
>whenever I try that, It says that "there is no variable named rand".
>This also happens when trying to get the Square Root.  Please
>help.

    Look at the documentation on whrandom for simple random numbers.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From chris at onca.catsden.net  Mon Feb  5 20:48:37 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Mon, 5 Feb 2001 11:48:37 -0800 (PST)
Subject: Can't Generate Random #
In-Reply-To: <95mu0p$aqc$1@nnrp1.deja.com>
Message-ID: <Pine.LNX.4.30.0102051145180.27504-100000@onca.catsden.net>

On Mon, 5 Feb 2001 jimmypop181 at my-deja.com wrote:

> Does somebody know how to generate a random number in
> Python?  I read the docs and it says rand(x, y) should do it.  But
> whenever I try that, It says that "there is no variable named rand".
> This also happens when trying to get the Square Root.  Please
> help.

Are you sure you're reading the /python/ documentation?? :) I've no idea
where you got 'rand' from.

import whrandom

r = whrandom.randint(a,b)  # Depricated
r = whrandom.randrange(a,b) # Better, gives you a random integer [a,b)

import math
i = math.sqrt(j)

or

from math import *
i = sqrt(j)



   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From fredrik at effbot.org  Mon Feb  5 20:57:09 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 05 Feb 2001 19:57:09 GMT
Subject: Can't Generate Random #
References: <95mu0p$aqc$1@nnrp1.deja.com>
Message-ID: <pADf6.11579$Qb7.1749241@newsb.telia.net>

jimmypop181 at my-deja.com wrote:
> Does somebody know how to generate a random number in
> Python?  I read the docs and it says rand(x, y) should do it.

what docs?  the rand module disappeared from the
language years ago...

use the "random" module instead:

    >>> import random
    >>> random.random()
    0.78562250893885366
    >>> random.randrange(0, 10)
    4

> This also happens when trying to get the Square Root.

you mean "sqrt"?  did you import the "math" module first?
did you use the full name of the function (math.sqrt)?

    >>> import math
    >>> math.sqrt(2)
    1.4142135623730951

(btw, have you read the tutorial at www.python.org?)

Cheers /F




From aleaxit at yahoo.com  Mon Feb  5 20:57:51 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 5 Feb 2001 20:57:51 +0100
Subject: Can't Generate Random #
References: <95mu0p$aqc$1@nnrp1.deja.com>
Message-ID: <95n0jv082k@news2.newsguy.com>

<jimmypop181 at my-deja.com> wrote in message
news:95mu0p$aqc$1 at nnrp1.deja.com...
> Does somebody know how to generate a random number in
> Python?  I read the docs and it says rand(x, y) should do it.  But

D:\MinGw32>python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import random
>>> random.random()
0.56534112337917264
>>> for i in range(10): print random.randint(1,7),
...
6 3 5 4 2 1 3 2 1 4
>>>

> whenever I try that, It says that "there is no variable named rand".
> This also happens when trying to get the Square Root.  Please

>>> import math
>>> math.sqrt(2)
1.4142135623730951
>>>

In brief: you must import the module containing the stuff you
need, in general.  Some functions are 'built-in' (no import is
needed), but most are in appropriate modules.


Alex






From DavidLNonOSpammy at raqia.com  Mon Feb  5 22:17:33 2001
From: DavidLNonOSpammy at raqia.com (David Lees)
Date: Mon, 05 Feb 2001 16:17:33 -0500
Subject: Can't Generate Random #
References: <95mu0p$aqc$1@nnrp1.deja.com>
Message-ID: <3A7F186D.B0078F6D@raqia.com>

I am not sure what doc you refer to, but look at the documentation for
the random and whrandom modules.  See:
http://www.python.org/doc/current/lib/rng-objects.html
http://www.python.org/doc/current/lib/module-whrandom.html

If you want a uniform real randome number between x and y try
import whrandom
whrandom.random(x,y)

Good luck,


david lees


jimmypop181 at my-deja.com wrote:
> 
> Does somebody know how to generate a random number in
> Python?  I read the docs and it says rand(x, y) should do it.  But
> whenever I try that, It says that "there is no variable named rand".
> This also happens when trying to get the Square Root.  Please
> help.
> 
> Sent via Deja.com
> http://www.deja.com/


From jimmypop181 at my-deja.com  Tue Feb  6 17:52:21 2001
From: jimmypop181 at my-deja.com (jimmypop181 at my-deja.com)
Date: Tue, 06 Feb 2001 16:52:21 GMT
Subject: Can't Generate Random #
References: <95mu0p$aqc$1@nnrp1.deja.com>
Message-ID: <95pa3v$bud$1@nnrp1.deja.com>

Thanks, I'm fairly new with Python and I didn't know that you had to
put "whrandom."random() or "math."sqrt().


Sent via Deja.com
http://www.deja.com/


From mwh21 at cam.ac.uk  Tue Feb  6 18:04:52 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Feb 2001 17:04:52 +0000
Subject: Can't Generate Random #
References: <95mu0p$aqc$1@nnrp1.deja.com> <95pa3v$bud$1@nnrp1.deja.com>
Message-ID: <m3vgqn3fpn.fsf@atrus.jesus.cam.ac.uk>

jimmypop181 at my-deja.com writes:

> Thanks, I'm fairly new with Python and I didn't know that you had to
> put "whrandom."random() or "math."sqrt().

You should write random.random() btw. Note the lack of "wh".

Cheers,
M.

-- 
  I'm not particularly fond of singing GSTQ because she stands for
  some things I don't, but it's not really worth letting politics
  getting in the way of a good bawling.
                                            -- Dan Sheppard, ucam.chat


From kopfarzt at my-deja.com  Mon Feb  5 20:36:25 2001
From: kopfarzt at my-deja.com (kopfarzt at my-deja.com)
Date: Mon, 05 Feb 2001 19:36:25 GMT
Subject: Python 2.1a[12] and MySQLdb 0.3.2
Message-ID: <95mvbp$c2r$1@nnrp1.deja.com>

Hi,

I get SEGFAULTs (backtrace included) when fetching more than one row
with Python 2.1, MySQLdb 0.3.2 and mxDateTime from a table that contains
a date. I tried to reduce the problem to the max:

I got this table:
create table test (some_date datetime);
insert into test values ('2000-01-01');
insert into test values ('2000-01-02');

and this testprogram
----------------------------------------------------------------
import MySQLdb
import mxDateTime

db = MySQLdb.connect (host='localhost', db='test',
                      user='test', passwd='test')
cr = db.cursor ()
cr.execute ('select * from test')
rw = cr.fetchone ()
print type (rw[0]), rw[0]
----------------------------------------------------------------

will print:

<type 'DateTime'> 2000-01-01 00:00:00.00
Segmentation fault (core dumped)

my setup is:

linux 2.2.12
python 2.1a1 (and 2.1a2)
mysql 3.22.32
mxDateTime 1.3.0

without any special configuration options. I also tried Python 1.5.2,
1.6 and 2.0 where it works.

Anybody else experiencing similar problems or got a solution?

Thanks in advance
Peter


The gdb backtrace is:
#0  0x400ae352 in chunk_free (ar_ptr=0x40145d00, p=0x80f2af0) at
malloc.c:3048
#1  0x400ae157 in __libc_free (mem=0x80f2c38) at malloc.c:2950
#2  0x808f08c in string_dealloc (op=0x80f2c38) at
Objects/stringobject.c:235
#3  0x805a45e in code_dealloc (co=0x80f2e90) at Python/compile.c:89
#4  0x8084703 in func_dealloc (op=0x813efa4) at Objects/funcobject.c:249
#5  0x808b42d in dict_dealloc (mp=0x813430c) at Objects/dictobject.c:602
#6  0x807c219 in class_dealloc (op=0x813454c) at
Objects/classobject.c:116
#7  0x808aec6 in insertdict (mp=0x80f3a64, key=0x811f8b0,
hash=-113409806,
    value=0x80bb42c) at Objects/dictobject.c:364
#8  0x808b190 in PyDict_SetItem (op=0x80f3a64, key=0x811f8b0,
value=0x80bb42c)
    at Objects/dictobject.c:498
#9  0x808cbbc in _PyModule_Clear (m=0x80f65e4) at
Objects/moduleobject.c:118
#10 0x8065635 in PyImport_Cleanup () at Python/import.c:291
#11 0x806b428 in Py_Finalize () at Python/pythonrun.c:212
#12 0x8051c83 in Py_Main (argc=1, argv=0xbffff764) at Modules/main.c:302
#13 0x8051670 in main (argc=2, argv=0xbffff764) at
Modules/python.c:10


Sent via Deja.com
http://www.deja.com/


From seefeld at sympatico.ca  Mon Feb  5 20:43:08 2001
From: seefeld at sympatico.ca (Stefan Seefeld)
Date: Mon, 05 Feb 2001 19:43:08 GMT
Subject: passing keyword arguments that are themselfs python keywords
Message-ID: <3A7F109D.60FD712B@sympatico.ca>

Hi,

I'm trying to write a python function that can take keywords such as 'class'.
Is that possible ?

I'd like to be able to define something like:

def entity(type, body, **attr):
    print '<' + type + ' ' + string.join(map(lambda item:', %s="%s"'%item, attr.items())) + '>'
    print body
    print '</type>'

and would like to call it like

entity('div', text, class='header')

Is there any way to achieve the desired effect ?

Thanks,	Stefan


From chris at onca.catsden.net  Mon Feb  5 20:54:45 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Mon, 5 Feb 2001 11:54:45 -0800 (PST)
Subject: passing keyword arguments that are themselfs python keywords
In-Reply-To: <3A7F109D.60FD712B@sympatico.ca>
Message-ID: <Pine.LNX.4.30.0102051152080.27504-100000@onca.catsden.net>

On Mon, 5 Feb 2001, Stefan Seefeld wrote:

> Hi,
>
> I'm trying to write a python function that can take keywords such as 'class'.
> Is that possible ?
>
> I'd like to be able to define something like:
>
> def entity(type, body, **attr):
>     print '<' + type + ' ' + string.join(map(lambda item:', %s="%s"'%item, attr.items())) + '>'
>     print body
>     print '</type>'
>
> and would like to call it like
>
> entity('div', text, class='header')
>
> Is there any way to achieve the desired effect ?

class is a python keyword, so the short answer is 'no'.

However, if you really /must/ use 'class' as a keyword, you could do
something like this (python-2.0 only)

entity ( 'div', test, **{'class':'header'} )


   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From fredrik at effbot.org  Mon Feb  5 20:57:09 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 05 Feb 2001 19:57:09 GMT
Subject: passing keyword arguments that are themselfs python keywords
References: <3A7F109D.60FD712B@sympatico.ca>
Message-ID: <pADf6.11578$Qb7.1749034@newsb.telia.net>

Stefan Seefeld wrote:
> I'm trying to write a python function that can take keywords such as 'class'.
> Is that possible ?

nope.

the usual workaround is to use a trailing underscore,
and strip it away inside your function:

    entity('div', text, class_='header')

or you could force your users to use ** syntax and
an ordinary dictionary:

    entity('div', text, **{'class': 'header'})

Cheers /F




From tismer at tismer.com  Mon Feb  5 21:08:17 2001
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 05 Feb 2001 21:08:17 +0100
Subject: Stackless Python 2.0 Minor Update ("Korea Edition")
Message-ID: <3A7F0831.E98229CD@tismer.com>

This is a bugfix release for Stackless Python 2.0 .

print >> sys.stdout, 1, 2, 3
  no longer crashes.
map(eval, ("a", "b"))
  now has access to locals of the enclosing function.

Bernd Rinn supplied RPM files for Linux! Many thanks to Bernd.

Please contact me if you have any problems with downloads
or if you find more bugs. (There ain't m^hany)

p.s.:
June, is my attempt to write "Stackless Python" correct?
http://www.stackless.com/

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From rvprasad at cis.ksu.edu  Mon Feb  5 21:15:05 2001
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 05 Feb 2001 14:15:05 -0600
Subject: os.popen2
Message-ID: <m3snlsrinq.fsf@boss.dreamsoft.com>

Hi,

The following code stalls

import os
i = os.popen2("/bin/bash")
i[0].write("ls -lR")
i[1].read()
-----------stalls here

Am I missing something the way it should behave or isn't it possible to execute
a process using popenX, keep it alive and interact with it?

waiting for reply,

-- 
Venkatesh Prasad Ranganath


From donn at u.washington.edu  Mon Feb  5 22:44:51 2001
From: donn at u.washington.edu (Donn Cave)
Date: 5 Feb 2001 21:44:51 GMT
Subject: os.popen2
References: <m3snlsrinq.fsf@boss.dreamsoft.com>
Message-ID: <95n6sj$dei$1@nntp6.u.washington.edu>

Quoth Venkatesh Prasad Ranganath <rvprasad at cis.ksu.edu>:
| Hi,
|
| The following code stalls
|
| import os
| i = os.popen2("/bin/bash")
| i[0].write("ls -lR")
| i[1].read()
| -----------stalls here
|
| Am I missing something the way it should behave or isn't it possible to execute
| a process using popenX, keep it alive and interact with it?

Yes, you're missing a couple of things, yes, you can interact with bash
this way, no, you can't interact this way with just any process.

To get a better sense of what's going on in your program, it might help
you to use a shell command that has some effect besides output to the
other pipe.  For example, have it write to a file - instead of 'ls -lR',
for example, try 'ls -lR > xxx' (or 'date > xxx', you really don't need
much output to start with.)  Now you can see specifically whether the
shell command executed, before you waste a lot of time asking why you
don't see its output.  You can also use "ps" to see whether the shell
process is still alive and waiting for something.

Now the first thing you need for your shell command lines is a trailing
newline:  'ls -lR\n', just like any file.  The shell doesn't execute
a command until it has read it, and it hasn't finished reading it until
it encounters the trailing newline.  Next, you have to actually write
the command line to the pipe, from the C library stdio buffer it's
sitting in now  - i[0].flush().  

At this point, I expect you will see evidence that the command actually
ran, and you should now change your command back to write to output -
but you still won't have any output.  This is because i[1].read()
is trying to read up everything, and the pipe is still open so there
could be more coming.  Another C library notion.  I personally would
just ditch the file object / C stdio stuff and use the UNIX file
descriptor directly - see fileobject fileno(), posix read() and write().

All right, now if you have been able to fix all this up, you should
be able to write and read one shell command after another.  It may
be a bit tricky to figure out how much output to expect from a shell
command, and when it's finished and ready for the next command, but
you can probably figure something out.  But if you take bash out of
there and replace it with almost any other application, you'll find
that you're back where you started, because most of these applications
use the same C library output and they're back there at step 1.b.,
without any way to get them to flush their buffer.

If there's any solution at all, it's to use a "pty" device instead
of pipes.  The pty library module has support for that, especially
in Python 2.0 and on the more common platforms.  The pty master fd
you get back will be a UNIX file descriptor that handles both reading
and writing.

	Donn Cave, donn at u.washington.edu


From cjensen at bioeng.ucsd.edu  Wed Feb  7 18:39:19 2001
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Wed, 07 Feb 2001 09:39:19 -0800
Subject: os.popen2
References: <m3snlsrinq.fsf@boss.dreamsoft.com>
Message-ID: <3A818847.D83E6EAB@bioeng.ucsd.edu>

Venkatesh Prasad Ranganath wrote:
> 
> Hi,
> 
> The following code stalls
> 
> import os
> i = os.popen2("/bin/bash")
> i[0].write("ls -lR")
> i[1].read()
> -----------stalls here
> 
> Am I missing something the way it should behave or isn't it possible to execute
> a process using popenX, keep it alive and interact with it?
> 
> waiting for reply,
> 
> --
> Venkatesh Prasad Ranganath

just a guess, try putting a carrage return at the end.

i[0].write("ls -lR\n")

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From junaftnoon at nospamplzyahoo.com  Mon Feb  5 21:57:00 2001
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Tue, 6 Feb 2001 05:57:00 +0900
Subject: 2nd Python Seminar in Seoul Was a Success
Message-ID: <95n463$anj$1@news.nuri.net>

Hello Python users all over the world,

We are happy to tell you that the 2nd Python Seminar in Seoul, South Korea
on Feb. 2 was a great success. We had more than 600 people on the day, and a
few had to go back home in the morning because we didn't have enough seats
for them. It was such a boom. A few reporters came to cover the seminar and
interviewed the people. There'll be several articles about the seminar in
some magazines next month.

We expect our next seminar will be much larger than this and want to
participate in the international community.

Thank you for keeping up and improving Python, and without you guys Korea
Python Users Group and its seminar couldn't have been this successful.




From m.faassen at vet.uu.nl  Wed Feb  7 01:01:53 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 7 Feb 2001 00:01:53 GMT
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net>
Message-ID: <95q39h$hji$5@newshost.accu.uu.nl>

June Kim <junaftnoon at nospamplzyahoo.com> wrote:
> Hello Python users all over the world,

> We are happy to tell you that the 2nd Python Seminar in Seoul, South Korea
> on Feb. 2 was a great success. We had more than 600 people on the day, and a
> few had to go back home in the morning because we didn't have enough seats
> for them. It was such a boom. A few reporters came to cover the seminar and
> interviewed the people. There'll be several articles about the seminar in
> some magazines next month.

Woah, we were happy in Europe to have about 70 people attending our
EuroZopeCon a few weeks ago, and you people get 600 people coming to
see about Stackless Python? What are you doing in Korea that we're
not doing in Europe? :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From junaftnoon at nospamplzyahoo.com  Wed Feb  7 02:36:11 2001
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Wed, 7 Feb 2001 10:36:11 +0900
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl>
Message-ID: <95q8tj$1c8$1@news.nuri.net>

"Martijn Faassen" <m.faassen at vet.uu.nl> wrote in message
news:95q39h$hji$5 at newshost.accu.uu.nl...
> June Kim <junaftnoon at nospamplzyahoo.com> wrote:
> > Hello Python users all over the world,
>
> > We are happy to tell you that the 2nd Python Seminar in Seoul, South
Korea
> > on Feb. 2 was a great success. We had more than 600 people on the day,
and a
> > few had to go back home in the morning because we didn't have enough
seats
> > for them. It was such a boom. A few reporters came to cover the seminar
and
> > interviewed the people. There'll be several articles about the seminar
in
> > some magazines next month.
>
> Woah, we were happy in Europe to have about 70 people attending our
> EuroZopeCon a few weeks ago, and you people get 600 people coming to
> see about Stackless Python? What are you doing in Korea that we're

It wasn't merely about SP. It was a Python seminar: ZOPE, wxPython, C/API,
Internet programming, XML, URL, and whole lot more.

> not doing in Europe? :)
>

We gave out Python Reference Library to the people and had a reception at
the
end; it wasn't totally free, though. (but we had sponsorship from several
companies)

Warmest regards,

June.

p.s. the coverage of Python in one of the four major newspapers a few days
b4
the seminar and developers' magazine's competitvely carrying articles about
Python seem to be major factors of our success.




From johngrayson at home.com  Wed Feb  7 13:57:23 2001
From: johngrayson at home.com (John Grayson)
Date: Wed, 07 Feb 2001 12:57:23 GMT
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net>
Message-ID: <95rgng$8ul$1@nnrp1.deja.com>

In article <95q8tj$1c8$1 at news.nuri.net>,
  "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:
>
>
>

I received a package from my publisher at Manning last night:
Inside was the Korean translation of Python and Tkinter Programming!
It is published by Infobook ISBN 89-8054-424-3.

This is great news for Python -- my only fear is that the frequent
Monty Python quips in the book translate to something offensive
in Korean...

    John Grayson


Sent via Deja.com
http://www.deja.com/


From junaftnoon at nospamplzyahoo.com  Wed Feb  7 19:24:38 2001
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Thu, 8 Feb 2001 03:24:38 +0900
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net> <95rgng$8ul$1@nnrp1.deja.com>
Message-ID: <95s40f$pui$1@news.nuri.net>

"John Grayson" <johngrayson at home.com> wrote in message
news:95rgng$8ul$1 at nnrp1.deja.com...
> In article <95q8tj$1c8$1 at news.nuri.net>,
>   "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:
> >
> >
> >
>
> I received a package from my publisher at Manning last night:
> Inside was the Korean translation of Python and Tkinter Programming!
> It is published by Infobook ISBN 89-8054-424-3.
>

Yes, we now have Learning Python in Korean published on the seminar day Feb.
2. and Python and Tkinter Programming a few weeks ago. (I'm sure there are
some more translations and some original works in Korean as well.)

> This is great news for Python -- my only fear is that the frequent
> Monty Python quips in the book translate to something offensive
> in Korean...

I reckon the people who are reading the translated version of the quips
would think them rather pedantic and esoteric than offensive; or very
valueable and profound about programming than laughable and comedic.

Nonetheless, it wouldn't hurt anyone's heart at least -- it would hurt some
of the reader's brains; they would struggle to interpret it.

Best regards,

June Kim

ps. I have a copy of your book in English and I like it pretty much. thank
you John for this great book.



From n8gray at caltech.edu.is.my.email.address  Sat Feb 10 07:47:21 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Fri, 9 Feb 2001 22:47:21 -0800
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net> <95rgng$8ul$1@nnrp1.deja.com> <95s40f$pui$1@news.nuri.net>
Message-ID: <962ogk$k1b@gap.cco.caltech.edu>

> "John Grayson" <johngrayson at home.com> wrote in message
> news:95rgng$8ul$1 at nnrp1.deja.com...
> > This is great news for Python -- my only fear is that the frequent
> > Monty Python quips in the book translate to something offensive
> > in Korean...

My hovercraft is full of eels.

Would you like to go back to my place, bouncy bouncy.

Has-the-Korean-peninsula-been-blessed-with-Spam?-ly y'rs,

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From akuchlin at mems-exchange.org  Wed Feb  7 21:33:30 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 07 Feb 2001 15:33:30 -0500
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net> <95rgng$8ul$1@nnrp1.deja.com>
Message-ID: <3dn1bygrmt.fsf@ute.cnri.reston.va.us>

John Grayson <johngrayson at home.com> writes:
> I received a package from my publisher at Manning last night:
> Inside was the Korean translation of Python and Tkinter Programming!
> It is published by Infobook ISBN 89-8054-424-3.

If people can point me at information about the Korean translations,
I'll happily add them to the list of non-English books in the Python
bookstore.  

--amk



From junaftnoon at nospamplzyahoo.com  Wed Feb  7 23:01:05 2001
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Thu, 8 Feb 2001 07:01:05 +0900
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net> <95rgng$8ul$1@nnrp1.deja.com> <3dn1bygrmt.fsf@ute.cnri.reston.va.us>
Message-ID: <95sgmd$j8v$1@news.nuri.net>

"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3dn1bygrmt.fsf at ute.cnri.reston.va.us...
> John Grayson <johngrayson at home.com> writes:
> > I received a package from my publisher at Manning last night:
> > Inside was the Korean translation of Python and Tkinter Programming!
> > It is published by Infobook ISBN 89-8054-424-3.
>
> If people can point me at information about the Korean translations,
> I'll happily add them to the list of non-English books in the Python
> bookstore.
>
> --amk
>

There are a few Korean translations and original works as well. Below are
the book titles and original publishers, followed by the translators, Korean
publisher, and the URL of the web page for the book. (This list isn't
comprehensive but it is all I know of, AFAIK)

Python Pocket Reference, O'Reilly, translated by Kim, Tae-Sun. Hanbit Media
http://www.hanbitbook.co.kr/look.htm?book_code=001016-00001

Learning Python, O'Reilly, translated by Lee, Kang-Sung. Hanbit Media
http://www.hanbitbook.co.kr/look.htm?book_code=010201-00001
(The translator is the current president of Korea Python Users Group.)

Python and Tkinter Programming, Manning, translated by Ryu Jung-Wook.
Infobook.
http://www.infobook.co.kr/cgi-bin/Vmk_BookInfo.cgi?bookinfo_but=????&in_book
id=ISBN_89_8054_424_3
(This URL includes Korean characters, which means you couldn't access the
page. In this case, follow this alternative URL of an online book store in
Korea :
http://www.wowbook.com/generic/book/info/book_detail.asp?isbn=ISBN1-884777-8
1-3 )

Internet Programming with Python, (by Aaron & Guido), translated by Kwak,
Joon-Ki. Samkakhyoung. (translated in 1997 but out of print now)
http://www.wowbook.com/generic/book/info/book_detail.asp?isbn=ISBN89-7467-43
5-1

About XML, (originally in Korean), Park Jae-Ho. Youngjin.
https://www.youngjin.com/shop/ycom-s-global002-pre.asp?url=ycom-s-002-01-pre
.asp&pcode=9788931415575
http://www.wowbook.com/generic/book/info/book_detail.asp?isbn=ISBN89-314-155
7-5
(This book is about XML but uses Python substantially)

In addition to these, there are a number of on-going works to write about
Python; moreover, I'm leading an open content project(in Korean), with
collective authorship, for writing An Introduction to Computer Programming
with Python -- it is geared towards complete newbies in computer programming
and it's in the spirit of CP4E. I hope open content projects would evolve
along with Python.

Best regards,

June




From junaftnoon at nospamplzyahoo.com  Thu Feb  8 03:46:13 2001
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Thu, 8 Feb 2001 11:46:13 +0900
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net> <95rgng$8ul$1@nnrp1.deja.com> <3dn1bygrmt.fsf@ute.cnri.reston.va.us> <95sgmd$j8v$1@news.nuri.net>
Message-ID: <95t1cu$7go$1@news.nuri.net>

"June Kim" <junaftnoon at nospamplzyahoo.com> wrote in message
news:95sgmd$j8v$1 at news.nuri.net...
[snip]
> Learning Python, O'Reilly, translated by Lee, Kang-Sung. Hanbit Media
> http://www.hanbitbook.co.kr/look.htm?book_code=010201-00001
> (The translator is the current president of Korea Python Users Group.)
[snip]

sorry. The name of the translator should've been "Lee, Gang Seong", instead.





From m.faassen at vet.uu.nl  Wed Feb  7 15:07:57 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 7 Feb 2001 14:07:57 GMT
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net>
Message-ID: <95rkru$k1s$1@newshost.accu.uu.nl>

June Kim <junaftnoon at nospamplzyahoo.com> wrote:
[me]
>> Woah, we were happy in Europe to have about 70 people attending our
>> EuroZopeCon a few weeks ago, and you people get 600 people coming to
>> see about Stackless Python? What are you doing in Korea that we're

> It wasn't merely about SP. It was a Python seminar: ZOPE, wxPython, C/API,
> Internet programming, XML, URL, and whole lot more.

Yeah, I figured actually, it just sounded funnier to just say SP, as that
was what I initially thought -- I shall strive for more accuracy in my
next post. :)

>> not doing in Europe? :)

> We gave out Python Reference Library to the people and had a reception at
> the end; it wasn't totally free, though. (but we had sponsorship from several
> companies)

We were totally free, and did have sponsorship. Anyway, more people couldn't
have fitted well in the room that we had, so in retrospect we must've been
glad 600 people did not show up. :)

[snip]
> p.s. the coverage of Python in one of the four major newspapers a few days
> b4 the seminar and developers' magazine's competitvely carrying articles about
> Python seem to be major factors of our success.

We shall attempt to strive for the same effect here, then! Anyway, you
people have my respect!

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From jcollin at exis.net  Mon Feb  5 22:22:30 2001
From: jcollin at exis.net (Jay Collins)
Date: Mon, 5 Feb 2001 16:22:30 -0500
Subject: Processing on a string
Message-ID: <Pine.LNX.4.21.0102051620000.9855-100000@tarpon.exis.net>

Hello,

 I'm trying to do the following:

I have a list that is all the lines from the password file.

if I did print p[0] # p being a line in /etc/passwd
I would get:

uucp:x:10:14:uucp:/var/spool/uucp:

or whatever. My question is how do process this list into another
list with just everything before the first ":". I'm not quite sure how to
do this. 



From fredrik at effbot.org  Mon Feb  5 22:30:48 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 05 Feb 2001 21:30:48 GMT
Subject: Processing on a string
References: <Pine.LNX.4.21.0102051620000.9855-100000@tarpon.exis.net>
Message-ID: <cYEf6.12459$AH6.1925394@newsc.telia.net>

Jay Collins wrote:
> I have a list that is all the lines from the password file.
>
> if I did print p[0] # p being a line in /etc/passwd
> I would get:
>
> uucp:x:10:14:uucp:/var/spool/uucp:

how about:

    import pwd
    mylist = [entry[0] for entry in pwd.getpwall()]

> or whatever. My question is how do process this list into another
> list with just everything before the first ":". I'm not quite sure how to
> do this.

...but if you insist on reading /etc/passwd yourself,
you could do:

    passwd = open("/etc/passwd").readlines()
    mylist = [entry.split(":", 1)[0] for entry in passwd]

hope this helps!

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From stadt at cs.utwente.nl  Mon Feb  5 22:53:29 2001
From: stadt at cs.utwente.nl (Richard van de Stadt)
Date: Mon, 05 Feb 2001 22:53:29 +0100
Subject: cgi.FieldStorage() problem
Message-ID: <3A7F20D9.4BE57828@cs.utwente.nl>

Hi,

I have a problem, which seems to happen with 2.0, and not with 1.5.2

I have a webform (with quite some data in it) which is "POST"ed
to a cgi-script that calls

cgi.FieldStorage()

It fails. This exception is what I get if I "try" it:

Traceback (most recent call last):
  File "/...../cgi-bin/submitAssignRP.py", line 144, in main
  File "/opt/python/2.0/sparc-sunos5.6/lib/python2.0/cgi.py", line 504, in __init__
  File "/opt/python/2.0/sparc-sunos5.6/lib/python2.0/cgi.py", line 593, in read_multi
  File "/opt/python/2.0/sparc-sunos5.6/lib/python2.0/cgi.py", line 506, in __init__
  File "/opt/python/2.0/sparc-sunos5.6/lib/python2.0/cgi.py", line 603, in read_single
  File "/opt/python/2.0/sparc-sunos5.6/lib/python2.0/cgi.py", line 623, in read_lines
  File "/opt/python/2.0/sparc-sunos5.6/lib/python2.0/cgi.py", line 713, in make_file
  File "/opt/python/2.0/sparc-sunos5.6/lib/python2.0/tempfile.py", line 144, in TemporaryFile
OSError: [Errno 24] Too many open files: '/var/tmp/@24739.61'

Would anyone know what is causing this? Is it python, is it the OS?
And what's more, if there is a solution for this? 

Richard.


From stadt at cs.utwente.nl  Mon Feb  5 23:31:59 2001
From: stadt at cs.utwente.nl (Richard van de Stadt)
Date: Mon, 05 Feb 2001 23:31:59 +0100
Subject: cgi.FieldStorage() problem
References: <3A7F20D9.4BE57828@cs.utwente.nl>
Message-ID: <3A7F29DF.17D5C91C@cs.utwente.nl>

Richard van de Stadt wrote:
> 
> Hi,
> 
> I have a problem, which seems to happen with 2.0, and not with 1.5.2

Not true, Python 1.5.2. has the same problem. On another system this does
not happen, so I assume it's a kernel setting.

Richard.


From erno-news at erno.iki.fi  Tue Feb  6 02:57:46 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 06 Feb 2001 03:57:46 +0200
Subject: cgi.FieldStorage() problem
References: <3A7F20D9.4BE57828@cs.utwente.nl>
Message-ID: <kuelxc3751.fsf@lasipalatsi.fi>

In article <3A7F20D9.4BE57828 at cs.utwente.nl>, Richard van de Stadt
<stadt at cs.utwente.nl> writes:

| OSError: [Errno 24] Too many open files: '/var/tmp/@24739.61'

you probably just need to increase the resource limits for the account
that your http daemon runs under. i don't know what kind of startup
procedure you're using for the daemon so it's hard to say exactly how
to fix it, but the /bin/sh command for manipulating resource limits is
"ulimit" ("limit" with csh descendants).

if it still blows up after you've upped the limit, it might
be a problem with the cgi module going haywire and opening
new files boundlessly, but that seems pretty unlikely.

  -- erno


From stadt at cs.utwente.nl  Tue Feb  6 10:51:53 2001
From: stadt at cs.utwente.nl (Richard van de Stadt)
Date: Tue, 06 Feb 2001 10:51:53 +0100
Subject: cgi.FieldStorage() problem
References: <3A7F20D9.4BE57828@cs.utwente.nl> <kuelxc3751.fsf@lasipalatsi.fi>
Message-ID: <3A7FC939.DAC10B75@cs.utwente.nl>

Erno Kuusela wrote:
> 
> In article <3A7F20D9.4BE57828 at cs.utwente.nl>, Richard van de Stadt
> <stadt at cs.utwente.nl> writes:
> 
> | OSError: [Errno 24] Too many open files: '/var/tmp/@24739.61'
> 
> you probably just need to increase the resource limits for the account
> that your http daemon runs under. i don't know what kind of startup
> procedure you're using for the daemon so it's hard to say exactly how
> to fix it, but the /bin/sh command for manipulating resource limits is
> "ulimit" ("limit" with csh descendants).
> 
> if it still blows up after you've upped the limit, it might
> be a problem with the cgi module going haywire and opening
> new files boundlessly, but that seems pretty unlikely.

But still, it looks like that, although only 1 at the time.
The webform it gets to work with has 140 <input ...> lines,
while the number of temporary files that were created is 61.
The limit is probably set at 64, If you count the opening of
the script, and some modules, this makes sense.

It seems like each <input ...> line causes one temporary file to be
opened, although only 1 exists at the same time. Does cgi.FieldStorage()
work like that???

calling ulimit from within the script, which should, I think, do what you
suggest:

ulimitCommand = "/usr/bin/ulimit -n 512"
os.system (ulimitCommand)

doesn't help.

Richard.


From stadt at cs.utwente.nl  Tue Feb  6 12:17:29 2001
From: stadt at cs.utwente.nl (Richard van de Stadt)
Date: Tue, 06 Feb 2001 12:17:29 +0100
Subject: tempfile(s) problem (Re: cgi.FieldStorage() problem)
References: <3A7F20D9.4BE57828@cs.utwente.nl> <kuelxc3751.fsf@lasipalatsi.fi> <3A7FC939.DAC10B75@cs.utwente.nl>
Message-ID: <3A7FDD49.201783F@cs.utwente.nl>

Richard van de Stadt wrote:
> 
> Erno Kuusela wrote:
> >
> > In article <3A7F20D9.4BE57828 at cs.utwente.nl>, Richard van de Stadt
> > <stadt at cs.utwente.nl> writes:
> >
> > | OSError: [Errno 24] Too many open files: '/var/tmp/@24739.61'
> >
> > you probably just need to increase the resource limits for the account
> > that your http daemon runs under. i don't know what kind of startup
> > procedure you're using for the daemon so it's hard to say exactly how
> > to fix it, but the /bin/sh command for manipulating resource limits is
> > "ulimit" ("limit" with csh descendants).
> >
> > if it still blows up after you've upped the limit, it might
> > be a problem with the cgi module going haywire and opening
> > new files boundlessly, but that seems pretty unlikely.
> 
> But still, it looks like that, although only 1 at the time.
> The webform it gets to work with has 140 <input ...> lines,
> while the number of temporary files that were created is 61.
> The limit is probably set at 64, If you count the opening of
> the script, and some modules, this makes sense.
> 
> It seems like each <input ...> line causes one temporary file to be
> opened, although only 1 exists at the same time. Does cgi.FieldStorage()
> work like that???
> 
> calling ulimit from within the script, which should, I think, do what you
> suggest:
> 
> ulimitCommand = "/usr/bin/ulimit -n 512"
> os.system (ulimitCommand)
> 
> doesn't help.
> 
> Richard.

I copied the data to another system where ulimit -n shows the number 260 for
the maximum number of open files, and there it works!

So, indeed it seems that cgi.FieldStorage() creates a temporary file for each
<input> line of a webform, only one at the time, however, each time with a
different file descriptor. I think this looks bad: "You can use this program, but
in order to run it, Python requires you to change your system settings"...

(I'm using Python for CyberChair, an online paper submission and reviewing system.
The webform under discussion has one input field for each reviewer, stating the
papers he or she is supposed to be reviewing. ICML'01 has almost 140 reviewers...).

Richard.
To boldly go where no webform has gone before :-)


From erno-news at erno.iki.fi  Tue Feb  6 12:28:24 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 06 Feb 2001 13:28:24 +0200
Subject: cgi.FieldStorage() problem
References: <3A7F20D9.4BE57828@cs.utwente.nl> <kuelxc3751.fsf@lasipalatsi.fi> <3A7FC939.DAC10B75@cs.utwente.nl>
Message-ID: <ku7l342gpz.fsf@lasipalatsi.fi>

In article <3A7FC939.DAC10B75 at cs.utwente.nl>, Richard van de Stadt
<stadt at cs.utwente.nl> writes:

| calling ulimit from within the script, which should, I think, do what you
| suggest:

| ulimitCommand = "/usr/bin/ulimit -n 512"
| os.system (ulimitCommand)

that can't work. it's like trying to run os.system('/usr/bin/cd /tmp')
to change directory. do it in the shell you start httpd from, or use
the resource module, or something.

  -- erno


From stadt at cs.utwente.nl  Tue Feb  6 13:25:46 2001
From: stadt at cs.utwente.nl (Richard van de Stadt)
Date: Tue, 06 Feb 2001 13:25:46 +0100
Subject: cgi.FieldStorage() problem
References: <3A7F20D9.4BE57828@cs.utwente.nl> <kuelxc3751.fsf@lasipalatsi.fi> <3A7FC939.DAC10B75@cs.utwente.nl> <ku7l342gpz.fsf@lasipalatsi.fi>
Message-ID: <3A7FED4A.7F4918EA@cs.utwente.nl>

Erno Kuusela wrote:
> 
> In article <3A7FC939.DAC10B75 at cs.utwente.nl>, Richard van de Stadt
> <stadt at cs.utwente.nl> writes:
> 
> | calling ulimit from within the script, which should, I think, do what you
> | suggest:
> 
> | ulimitCommand = "/usr/bin/ulimit -n 512"
> | os.system (ulimitCommand)
> 
> that can't work. it's like trying to run os.system('/usr/bin/cd /tmp')
> to change directory. do it in the shell you start httpd from, or use
> the resource module, or something.

I don't start the httpd. It is not in my control... I've asked the
sysadmins of the server it concerns (at Purdue Univ.) to have a look at
it.

Richard.


From swun at esec.com.au  Mon Feb  5 23:18:51 2001
From: swun at esec.com.au (Sam Wun)
Date: Tue, 06 Feb 2001 09:18:51 +1100
Subject: sorting on IP addresses
Message-ID: <3A7F26CB.BE85C96F@esec.com.au>

Hi,

Does anyone know what is the quickly way to sort a list of IP addresses?

ie. 203.21.254.89 should be larger than 203.21.254.9

Thanks
Sam



From chris at onca.catsden.net  Mon Feb  5 23:23:15 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Mon, 5 Feb 2001 14:23:15 -0800 (PST)
Subject: sorting on IP addresses
In-Reply-To: <3A7F26CB.BE85C96F@esec.com.au>
Message-ID: <Pine.LNX.4.30.0102051415240.27504-100000@onca.catsden.net>

On Tue, 6 Feb 2001, Sam Wun wrote:

> Hi,
>
> Does anyone know what is the quickly way to sort a list of IP addresses?
>
> ie. 203.21.254.89 should be larger than 203.21.254.9

This may not be the fastest way, but it works, and its elegant :)

import string

def cmp_ipaddress ( ip1, ip2 ):
	parts1 = map(lambda x:int(x), string.split(ip1,'.'))
	parts2 = map(lambda x:int(x), string.split(ip2,'.'))
	comparisons = map ( lambda x,y: cmp(x,y), parts1, parts2 )
	return reduce ( lambda x,y: x or y, comparisons )

ips = [ '203.21.254.89', '203.21.254.9' ]

ips.sort ( cmp_ipaddress )

print `ips`

-> ['203.21.254.9', '203.21.254.89']


   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From swun at esec.com.au  Mon Feb  5 23:42:42 2001
From: swun at esec.com.au (Sam Wun)
Date: Tue, 06 Feb 2001 09:42:42 +1100
Subject: sorting on IP addresses
References: <Pine.LNX.4.30.0102051415240.27504-100000@onca.catsden.net>
Message-ID: <3A7F2C62.79E828BC@esec.com.au>

Many thanks.
sam
chris at onca.catsden.net wrote:

> On Tue, 6 Feb 2001, Sam Wun wrote:
>
> > Hi,
> >
> > Does anyone know what is the quickly way to sort a list of IP addresses?
> >
> > ie. 203.21.254.89 should be larger than 203.21.254.9
>
> This may not be the fastest way, but it works, and its elegant :)
>
> import string
>
> def cmp_ipaddress ( ip1, ip2 ):
>         parts1 = map(lambda x:int(x), string.split(ip1,'.'))
>         parts2 = map(lambda x:int(x), string.split(ip2,'.'))
>         comparisons = map ( lambda x,y: cmp(x,y), parts1, parts2 )
>         return reduce ( lambda x,y: x or y, comparisons )
>
> ips = [ '203.21.254.89', '203.21.254.9' ]
>
> ips.sort ( cmp_ipaddress )
>
> print `ips`
>
> -> ['203.21.254.9', '203.21.254.89']
>
>    ("`-/")_.-'"``-._        Ch'marr, a.k.a.
>     . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
>    (v_,)'  _  )`-.\  ``-'
>   _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
> ((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++

--
Sam Wun                         Firewalls / Security
Senior Software Engineer        Electronic Commerce
eSec Limited                    Phone: +61 3 8371 5376
mailto:swun at eSec.com.au         C++/JAVA/UNIX/OOP/OOD


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010206/dd442030/attachment.htm>

From che at debian.org  Mon Feb  5 23:25:06 2001
From: che at debian.org (Ben Gertzfield)
Date: 05 Feb 2001 14:25:06 -0800
Subject: sorting on IP addresses
References: <3A7F26CB.BE85C96F@esec.com.au>
Message-ID: <yttvgqovkcd.fsf@gilgamesh.cse.ucsc.edu>

>>>>> "Sam" == Sam Wun <swun at esec.com.au> writes:

    Sam> Hi, Does anyone know what is the quickly way to sort a list
    Sam> of IP addresses?

    Sam> ie. 203.21.254.89 should be larger than 203.21.254.9

If you're using Python 1.6 or greater, the socket module includes
inet_aton and inet_ntoa functions, used for converting IP addresses
from 123.45.67.89 format to the 4-byte "packed IP" representation.

You can then just use a normal sort on the packed IPs.

Ben

-- 
Brought to you by the letters P and R and the number 1.
"Atatakaku natta kara, otaku to uchi no kazoku de, nani ka shimasen ka?"
Debian GNU/Linux maintainer of Gimp and GTK+ -- http://www.debian.org/


From jacobkm at cats.ucsc.edu  Mon Feb  5 23:46:07 2001
From: jacobkm at cats.ucsc.edu (Jacob Kaplan-Moss)
Date: Mon, 05 Feb 2001 14:46:07 -0800
Subject: sorting on IP addresses
References: <3A7F26CB.BE85C96F@esec.com.au>
Message-ID: <jacobkm-71251E.14460705022001@news.ucsc.edu>

In article <3A7F26CB.BE85C96F at esec.com.au>, Sam Wun <swun at esec.com.au> 
wrote:
> Hi,
> 
> Does anyone know what is the quickly way to sort a list of IP addresses?
> 
> ie. 203.21.254.89 should be larger than 203.21.254.9
> 
> Thanks
> Sam
> 

Sam --

Not sure about quick, but this is the way I've been doing it.  If any of 
the gurus out there would like to point out a better method, I would be 
happy to read it.

iplist = ["203.21.254.89",
        "203.21.254.9",
        "1.2.3.4",
        "1.2.3.5",
        "211.1.50.124",
        "207.1.23.4",
        "192.168.1.101",
        "192.168.1.1"]

def ip_compare( a, b ):
   """a and b are valid IP addresses of the form w.x.y.z."""
   
   la = [int(n) for n in a.split(".")]
   lb = [int(n) for n in b.split(".")]
   
   for pair in zip(la, lb):
      if pair[0] > pair[1]:
         return -1
      elif pair[0] < pair[1]:
         return 1
   
   return 0
   
iplist.sort(ip_compare)
print iplist

Hope that helps, 

Jacob


From f8dy at my-deja.com  Tue Feb  6 17:33:18 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Tue, 06 Feb 2001 16:33:18 GMT
Subject: sorting on IP addresses
References: <3A7F26CB.BE85C96F@esec.com.au> <jacobkm-71251E.14460705022001@news.ucsc.edu>
Message-ID: <95p90e$aot$1@nnrp1.deja.com>

In article <jacobkm-71251E.14460705022001 at news.ucsc.edu>,
  Jacob Kaplan-Moss <jacobkm at cats.ucsc.edu> wrote:
> In article <3A7F26CB.BE85C96F at esec.com.au>, Sam Wun
<swun at esec.com.au>
> wrote:
> > Does anyone know what is the quickly way to sort a list of IP
addresses?
> >
> > ie. 203.21.254.89 should be larger than 203.21.254.9

A quick-and-dirty timing test reveals that mapping the IP addresses to
packed IP addresses (via socket.inet_aton), sorting that list, and
mapping the result back to unpacked IP addresses (via socket.inet_ntoa)
is about much, much faster than defining a custom sort routine.

def packunpack(li):
    import socket
    packed = [socket.inet_aton(addr) for addr in li]
    packed.reverse()
    packed.sort()
    li = [socket.inet_ntoa(addr) for addr in packed]

def cmplambda(li):
    def cmp_ipaddress ( ip1, ip2 ):
        import string
        parts1 = map(lambda x:int(x), string.split(ip1,'.'))
        parts2 = map(lambda x:int(x), string.split(ip2,'.'))
        comparisons = map ( lambda x,y: cmp(x,y), parts1, parts2 )
        return reduce ( lambda x,y: x or y, comparisons )
    li.sort(cmp_ipaddress)

def cmpfor(li):
    def ip_compare( a, b ):
      """a and b are valid IP addresses of the form w.x.y.z."""
      la = [int(n) for n in a.split(".")]
      lb = [int(n) for n in b.split(".")]
      for pair in zip(la, lb):
          if pair[0] > pair[1]:
            return -1
          elif pair[0] < pair[1]:
            return 1
      return 0
    li.sort(ip_compare)

def makelist():
    import random
    li = []
    for i in range(10000):
        li.append("%s.%s.%s.%s" % (random.randint(0, 255),
random.randint(0, 255), random.randint(0, 255), random.randint(0, 255)))
    return li

def timedsort(li, funcname):
    import time
    func = eval(funcname)
    startTime = time.time()
    func(li)
    endTime = time.time()
    print funcname.ljust(20), str(endTime - startTime)

if __name__ == "__main__":
    li = makelist()
    timedsort(li[:], "packunpack")
    timedsort(li[:], "cmplambda")
    timedsort(li[:], "cmpfor")

##Output (smaller numbers are better):
packunpack           0.230000019073
cmplambda            15.3830000162
cmpfor               9.73399996758

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From maxm at normik.dk  Wed Feb  7 09:33:26 2001
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Wed, 7 Feb 2001 09:33:26 +0100 
Subject: sorting on IP addresses
Message-ID: <7BD10B680501D411B9DF009027E06F32197BE0@exchange>

From: Sam Wun [mailto:swun at esec.com.au]

>Does anyone know what is the quickly way to sort a list of IP addresses?

>ie. 203.21.254.89 should be larger than 203.21.254.9

it probably doesn't get any quikclier than this:

-----------------------------------

from string import split, join

ips = """
192.168.1.1
192.168.1.2
192.168.1.3
192.169.1.5
127.1.2.3
192.167.1.6
191.168.1.4
193.168.1.2
127.1.1.1
194.168.1.8
195.168.1.2
"""

ipList = split(ips, '\n')

ipTupleList = []
for ip in ipList:
    if ip != '': # no empty lines
        n1, n2, n3, n4 = tuple(split(ip, '.'))
        ipTupleList.append((int(n1), int(n2), int(n3), int(n4)))
ipTupleList.sort()

for ipTuple in ipTupleList:
    print "%s.%s.%s.%s" % ipTuple

-----------------------------------

Regards Max M



From f8dy at my-deja.com  Wed Feb  7 15:29:59 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Wed, 07 Feb 2001 14:29:59 GMT
Subject: sorting on IP addresses
References: <mailman.981535000.18754.python-list@python.org>
Message-ID: <95rm4s$dom$1@nnrp1.deja.com>

In article <mailman.981535000.18754.python-list at python.org>,
  =?iso-8859-1?Q?Max_M=F8ller_Rasmussen?= <maxm at normik.dk> wrote:
> it probably doesn't get any quikclier than this:
>
> -----------------------------------
>
> from string import split, join
>
> ips = """
> 192.168.1.1
> 192.168.1.2
> 192.168.1.3
> 192.169.1.5
> 127.1.2.3
> 192.167.1.6
> 191.168.1.4
> 193.168.1.2
> 127.1.1.1
> 194.168.1.8
> 195.168.1.2
> """
>
> ipList = split(ips, '\n')
>
> ipTupleList = []
> for ip in ipList:
>     if ip != '': # no empty lines
>         n1, n2, n3, n4 = tuple(split(ip, '.'))
>         ipTupleList.append((int(n1), int(n2), int(n3), int(n4)))
> ipTupleList.sort()
>
> for ipTuple in ipTupleList:
>     print "%s.%s.%s.%s" % ipTuple

Packing/unpacking the IP addresses via socket.inet_aton/inet_ntoa is
still faster, probably because more of the work is done in a natively
compiled module.

Timing test of various methods:
  http://deja.com/threadmsg_md.xp?AN=724795461.1

The above method takes 0.69 seconds on my machine to sort a list of
10000 IP addresses, whereas the pack/unpack method takes only 0.23
seconds.

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From gradha at iname.com  Mon Feb  5 23:38:01 2001
From: gradha at iname.com (gradha at iname.com)
Date: Mon, 5 Feb 2001 23:38:01 +0100
Subject: Is comp.lang.python.announce dead?
Message-ID: <90an59.4hk.ln@127.0.0.1>

Hi.

Just today I fetched the news and got around 10 "announce" messages, in
this newsgroup instead of comp.lang.python.announce. Does this mean that
the latter is officially dead?

-- 

 Grzegorz Adam Hankiewicz   gradha at iname.com - http://gradha.infierno.org
 Other web pages:  http://glub.ehu.es/  -  http://welcome.to/gogosoftware/


From mwh21 at cam.ac.uk  Tue Feb  6 00:38:32 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Feb 2001 23:38:32 +0000
Subject: Is comp.lang.python.announce dead?
References: <90an59.4hk.ln@127.0.0.1>
Message-ID: <m366io4s5j.fsf@atrus.jesus.cam.ac.uk>

gradha at iname.com writes:

> Just today I fetched the news and got around 10 "announce" messages,
> in this newsgroup instead of comp.lang.python.announce. Does this
> mean that the latter is officially dead?

No, it's alive.  I see 23 messages in it at the moment.  It *was* dead
for quite a long time, but it's back again.

Cheers,
M.
(part of the comp.lang.python.announce moderation team)

-- 
  Arrrrgh, the braindamage!  It's not unlike the massively
  non-brilliant decision to use the period in abbreviations 
  as well as a sentence terminator.  Had these people no 
  imagination at _all_?                 -- Erik Naggum, comp.lang.lisp


From moshez at zadka.site.co.il  Tue Feb  6 12:01:08 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Tue,  6 Feb 2001 13:01:08 +0200 (IST)
Subject: Is comp.lang.python.announce dead?
In-Reply-To: <90an59.4hk.ln@127.0.0.1>
References: <90an59.4hk.ln@127.0.0.1>
Message-ID: <20010206110108.88F2BA840@darjeeling.zadka.site.co.il>

On Mon, 5 Feb 2001 23:38:01 +0100, gradha at iname.com wrote:
> Hi.
> 
> Just today I fetched the news and got around 10 "announce" messages, in
> this newsgroup instead of comp.lang.python.announce. Does this mean that
> the latter is officially dead?

Hardly. c.l.p.a is very much alive and being used. Perhaps less often
then it should be. Consider this a call to announce stuff on c.l.p.a --
we have a 24/7 team of moderators around the world, ready to approve
postings.
-- 
For public key: finger moshez at debian.org | gpg --import
<doogie> Debian - All the power, without the silly hat.
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From barry at digicool.com  Fri Feb  9 16:26:04 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Fri, 9 Feb 2001 10:26:04 -0500
Subject: Is comp.lang.python.announce dead?
References: <90an59.4hk.ln@127.0.0.1>
Message-ID: <14980.3084.424354.137797@anthem.wooz.org>

>>>>> "gradha" ==   <gradha at iname.com> writes:

    gradha> Just today I fetched the news and got around 10 "announce"
    gradha> messages, in this newsgroup instead of
    gradha> comp.lang.python.announce. Does this mean that the latter
    gradha> is officially dead?

c.l.py.a is absolutely live again, and as mentioned in a different
follow, there is a team of moderators who are helping to approve
messages for the newsgroup.

There was a separate and temporary outage of the moderation gateway as
we changed mail daemons on mail.python.org (for reasons I won't go
into).  That problem should be fixed now.

Please post your announcements to c.l.py.a!  Also, register your
programs with Parnassus and those announcements can be automatically
forwarded to the newsgroup.

-Barry



From NOSPAM at pacificnet.net  Mon Feb  5 23:54:20 2001
From: NOSPAM at pacificnet.net (Ian L)
Date: Mon, 05 Feb 2001 22:54:20 GMT
Subject: gdchart problem
Message-ID: <waGf6.2744$iM6.373896@newsread1.prod.itd.earthlink.net>

i just installed gdchart http://www.fred.net/brv/chart/ along with the
python interface.

I've gotten it to work, sort of. When i run my file from the command line,
it works fine. If i then try to run the same file from a web browser, it
doesnt work. Or rather, the code on the page executes without any errors.
But it doesnt generate the image.

I assume this is a permission problem. I've tried giving full permissions to
both the file and directory, but it still only works when i run it from the
command line.

can anyone give me some suggestions as to what i could check to get it
working from the browser?




From dan at eevolved.com  Tue Feb  6 03:14:39 2001
From: dan at eevolved.com (Dan Parisien)
Date: Tue, 06 Feb 2001 02:14:39 GMT
Subject: importing text code
Message-ID: <j6Jf6.204847$f36.8214792@news20.bellglobal.com>

Is it possible to import from a string in memory instead of a file? I can't 
find information on how to do it...

thanks, 
Dan


From scarblac at pino.selwerd.nl  Tue Feb  6 11:00:13 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 6 Feb 2001 10:00:13 GMT
Subject: importing text code
References: <j6Jf6.204847$f36.8214792@news20.bellglobal.com>
Message-ID: <slrn97vib5.siv.scarblac@pino.selwerd.nl>

Dan Parisien <dan at eevolved.com> wrote in comp.lang.python:
> Is it possible to import from a string in memory instead of a file? I can't 
> find information on how to do it...

I asked that last year and the effbot replied thusly:

# import module from string
# fredrik lundh, may 2000

import imp, sys

def do_import(name, source):
   module = imp.new_module(name)
   sys.modules[name] = module
   exec source in vars(module)
   return module

script = """
def hello():
   print 'hello you too'
"""

do_import("mymodule", script)

import mymodule
mymodule.hello()

-- 
Remco Gerlich


From jakegren at home.com  Tue Feb  6 03:55:37 2001
From: jakegren at home.com (jakegren at home.com)
Date: Tue, 06 Feb 2001 02:55:37 GMT
Subject: New Instance for Pmw.Notebook()?
Message-ID: <3A7F12E1.14EEAC60@home.com>

I'm having difficulty creating a Pmw.Notebook() outside of the root
window. I would like to click on a menu item and have a new window
created with the Pmw.Notebook(). This is what I have:

Button4.menu.add_command(label='Properties...',command=TestClick)
...
def TestClick():
    notebook=Pmw.NoteBook()
    notebook.pack()
    page=notebook.add('My Tab')
    entry=Entry(page,width=20)
    entry.pack()

This creates a Pmw.Notebook() immediately below the main menu. I know
I have to create a class to do what I want (create a new instance), but
I'm not sure how to start... Any suggestions?
--
Jake



From doughellmann at bigfoot.com  Tue Feb  6 13:35:35 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Tue, 06 Feb 2001 12:35:35 GMT
Subject: New Instance for Pmw.Notebook()?
References: <3A7F12E1.14EEAC60@home.com>
Message-ID: <rcSf6.56287$m5.3118370@news1.rdc1.ga.home.com>

To make the notebook appear in a new dialog, you have to instantiate the
new dialog and then specify that widget as the parent for the notebook
you create.  By default, when you pass no arguments, the parent is
assumed to be the root window and you get the behavior you're seeing.

Doug

In article <3A7F12E1.14EEAC60 at home.com>, jakegren at home.com wrote:

> I'm having difficulty creating a Pmw.Notebook() outside of the root
> window. I would like to click on a menu item and have a new window
> created with the Pmw.Notebook(). This is what I have:
> 
> Button4.menu.add_command(label='Properties...',command=TestClick)
> ...
> def TestClick():
>     notebook=Pmw.NoteBook() notebook.pack() page=notebook.add('My Tab')
>     entry=Entry(page,width=20) entry.pack()
> 
> This creates a Pmw.Notebook() immediately below the main menu. I know I
> have to create a class to do what I want (create a new instance), but
> I'm not sure how to start... Any suggestions?
> --
> Jake
>


From jakegren at home.com  Wed Feb  7 04:29:46 2001
From: jakegren at home.com (Jake Gren)
Date: Wed, 07 Feb 2001 03:29:46 GMT
Subject: New Instance for Pmw.Notebook()?
References: <3A7F12E1.14EEAC60@home.com> <rcSf6.56287$m5.3118370@news1.rdc1.ga.home.com>
Message-ID: <3A806C61.1085F97B@home.com>

Hmm. Ok, my class knowledge is limited (I'm working on it...), but I have to
call a class object, right? Each time a class is called, it generates and
returnsa new instance object. This is where I grind to a halt... I can
follow the simple class examples in "Learning Python," but I'm not sure how
to create a class for the "thing" I want to do below. Can you assist or
suggest a book or document? From what you mention below it doesn't seem like
a difficult task, but I'm lost... Thanks!
--
Jake

Doug Hellmann wrote:

> To make the notebook appear in a new dialog, you have to instantiate the
> new dialog and then specify that widget as the parent for the notebook
> you create.  By default, when you pass no arguments, the parent is
> assumed to be the root window and you get the behavior you're seeing.
>
> Doug
>
> In article <3A7F12E1.14EEAC60 at home.com>, jakegren at home.com wrote:
>
> > I'm having difficulty creating a Pmw.Notebook() outside of the root
> > window. I would like to click on a menu item and have a new window
> > created with the Pmw.Notebook(). This is what I have:
> >
> > Button4.menu.add_command(label='Properties...',command=TestClick)
> > ...
> > def TestClick():
> >     notebook=Pmw.NoteBook() notebook.pack() page=notebook.add('My Tab')
> >     entry=Entry(page,width=20) entry.pack()
> >
> > This creates a Pmw.Notebook() immediately below the main menu. I know I
> > have to create a class to do what I want (create a new instance), but
> > I'm not sure how to start... Any suggestions?
> > --
> > Jake



From fredrik at effbot.org  Wed Feb  7 08:30:42 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 07 Feb 2001 07:30:42 GMT
Subject: New Instance for Pmw.Notebook()?
References: <3A7F12E1.14EEAC60@home.com>
Message-ID: <CQ6g6.13181$AH6.1981433@newsc.telia.net>

jakegren at home.com wrote:
> I'm having difficulty creating a Pmw.Notebook() outside of the root
> window. I would like to click on a menu item and have a new window
> created with the Pmw.Notebook().

try this:

> def TestClick():
+     top = Toplevel() # create a new toplevel window
+     notebook=Pmw.NoteBook(top) # create in new toplevel window
-     notebook=Pmw.NoteBook() # create in root window
>     notebook.pack()
>     page=notebook.add('My Tab')
>     entry=Entry(page,width=20)
>     entry.pack()

for more info on toplevel windows, see:
http://www.pythonware.com/library/tkinter/introduction/application-windows.htm
http://www.pythonware.com/library/tkinter/introduction/toplevel.htm

Cheers /F




From dan at eevolved.com  Tue Feb  6 06:26:52 2001
From: dan at eevolved.com (Dan Parisien)
Date: Tue, 06 Feb 2001 05:26:52 GMT
Subject: do...while loops
Message-ID: <wWLf6.59035$Pm2.1271675@news20.bellglobal.com>

why don't do...while loops exist in python? Is there are better way to get 
to that functionality?

e.g.
file = open(...)
do:
        line = file.readline()
while line != ""


Do I have to write a PEP or read some more documentation? ;-)

Thanks,
Dan


From trotts at mit.edu  Tue Feb  6 06:48:27 2001
From: trotts at mit.edu (Issac Trotts)
Date: Mon, 5 Feb 2001 21:48:27 -0800
Subject: do...while loops
In-Reply-To: <wWLf6.59035$Pm2.1271675@news20.bellglobal.com>; from dan@eevolved.com on Tue, Feb 06, 2001 at 05:26:52AM +0000
References: <wWLf6.59035$Pm2.1271675@news20.bellglobal.com>
Message-ID: <20010205214827.A5294@llnl.gov>

How about this: 

file = open(...)
while(1): 
    line = file.readline()
    if(line == ""): break

-Issac

Dan Parisien (dan at eevolved.com) wrote:

> why don't do...while loops exist in python? Is there are better way to get 
> to that functionality?
> 
> e.g.
> file = open(...)
> do:
>         line = file.readline()
> while line != ""
> 
> 
> Do I have to write a PEP or read some more documentation? ;-)
> 
> Thanks,
> Dan
> -- 
> http://mail.python.org/mailman/listinfo/python-list



From dan at eevolved.com  Tue Feb  6 09:44:01 2001
From: dan at eevolved.com (Dan Parisien)
Date: Tue, 06 Feb 2001 08:44:01 GMT
Subject: do...while loops
References: <wWLf6.59035$Pm2.1271675@news20.bellglobal.com> <mailman.981438629.14737.python-list@python.org>
Message-ID: <lPOf6.205292$f36.8273714@news20.bellglobal.com>

Issac Trotts wrote:

> How about this:
> 
> file = open(...)
> while(1):
>     line = file.readline()
>     if(line == ""): break
> 
> -Issac

Oh yeah... breaks. 
*sigh* (not a big fan...)

Thank you Issac,
Dan


From fredrik at effbot.org  Tue Feb  6 09:51:49 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 06 Feb 2001 08:51:49 GMT
Subject: do...while loops
References: <wWLf6.59035$Pm2.1271675@news20.bellglobal.com> <mailman.981438629.14737.python-list@python.org>
Message-ID: <FWOf6.13030$AH6.1939889@newsc.telia.net>

Issac Trotts wrote:
> How about this:
>
> file = open(...)
> while(1):
>     line = file.readline()
>     if(line == ""): break

usually spelled:

    while 1:
        line = file.readline()
        if not line:
            break
        ...

(using parens after statements is bad style in Python,
and so is using == to test for an empty sequence...)

in 2.1, it's better spelled as:

    for line in file.xreadlines():
        ...

Cheers /F




From scarblac at pino.selwerd.nl  Tue Feb  6 12:28:13 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 6 Feb 2001 11:28:13 GMT
Subject: do...while loops
References: <wWLf6.59035$Pm2.1271675@news20.bellglobal.com>
Message-ID: <slrn97vng6.snm.scarblac@pino.selwerd.nl>

Dan Parisien <dan at eevolved.com> wrote in comp.lang.python:
> why don't do...while loops exist in python? Is there are better way to get 
> to that functionality?
> 
> e.g.
> file = open(...)
> do:
>         line = file.readline()
> while line != ""
>  
> Do I have to write a PEP or read some more documentation? ;-)

This has been discussed to death, over and over again. Some of the reasons
cited are that the construct is redundant, there is no way to indent it in a
way consistent with the rest of Python, and Guido doesn't like it (all of
those may or may not be true...).

The Python idiom in the general case is:

while 1:
   line = file.readline()
   ...
   if not line:
      break
   ...
   
But in the specific case of reading lines from files, this is neater:

for line in file.readlines():
   ...
   
Which reads the whole file into memory at once. For the case your files are
too big, the upcoming Python 2.1 will have the .xreadlines() method which
works the same way but does not load everything at once.

At the moment, the fileinput module gives that functionality:

import fileinput
for line in fileinput.input("filename"):
   ...
   

HTH.

-- 
Remco Gerlich


From LangaK at discoveryhealth.co.za  Tue Feb  6 14:00:33 2001
From: LangaK at discoveryhealth.co.za (Langa Kentane)
Date: Tue, 6 Feb 2001 15:00:33 +0200 
Subject: do...while loops
Message-ID: <A69BC9CA4B39D411BF2200105A45B45B075B8D3D@dhexchange.discoveryhealth.co.za>

Why is is it while 1:
would while 2:
or 	while 3: work?

newbie

-----Original Message-----
From: scarblac at pino.selwerd.nl [mailto:scarblac at pino.selwerd.nl]
Sent: 06 February 2001 1:28 PM
To: python-list at python.org
Subject: Re: do...while loops


Dan Parisien <dan at eevolved.com> wrote in comp.lang.python:
> why don't do...while loops exist in python? Is there are better way to get

> to that functionality?
> 
> e.g.
> file = open(...)
> do:
>         line = file.readline()
> while line != ""
>  
> Do I have to write a PEP or read some more documentation? ;-)

This has been discussed to death, over and over again. Some of the reasons
cited are that the construct is redundant, there is no way to indent it in a
way consistent with the rest of Python, and Guido doesn't like it (all of
those may or may not be true...).

The Python idiom in the general case is:

while 1:
   line = file.readline()
   ...
   if not line:
      break
   ...
   
But in the specific case of reading lines from files, this is neater:

for line in file.readlines():
   ...
   
Which reads the whole file into memory at once. For the case your files are
too big, the upcoming Python 2.1 will have the .xreadlines() method which
works the same way but does not load everything at once.

At the moment, the fileinput module gives that functionality:

import fileinput
for line in fileinput.input("filename"):
   ...
   

HTH.

-- 
Remco Gerlich
-- 
http://mail.python.org/mailman/listinfo/python-list



From jurgen.defurne at philips.com  Tue Feb  6 14:27:14 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Tue, 6 Feb 2001 14:27:14 +0100
Subject: do...while loops
Message-ID: <0056900015700115000002L052*@MHS>

If I can't use readlines, I usually do it like this :

fh = open(...)

line = fh.readline()
while line:
    # Process line here
    line = fh.readline()

No break necessary...

Jurgen


From sholden at holdenweb.com  Tue Feb  6 14:28:54 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 6 Feb 2001 08:28:54 -0500
Subject: do...while loops
References: <mailman.981465338.22824.python-list@python.org>
Message-ID: <92Tf6.2689$x4.52221@e420r-atl2.usenetserver.com>

"Langa Kentane" <LangaK at discoveryhealth.co.za> wrote in message
news:mailman.981465338.22824.python-list at python.org...
> Why is is it while 1:
> would while 2:
> or while 3: work?
>
> newbie
>
Absolutely. Python has many "true" values, and far fewer "false" ones.

The following is a (non-exclusive) list of values which are false in a
Boolean context:

    0
    ()
    []
    {}
    ""
    None

The idea behind "while 1:" is simply to generate an infinite loop (which is
usually terminated by a break inside the loop body). Therefore "while 2:" or
"while 3:" or "while 'something':" or "while [42]:", and many others, will
work exactly the same way.

However, Python users will recognise "while 1:" instantly, and spend three
weeks discussing on c.l.py why this is or is not the best way to express the
concept.

regards
 Steve




From grey at despair.rpglink.com  Tue Feb  6 14:43:03 2001
From: grey at despair.rpglink.com (Steve Lamb)
Date: Tue, 06 Feb 2001 13:43:03 -0000
Subject: do...while loops
References: <mailman.981465338.22824.python-list@python.org> <92Tf6.2689$x4.52221@e420r-atl2.usenetserver.com>
Message-ID: <slrn97vvr7.lcl.grey@teleute.dmiyu.org>

On Tue, 6 Feb 2001 08:28:54 -0500, Steve Holden <sholden at holdenweb.com> wrote:
>However, Python users will recognise "while 1:" instantly, and spend three
>weeks discussing on c.l.py why this is or is not the best way to express the
>concept.

    Henceforth I declare that the "while 1:" idiom no longer be used.  From
this day forward a new idiom will be used instead to denote an infinite loop.
When one is needed, please use "while 'this thread':" since we all know it
will never die thus providing instant recognition of its infinite nature.

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From wware at world.std.com  Tue Feb  6 14:56:29 2001
From: wware at world.std.com (Will Ware)
Date: Tue, 6 Feb 2001 13:56:29 GMT
Subject: do...while loops
References: <mailman.981465338.22824.python-list@python.org> <92Tf6.2689$x4.52221@e420r-atl2.usenetserver.com> <slrn97vvr7.lcl.grey@teleute.dmiyu.org>
Message-ID: <G8C9E5.5Hp@world.std.com>

Steve Lamb (grey at despair.rpglink.com) wrote:
> Henceforth I declare that the "while 1:" idiom no longer be used...
> please use "while 'this thread':" since we all know it will never die...

Maybe something a little more general, perhaps involving a new keyword.

whileWhiningAbout "indentation":
    ...do stuff...

whileWhiningAbout "lambda":
    ...do stuff...

whileWhiningAbout "rich comparisons":
    ...do stuff...

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From danielk at aracnet.com  Tue Feb  6 15:52:39 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Tue, 06 Feb 2001 06:52:39 -0800
Subject: do...while loops
References: <mailman.981465338.22824.python-list@python.org>
Message-ID: <di308tgio7grhpqgti7quv0l4mmo5fmj5p@4ax.com>

On Tue, 6 Feb 2001 15:00:33 +0200, Langa Kentane <LangaK at discoveryhealth.co.za>
wrote:

>Why is is it while 1:
>would while 2:
>or 	while 3: work?

Of course it will work. The only integer that doesn't test for 'true' is 0 (or
-0).

Coming from a Smalltalk background I don't like using integer literals for
this. Instead I define a module like

	system_constants.py
	true = 1
	false = 0
etc...

and then write the loop as:

	while true:

Daniel Klein
Portland OR USA



From dan at eevolved.com  Wed Feb  7 01:34:35 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 07 Feb 2001 00:34:35 GMT
Subject: do...while loops
References: <mailman.981465338.22824.python-list@python.org>
Message-ID: <vK0g6.65590$Pm2.1354846@news20.bellglobal.com>

Langa Kentane wrote:

> Why is is it while 1:
> would while 2:
> or    while 3: work?

yes :)

any non-zero value means true, so saying while 1: means 'always'



From SBrunning at trisystems.co.uk  Tue Feb  6 14:41:50 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 6 Feb 2001 13:41:50 -0000 
Subject: do...while loops
Message-ID: <31575A892FF6D1118F5800600846864D5B17BB@intrepid>

> From:	Langa Kentane [SMTP:LangaK at discoveryhealth.co.za]
> Why is is it while 1:
 
Because 1 is considered a true value, so this while loop will run forever.
Or until broken, hopefully!
 
> would while 2:
> or 	while 3: work?
 
Yup - any non-zero number is considered true. They wouldn't do anything
different, though, and they just might confuse someone. (Why 2?)
 
Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From aahz at panix.com  Tue Feb  6 17:01:51 2001
From: aahz at panix.com (Aahz Maruch)
Date: 6 Feb 2001 08:01:51 -0800
Subject: do...while loops
References: <mailman.981466899.30238.python-list@python.org>
Message-ID: <95p75f$85m$1@panix3.panix.com>

In article <mailman.981466899.30238.python-list at python.org>,
Simon Brunning  <SBrunning at trisystems.co.uk> wrote:
> From:	Langa Kentane [SMTP:LangaK at discoveryhealth.co.za]
>>
>> Why is is it while 1:
>> would while 2:
>> or 	while 3: work?
> 
>Yup - any non-zero number is considered true. They wouldn't do anything
>different, though, and they just might confuse someone. (Why 2?)

Well, let's split the difference:

while 2.5:
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Life as a contractor: when they tell you it's going to rain, sometimes
it drizzles and sometimes it pours.


From joconnor at cybermesa.com  Tue Feb  6 17:15:21 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 06 Feb 2001 16:15:21 +0000
Subject: do...while loops
References: <mailman.981466899.30238.python-list@python.org> <95p75f$85m$1@panix3.panix.com>
Message-ID: <3A802319.46EBCA35@cybermesa.com>

Aahz Maruch wrote:
> 
> In article <mailman.981466899.30238.python-list at python.org>,
> Simon Brunning  <SBrunning at trisystems.co.uk> wrote:
> > From: Langa Kentane [SMTP:LangaK at discoveryhealth.co.za]
> >>
> >> Why is is it while 1:
> >> would while 2:
> >> or   while 3: work?
> >
> >Yup - any non-zero number is considered true. They wouldn't do anything
> >different, though, and they just might confuse someone. (Why 2?)
> 
> Well, let's split the difference:
> 
> while 2.5:


Umm..then how about a counter?  Would this work?

x = 100
while x:
	print x
	x -= 1

Yeah, just tried it...it works.

Umm...anyone really like it? :)
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython


From dsh8290 at rit.edu  Tue Feb  6 20:15:16 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 6 Feb 2001 14:15:16 -0500
Subject: do...while loops
In-Reply-To: <3A802319.46EBCA35@cybermesa.com>; from joconnor@cybermesa.com on Tue, Feb 06, 2001 at 04:15:21PM +0000
References: <mailman.981466899.30238.python-list@python.org> <95p75f$85m$1@panix3.panix.com> <3A802319.46EBCA35@cybermesa.com>
Message-ID: <20010206141516.A2046@harmony.cs.rit.edu>

On Tue, Feb 06, 2001 at 04:15:21PM +0000, Jay O'Connor wrote:
| Aahz Maruch wrote:
| > 
| > In article <mailman.981466899.30238.python-list at python.org>,
| > Simon Brunning  <SBrunning at trisystems.co.uk> wrote:
| > > From: Langa Kentane [SMTP:LangaK at discoveryhealth.co.za]
| > >>
| > >> Why is is it while 1:
| > >> would while 2:
| > >> or   while 3: work?
| > >
| > >Yup - any non-zero number is considered true. They wouldn't do anything
| > >different, though, and they just might confuse someone. (Why 2?)
| > 
| > Well, let's split the difference:
| > 
| > while 2.5:

Shouldn't that be 1.5 ?  ;-)

| 
| 
| Umm..then how about a counter?  Would this work?
| 
| x = 100
| while x:
| 	print x
| 	x -= 1
| 
| Yeah, just tried it...it works.
| 
| Umm...anyone really like it? :)

for x in range( 100 , 0 , -1 ) :
	print x


With for you only have 1 one line that describes the entire loop,
instead of 3 lines that are spread about.  You also have no chance of
forgetting to (de/in)crement the counter when you get to the end of
the loop body.

-D




From joconnor at cybermesa.com  Tue Feb  6 20:43:29 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 06 Feb 2001 19:43:29 GMT
Subject: do...while loops
References: <mailman.981466899.30238.python-list@python.org> <95p75f$85m$1@panix3.panix.com> <3A802319.46EBCA35@cybermesa.com> <mailman.981487000.23183.python-list@python.org>
Message-ID: <3a80537c.8644434@news.cybermesa.com>

On Tue, 6 Feb 2001 14:15:16 -0500, D-Man <dsh8290 at rit.edu> wrote:

>On Tue, Feb 06, 2001 at 04:15:21PM +0000, Jay O'Connor wrote:

>| Umm..then how about a counter?  Would this work?
>| 
>| x = 100
>| while x:
>| 	print x
>| 	x -= 1
>| 
>| Yeah, just tried it...it works.
>| 
>| Umm...anyone really like it? :)
>
>for x in range( 100 , 0 , -1 ) :
>	print x
>
>
>With for you only have 1 one line that describes the entire loop,
>instead of 3 lines that are spread about.  You also have no chance of
>forgetting to (de/in)crement the counter when you get to the end of
>the loop body.

Oh, I agree totally, for the reasons you mention.  I thought it was
goofy which is why I put the smiley 

I was just trying to think of a practical (?) use for doing a "while"
loop with a number other than 1

Take care,
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum - http://pub1.ezboard.com/fobjectorienteddevelopmentpython

"God himself plays on the bass strings first, when he tunes the soul"


From wware at world.std.com  Tue Feb  6 21:41:14 2001
From: wware at world.std.com (Will Ware)
Date: Tue, 6 Feb 2001 20:41:14 GMT
Subject: do...while loops
References: <mailman.981466899.30238.python-list@python.org> <95p75f$85m$1@panix3.panix.com> <3A802319.46EBCA35@cybermesa.com> <mailman.981487000.23183.python-list@python.org> <3a80537c.8644434@news.cybermesa.com>
Message-ID: <G8Cs4r.GxE@world.std.com>

Jay O'Connor (joconnor at cybermesa.com) wrote:
> I was just trying to think of a practical (?) use for doing a "while"
> loop with a number other than 1

It could make sense if the "increment" operation were something tricky
and convoluted. I dunno, maybe something like

x = some_value
while x > some_threshold:
  do stuff with x
  x = 0.8 ** math.cos(x)

If the series of x values can't easily be generated by plugging a
sequence of integers into a function, then you might have a case for
using while. You'd be lucky, though, if your terminating condition
were as simple as x != 0.

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From mlh at idi.ntnu.no  Thu Feb  8 13:42:42 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 8 Feb 2001 13:42:42 +0100
Subject: do...while loops
References: <mailman.981466899.30238.python-list@python.org> <95p75f$85m$1@panix3.panix.com> <3A802319.46EBCA35@cybermesa.com> <mailman.981487000.23183.python-list@python.org>
Message-ID: <95u4do$jem$1@tyfon.itea.ntnu.no>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.981487000.23183.python-list at python.org...
> On Tue, Feb 06, 2001 at 04:15:21PM +0000, Jay O'Connor wrote:
> | Aahz Maruch wrote:
[...]
> | > >>
> | > >> Why is is it while 1:
> | > >> would while 2:
> | > >> or   while 3: work?
> | > >
> | > >Yup - any non-zero number is considered true. They wouldn't do
anything
> | > >different, though, and they just might confuse someone. (Why 2?)
> | >
> | > Well, let's split the difference:
> | >
> | > while 2.5:
>
> Shouldn't that be 1.5 ?  ;-)

Actually, the average of 1, 2, and 3 (the values in question) should
be 2, shouldn't it? :)

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From gregj at pobox.com  Tue Feb  6 21:25:50 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Tue, 06 Feb 2001 20:25:50 GMT
Subject: do...while loops
References: <mailman.981466899.30238.python-list@python.org> <95p75f$85m$1@panix3.panix.com> <3A802319.46EBCA35@cybermesa.com>
Message-ID: <95pmjp$p3b$1@nnrp1.deja.com>

In article <3A802319.46EBCA35 at cybermesa.com>, Jay O'Connor
<joconnor at cybermesa.com> wrote:

> Umm..then how about a counter?  Would this work?
>
> x = 100
> while x:
> 	print x
> 	x -= 1
>
> Yeah, just tried it...it works.
>
> Umm...anyone really like it? :)

That's a common C idiom, but I've always disliked it because x is
either a logical or arithmetic value depending on which statement you
are looking at. I prefer:

  while x > 0:
    ...
    x -= 1

To me this has several advantages:

1. There is absolutely no question about what the loop is doing:
counting down to zero, not waiting for some flag named x to be false.

2. If x is initially negative the loop won't even get started, whereas

  x = -1
  ...
  while x:
    ...
    x -= 1

Will loop for a long time and then crash.

3. The explicit comparison will still work if x is changed from an
integer to a floating point type. This kind of bug is particularly hard
to find:

  x = (((11.0 / 3.0) / 3.0) * 3.0) * 3.0  # 10.999999999999998
  ...
  while x:
      print x
      x -= 1

x will never be exactly equal to 0 because of floating-point
representation and rounding errors. The example above will even print
the sequence 11.0, 10.0, 9.0, ... 1.0, so it may not be immediately
obvious why the loop doesn't terminate.

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From joconnor at cybermesa.com  Tue Feb  6 22:11:59 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 06 Feb 2001 21:11:59 GMT
Subject: do...while loops
References: <mailman.981466899.30238.python-list@python.org> <95p75f$85m$1@panix3.panix.com> <3A802319.46EBCA35@cybermesa.com> <95pmjp$p3b$1@nnrp1.deja.com>
Message-ID: <3a806855.3522055@news.cybermesa.com>

On Tue, 06 Feb 2001 20:25:50 GMT, Greg Jorgensen <gregj at pobox.com>
wrote:

>That's a common C idiom, but I've always disliked it

Me too.  I was being silly.  I much prefer to loop over sequences,
even sequences of numbers
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum - http://pub1.ezboard.com/fobjectorienteddevelopmentpython

"God himself plays on the bass strings first, when he tunes the soul"


From stevena at neosynapse.net  Tue Feb  6 08:01:04 2001
From: stevena at neosynapse.net (Steven D. Arnold)
Date: Tue, 06 Feb 2001 02:01:04 -0500
Subject: using a func param in a nested function
Message-ID: <5.0.2.1.2.20010206015719.0409fcb8@phear.dementian.com>

Hi,

I know this is illegal:

         def a(self, *x):
                 def b(m, n):
                         global x
                         for pref in x:

The documentation says that you can't use global with a name defined as a 
formal parameter in a function.  The question, then, is how can I use x in 
function b, other than by passing it as a parameter?  I've already tried this:

         def a(self, *x):
                 y = deepcopy(x)
                 def b(m, n):
                         global y
                         for pref in y:

That approach doesn't work either.  Ideas?


--
Steven D. Arnold          Que quiero sera         stevena at neosynapse.net
"He was part of my dream, of course -- but then  I was part of his dream
too."                                                   -- Lewis Carroll




From ivnowa at hvision.nl  Tue Feb  6 09:44:19 2001
From: ivnowa at hvision.nl (Hans Nowak)
Date: Tue, 06 Feb 2001 08:44:19 GMT
Subject: using a func param in a nested function
References: <mailman.981442959.30419.python-list@python.org>
Message-ID: <3A7FB77E.5BF6@hvision.nl>

Steven D. Arnold wrote:
> 
> Hi,
> 
> I know this is illegal:
> 
>          def a(self, *x):
>                  def b(m, n):
>                          global x
>                          for pref in x:
> 
> The documentation says that you can't use global with a name defined as a
> formal parameter in a function.  The question, then, is how can I use x in
> function b, other than by passing it as a parameter?  

As far as I understand, the new scoping rules in Python 2.1 will fix
this. Take a look at this document:

  http://amk.ca/python/2.1/

--Hans Nowak


From swun at esec.com.au  Tue Feb  6 08:20:26 2001
From: swun at esec.com.au (Sam Wun)
Date: Tue, 06 Feb 2001 18:20:26 +1100
Subject: how to send email in python?
Message-ID: <3A7FA5BA.82FB7593@esec.com.au>

hi,

does anyone knows? I am using python 1.6.

Thansk
Sam



From chris at onca.catsden.net  Tue Feb  6 08:24:35 2001
From: chris at onca.catsden.net (chris at onca.catsden.net)
Date: Mon, 5 Feb 2001 23:24:35 -0800 (PST)
Subject: how to send email in python?
In-Reply-To: <3A7FA5BA.82FB7593@esec.com.au>
Message-ID: <Pine.LNX.4.30.0102052322480.29293-100000@onca.catsden.net>

On Tue, 6 Feb 2001, Sam Wun wrote:

> hi,
>
> does anyone knows? I am using python 1.6.

import os

f = os.popen ( "/usr/sbin/sendmail", "w" )

f.write ( """\
From: your address
To: recipient address
Subject: spam

spam spam wonderful spam spam spam
""" )


If you dont have sendmail on your system, or some other MTA, then you can
use smtplib to send it directly to an off-site MTA. I can get you some
simple code for that, but the above is the preferred way.


   ("`-/")_.-'"``-._        Ch'marr, a.k.a.
    . . `; -._    )-;-,_`)  Chris Cogdon <chmarr at furry.org.au>
   (v_,)'  _  )`-.\  ``-'
  _.- _..-_/ / ((.'       FC1.3: FFH3cmA+>++C++D++H++M++P++R++T+++WZ++Sm++
((,.-'   ((,/   fL               RLCT acl+++d++e+f+++h++i++++jp-sm++




From sheila at spamcop.net  Tue Feb  6 08:37:23 2001
From: sheila at spamcop.net (Sheila King)
Date: Tue, 06 Feb 2001 07:37:23 GMT
Subject: how to send email in python?
References: <3A7FA5BA.82FB7593@esec.com.au>
Message-ID: <h7av7t8ssi6v9djasb3mo3n7d7vigp16a1@4ax.com>

On Tue, 06 Feb 2001 18:20:26 +1100, Sam Wun <swun at esec.com.au> wrote in
comp.lang.python in article <3A7FA5BA.82FB7593 at esec.com.au>:

:does anyone knows? I am using python 1.6.


Here is one way:

-------------------------------------------
import smtplib

def readManyLines():
    """Read lines from the user until they enter a line with a single
period, and return a single string."""
    result = ""
    while 1:
        line = raw_input()
        if line == '.': return result
        result = result + line + "\n"

servername=raw_input("SMTP server name: ")
Fromaddress=raw_input("From address: ")
Toaddress=raw_input("To address: ")
print "Enter your message. To end, enter a line that begins with a"
print "period and has no other characters: "
MessageText=readManyLines()
server=smtplib.SMTP(servername)
server.sendmail(Fromaddress, Toaddress, MessageText)
server.quit()

-------------------------------------------

Note: This is very crude.

When it runs, it will prompt you for a smtp server. You input that data. This
doesn't handle any type of authorization (I don't know how SMTP authorization
is done in Python. None of the examples for SMTP that I've seen deal with
authorization.)'

Then it prompts you for a From and To address. These are not the headers that
go in the message body. They are part of the message envelope.

When you enter the message body, you will have to type all the e-mail headers
yourself.

Enter a period at the beginning of the line to get the program to finish
execution (stop looping) and send off the e-mail.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From grante at visi.com  Fri Feb  9 22:18:53 2001
From: grante at visi.com (Grant Edwards)
Date: Fri, 09 Feb 2001 21:18:53 GMT
Subject: how to send email in python?
References: <3A7FA5BA.82FB7593@esec.com.au> <h7av7t8ssi6v9djasb3mo3n7d7vigp16a1@4ax.com>
Message-ID: <19Zg6.1876$a7.67822@ruti.visi.com>

In article <h7av7t8ssi6v9djasb3mo3n7d7vigp16a1 at 4ax.com>, Sheila King wrote:
>On Tue, 06 Feb 2001 18:20:26 +1100, Sam Wun <swun at esec.com.au> wrote in
>comp.lang.python in article <3A7FA5BA.82FB7593 at esec.com.au>:
>
>:does anyone knows? I am using python 1.6.
>
>
>Here is one way:
>
>-------------------------------------------
>import smtplib

[...]

If you want to get fancy, you can use the MimeWriter module to
prepare the message.  That way you can do attachments and all
that sort of stuff.

-- 
Grant Edwards                   grante             Yow!  Hello. Just walk
                                  at               along and try NOT to think
                               visi.com            about your INTESTINES being
                                                   almost FORTY YARDS LONG!!


From f8dy at my-deja.com  Tue Feb  6 14:02:37 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Tue, 06 Feb 2001 13:02:37 GMT
Subject: how to send email in python?
References: <3A7FA5BA.82FB7593@esec.com.au>
Message-ID: <95osl9$vds$1@nnrp1.deja.com>

In article <3A7FA5BA.82FB7593 at esec.com.au>,
  Sam Wun <swun at esec.com.au> wrote:
> does anyone knows? I am using python 1.6.

You don't specify which platform you're on, so I won't assume you have
sendmail or anything easy like that, so we'll do it the hard (and
cross-platform) way.

from smtplib import SMTP def sendmail(smtpserver, fromaddr, toaddr, subject,
msg):  smtpsock = SMTP(smtpserver)  msg = "From: Your name
<%(fromaddr)s\nSubject: %(subject)s\n\n%(msg)s" % vars()  rc =
smtpsock.sendmail(fromaddr, [toaddr], msg)  smtpsock.quit()  return rc

## If your mail server requires POP authentication (like Yahoo! Mail does),
you can't call the above function directly; call this one instead: from
poplib import POP3 def sendmailwithPOPauth(popserver, popuser, poppassword,
smtpserver, fromaddr, toaddr, subject, msg):  popsock = POP3(popserver) 
popsock.user(popuser)  popsock.pass_(poppassword) # note weird function name;
"pass" is reserved word in Python  popsock.quit()  return
sendmail(smtpserver, fromaddr, toaddr, subject, msg)

Hope this helps.

-M
--
You're smart; why haven't you learned Python yet?  http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From f8dy at my-deja.com  Tue Feb  6 15:10:13 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Tue, 06 Feb 2001 14:10:13 GMT
Subject: how to send email in python?
References: <3A7FA5BA.82FB7593@esec.com.au> <95osl9$vds$1@nnrp1.deja.com>
Message-ID: <95p0jv$2so$1@nnrp1.deja.com>

In article <95osl9$vds$1 at nnrp1.deja.com>,
  Mark Pilgrim <f8dy at my-deja.com> wrote:
> In article <3A7FA5BA.82FB7593 at esec.com.au>,
>   Sam Wun <swun at esec.com.au> wrote:
> > does anyone knows? I am using python 1.6.
>
> You don't specify which platform you're on, so I won't assume you have
> sendmail or anything easy like that, so we'll do it the hard (and
> cross-platform) way.
>

OK, let's try this again, this time with !@#$% hard line breaks:

from smtplib import SMTP
def sendmail(smtpserver, fromaddr, toaddr, subject, msg):
  smtpsock = SMTP(smtpserver)
  msg = "From: Your name <%(fromaddr)s\nSubject: %(subject)s\n\n%(msg)
s" % vars()
  rc = smtpsock.sendmail(fromaddr, [toaddr], msg)
  smtpsock.quit()
  return rc

## If your mail server requires POP authentication
## (like Yahoo! Mail does), you can't call the
## above function directly; call this one instead:
from poplib import POP3
def sendmailwithPOPauth( \
    popserver, popuser, poppassword, smtpserver, \
    fromaddr, toaddr, subject, msg):
  popsock = POP3(popserver)
  popsock.user(popuser)
  popsock.pass_(poppassword) # note weird function name
  popsock.quit()
  return sendmail(smtpserver, fromaddr, toaddr, subject, msg)

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From mj at muc.das-werk.de  Tue Feb  6 08:37:04 2001
From: mj at muc.das-werk.de (Michael Jonas)
Date: Tue, 6 Feb 2001 08:37:04 +0100
Subject: Parsing C Code
Message-ID: <NDBBICIKFJMPFCCMBLNIIEKNCDAA.mj@muc.das-werk.de>

Hi,
I need to parse code which is formatted like C.
Is there a bison-like module? or just stick to re and do it myself?!


Michael

-----------------------------------------
Michael Jonas
Head of R&D
Das Werk Digitale Bildverarbeitungs GmbH
mail:  mj at muc.das-werk.de
phone: +49(89)368148-700
fax:   +49(89)368148-111
-----------------------------------------



From hat at se-46.wpa.wtb.tue.nl  Tue Feb  6 12:36:00 2001
From: hat at se-46.wpa.wtb.tue.nl (hat at se-46.wpa.wtb.tue.nl)
Date: Tue, 6 Feb 2001 12:36:00 +0100
Subject: Parsing C Code
In-Reply-To: <mailman.981445298.6855.python-list@python.org>
References: <mailman.981445298.6855.python-list@python.org>
Message-ID: <200102061136.MAA12238@se-46.wpa.wtb.tue.nl>

In comp.lang.python, you wrote:
>Hi,
>I need to parse code which is formatted like C.
>Is there a bison-like module? or just stick to re and do it myself?!

spark should work (http://www.csr.uvic.ca/~aycock/python/)

Also, there is another python package called

kwP<something> (I thought it is kwParse)



Albert
-- 
Signature is being repaired, sorry for the inconvenience



From slinkp23 at yahoo.com  Tue Feb  6 08:54:51 2001
From: slinkp23 at yahoo.com (Paul Winkler)
Date: Tue, 06 Feb 2001 07:54:51 GMT
Subject: MIDI extensions (WIN32, Mac)?
References: <3A3FACEF.DAF77229@cableinet.co.uk> <3A3FE5C3.76B4@loop.com> <3A3FF97B.BBF116C3@cableinet.co.uk>
Message-ID: <3A7FADE4.16B3AFF5@yahoo.com>

Richard Dobson wrote:
(snip)
> I want to be
> able to at least write MIDI messages to the MIDI OUT port, and hopefully
> to be able to read them from MIDI IN too. Then I could write a simple
> MIDI filter in Python which would, for example, receive live note data
> from MIDI IN, transpose it by some interval, and ouptut the result.

Hi Richard, fancy meeting you here.
I'm afraid I have no idea how to get access to midi devices in Windows
or Mac.

You may want to look at this guy's project:
http://quitte.de/midithing.html

In Linux it's easy to read or write raw midi data:

input = open('/dev/midi00', 'r')
output = open('/dev/midi01', 'w')

while 1:
	some_notes = MidiBytes(input.read())
	for foo in some_notes:
		foo.transpose(5.0)
		foo.channel = foo.channel + 1	
	output.write(some_notes)
	output.flush()

Implementing the MidiByte class and guessing what the methods would do
is left as an exercise to the reader. Don't ask me, I just made them
up.

Also I should say that I have not tested output because I don't have
any external synths. I have read midi input from a keyboard and
printed information about it. Didn't get much farther than that - MIDI
isn't  a big interest of mine.


Good luck,

PW

.................    paul winkler    ..................
slinkP arts:   music, sound, illustration, design, etc.
           web page:  http://www.slinkp.com
      A member of ARMS:   http://www.reacharms.com


From wware at world.std.com  Tue Feb  6 14:44:55 2001
From: wware at world.std.com (Will Ware)
Date: Tue, 6 Feb 2001 13:44:55 GMT
Subject: MIDI extensions (WIN32, Mac)?
References: <3A3FACEF.DAF77229@cableinet.co.uk> <3A3FE5C3.76B4@loop.com> <3A3FF97B.BBF116C3@cableinet.co.uk> <3A7FADE4.16B3AFF5@yahoo.com>
Message-ID: <G8C8uv.L3B@world.std.com>

Richard Dobson wrote:
(snip)
> I want to be
> able to at least write MIDI messages to the MIDI OUT port, and hopefully
> to be able to read them from MIDI IN too. Then I could write a simple
> MIDI filter in Python which would, for example, receive live note data
> from MIDI IN, transpose it by some interval, and ouptut the result.

I haven't done anything with reading and writing MIDI ports, but I've
tinkered a bit with a MIDI parser, which might be useful:
http://www.deja.com/getdoc.xp?AN=711228575&fmt=text

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From graham at nospam.microtonal.co.uk  Thu Feb 15 00:25:00 2001
From: graham at nospam.microtonal.co.uk (Graham Breed)
Date: Wed, 14 Feb 2001 23:25 +0000 (GMT Standard Time)
Subject: MIDI extensions (WIN32, Mac)?
References: <3A7FADE4.16B3AFF5@yahoo.com>
Message-ID: <memo.20010214232533.53913C@gbreed.compulink.co.uk>

Below is some code that actually works.  I don't think it's fully 
compliant with the MIDI standard, but you can check that.  There's no 
obvious problem with speed.

Note: do not use input.read() without an argument.  It will block 
indefinitely.

I have written code in C to use the MIDI ports in Windows.  I don't think 
it can be done directly in Python, but I would be interested in working on 
a C module.

I have no idea about Macintosh.  Hopefully, OSX will behave like Linux in 
this respect.


import string

#messages
noteOff = 0x80
noteOn = 0x90
polyAftertouch = 0xA0
controlChange = 0xB0
programChange = 0xC0
monoAftertouch = 0xD0
pitchBend = 0xE0
beginSysEx = 0xF0
songPositionPointer = 0xF2
songSelect = 0xF3
tuneRequest = 0xF6
endSysEx = 0xF7
timingClock = 0xF8
start = 0xFA
cont = 0xFB
stop = 0xFC
activeSensing = 0xFE
systemReset = 0xFF

messagesWithNotes = (noteOn, noteOff, polyAftertouch)

messageNames = {
    noteOff: 'note off',
    noteOn: 'note on',
    polyAftertouch: 'poly aftertouch',
    controlChange: 'control change',
    programChange: 'program change',
    monoAftertouch: 'mono aftertouch',
    pitchBend: 'pitch bend',
    beginSysEx: 'begin SysEx',
    endSysEx: 'end SysEx',
    songPositionPointer: 'song position pointer',
    songSelect: 'song select',
    tuneRequest: 'tune request',
    endSysEx: 'end SysEx',
    timingClock: 'timing clock',
    start: 'start',
    cont: 'continue',
    stop: 'stop',
    activeSensing: 'active sensing',
    systemReset: 'system reset'}

nDataBytes = {
  noteOff: 2,
  noteOn: 2,
  polyAftertouch: 2,
  controlChange: 2,
  programChange: 1,
  monoAftertouch: 1,
  pitchBend: 2}

noteNames = ('C','C#','D','Eb','E','F','F#',
    'G','GA','A','Bb','B')

def isData(byte):
  return byte<0x80;

def isChannelMessage(byte):
  return 0x80<=byte<0xF0;

def formatNote(noteNum):
  return noteNames[noteNum%12] + str(noteNum/12 - 1)

def sendMessage(output, message):
  output.write(string.join(map(chr, message),''))
  try:
    output.flush()
  except:
    print "couldn't flush"
  


midiIn = open('/dev/midi00', 'r')
midiOut = open('/dev/midi1','w')

messageNumber = ''
data = []
dataLength = 0
channel=-1
while 1:
  inputData = midiIn.read(1)
  if len(inputData)>1 :
    print 'read to much '+map(hex, map(ord, inputData))
  else:
    datum = ord(inputData)
    if isData(datum):
      data.append(datum)
      if len(data)==dataLength:
        print messageNames[messageNumber], 'channel', channel, 
        if messageNumber in messagesWithNotes:
          print formatNote(data[0]), data[1]
        else:
          print string.join(map(hex, data))
        sendMessage(midiOut, [messageNumber+channel]+data)
        data = []
    elif isChannelMessage(datum):
      channel = datum%16
      messageNumber = datum-channel
      dataLength = nDataBytes[messageNumber]
      if data:
        print 'residual data '+string.join(map(hex, data))
        data = []
    elif datum==beginSysEx:
      messageNumber=beginSysEx
      dataLength = -1
      data = [] # should this be different to the normal data?
    elif datum==endSysEx:
      if messageNumber != beginSysEx:
        print 'end with no beginning'
      formattedData = []
      for byte in data:
        formattedData.append(hex(byte)[-2:])
      print 'SysEx', string.replace(string.join(
          formattedData,''),'x','0')
      sendMessage(midiOut, [beginSysEx]+data+[endSysEx])
      data = []
    else:
      print messageNames.get(datum, 'undefined')
      sendMessage(midiOut, datum)
      


From lawick at wanadoo.nl  Tue Feb  6 10:54:57 2001
From: lawick at wanadoo.nl (lawick at wanadoo.nl)
Date: 06 Feb 2001 09:54:57 GMT
Subject: SOAP
Message-ID: <3a7fc9f1$0$11094@reader4>

Is there a SOAP module available for Python?

Cheers,

Joost

-- 

Joost van Lawick
E: lawick at wanadoo.nl
W: http://home.wanadoo.nl/lawick


From phd at phd.pp.ru  Tue Feb  6 11:20:45 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 6 Feb 2001 13:20:45 +0300 (MSK)
Subject: SOAP
In-Reply-To: <3a7fc9f1$0$11094@reader4>
Message-ID: <Pine.LNX.4.30.0102061320210.3861-100000@fep132.fep.ru>

On 6 Feb 2001 lawick at wanadoo.nl wrote:
> Is there a SOAP module available for Python?

   Yes, soaplib.py. Search the Net. (Hint: PythonWare)

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From jhauser at ifm.uni-kiel.de  Tue Feb  6 11:23:20 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 06 Feb 2001 11:23:20 +0100
Subject: SOAP
References: <3a7fc9f1$0$11094@reader4>
Message-ID: <87snlshzzb.fsf@lisboa.ifm.uni-kiel.de>

For SOAP and xml-rpc look at

http://www.pythonware.com/products/soap/index.htm

HTH,
__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From glenfant at equod.com.nospam  Tue Feb  6 11:29:11 2001
From: glenfant at equod.com.nospam (Gilles Lenfant)
Date: Tue, 6 Feb 2001 11:29:11 +0100
Subject: SOAP
References: <3a7fc9f1$0$11094@reader4>
Message-ID: <95ojau$1ip$1@reader1.imaginet.fr>

Yes,

A SOAP client module here:

http://www.pythonware.com/downloads/index.htm#soap


<lawick at wanadoo.nl> a ?crit dans le message news:
3a7fc9f1$0$11094 at reader4...
> Is there a SOAP module available for Python?
>
> Cheers,
>
> Joost
>
> --
>
> Joost van Lawick
> E: lawick at wanadoo.nl
> W: http://home.wanadoo.nl/lawick



From immoos at hotkey.net.au  Tue Feb  6 11:10:56 2001
From: immoos at hotkey.net.au (Philip Immoos)
Date: Tue, 06 Feb 2001 20:10:56 +1000
Subject: Fetchmailconf no libtix4.1.8.0.so
Message-ID: <3A7FCDAF.1EB1BF1A@hotkey.net.au>

I am Trying to set up fetchmail. I have installed fetchmail
fetchmailconf python tk.
When I run fetchmailconf I get an error.

    Traceback (innermost last):
        File "/usr/bin/fetchmailconf", line 9 in ?
            from Tkinter import *
    ImportError: No module named Tkinter

I have Installed Tkinter-1.5-7.i386.rpm and Tkinter-1.5.2-13.i368.rpm.
When I install them I get

libtix4.1.8.0.so is needed by Tkinter-1.5.2-?????????

As I have installed Tkinter Where do I go ????

Regards Philip Immoos



From cg at schlund.de  Tue Feb  6 11:11:53 2001
From: cg at schlund.de (Carsten Gaebler)
Date: Tue, 06 Feb 2001 11:11:53 +0100
Subject: flock() question
Message-ID: <3A7FCDE9.2AD808CD@schlund.de>

Hi there!

It seems I don't understand file locking. In the following example the
parent process locks a file and then the child process locks the same file
but does not wait until the parent unlocks it. What am I doing wrong?

import fcntl, FCNTL, os, time

f = open("/tmp/locktest", "w")

pid = os.fork()
if pid == 0:
        time.sleep(1)
        print "Child: locking file"
        fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
        print "Child: file locked"
        time.sleep(5)
        os._exit(0)

else:
        print "Parent: locking file"
        fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
        print "Parent: file locked"
        time.sleep(10)
        fcntl.flock(f.fileno(), FCNTL.LOCK_UN)



Regards
Carsten.


From stephen_purcell at yahoo.com  Tue Feb  6 11:44:36 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 6 Feb 2001 11:44:36 +0100
Subject: flock() question
In-Reply-To: <3A7FCDE9.2AD808CD@schlund.de>; from cg@schlund.de on Tue, Feb 06, 2001 at 11:11:53AM +0100
References: <3A7FCDE9.2AD808CD@schlund.de>
Message-ID: <20010206114436.A21002@freedom.puma-ag.com>

Carsten Gaebler wrote:
> It seems I don't understand file locking. In the following example the
> parent process locks a file and then the child process locks the same file
> but does not wait until the parent unlocks it. What am I doing wrong?

>From the 'flock' man page:

  "A file is locked (i.e., the inode), not the file  descriptor."

So the correct procedure is to open the lock file separately in each process:

pid = os.fork()
if pid == 0:
        f = open("/tmp/locktest", "w")
        time.sleep(1)
        print "Child: locking file"
        fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
        print "Child: file locked"
        time.sleep(5)
        os._exit(0)

else:
        f = open("/tmp/locktest", "w")
        print "Parent: locking file"
        fcntl.flock(f.fileno(), FCNTL.LOCK_EX)
        print "Parent: file locked"
        time.sleep(10)
        fcntl.flock(f.fileno(), FCNTL.LOCK_UN)


-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From Milan.Gardian at leibinger.com  Tue Feb  6 11:16:37 2001
From: Milan.Gardian at leibinger.com (Milan.Gardian at leibinger.com)
Date: Tue, 06 Feb 2001 10:16:37 GMT
Subject: Q: Python regular expression \Z delimiter
Message-ID: <95oiu2$o7v$1@nnrp1.deja.com>

Hello,

I have the following problem with Python's re module:

When using multiline mode (re.MULTILINE alias re.M) together with
single-line mode (re.DOTALL alias re.S), the regular expression should
evaluate metacharacter \Z as the end of string regardless of any
embedded newlines in the processed string (unlike $).

Unfortunately it does not work for me this way... Please consider those
examples:
---
#Perl: r1.pl
$txt = "Hello\nWorld\n";
$txt =~ /(.*?\Z)/ms;
print "$1\n";
---
#Python: r1.py
import re;
txt = "Hello\nWorld\n";
reg = re.compile(r'(.*?\Z)', re.M | re.S);
res = reg.search(txt);
print res.group(1);
---
They should both produce the same result because they use the same
regular expression with the same modifiers (ms). Perl behaves as
expected (matches the string until the end):
    C:\Temp>r1.pl
    Hello
    World

    C:\Temp>

On the other hand, Python behaves differently (matches the string only
until the first line-delimiter):
    C:\Temp>r1.py
    Hello

    C:\Temp>

The \Z metacharacter evidently does not match at the end of string (as
it should), but at the end of line (i.e. behaves exactly like '$'
should and does in multiline mode). When using '$' delimiter instead,
both scripts behave alike:
---
#Perl: r2.pl
$txt = "Hello\nWorld\n";
$txt =~ /(.*?$)/ms;
print "$1\n";
---
#Python: r2.py
import re;
txt = "Hello\nWorld\n";
reg = re.compile(r'(.*?$)', re.M | re.S);
res = reg.search(txt);
print res.group(1);
---
    C:\Temp>r2.pl
    Hello

    C:\Temp>r2.py
    Hello

Could anybody explain this behavior to me please? Perhaps I do
something wrong, but currently it seems to me this behavior is a bug in
Python's implementation of re.

Regards,
	Milan G.


Sent via Deja.com
http://www.deja.com/


From Milan.Gardian at leibinger.com  Tue Feb  6 11:20:58 2001
From: Milan.Gardian at leibinger.com (Milan.Gardian at leibinger.com)
Date: Tue, 06 Feb 2001 10:20:58 GMT
Subject: Q: Python regular expression \Z delimiter
Message-ID: <95oj66$oi0$1@nnrp1.deja.com>

Hello,

I have the following problem with Python's re module:

When using multiline mode (re.MULTILINE alias re.M) together with
single-line mode (re.DOTALL alias re.S), the regular expression should
evaluate metacharacter \Z as the end of string regardless of any
embedded newlines in the processed string (unlike $).

Unfortunately it does not work for me this way... Please consider those
examples:
---
#Perl: r1.pl
$txt = "Hello\nWorld\n";
$txt =~ /(.*?\Z)/ms;
print "$1\n";
---
#Python: r1.py
import re;
txt = "Hello\nWorld\n";
reg = re.compile(r'(.*?\Z)', re.M | re.S);
res = reg.search(txt);
print res.group(1);
---
They should both produce the same result because they use the same
regular expression with the same modifiers (ms). Perl behaves as
expected (matches the string until the end):
    C:\Temp>r1.pl
    Hello
    World

    C:\Temp>

On the other hand, Python behaves differently (matches the string only
until the first line-delimiter):
    C:\Temp>r1.py
    Hello

    C:\Temp>

The \Z metacharacter evidently does not match at the end of string (as
it should), but at the end of line (i.e. behaves exactly like '$'
should and does in multiline mode). When using '$' delimiter instead,
both scripts behave alike:
---
#Perl: r2.pl
$txt = "Hello\nWorld\n";
$txt =~ /(.*?$)/ms;
print "$1\n";
---
#Python: r2.py
import re;
txt = "Hello\nWorld\n";
reg = re.compile(r'(.*?$)', re.M | re.S);
res = reg.search(txt);
print res.group(1);
---
    C:\Temp>r2.pl
    Hello

    C:\Temp>r2.py
    Hello

Could anybody explain this behavior to me please? Perhaps I do
something wrong, but currently it seems to me this behavior is a bug in
Python's implementation of re.

Regards,
	Milan G.


Sent via Deja.com
http://www.deja.com/


From fredrik at pythonware.com  Tue Feb  6 13:54:45 2001
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Tue, 06 Feb 2001 12:54:45 GMT
Subject: Q: Python regular expression \Z delimiter
References: <95oj66$oi0$1@nnrp1.deja.com>
Message-ID: <95os6k$v0b$1@nnrp1.deja.com>

Milan Gardian wrote:
> Could anybody explain this behavior to me please? Perhaps I do
> something wrong, but currently it seems to me this behavior is
> a bug in Python's implementation of re.

It's a bug.  It has been fixed in 2.1.

Cheers /F


Sent via Deja.com
http://www.deja.com/


From sgriggs at pobox.com  Tue Feb  6 13:16:09 2001
From: sgriggs at pobox.com (sgriggs at pobox.com)
Date: Tue, 06 Feb 2001 12:16:09 GMT
Subject: Pyt;hon 2.1a2 compile problems
Message-ID: <95opu4$t8k$1@nnrp1.deja.com>

Is anyone else having problems getting Python 2.1a2 to compile on
Solaris? I'm using Solaris 2.8 with GCC 2.95.2 and just can't seem to
get it to work. I can compile Python 2.0 on this same box. I'm just
wondering if it's just me at this point. Thanks.

Scott Griggs


Sent via Deja.com
http://www.deja.com/


From sdm7g at virginia.edu  Tue Feb  6 15:36:00 2001
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 6 Feb 2001 09:36:00 -0500 (EST)
Subject: Pyt;hon 2.1a2 compile problems
In-Reply-To: <95opu4$t8k$1@nnrp1.deja.com>
Message-ID: <Pine.A32.3.90.1010206092830.5798B@elvis.med.Virginia.EDU>

On Tue, 6 Feb 2001 sgriggs at pobox.com wrote:

> Is anyone else having problems getting Python 2.1a2 to compile on
> Solaris? I'm using Solaris 2.8 with GCC 2.95.2 and just can't seem to
> get it to work. I can compile Python 2.0 on this same box. I'm just
> wondering if it's just me at this point. Thanks.

There have been a lot of changes to the Makefile, Setup and the
rest of the build process -- and besides, this is an alpha release.

Odds are it's not just you. 

How about some details about how exactly it doesn't compile ?

What is it doing differently at the point of failure than what
it did building 2.0 ? 

You could also check the bug tracker at:
	http://sourceforge.net/bugs/?group_id=5470
to see if it's already been reported (and if not, post it there).


---|  Steven D. Majewski   (804-982-0831)  <sdm7g at Virginia.EDU>  |---
---|  Department of Molecular Physiology and Biological Physics  |---
---|  University of Virginia             Health Sciences Center  |---
---|  P.O. Box 10011            Charlottesville, VA  22906-0011  |---
		"All operating systems want to be unix, 
		 All programming languages want to be lisp." 




From sgriggs at pobox.com  Tue Feb  6 19:30:53 2001
From: sgriggs at pobox.com (sgriggs at pobox.com)
Date: Tue, 06 Feb 2001 18:30:53 GMT
Subject: Pyt;hon 2.1a2 compile problems
References: <mailman.981470260.11991.python-list@python.org>
Message-ID: <95pfsu$ht1$1@nnrp1.deja.com>

Here's a small amount of the output. I've also tried to compile on 2.6
and 2.7 with no luck.

building 'sunaudiodev' extension
skipping /var/tmp/Python-2.1a2/Modules/sunaudiodev.c
(build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o up-to-date)
gcc -shared build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
-L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.1/sunaudiodev.so
Text relocation remains                         referenced
    against symbol                  offset      in file
<unknown>                           0x174
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x17c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x398
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x39c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x70c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x714
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x718
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x71c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x734
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x73c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x760
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x768
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x78c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x798
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x7e0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x7e4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x98
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x9c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x154
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x158
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x238
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x250
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x2ec
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x2f0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x354
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x35c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x3f8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x3fc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x468
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x46c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x4ec
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x4f0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x55c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x560
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x5d8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x5dc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x648
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x64c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x820
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
SunAudioError                       0x82c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0xc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x20
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x2c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x38
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x4c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x58
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x6c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x78
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x8c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0xa4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0xac
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0xb8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0xc0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0xd0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0xd4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x108
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x10c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x1d8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x1e4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x288
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x290
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x308
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x318
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x3b4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x3bc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x414
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x420
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x484
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x490
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x508
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x514
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x578
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x584
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x5b8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x5c0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x5f4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x600
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x664
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x670
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x6c4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x6d0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x7b0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x7b8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x7dc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x7f8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x808
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x810
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x830
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
<unknown>                           0x83c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyMember_Get                        0x76c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
open                                0x128
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
open                                0x140
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyString_AsString                   0x210
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrnoWithFilename      0x15c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
strlen                              0xd8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyDict_SetItemString                0x838
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
write                               0x2b8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
Py_FindMethod                       0x720
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_NoMemory                      0xf4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
.udiv                               0x4d0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyString_FromStringAndSize          0x1fc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
close                               0x1a4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
close                               0x1bc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
close                               0x694
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
Py_BuildValue                       0x5c8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x10
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x1e8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x298
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x314
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x41c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x48c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x510
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x580
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x5fc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x66c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyArg_Parse                         0x6cc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyInt_FromLong                      0x458
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyInt_FromLong                      0x4dc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyInt_FromLong                      0x6e0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_PyObject_Del                       0x1c4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x2c8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x2d8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x2dc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x2e8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x3e0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x3e4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x3e8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x3f4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x544
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x548
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x54c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x558
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x630
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x634
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x638
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x644
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x690
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x6a4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x6a8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x6ac
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_Py_NoneStruct                      0x6b4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
malloc                              0xe0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
strcat                              0x110
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
strcmp                              0x28
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
strcmp                              0x48
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
strcmp                              0x68
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
strcmp                              0x88
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetString                     0xa8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetString                     0x3b8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetString                     0x7b4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
strcpy                              0x100
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
getenv                              0xbc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x24c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x2f4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x358
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x400
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x470
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x4f4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x564
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x5e0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_SetFromErrno                  0x650
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
free                                0x16c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
free                                0x74c
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyModule_GetDict                    0x7fc
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyErr_NewException                  0x814
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_PyObject_New                       0x178
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
_PyObject_New                       0x738
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyExc_TypeError                     0x3a8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyExc_TypeError                     0x3ac
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyExc_TypeError                     0x7a4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyExc_TypeError                     0x7a8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
.umul                               0x4c4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
read                                0x228
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ioctl                               0x344
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ioctl                               0x3d0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ioctl                               0x440
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ioctl                               0x4b0
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ioctl                               0x534
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ioctl                               0x5a8
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ioctl                               0x620
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
PyMember_Set                        0x794
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
Py_InitModule4                      0x7f4
build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
ld: fatal: relocations remain against allocatable but non-writable
sections
collect2: ld returned 1 exit status
WARNING: building of extension "sunaudiodev" failed: command 'gcc'
failed with exit status 1


Sent via Deja.com
http://www.deja.com/


From akuchlin at mems-exchange.org  Tue Feb  6 21:56:02 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 06 Feb 2001 15:56:02 -0500
Subject: Python 2.1a2 compile problems
References: <mailman.981470260.11991.python-list@python.org> <95pfsu$ht1$1@nnrp1.deja.com>
Message-ID: <3dpugvh6ot.fsf_-_@ute.cnri.reston.va.us>

sgriggs at pobox.com writes:
> Here's a small amount of the output. I've also tried to compile on 2.6
> and 2.7 with no luck.
> 
> building 'sunaudiodev' extension
> skipping /var/tmp/Python-2.1a2/Modules/sunaudiodev.c
> (build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o up-to-date)
> gcc -shared build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
> -L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.1/sunaudiodev.so
> Text relocation remains                         referenced
>     against symbol                  offset      in file
> <unknown>                           0x174
> build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o
> <unknown>                           0x17c
> build/temp.solaris-2.8-sun4u-2.1/sunaudiodev.o

Confirmed.  This isn't specific to the sunaudiodev module; the struct
module doesn't compile for me for similar problems.  The errors are
triggered by the gcc -shared invocation, like this:

gcc -shared build/temp.solaris-2.6-sun4u-2.1/sunaudiodev.o -L/usr/local/lib -o build/lib.solaris-2.6-sun4u-2.1/sunaudiodev.so

Did some compiler argument change?

--amk


From sgriggs at pobox.com  Wed Feb  7 00:26:18 2001
From: sgriggs at pobox.com (sgriggs at pobox.com)
Date: Tue, 06 Feb 2001 23:26:18 GMT
Subject: Python 2.1a2 compile problems
References: <mailman.981470260.11991.python-list@python.org> <95pfsu$ht1$1@nnrp1.deja.com> <3dpugvh6ot.fsf_-_@ute.cnri.reston.va.us>
Message-ID: <95q16e$3go$1@nnrp1.deja.com>

Thank you for the pointer Andrew. I have changed the SunOS entry in
configure and now I can run configure and make and it works. Also,
after all that, changes in the makefile have broken make install. I
fixed that. Let me figure out how to make a patch and I'll submit it
python.sourceforge.net. Thanks again. I can post my changes here if
anyone cares.

Scott Griggs


Sent via Deja.com
http://www.deja.com/


From mlh at idi.ntnu.no  Thu Feb  8 14:30:22 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Thu, 8 Feb 2001 14:30:22 +0100
Subject: Python 2.1a2 compile problems
References: <mailman.981470260.11991.python-list@python.org> <95pfsu$ht1$1@nnrp1.deja.com> <3dpugvh6ot.fsf_-_@ute.cnri.reston.va.us> <95q16e$3go$1@nnrp1.deja.com>
Message-ID: <95u774$ktc$1@tyfon.itea.ntnu.no>

> Thanks again. I can post my changes here if
> anyone cares.
>

Please do!

I'm having the same problems, and was prepared to wait
for a beta... ;)

> Scott Griggs

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From akuchlin at mems-exchange.org  Thu Feb  8 16:36:16 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 08 Feb 2001 10:36:16 -0500
Subject: Python 2.1a2 compile problems
References: <mailman.981470260.11991.python-list@python.org> <95pfsu$ht1$1@nnrp1.deja.com> <3dpugvh6ot.fsf_-_@ute.cnri.reston.va.us> <95q16e$3go$1@nnrp1.deja.com> <95u774$ktc$1@tyfon.itea.ntnu.no>
Message-ID: <3dy9vh5gr3.fsf@ute.cnri.reston.va.us>

"Magnus Lie Hetland" <mlh at idi.ntnu.no> writes:
> I'm having the same problems, and was prepared to wait
> for a beta... ;)

Erm... you do know that problems don't get fixed unless they actually
get reported, right?

--amk


From sgriggs at pobox.com  Fri Feb  9 17:23:24 2001
From: sgriggs at pobox.com (sgriggs at pobox.com)
Date: Fri, 09 Feb 2001 16:23:24 GMT
Subject: Python 2.1a2 compile problems
References: <mailman.981470260.11991.python-list@python.org> <95pfsu$ht1$1@nnrp1.deja.com> <3dpugvh6ot.fsf_-_@ute.cnri.reston.va.us> <95q16e$3go$1@nnrp1.deja.com> <95u774$ktc$1@tyfon.itea.ntnu.no>
Message-ID: <9615hf$4u2$1@nnrp1.deja.com>

I put the patches on sourceforge in python's patch area.

Scott Griggs


Sent via Deja.com
http://www.deja.com/


From slhath at home.com.nospam  Tue Feb  6 14:37:23 2001
From: slhath at home.com.nospam (Scott Hathaway)
Date: Tue, 06 Feb 2001 13:37:23 GMT
Subject: smtp mail with attachment
Message-ID: <n6Tf6.5629$tv5.438444@news.flash.net>

How do I send an email with an attachment with python (I need a cross
platform solution for Windows and Unix)?

Thanks,
Scott




From sholden at holdenweb.com  Tue Feb  6 15:20:53 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 6 Feb 2001 09:20:53 -0500
Subject: smtp mail with attachment
References: <n6Tf6.5629$tv5.438444@news.flash.net>
Message-ID: <1PTf6.34250$o5.413519@e420r-atl1.usenetserver.com>

"Scott Hathaway" <slhath at home.com.nospam> wrote in message
news:n6Tf6.5629$tv5.438444 at news.flash.net...
> How do I send an email with an attachment with python (I need a cross
> platform solution for Windows and Unix)?
>
> Thanks,
> Scott
>
>
Scott:

There are several ways you could do this, depending on your setup.
Basically:

1.    Generate the email with attachment; then

2.    Send it out.

The first step can be done with several libraries. I myself frequently use
the mimecntl module, available from the Vaults of Parnassus, in which you
would do something like the following:

        f = mimecntl.MIME_document("", type='text/plain')
        f.write("""Dear Customer,

Thank you for your request for services. We have performed these
services, and attach our invoice accordingly.

Sincerely
Holden Web""")
        rtf = mimecntl.MIME_recoder()
        inf = open("Invoice.rtf", "rb")
        rtfstream = inf.read()
        rtf.write(rtfstream)
        rtf['content-type'] = mimecntl.MIMEField('content-type',
'application/rtf')
        rtf['content-length'] = mimecntl.MIMEField('content-length',
len(rtfstream))
        rtf['content-disposition'] =
mimecntl.MIMEField('content-disposition',
                        'attachment',
                        filename=attachname)
        rtf = rtf.encode('base64')
        #
        # Wrap the message, and its attachment, in a single MIME envelope
        #
        msg = mimecntl.MIME_document( (f, rtf),
                From="sholden at holdenweb.com",
                To="Lucky.Client at clientorganization.com",
                Subject = 'Services Invoice for You')

The second step would be to use smtplib to send the mail out via your local
mail system (presumably the same one as you configure your mail agent to use
for sending). This would go something like:

    smtp = smtplib.SMTP("your.mail.host")
    smtp.sendmail("sholden at holdenweb.com",
         "Lucky.Client at clientorganization.com", msg.dump())
    smtp.quit()

There are many ways to skin this particular cat. You probably also want to
take a look at the mimetools and rfc822 modules, which are a part of the
standard distribution, and can also cope with what you want to do.

Note that this code has been cut and farily extensively modified from a
working program which sends invoices out by e-mail (for a client), so I
cannot guarantee its integrity. I think it should give you some ideas,
though.

regards
 Steve




From f8dy at my-deja.com  Tue Feb  6 15:29:39 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Tue, 06 Feb 2001 14:29:39 GMT
Subject: smtp mail with attachment
References: <n6Tf6.5629$tv5.438444@news.flash.net>
Message-ID: <95p1o9$3rp$1@nnrp1.deja.com>

In article <n6Tf6.5629$tv5.438444 at news.flash.net>,
  "Scott Hathaway" <slhath at home.com.nospam> wrote:
> How do I send an email with an attachment with python (I need a cross
> platform solution for Windows and Unix)?

Found this on the Python Knowledge Base (a wonderful resource, if you
don't know about it):
http://www.faqts.com/knowledge_base/view.phtml/aid/3992/fid/380

The post references a script available via FTP, but the URL is no
longer valid; here is the correct URL:
ftp://ftp.visi.com/users/grante/python/mfetch.py

The script contains some Windows-specific stuff, but I believe the
class you want (SMTPWriter) is cross-platform.

Hope this helps.

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From grante at visi.com  Fri Feb  9 22:23:45 2001
From: grante at visi.com (Grant Edwards)
Date: Fri, 09 Feb 2001 21:23:45 GMT
Subject: smtp mail with attachment
References: <n6Tf6.5629$tv5.438444@news.flash.net> <95p1o9$3rp$1@nnrp1.deja.com>
Message-ID: <BdZg6.1877$a7.68055@ruti.visi.com>

In article <95p1o9$3rp$1 at nnrp1.deja.com>, Mark Pilgrim wrote:

>> How do I send an email with an attachment with python (I need a cross
>> platform solution for Windows and Unix)?
>
>Found this on the Python Knowledge Base (a wonderful resource, if you
>don't know about it):
>http://www.faqts.com/knowledge_base/view.phtml/aid/3992/fid/380
>
>The post references a script available via FTP, but the URL is no
>longer valid; here is the correct URL:
>ftp://ftp.visi.com/users/grante/python/mfetch.py
>
>The script contains some Windows-specific stuff, but I believe the
>class you want (SMTPWriter) is cross-platform.

The SMTPWriter class was developed on Linux, and should work on
any system which supports smtplib and MimeWriter.  It's
basically just a wrapper around those two modules.

The other stuff in that script uses Microsoft COM/MAPI and only
runs under Windows, but it's concerned with fetching messages
from Exchange Server.

-- 
Grant Edwards                   grante             Yow!  I have a TINY BOWL in
                                  at               my HEAD
                               visi.com            


From sandipan at vsnl.com  Tue Feb  6 15:35:10 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Tue, 6 Feb 2001 20:05:10 +0530
Subject: smtp mail with attachment
References: <n6Tf6.5629$tv5.438444@news.flash.net>
Message-ID: <00e001c0904a$02831aa0$020000c3@node02>

AFAIK, and I could be ignorant, but you just have to construct the mail with
MimeWriter:
With content-type as multi-part (because the email will have more than one
part (itself) - the attachment(s))
And then, in one of the parts, put in the attachment after declaring its
type.
Please note that if you are going to need to send binary attachments, base64
will come in use to encode it first. As in:
Content-Type: whatever/whatever
Content-Transfer-Encoding: base64
You can use smtplib to send the mime you have built then.

Also, you can have other parts of the email refer to this attachment by
giving it an identifier (URL sort of)

Works fine on both Unix and Windows.

HTH. Regards,

Sandipan
----- Original Message -----
From: "Scott Hathaway" <slhath at home.com.nospam>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Tuesday, February 06, 2001 7:07 PM
Subject: smtp mail with attachment


> How do I send an email with an attachment with python (I need a cross
> platform solution for Windows and Unix)?
>
> Thanks,
> Scott
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>




From maxm at normik.dk  Wed Feb  7 09:19:31 2001
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Wed, 7 Feb 2001 09:19:31 +0100 
Subject: smtp mail with attachment
Message-ID: <7BD10B680501D411B9DF009027E06F32197BDF@exchange>

From: Scott Hathaway [mailto:slhath at home.com.nospam]

>How do I send an email with an attachment with python (I need a cross
>platform solution for Windows and Unix)?

Here is a class I have partly written, partly pilfered from someone else.

It works nicely, and acts like an ordinary mail client.

regards Max M

-------------- next part --------------
A non-text attachment was scrubbed...
Name: mxmMail.py
Type: application/octet-stream
Size: 6599 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010207/2b4f478d/attachment.obj>

From slhath at home.com.nospam  Wed Feb  7 17:26:59 2001
From: slhath at home.com.nospam (Scott Hathaway)
Date: Wed, 07 Feb 2001 16:26:59 GMT
Subject: smtp mail with attachment
References: <mailman.981534159.15499.python-list@python.org>
Message-ID: <nHeg6.131$y03.12604@news.flash.net>

Thanks!

Scott

"Max M?ller Rasmussen" <maxm at normik.dk> wrote in message
news:mailman.981534159.15499.python-list at python.org...
> From: Scott Hathaway [mailto:slhath at home.com.nospam]
>
> >How do I send an email with an attachment with python (I need a cross
> >platform solution for Windows and Unix)?
>
> Here is a class I have partly written, partly pilfered from someone else.
>
> It works nicely, and acts like an ordinary mail client.
>
> regards Max M
>
>




From barry at digicool.com  Fri Feb  9 16:32:22 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Fri, 9 Feb 2001 10:32:22 -0500
Subject: smtp mail with attachment
References: <n6Tf6.5629$tv5.438444@news.flash.net>
Message-ID: <14980.3462.924312.970597@anthem.wooz.org>

>>>>> "SH" == Scott Hathaway <slhath at home.com.nospam> writes:

    SH> How do I send an email with an attachment with python (I need
    SH> a cross platform solution for Windows and Unix)?

Other people have made good suggestions in separate followups.  I just
add that I'm working on a new library to deal with MIME documents
(well, any rfc822 style message, but MIME is the really interesting
bit).  I will be using this library in the next release of Mailman,
and hope eventually to get it into Python's standard library.  For
now, you can find out more at:

    http://barry.wooz.org/software/pyware.html

Look for `mimelib'.  I will likely have a 0.2 release sometime soon.

-Barry



From pearu at cens.ioc.ee  Tue Feb  6 14:42:55 2001
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Tue, 6 Feb 2001 15:42:55 +0200
Subject: list.index(..) -> TypeError bug or feature?
Message-ID: <Pine.LNX.4.21.0102061534380.29543-100000@kev.ioc.ee>

In Python 2.1a2 I get TypeError exception from list index() method even if
the list contains given object:

>>> from gmpy import mpz
>>> a = [mpz(1),[]]
>>> a.index([])
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: coercion to gmpy.mpz type failed

while in Python 2.0b2 it works:

>>> a = [mpz(1),[]]
>>> a.index([])
1

Is this Python 2.1a2 bug or gmpy bug? Or my bug and Python 2.1 feature?

Thanks,
	Pearu



From fredrik at effbot.org  Tue Feb  6 18:53:47 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Tue, 06 Feb 2001 17:53:47 GMT
Subject: list.index(..) -> TypeError bug or feature?
References: <Pine.LNX.4.21.0102061534380.29543-100000@kev.ioc.ee>
Message-ID: <LSWf6.13091$AH6.1954273@newsc.telia.net>

Pearu Peterson wrote:
> In Python 2.1a2 I get TypeError exception from list index() method even if
> the list contains given object:
>
> >>> from gmpy import mpz
> >>> a = [mpz(1),[]]
> >>> a.index([])
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: coercion to gmpy.mpz type failed
>
> while in Python 2.0b2 it works:
>
> >>> a = [mpz(1),[]]
> >>> a.index([])
> 1

looks like mpz's coercion code doesn't work well under
2.1's new coercion model:

    http://python.sourceforge.net/peps/pep-0208.html

possibly when combined with the new comparision model:

    http://python.sourceforge.net/peps/pep-0207.html

I'll leave it to the PEP authors to figure out what's wrong
here (2.1, gmpy, or "sorry, you cannot do that")

Cheers /F




From khaghdad23 at yahoo.com  Tue Feb  6 15:34:22 2001
From: khaghdad23 at yahoo.com (Kian Haghdad)
Date: Tue, 6 Feb 2001 09:34:22 -0500
Subject: Software engineer
Message-ID: <20010206143420.OIGB27416.tomts5-srv.bellnexxia.net@kian>

Hi:

I have got your email from the Web site and I am very interested in working for your company. I 
have a BS in Electrical Engineering and computer science and worked on my Master degree 
in Telecommunication engineering (not finished due to the lucrative market!). I have also 7++ 
years of experience with software development, Internet development and hardware. I have 
worked for many companies both in Iran and in the United States using different development 
environments.
I am looking for a position which utilizes my experience. I am open to both Permanent position 
or contract to permanent positions.

I am Canadian resident and have permission to work in Canada.
I have attached my resume and sincerely appreciate your attention, thank you.

Sincerely,

Kian Haghdad

Here is my resume:
Kian Haghdad
8 Kingsbridge Crt., Apt. 508 
Toronto, Ontario M2R 1L5
Home: (416) 630-7066
Fax:  (416) 630-7991
Email: khaghdad23 at yahoo.com

I have more than 7+ years of experience in application development, Database Management, 
Internet development. I have also been involved in hardware design. My software 
development experience is mainly in Visual Basic, Visual C++, Visual J++, Visual InterDev, 
COM, DCOM, MFC, SDKs, MS-Access, MS-SQL, FoxPro,
Java script and VB script. Over the years I have been responsible for the 
Development of many applications from design to commercial release.

EDUCATION:

Sharif University OF Technology, Tehran (The best technical school in Iran)	9/87-6/92
B.S. Degree: Electrical Engineering and computer science
Worked on M.S. degree in Telecommunication engineering

EXPERIENCE:

01Communique, Mississauga, Ontario	12/00-Present
Involved in the development of a web supported version of the communicating Interface
using Active Server Pages, JavaScript, VB Script and browser compatibility, Visual C++6.0,
Borland C++, MSMQ, TCP/IP. 
Application increases the Modem compatibility for the wide variety of connections.
It also provides additional capabilities for the desktop version on the web.

American SkySat, Walnut Creek, California	5/00-Present
* Involved in the development of a web application in Visual InterDev 6.0 using Active Server 
Pages (ASP), Microsoft E-Commerce, SQL 7.0, XML, Visual C++ 6.0 and Visual Basic 6.0, 
Visual J++ 6.0, COM (ATL), DCOM, JavaScript and VB Script. The web server was Microsoft 
Internet Information Server (IIS), Microsoft site server 3.0, with Microsoft E-Commerce edition 
3.0 and FrontPage extension running under the NT Server The web application provides NT 
related services online. The user can register and buy services and also allows the user to 
troubleshoot and configure a system at real time online. The web site also provides the 
capability of chatting online, customer support online, statistical analysis online and transaction 
online.

* Developed a database management application for a medical health care center, using 
Visual Basic 6.0, SQL 7.0, Crystal report and Access 2000. (Windows 98 and NT)

American Computech, Pleasanton, California	3/98-5/00
Developed different applications using Visual C++, Visual J++, and Access.

* Involved in the development of a CRM (Customer Relation Management)web application in 
Visual InterDev 6.0 using Active Server Pages (ASP), Microsoft E-Commerce, SQL 7.0, XML, 
Visual Basic 6.0, Visual J++ 6.0, COM, DCOM, JavaScript and VB Script. (Windows 98 and 
NT)

Developed an application in Visual Basic 5.0, Access 97 and Crystal reports (using ODBC) for 
financial analyses department of FHP (Concord, California). The application is capable of 
importing data from other systems, analyzing and processing the data and generating hundreds 
of reports base on the original data.

*Developed an MDI application in Visual C++5,0 and Access 97 for windows called ID maker. 
The application is designed to drive a digital camera, capture a picture and generate different 
ID cards using OLE (marketing for schools and clubs). The Cannon digital camera is controlled 
via DLL functions calls to the camera's driver. The camera is connected via a parallel port. I 
have also developed several OCXs for this project.

Developed an application in Visual Basic 5.0, Access 97 and Crystal reports (using ODBC). 
The application is a very large information management system for CBI (City Building Inc. in 
San Francisco). The application contains 28 relational tables large number of queries forms 
and reports.

Tavanir Co., Tehran, Iran	9/95-3/98
*Developed an application in Visual C++ 5.0 intended for network control managements. The 
application is able to get information such as voltage, active and reactive power from power 
plants and transmission stations and specifies transmission state using power flow method after 
a fault occurs on the transmission line. 

* Developed an application for voltage drop on the net. The application was developed in 
Visual Basic 5.0 using graphical OCXs.

* Developed hardware for the high voltage network.

*Developed an application in Visual C++ 4.0. The application checks the transmitting power on 
the lines and controls the system's steady state by alarm system.

*Developed an application in Visual C++ 4.0 set the protection relays in substations to protect 
the network's reliability.

Pishro Co., Tehran, Iran	11/94-6/95
Designed, developed and implemented electrical systems using computer aided software. The 
electrical systems were intended for both indoors (Buildings) and outdoors (Airports and 
Terminals).
The systems were designed and based on the standard and regulation of electrical system.

Pardis Tower Co, Tehran, Iran	1/93-11/94
Designed simulators for training purpose.
*Developed an application in  C++ for  cycloconvertor simulator. The  
cycloconvertor is used in the starters of synchronous motors.
*Developed application using C++  for control speed of  DC motors. The application
 gets input voltage to the motor, speed of motor and rotor's current. It can control
 the motor speed based on requested speed.
*Developed application using C++ for step motor's control speed. In order to change the angle 
of dish antenna, the application orders to step motor to change the rotor's situation.

SKILLS:
Software development
Visual C++, Visual Basic, Visual InterDev, Visual J++, MFC, SDKs, COM, DCOM, ATL ActiveXs 
, C++, Fortran, Intel  8085 Assembler Access.
Internet development
Visual InterDev 6.0, Active Server Pages (ASP), SQL 7.0, COM, DCOM, Java script. 
Database development
MS-SQL, Access, FoxPro
Operating System Used
NT, MS-Win, UNIX, VAX, MS-DOS

-------------- next part --------------
A non-text attachment was scrubbed...
Name: kian Resume U_C.doc
Type: application/msword
Size: 14325 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010206/b590f249/attachment.doc>

From Randy.L.Kemp at motorola.com  Tue Feb  6 15:42:18 2001
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Tue, 6 Feb 2001 08:42:18 -0600 
Subject: Software engineer
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84DE7@il93exp01.css.mot.com>

Haghdad:
   I think you are addressing the wrong list, since the Python mailing list is a non profit organization.  Of course, you may have been trying to influence the readers of these mailing lists.  If you were to indulge in a sharp display of Python code, you would catch their attention off the bat.  Seriously -- there are many good, free internet sites where you can post your resume, such as monster.com, headhunter, etc.  Good luck.
Randy

-----Original Message-----
From: Kian Haghdad [mailto:khaghdad23 at yahoo.com]
Sent: Tuesday, February 06, 2001 8:34 AM
To: python-list at python.org
Subject: Software engineer


Hi:

I have got your email from the Web site and I am very interested in working for your company. I 
have a BS in Electrical Engineering and computer science and worked on my Master degree 
in Telecommunication engineering (not finished due to the lucrative market!). I have also 7++ 
years of experience with software development, Internet development and hardware. I have 
worked for many companies both in Iran and in the United States using different development 
environments.
I am looking for a position which utilizes my experience. I am open to both Permanent position 
or contract to permanent positions.

I am Canadian resident and have permission to work in Canada.
I have attached my resume and sincerely appreciate your attention, thank you.

Sincerely,

Kian Haghdad

Here is my resume:
Kian Haghdad
8 Kingsbridge Crt., Apt. 508 
Toronto, Ontario M2R 1L5
Home: (416) 630-7066
Fax:  (416) 630-7991
Email: khaghdad23 at yahoo.com

I have more than 7+ years of experience in application development, Database Management, 
Internet development. I have also been involved in hardware design. My software 
development experience is mainly in Visual Basic, Visual C++, Visual J++, Visual InterDev, 
COM, DCOM, MFC, SDKs, MS-Access, MS-SQL, FoxPro,
Java script and VB script. Over the years I have been responsible for the 
Development of many applications from design to commercial release.

EDUCATION:

Sharif University OF Technology, Tehran (The best technical school in Iran)	9/87-6/92
B.S. Degree: Electrical Engineering and computer science
Worked on M.S. degree in Telecommunication engineering

EXPERIENCE:

01Communique, Mississauga, Ontario	12/00-Present
Involved in the development of a web supported version of the communicating Interface
using Active Server Pages, JavaScript, VB Script and browser compatibility, Visual C++6.0,
Borland C++, MSMQ, TCP/IP. 
Application increases the Modem compatibility for the wide variety of connections.
It also provides additional capabilities for the desktop version on the web.

American SkySat, Walnut Creek, California	5/00-Present
* Involved in the development of a web application in Visual InterDev 6.0 using Active Server 
Pages (ASP), Microsoft E-Commerce, SQL 7.0, XML, Visual C++ 6.0 and Visual Basic 6.0, 
Visual J++ 6.0, COM (ATL), DCOM, JavaScript and VB Script. The web server was Microsoft 
Internet Information Server (IIS), Microsoft site server 3.0, with Microsoft E-Commerce edition 
3.0 and FrontPage extension running under the NT Server The web application provides NT 
related services online. The user can register and buy services and also allows the user to 
troubleshoot and configure a system at real time online. The web site also provides the 
capability of chatting online, customer support online, statistical analysis online and transaction 
online.

* Developed a database management application for a medical health care center, using 
Visual Basic 6.0, SQL 7.0, Crystal report and Access 2000. (Windows 98 and NT)

American Computech, Pleasanton, California	3/98-5/00
Developed different applications using Visual C++, Visual J++, and Access.

* Involved in the development of a CRM (Customer Relation Management)web application in 
Visual InterDev 6.0 using Active Server Pages (ASP), Microsoft E-Commerce, SQL 7.0, XML, 
Visual Basic 6.0, Visual J++ 6.0, COM, DCOM, JavaScript and VB Script. (Windows 98 and 
NT)

Developed an application in Visual Basic 5.0, Access 97 and Crystal reports (using ODBC) for 
financial analyses department of FHP (Concord, California). The application is capable of 
importing data from other systems, analyzing and processing the data and generating hundreds 
of reports base on the original data.

*Developed an MDI application in Visual C++5,0 and Access 97 for windows called ID maker. 
The application is designed to drive a digital camera, capture a picture and generate different 
ID cards using OLE (marketing for schools and clubs). The Cannon digital camera is controlled 
via DLL functions calls to the camera's driver. The camera is connected via a parallel port. I 
have also developed several OCXs for this project.

Developed an application in Visual Basic 5.0, Access 97 and Crystal reports (using ODBC). 
The application is a very large information management system for CBI (City Building Inc. in 
San Francisco). The application contains 28 relational tables large number of queries forms 
and reports.

Tavanir Co., Tehran, Iran	9/95-3/98
*Developed an application in Visual C++ 5.0 intended for network control managements. The 
application is able to get information such as voltage, active and reactive power from power 
plants and transmission stations and specifies transmission state using power flow method after 
a fault occurs on the transmission line. 

* Developed an application for voltage drop on the net. The application was developed in 
Visual Basic 5.0 using graphical OCXs.

* Developed hardware for the high voltage network.

*Developed an application in Visual C++ 4.0. The application checks the transmitting power on 
the lines and controls the system's steady state by alarm system.

*Developed an application in Visual C++ 4.0 set the protection relays in substations to protect 
the network's reliability.

Pishro Co., Tehran, Iran	11/94-6/95
Designed, developed and implemented electrical systems using computer aided software. The 
electrical systems were intended for both indoors (Buildings) and outdoors (Airports and 
Terminals).
The systems were designed and based on the standard and regulation of electrical system.

Pardis Tower Co, Tehran, Iran	1/93-11/94
Designed simulators for training purpose.
*Developed an application in  C++ for  cycloconvertor simulator. The  
cycloconvertor is used in the starters of synchronous motors.
*Developed application using C++  for control speed of  DC motors. The application
 gets input voltage to the motor, speed of motor and rotor's current. It can control
 the motor speed based on requested speed.
*Developed application using C++ for step motor's control speed. In order to change the angle 
of dish antenna, the application orders to step motor to change the rotor's situation.

SKILLS:
Software development
Visual C++, Visual Basic, Visual InterDev, Visual J++, MFC, SDKs, COM, DCOM, ATL ActiveXs 
, C++, Fortran, Intel  8085 Assembler Access.
Internet development
Visual InterDev 6.0, Active Server Pages (ASP), SQL 7.0, COM, DCOM, Java script. 
Database development
MS-SQL, Access, FoxPro
Operating System Used
NT, MS-Win, UNIX, VAX, MS-DOS




From phd at phd.pp.ru  Tue Feb  6 15:50:49 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 6 Feb 2001 17:50:49 +0300 (MSK)
Subject: Software engineer
In-Reply-To: <69BC4C6AE83ED311BC9400805FA7B14AF84DE7@il93exp01.css.mot.com>
Message-ID: <Pine.LNX.4.30.0102061748220.7668-100000@fep132.fep.ru>

On Tue, 6 Feb 2001, Kemp Randy-W18971 wrote:
>    I think you are addressing the wrong list, since the Python mailing
>    list is a non profit organization.

   Kemp, it was clearly spam. The boy's trying to sent the message without
writing proper "Hello, Pythoneers" to a lot of newsgroups/lists. I already
sent spam complain to his provider.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From juergen at yellowbull.de  Tue Feb  6 16:11:05 2001
From: juergen at yellowbull.de (Juergen)
Date: Tue, 06 Feb 2001 15:11:05 GMT
Subject: cgi.py and HTML 4.0 compliancy
Message-ID: <95p461$62h$1@nnrp1.deja.com>

Hello,

I am using PYTHON 2.0 and had a slight problem with the cgi module in
combination with HTML 4.0 compliant HTML code.

according to "validator.w3.org" the following URL is not HTML 4
compliant:
   <a href="mycgiscript.py?firstParam=Bill&secondParam=Whitters">Some
Text</a>

it seems that the "&" character between firstParam and secondParam is
not allowed.
I think the correct form of the above URL would be:
   <a href="mycgiscript.py?firstParam=Bill%26secondParam=Whitters">Some
Text</a>
but I am not sure. At least the validator doesn't complain anymore.

The problem with this URL is, that the cgi modules FieldStorage class
can't handle this format properly.

As a workaround I modified cgi.parse_qsl by adding a new line:

def parse_qsl(qs, keep_blank_values=0, strict_parsing=0):
    """Parse ....... intended.
    """
    qs = qs.replace('%26','&')    # had to add this line to make it HTML
4.0 compliant.
    pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
    ...

For my purpose this is good enough. I just want to generate HTML code
that passes w3's validator.

I wonder if some Python core developer should take care of this or if
the HTML validator needs to be fixed.

Any opinions ?

Juergen


Sent via Deja.com
http://www.deja.com/


From michael at stroeder.com  Tue Feb  6 16:44:06 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Tue, 06 Feb 2001 16:44:06 +0100
Subject: cgi.py and HTML 4.0 compliancy
References: <95p461$62h$1@nnrp1.deja.com>
Message-ID: <3A801BC6.6ED77FEA@stroeder.com>

Juergen wrote:
> 
> according to "validator.w3.org" the following URL is not HTML 4
> compliant:
>    <a href="mycgiscript.py?firstParam=Bill&secondParam=Whitters">Some
> Text</a>

Maybe HTML4 mandates to use a form.

<form>
  <input type="hidden" name="firstParam" value="Bill">
  <input type="hidden" name="secondParam" value="Whitters">
  <input type="submit" value="Some Text">
</form>

> I think the correct form of the above URL would be:
>    <a href="mycgiscript.py?firstParam=Bill%26secondParam=Whitters">Some
> Text</a>
> but I am not sure. At least the validator doesn't complain anymore.

Well, this is not the same as the URL above.

> The problem with this URL is, that the cgi modules FieldStorage class
> can't handle this format properly.

Yes, off course. You URL-quoted the ampersand ('&') which is
necessary to separate the input fields.

> As a workaround I modified cgi.parse_qsl by adding a new line:

Don't do that. It violates the standard and works only if your
parameter values does not contain an ampersand.

> I wonder if some Python core developer should take care of this or if
> the HTML validator needs to be fixed.

Both are ok. IMHO your HTML has to be fixed.

Ciao, Michael.


From f8dy at my-deja.com  Tue Feb  6 17:39:45 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Tue, 06 Feb 2001 16:39:45 GMT
Subject: cgi.py and HTML 4.0 compliancy
References: <95p461$62h$1@nnrp1.deja.com> <3A801BC6.6ED77FEA@stroeder.com>
Message-ID: <95p9cf$b0g$1@nnrp1.deja.com>

In article <3A801BC6.6ED77FEA at stroeder.com>,
  Michael =?iso-8859-1?Q?Str=F6der?= <michael at stroeder.com> wrote:
> Juergen wrote:
> >
> > according to "validator.w3.org" the following URL is not HTML 4
> > compliant:
> >    <a href="mycgiscript.py?
firstParam=Bill&secondParam=Whitters">Some
> > Text</a>

Technically, this is not a valid URL.  The "&" must be written
as "&amp;".  Try using the HTML validator at http://htmlhelp.com/.  I
have found that it gives more descriptive error messages than the w3
validator, including this (very common) error.

>
> Maybe HTML4 mandates to use a form.
>

This is not true; URLs can contain parameters, as long as they are
properly URL-encoded.

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From ben.hutchings at roundpoint.com  Tue Feb  6 23:10:07 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 06 Feb 2001 14:10:07 -0800
Subject: cgi.py and HTML 4.0 compliancy
References: <95p461$62h$1@nnrp1.deja.com> <3A801BC6.6ED77FEA@stroeder.com> <95p9cf$b0g$1@nnrp1.deja.com>
Message-ID: <uofwfa2f4.fsf@roundpoint.com>

Mark Pilgrim <f8dy at my-deja.com> writes:

> In article <3A801BC6.6ED77FEA at stroeder.com>,
>   Michael =?iso-8859-1?Q?Str=F6der?= <michael at stroeder.com> wrote:
> > Juergen wrote:
> > >
> > > according to "validator.w3.org" the following URL is not HTML 4
> > > compliant:
> > >    <a href="mycgiscript.py?
> firstParam=Bill&secondParam=Whitters">Some
> > > Text</a>
> 
> Technically, this is not a valid URL.

It's a valid URL, but the value of the HREF attribute is interpreted
as PCDATA, i.e. ampersands within it introduce entities.

> The "&" must be written as "&amp;".

Right.

<snip>
> This is not true; URLs can contain parameters, as long as they are
> properly URL-encoded.

No, this is not URL-encoding.  URL-encoding (which is what Michael was
trying, replacing '&' with '%26') is a way of escaping characters that
are outside the range of legal characters for URLs or have special
meaning in URL syntax.  Since the '&' here is *meant* to have special
meaning as a query parameter separator, it must not be escaped.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From erno-news at erno.iki.fi  Tue Feb  6 17:32:32 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 06 Feb 2001 18:32:32 +0200
Subject: cgi.py and HTML 4.0 compliancy
References: <95p461$62h$1@nnrp1.deja.com>
Message-ID: <ku3ddr3h7j.fsf@lasipalatsi.fi>

In article <95p461$62h$1 at nnrp1.deja.com>, Juergen
<juergen at yellowbull.de> writes:

| Hello,
| I am using PYTHON 2.0 and had a slight problem with the cgi module in
| combination with HTML 4.0 compliant HTML code.

| according to "validator.w3.org" the following URL is not HTML 4
| compliant:
|    <a href="mycgiscript.py?firstParam=Bill&secondParam=Whitters">Some
| Text</a>

correct form:
<a href="mycgiscript.py?firstParam=Bill&amp;secondParam=Whitters">Some
Text</a>

  -- erno


From kalle at gnupung.net  Tue Feb  6 17:37:23 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Tue, 6 Feb 2001 17:37:23 +0100
Subject: cgi.py and HTML 4.0 compliancy
In-Reply-To: <95p461$62h$1@nnrp1.deja.com>; from juergen@yellowbull.de on Tue, Feb 06, 2001 at 03:11:05PM +0000
References: <95p461$62h$1@nnrp1.deja.com>
Message-ID: <20010206173723.A12766@father>

Sez Juergen:
> according to "validator.w3.org" the following URL is not HTML 4
> compliant:
>    <a href="mycgiscript.py?firstParam=Bill&secondParam=Whitters">Some
> Text</a>

Use
<a href="mycgiscript.py?firstParam=Bill&amp;secondParam=Whitters">Some Text</a>
instead.  '&amp;' is the character entity for '&'.  '&#38;' should work too,
I believe, but is perhaps less obvious.

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010206/768ce2a4/attachment.pgp>

From R.Brodie at rl.ac.uk  Tue Feb  6 18:03:59 2001
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Tue, 6 Feb 2001 17:03:59 -0000
Subject: cgi.py and HTML 4.0 compliancy
References: <95p461$62h$1@nnrp1.deja.com>
Message-ID: <95papp$110s@newton.cc.rl.ac.uk>

"Juergen" <juergen at yellowbull.de> wrote in message news:95p461$62h$1 at nnrp1.deja.com...

> it seems that the "&" character between firstParam and secondParam is
> not allowed.

A bare & is not allowed in HTML because it starts an entity. It should be
encoded as &amp; (you might have e.g. A href="users/J&uuml;rgen).

Even better, follow the HTML4 recommendation (B.2.2) and use
; (semicolon) as a delimiter. cgi.py supports this out of the box.






From stadt at cs.utwente.nl  Wed Feb  7 02:37:10 2001
From: stadt at cs.utwente.nl (Richard van de Stadt)
Date: Wed, 07 Feb 2001 02:37:10 +0100
Subject: cgi.py and HTML 4.0 compliancy
References: <95p461$62h$1@nnrp1.deja.com>
Message-ID: <3A80A6C6.90EF98EB@cs.utwente.nl>

Juergen wrote:
> 
> Hello,
> 
> I am using PYTHON 2.0 and had a slight problem with the cgi module in
> combination with HTML 4.0 compliant HTML code.
> 
> according to "validator.w3.org" the following URL is not HTML 4
> compliant:
>    <a href="mycgiscript.py?firstParam=Bill&secondParam=Whitters">Some
> Text</a>
> 
> it seems that the "&" character between firstParam and secondParam is
> not allowed.

I have always used '+' (without the quotes) to separate parameters. Has
always worked, still does (but I don't know if it's HTML 4.0).

Richard.


From juergen at yellowbull.de  Wed Feb  7 10:26:51 2001
From: juergen at yellowbull.de (Juergen)
Date: Wed, 07 Feb 2001 09:26:51 GMT
Subject: cgi.py and HTML 4.0 compliancy
References: <95p461$62h$1@nnrp1.deja.com> <3A80A6C6.90EF98EB@cs.utwente.nl>
Message-ID: <95r4cl$1f$1@nnrp1.deja.com>

Thanks a lot everybody,

I replaced '&' with '&amp;' and everything works perfect.

J?rgen


Sent via Deja.com
http://www.deja.com/


From hgg9140 at cola.ca.boeing.com  Tue Feb  6 16:34:59 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Tue, 6 Feb 2001 15:34:59 GMT
Subject: UDDI (XML discovery)?
Message-ID: <xqx1ytbrfj0.fsf@cola.ca.boeing.com>

Anyone working on a python rendition of UDDI?  There are some open
source java implementation which could server as templates if need be.


-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From sampe99 at my-deja.com  Tue Feb  6 17:10:09 2001
From: sampe99 at my-deja.com (sampe99 at my-deja.com)
Date: Tue, 06 Feb 2001 16:10:09 GMT
Subject: Binary numbers
Message-ID: <95p7ks$9e6$1@nnrp1.deja.com>

Does anyone know an easy way to get a list of binary numbers within a
certain range in python? E.g for
n=1 [0],[1]
n=2 [0,0],[0,1],[1,0],[1,1]
n=3 [0,0,0],[0,0,1] a.s.o

I need this for n<=18...

Thanks a lot in advance

/Sam


Sent via Deja.com
http://www.deja.com/


From f8dy at my-deja.com  Tue Feb  6 17:54:07 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Tue, 06 Feb 2001 16:54:07 GMT
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com>
Message-ID: <95pa79$c0j$1@nnrp1.deja.com>

In article <95p7ks$9e6$1 at nnrp1.deja.com>,
  sampe99 at my-deja.com wrote:
> Does anyone know an easy way to get a list of binary numbers within a
> certain range in python? E.g for
> n=1 [0],[1]
> n=2 [0,0],[0,1],[1,0],[1,1]
> n=3 [0,0,0],[0,0,1] a.s.o

>>> def b(l): return l>1 and [x+y for x in ['0','1'] for y in b(l-1)]
or ['0','1']
>>> b(1)
['0', '1']
>>> b(2)
['00', '01', '10', '11']
>>> b(3)
['000', '001', '010', '011', '100', '101', '110', '111']

Why do I get the feeling I'm doing your CS homework? ;)

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From sholden at holdenweb.com  Tue Feb  6 17:55:15 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 6 Feb 2001 11:55:15 -0500
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com>
Message-ID: <M3Wf6.34454$o5.418224@e420r-atl1.usenetserver.com>

<sampe99 at my-deja.com> wrote in message news:95p7ks$9e6$1 at nnrp1.deja.com...
> Does anyone know an easy way to get a list of binary numbers within a
> certain range in python? E.g for
> n=1 [0],[1]
> n=2 [0,0],[0,1],[1,0],[1,1]
> n=3 [0,0,0],[0,0,1] a.s.o
> 
> I need this for n<=18...
> 
> Thanks a lot in advance
> 
> /Sam
> 
Maybe this will help:

def bin(n):
    l = []
    while n > 0:
        l[:0] = [n%2]
        n /= 2
        if l == [] : return [0]
    return l

>>> bin(3)
[1, 1]
>>> bin(5)
[1, 0, 1]
>>> bin(11)
[1, 0, 1, 1]

regards
 Steve




From sholden at holdenweb.com  Tue Feb  6 18:44:00 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 6 Feb 2001 12:44:00 -0500
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com> <M3Wf6.34454$o5.418224@e420r-atl1.usenetserver.com>
Message-ID: <hNWf6.3604$x4.60565@e420r-atl2.usenetserver.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:M3Wf6.34454$o5.418224 at e420r-atl1.usenetserver.com...
> <sampe99 at my-deja.com> wrote in message news:95p7ks$9e6$1 at nnrp1.deja.com...
> > Does anyone know an easy way to get a list of binary numbers within a
> > certain range in python? E.g for
> > n=1 [0],[1]
> > n=2 [0,0],[0,1],[1,0],[1,1]
> > n=3 [0,0,0],[0,0,1] a.s.o
> >
> > I need this for n<=18...
> >
> > Thanks a lot in advance
> >
> > /Sam
> >
> Maybe this will help:
>
In my enthusiasm to present a tidy code listing I appear to have indented
one statement a little further than I should. I hate mailers that mess with
my layout! The function should have been:

def bin(n):
    l = []
    while n > 0:
        l[:0] = [n%2]
        n /= 2
    if l == [] : return [0]
    return l

regards
 Steve




From stephen_purcell at yahoo.com  Tue Feb  6 18:05:19 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 6 Feb 2001 18:05:19 +0100
Subject: Binary numbers
In-Reply-To: <95p7ks$9e6$1@nnrp1.deja.com>; from sampe99@my-deja.com on Tue, Feb 06, 2001 at 04:10:09PM +0000
References: <95p7ks$9e6$1@nnrp1.deja.com>
Message-ID: <20010206180519.A23598@freedom.puma-ag.com>

sampe99 at my-deja.com wrote:
> Does anyone know an easy way to get a list of binary numbers within a
> certain range in python? E.g for
> n=1 [0],[1]
> n=2 [0,0],[0,1],[1,0],[1,1]
> n=3 [0,0,0],[0,0,1] a.s.o
> 
> I need this for n<=18...

How about:

def binary(num):
    digits = [] 
    power = 0  
    while 2 ** power <= num:
        if 2 ** power & num:
            digits.append(1)
        else:
            digits.append(0)
        power = power + 1
    digits.reverse()     
    return digits   

for n in range(18):
    print n, '=', binary(n)


The output is as follows:

0 = []
1 = [1]
2 = [1, 0]
3 = [1, 1]
4 = [1, 0, 0]
5 = [1, 0, 1]
6 = [1, 1, 0]


The 'reverse' means that the list output has the most significant digits first,
which seems to be what you want.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From stephen_purcell at yahoo.com  Tue Feb  6 18:14:27 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 6 Feb 2001 18:14:27 +0100
Subject: Binary numbers
In-Reply-To: <20010206180519.A23598@freedom.puma-ag.com>; from stephen_purcell@yahoo.com on Tue, Feb 06, 2001 at 06:05:19PM +0100
References: <95p7ks$9e6$1@nnrp1.deja.com> <20010206180519.A23598@freedom.puma-ag.com>
Message-ID: <20010206181427.A23935@freedom.puma-ag.com>

Steve Purcell wrote:
> sampe99 at my-deja.com wrote:
> > Does anyone know an easy way to get a list of binary numbers within a
> > certain range in python? E.g for
> > n=1 [0],[1]
> > n=2 [0,0],[0,1],[1,0],[1,1]
> > n=3 [0,0,0],[0,0,1] a.s.o
> > 
> > I need this for n<=18...

Or even:

def binary(num):
    digits = map(lambda p, n=num: 2**p & n and 1, range(31))
    digits.reverse()                                        
    try: return digits[digits.index(1):]
    except: return [0]                  

for n in range(18):
    print n, '=', binary(n)


-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From ken at lt.com  Tue Feb  6 20:01:31 2001
From: ken at lt.com (Kenneth Loafman)
Date: Tue, 06 Feb 2001 13:01:31 -0600
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com>
Message-ID: <3A804A0B.E0445B7F@lt.com>

This is my second post on this subject.  Mail to this user bounces. 
Could be a troll, but since I'd already answered, here's what I said
(important to conserve electrons, don't you know)...

sampe99 at my-deja.com wrote:
> 
> Does anyone know an easy way to get a list of binary numbers within a
> certain range in python? E.g for
> n=1 [0],[1]
> n=2 [0,0],[0,1],[1,0],[1,1]
> n=3 [0,0,0],[0,0,1] a.s.o
> 
> I need this for n<=18...

With n=18, you're asking for
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 1],
...
[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1] 

n=18 would be a huge list of lists, (2^18) * 18 in size.

Assuming 4-byte integers, that would be (262144 * 18 * 4) = 18,874,368
which is 18MB in data alone.  Throw in the python list overhead and it
would probably occupy 25-30MB.  Of course, printing it out would be
impractical (useless?).

Any practical use for such a list?

...Ken



From sampe99 at my-deja.com  Wed Feb  7 17:47:28 2001
From: sampe99 at my-deja.com (sampe99 at my-deja.com)
Date: Wed, 07 Feb 2001 16:47:28 GMT
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com> <mailman.981486160.19141.python-list@python.org>
Message-ID: <95ru6u$lfg$1@nnrp1.deja.com>

In article <mailman.981486160.19141.python-list at python.org>,
  Kenneth Loafman <ken at lt.com> wrote:
> This is my second post on this subject.  Mail to this user bounces.
> Could be a troll, but since I'd already answered, here's what I said
> (important to conserve electrons, don't you know)...
>
> sampe99 at my-deja.com wrote:
> >
> > Does anyone know an easy way to get a list of binary numbers within
a
> > certain range in python? E.g for
> > n=1 [0],[1]
> > n=2 [0,0],[0,1],[1,0],[1,1]
> > n=3 [0,0,0],[0,0,1] a.s.o
> >
> > I need this for n<=18...
>
> With n=18, you're asking for
> [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
> [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],
> [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 1],
> ...
> [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]
>
> n=18 would be a huge list of lists, (2^18) * 18 in size.
>
> Assuming 4-byte integers, that would be (262144 * 18 * 4) = 18,874,368
> which is 18MB in data alone.  Throw in the python list overhead and it
> would probably occupy 25-30MB.  Of course, printing it out would be
> impractical (useless?).
>
> Any practical use for such a list?
>
> ...Ken
>
>

First, thanks to you all that presented excellent solutions to my
problem.

And to Ken. The basic problem is to make a transition probability
matrix (over one year) apply to a smaller time frame (like a week). To
do this I take the n:th root of the matrix which produces n different
solutions for each eigenvalue. I found python returns only one answer
to x**(1/n) and I therefor need to try all possible combinations of
positive and negative eigenvalues to find the correct solution. It
would also be preferred to try the complex roots too...

Do you have a better solution to this main problem I would be happy :)

Thanks again,

Sam


Sent via Deja.com
http://www.deja.com/


From ullrich at math.okstate.edu  Wed Feb  7 18:15:59 2001
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Wed, 07 Feb 2001 17:15:59 GMT
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com> <mailman.981486160.19141.python-list@python.org> <95ru6u$lfg$1@nnrp1.deja.com>
Message-ID: <95rvs3$n6n$1@nnrp1.deja.com>

In article <95ru6u$lfg$1 at nnrp1.deja.com>,
  sampe99 at my-deja.com wrote:
[...]
>
> First, thanks to you all that presented excellent solutions to my
> problem.
>
> And to Ken. The basic problem is to make a transition probability
> matrix (over one year) apply to a smaller time frame (like a week). To
> do this I take the n:th root of the matrix which produces n different
> solutions for each eigenvalue. I found python returns only one answer
> to x**(1/n) and I therefor need to try all possible combinations of
> positive and negative eigenvalues to find the correct solution. It
> would also be preferred to try the complex roots too...

How are you planning on deciding which one of the many n-th roots
of your matrix is the "right" one? (Is there some reason to
think that your matrix _has_ an n-th root which is also a
transition matrix???)

> Do you have a better solution to this main problem I would be happy :)
>
> Thanks again,
>
> Sam
>
> Sent via Deja.com
> http://www.deja.com/
>

--
Oh, dejanews lets you add a sig - that's useful...


Sent via Deja.com
http://www.deja.com/


From sampe99 at my-deja.com  Fri Feb  9 13:44:35 2001
From: sampe99 at my-deja.com (sampe99 at my-deja.com)
Date: Fri, 09 Feb 2001 12:44:35 GMT
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com> <mailman.981486160.19141.python-list@python.org> <95ru6u$lfg$1@nnrp1.deja.com> <95rvs3$n6n$1@nnrp1.deja.com>
Message-ID: <960onh$ov3$1@nnrp1.deja.com>



> How are you planning on deciding which one of the many n-th roots
> of your matrix is the "right" one? (Is there some reason to
> think that your matrix _has_ an n-th root which is also a
> transition matrix???)

All matrices with negative or complex elements can be excluded and
those left can be ranked after certain criterias. The question if there
exist a transition matrix or not among the roots is still left
unanswered. My intuitive feeling is that there exists one
unique "correct" matrix after my experiments on 3x3 and 4x4 matrices
(calculations by hand).

If I have a transistion matrix and want to transform it to a longer
time horizon I take the n:th power of it (under certain assumptions).


Sent via Deja.com
http://www.deja.com/


From ullrich at math.okstate.edu  Fri Feb  9 18:01:21 2001
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Fri, 09 Feb 2001 17:01:21 GMT
Subject: Binary numbers
References: <95p7ks$9e6$1@nnrp1.deja.com> <mailman.981486160.19141.python-list@python.org> <95ru6u$lfg$1@nnrp1.deja.com> <95rvs3$n6n$1@nnrp1.deja.com> <960onh$ov3$1@nnrp1.deja.com>
Message-ID: <9617oq$760$1@nnrp1.deja.com>

In article <960onh$ov3$1 at nnrp1.deja.com>,
  sampe99 at my-deja.com wrote:
>
>
> > How are you planning on deciding which one of the many n-th roots
> > of your matrix is the "right" one? (Is there some reason to
> > think that your matrix _has_ an n-th root which is also a
> > transition matrix???)
>
> All matrices with negative or complex elements can be excluded and
> those left can be ranked after certain criterias. The question if
there
> exist a transition matrix or not among the roots is still left
> unanswered. My intuitive feeling is that there exists one
> unique "correct" matrix

Ah.

> after my experiments on 3x3 and 4x4 matrices
> (calculations by hand).
>
> If I have a transistion matrix and want to transform it to a longer
> time horizon I take the n:th power of it (under certain assumptions).

A lot of things have n-th powers (for integers n) but no square
roots. I _think_ that a "transition matrix" is an nxn matrix
with all entries non-negative and with each row summing to 1.
Assuming so: A = [[0,1], [1,0]] is a transition matrix with
no real square root, much less a square root which is a transition
matrix.

(A has eigenvalues 1 and -1, so any square root of A will
have one real eigenvalue and one complex eigenvalue. A
real matrix can have complex eigenvalues but they come
in conjugate pairs, so the square roots of A cannot be real
matrices.)

> Sent via Deja.com
> http://www.deja.com/
>

--
Oh, dejanews lets you add a sig - that's useful...


Sent via Deja.com
http://www.deja.com/


From dbrueck at edgix.com  Tue Feb  6 17:45:36 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 6 Feb 2001 09:45:36 -0700
Subject: profile + threads?
Message-ID: <BLEOJGEMBNBNEAKMIFDHIEJOPIAA.dbrueck@edgix.com>

Has anybody had problems using the profile module with threads?

This program:
----------------------------
	import profile

	def yo():
	    for j in range(5):
	        print j,

	profile.run('yo()')
----------------------------

displays:

0 1 2 3 4          3 function calls in 0.070 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.000    0.000    0.000    0.000 <string>:1(?)
        1    0.000    0.000    0.000    0.000 p1.py:3(yo)
        0    0.000             0.000          profile:0(profiler)
        1    0.069    0.069    0.070    0.070 profile:0(yo())

but this version:

----------------------------
	import threading, profile, time

	def yo():
	    for j in range(5):
	        print j,

	def go():
	    threading.Thread(target=yo).start()
	    time.sleep(1)

	profile.run('go()')
----------------------------

displays:

0 1 2 3 4          14 function calls in 1.052 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.000    0.000    1.001    1.001 <string>:1(?)
        1    1.000    1.000    1.001    1.001 p1.py:7(go)
        1    0.050    0.050    1.052    1.052 profile:0(go())
        0    0.000             0.000          profile:0(profiler)
        1    0.000    0.000    0.000    0.000 threading.py:135(Condition)
        1    0.000    0.000    0.000    0.000 threading.py:140(__init__)
        1    0.000    0.000    0.000    0.000 threading.py:304(_newname)
        1    0.000    0.000    0.000    0.000 threading.py:321(__init__)
        1    0.000    0.000    0.000    0.000 threading.py:335(_set_daemon)
        1    0.001    0.001    0.001    0.001 threading.py:350(start)
        2    0.000    0.000    0.000    0.000 threading.py:39(__init__)
        1    0.000    0.000    0.000    0.000 threading.py:44(_note)
        1    0.000    0.000    0.000    0.000 threading.py:444(isDaemon)
        1    0.000    0.000    0.000    0.000
threading.py:516(currentThread)

The function yo() isn't mentioned anywhere. The single call to
threading.start shows up, but in the real program in which I'm seeing this
program has many threads, each of which calls many child functions, so I
can't rely on that.

Any ideas?

Thanks,
Dave




From btheld2 at my-deja.com  Tue Feb  6 19:49:55 2001
From: btheld2 at my-deja.com (btheld2 at my-deja.com)
Date: Tue, 06 Feb 2001 18:49:55 GMT
Subject: Using PyRun_SimpleString for blocks of Python Code
Message-ID: <95ph0e$it6$1@nnrp1.deja.com>

I am embedded Python in a C++ application and I need to send it blocks
of code.  For example:

if 1:
     print "Hi"

How do I do this using PyRun_SimpleString?  For this simple case, I
think I can send it "if 1: print "Hi"", but for more complex loops:

if 1:
     print "Hi"
     if 2:
           print "Bye"
           print "Bye"
     print "Hi2"
print "Done"

What do I do?


Sent via Deja.com
http://www.deja.com/


From root at rainerdeyke.com  Tue Feb  6 21:10:03 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 06 Feb 2001 20:10:03 GMT
Subject: Using PyRun_SimpleString for blocks of Python Code
References: <95ph0e$it6$1@nnrp1.deja.com>
Message-ID: <vSYf6.182682$ge4.63154986@news2.rdc2.tx.home.com>

<btheld2 at my-deja.com> wrote in message news:95ph0e$it6$1 at nnrp1.deja.com...
> I am embedded Python in a C++ application and I need to send it blocks
> of code.  For example:
>
> if 1:
>      print "Hi"
>
> How do I do this using PyRun_SimpleString?

Meet mister newline:

char *command =
  "if 1:\n"
  "    print \"Hi\"\n";

PyRun_SimpleString(command);


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From frank at quantiva.com  Tue Feb  6 21:09:13 2001
From: frank at quantiva.com (Frank Stolze)
Date: Tue, 6 Feb 2001 15:09:13 -0500 (EST)
Subject: [JOB AD] NJ - Python Jobs
Message-ID: <Pine.LNX.4.30.0102061508060.10315-100000@localhost.localdomain>

Company Profile:

We're onto something really big in network service management.  We have VC
backing and a clear view of the future.  Our stealth situation hides our
long history in this area and lets us focus on the goal.  Join the team of
innovative, visionary, enthusiastic, and passionate people that will shape
our future.  Share the opportunity to work with industry leading experts on
the design and creation of a next generation system.

Quantiva is looking for people who can quickly grasp new concepts, develop
new, original solutions to existing problems, and in general can "hit the
ground running."

Quantiva is located in Princeton, New Jersey.

Please send your resume to techjobs at quantiva.com.


1) Senior Network Software Engineer:

The Senior Network Software Engineer is responsible for the design and
development of a critical, major component of Quantiva's network service
management software. This position will involve understanding an existing,
smaller-scale implementation, and active participation in the design,
architecture, and implementation of the next stage of the product. More
details will be provided during the interview after the execution of a
non-disclosure agreement.

Requirements:

* Strong software design and development skills in a Unix environment.
* Strong network programming experience including a solid knowledge of
  protocols such as TCP/IP, HTTP, POP3, SMTP, DNS and concepts such as
  distributed systems, failover, routing, security, and firewalls.
* 4+ years of applications programming experience including Python and C++
  or Java.
* DBMS experience including database schemas, SQL, and database programming.
* Solid Unix experience (Solaris preferred) in a commercial environment.
* BS/MS/PhD in CS, EE, CE or related field.


2) Senior Software Engineer:

The Senior Software Engineer is responsible for the design and development
of a critical, major component of Quantiva's network service management
software. This position will involve understanding an existing, smaller-
scale implementation, and active participation in the design, architecture,
and implementation of the next stage of the product. More details will be
provided during the interview after the execution of a non-disclosure agreement.

Requirements:

* Strong software design and development skills in a Unix environment.
* Strong network and distributed systems programming experience.
* 4+ years of applications programming experience including Python and C++
  or Java.
* DBMS experience including database schemas, SQL, JDBC, and database
  programming.
* Solid Unix experience (Solaris preferred) in a commercial environment.
* BS/MS/PhD in CS, EE, CE or related field.





From burkhard.kayser at freenet.de  Tue Feb  6 22:06:15 2001
From: burkhard.kayser at freenet.de (Burkhard Kayser)
Date: Tue, 6 Feb 2001 22:06:15 +0100
Subject: timeoutsocket - error 10022
Message-ID: <95pp8k$o0r$02$1@news.t-online.com>

I want to user timeoutsocket together with telnetlib.

File "F:\PythonScripts\tn.py", line 13, in ?
    tn = telnetlib.Telnet(HOST)
  File "f:\python20\lib\telnetlib.py", line 122,
in __init__
    self.open(host, port)
  File "f:\python20\lib\telnetlib.py", line 139, in open

self.sock.connect((self.host, self.port))
  File "f:\python20\lib\timeoutsocket.py", line 210, in connect
    return
self.connect(addr, dumbhack=1)
  File "f:\python20\lib\timeoutsocket.py", line 182, in connect

sock.connect(addr)
  File "<string>", line 1, in connect
error: (10022, 'Invalid argument')

I seems that the value of addr is incorrect.
But if I test the script in by stepping through the code it works.
I'm running ActivePython build 202.

Has anyone had the same problem.





From gallegoj at ono.com  Tue Feb  6 22:13:13 2001
From: gallegoj at ono.com (José Luis Gallego)
Date: Tue, 06 Feb 2001 21:13:13 GMT
Subject: x = y
Message-ID: <JNZf6.2765$O31.16647@newsr01.ono.com>

Hi all,

I want to create a list x from a list y:

y = [[1,2,3],[4,5,6]]
x = y

How do I change individual elements of x without changing y?

Thanks in advance,

Jose Luis




From joconnor at cybermesa.com  Tue Feb  6 22:41:35 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 06 Feb 2001 21:41:35 GMT
Subject: x = y
References: <JNZf6.2765$O31.16647@newsr01.ono.com>
Message-ID: <3a806eed.5209768@news.cybermesa.com>

On Tue, 06 Feb 2001 21:13:13 GMT, "Jos? Luis Gallego"
<gallegoj at ono.com> wrote:

>Hi all,
>
>I want to create a list x from a list y:
>
>y = [[1,2,3],[4,5,6]]
>x = y
>
>How do I change individual elements of x without changing y?


Make a copy of the list

	x  = list (y)

x will contain the same elements as y, but will be a seperate list, so
you can update either indepently

>>> y = [[1,2,3],[4,5,6]]
>>> x = list (y)
>>> x
[[1, 2, 3], [4, 5, 6]]
>>> y
[[1, 2, 3], [4, 5, 6]]
>>> x [0] = "hi there"
>>> x
['hi there', [4, 5, 6]]
>>> y
[[1, 2, 3], [4, 5, 6]]
>>> 

Take care,


Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum - http://pub1.ezboard.com/fobjectorienteddevelopmentpython

"God himself plays on the bass strings first, when he tunes the soul"


From root at rainerdeyke.com  Tue Feb  6 22:49:21 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 06 Feb 2001 21:49:21 GMT
Subject: x = y
References: <JNZf6.2765$O31.16647@newsr01.ono.com> <3a806eed.5209768@news.cybermesa.com>
Message-ID: <Bj_f6.182846$ge4.63257732@news2.rdc2.tx.home.com>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:3a806eed.5209768 at news.cybermesa.com...
> On Tue, 06 Feb 2001 21:13:13 GMT, "Jos? Luis Gallego"
> <gallegoj at ono.com> wrote:
>
> >Hi all,
> >
> >I want to create a list x from a list y:
> >
> >y = [[1,2,3],[4,5,6]]
> >x = y
> >
> >How do I change individual elements of x without changing y?
>
>
> Make a copy of the list
>
> x  = list (y)

I think he wants a deep copy.  The 'copy' module might be useful here.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From bernie at pacific.net.hk  Tue Feb  6 22:58:26 2001
From: bernie at pacific.net.hk (Bernie)
Date: Wed, 07 Feb 2001 05:58:26 +0800
Subject: x = y
References: <JNZf6.2765$O31.16647@newsr01.ono.com>
Message-ID: <3A807382.298371E0@pacific.net.hk>

"Jos? Luis Gallego" wrote:

> Hi all,
>
> I want to create a list x from a list y:
>
> y = [[1,2,3],[4,5,6]]
> x = y
>
> How do I change individual elements of x without changing y?
>

I think you need the copy module.

import copy
y = [[1,2,3],[4,5,6]]
x = copy.copy(y)

This will do for a shadow copy.  See documentation for the copy module
in Python
Library Reference for furthur details

> Thanks in advance,
>
> Jose Luis

Bernie



From j_lockley at hotmail.com  Tue Feb  6 23:23:44 2001
From: j_lockley at hotmail.com (James Lockley)
Date: Wed, 7 Feb 2001 00:23:44 +0200
Subject: x = y
References: <JNZf6.2765$O31.16647@newsr01.ono.com> <3A807382.298371E0@pacific.net.hk>
Message-ID: <95pt8e$s54$1@news.adamastor.ac.za>

or just:

>>> x=y[:]
>>> x[1]='me'
>>> x
[[1, 2, 3], 'me']
>>> y
[[1, 2, 3], [4, 5, 6]]


cheers
james


"Bernie" <bernie at pacific.net.hk> wrote in message
news:3A807382.298371E0 at pacific.net.hk...
> "Jos? Luis Gallego" wrote:
>
> > Hi all,
> >
> > I want to create a list x from a list y:
> >
> > y = [[1,2,3],[4,5,6]]
> > x = y
> >
> > How do I change individual elements of x without changing y?
> >
>
> I think you need the copy module.
>
> import copy
> y = [[1,2,3],[4,5,6]]
> x = copy.copy(y)
>
> This will do for a shadow copy.  See documentation for the copy module
> in Python
> Library Reference for furthur details
>
> > Thanks in advance,
> >
> > Jose Luis
>
> Bernie
>




From aahz at panix.com  Wed Feb  7 03:04:41 2001
From: aahz at panix.com (Aahz Maruch)
Date: 6 Feb 2001 18:04:41 -0800
Subject: x = y
References: <JNZf6.2765$O31.16647@newsr01.ono.com> <3A807382.298371E0@pacific.net.hk> <95pt8e$s54$1@news.adamastor.ac.za>
Message-ID: <95qafp$a53$1@panix3.panix.com>

In article <95pt8e$s54$1 at news.adamastor.ac.za>,
James Lockley <j_lockley at hotmail.com> wrote:
>or just:
>
>>>> x=y[:]
>>>> x[1]='me'
>>>> x
>[[1, 2, 3], 'me']
>>>> y
>[[1, 2, 3], [4, 5, 6]]

But

>>> x[0][1]='you'
>>> x
[[1, 'you', 3], 'me']
>>> y
[[1, 'you', 3], [4, 5, 6]]
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Life as a contractor: when they tell you it's going to rain, sometimes
it drizzles and sometimes it pours.


From bernie at pacific.net.hk  Tue Feb  6 22:48:29 2001
From: bernie at pacific.net.hk (Bernie)
Date: Wed, 07 Feb 2001 05:48:29 +0800
Subject: Tkinter: How to remove stacking order of a widget in grid manager
Message-ID: <3A80712D.BAF8FC0D@pacific.net.hk>

Hi,

I have an application which requires reordering the widgets during
runtime.  I am using the grid manager and found two functions
grid_remove()
and grid_forget().  I thought grid_remove() should does the job, it
doesn't,
neither do grid_forget().

How can that be done?

Thanks in advance!


Below is the code fragment:

    # remove widget from Grid Display manager
    for w in widget_list
        w.grid_remove()

    # re-ordering the widget
    widget_list.sort( mySort)                # mySort() is not listed

    # display widgets in new orders
    col = range(5)
    row = 0
    for w in widget_list:
        for i in col:
            w.grid( column = i, row = row)
        row = row + 1





From s2mdalle at titan.vcu.edu  Wed Feb  7 01:31:43 2001
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Wed, 07 Feb 2001 00:31:43 GMT
Subject: Tkinter: How to remove stacking order of a widget in grid manager
References: <3A80712D.BAF8FC0D@pacific.net.hk>
Message-ID: <PH0g6.9255$5H.2228084@typhoon2.ba-dsg.net>

In article <3A80712D.BAF8FC0D at pacific.net.hk>, "Bernie"
<bernie at pacific.net.hk> wrote:

> Below is the code fragment:
> 
>     # remove widget from Grid Display manager
>     for w in widget_list
>         w.grid_remove()
> 
>     # re-ordering the widget
>     widget_list.sort( mySort)                # mySort() is not listed
> 
>     # display widgets in new orders
>     col = range(5) row = 0 for w in widget_list:
>         for i in col:
>             w.grid( column = i, row = row)
>         row = row + 1

I think what the problem is that it has to be done
in the parent, not the child.  So if 5 things are
grid() into a widget called wid, then

wid.grid_forget() will ungrid all of the items.
By calling grid_forget on the items you want to take
out of the grid, you're unpacking their subitems
but not them.

Example:

f = Frame(parent)
b = Button(f, text="Foobar")
b.grid(row=0, col=0)

if you then want to get rid of b, you do
f.grid_forget()

NOT
b.grid_forget()

Hope this helps.
-- 
David Allen
http://opop.nols.com/
----------------------------------------
"I must say that I find television very educational. The minute
somebody turns it on, I go to the library and read a book." 
		--Groucho Marx


From bernie at pacific.net.hk  Wed Feb  7 09:23:38 2001
From: bernie at pacific.net.hk (Bernie)
Date: Wed, 07 Feb 2001 16:23:38 +0800
Subject: Tkinter: How to remove stacking order of a widget in grid manager
References: <3A80712D.BAF8FC0D@pacific.net.hk> <PH0g6.9255$5H.2228084@typhoon2.ba-dsg.net>
Message-ID: <3A81060A.4ED32F2@pacific.net.hk>

Hi David,

I've try your suggestion but it still doesn't work.  Below
is the code I've used.

    root = Tkinter.Tk()
    frame = Tkinter.Frame( root)
    frame.pack()
    a = []
    for i in range(10):
        a.append( Tkinter.Button( frame, text=i))
        a[i].grid( column = 0, row = i)
    frame.grid_forget()                          # buttons did not
unmapped
    a.reverse()

    for i in range(10):
        a[i].grid( column = 0, row = i)
    # now the widgets are reordered but the stacking order did not
reverse

Bernie


David Allen wrote:

> In article <3A80712D.BAF8FC0D at pacific.net.hk>, "Bernie"
> <bernie at pacific.net.hk> wrote:
>
> > Below is the code fragment:
> >
> >     # remove widget from Grid Display manager
> >     for w in widget_list
> >         w.grid_remove()
> >
> >     # re-ordering the widget
> >     widget_list.sort( mySort)                # mySort() is not listed
>
> >
> >     # display widgets in new orders
> >     col = range(5) row = 0 for w in widget_list:
> >         for i in col:
> >             w.grid( column = i, row = row)
> >         row = row + 1
>
> I think what the problem is that it has to be done
> in the parent, not the child.  So if 5 things are
> grid() into a widget called wid, then
>
> wid.grid_forget() will ungrid all of the items.
> By calling grid_forget on the items you want to take
> out of the grid, you're unpacking their subitems
> but not them.
>
> Example:
>
> f = Frame(parent)
> b = Button(f, text="Foobar")
> b.grid(row=0, col=0)
>
> if you then want to get rid of b, you do
> f.grid_forget()
>
> NOT
> b.grid_forget()
>
> Hope this helps.
> --
> David Allen
> http://opop.nols.com/
> ----------------------------------------
> "I must say that I find television very educational. The minute
> somebody turns it on, I go to the library and read a book."
>                 --Groucho Marx

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
















From J.Jonkers at acriter.com  Wed Feb  7 09:50:12 2001
From: J.Jonkers at acriter.com (Johan Jonkers)
Date: Wed, 07 Feb 2001 09:50:12 +0100
Subject: Tkinter: How to remove stacking order of a widget in grid manager
References: <3A80712D.BAF8FC0D@pacific.net.hk> <PH0g6.9255$5H.2228084@typhoon2.ba-dsg.net> <3A81060A.4ED32F2@pacific.net.hk>
Message-ID: <3A810C44.EF8E9280@acriter.com>

I tried the grid_forget on a frame with a single button. It did not
dissapear. However, if you call the destroy() (or the forget()) method
on the button, it did dissapear.  Maybe this helps.

Johan
-- 
        -- Programmers don't die; they just GOSUB with no RETURN --


From bernie at pacific.net.hk  Wed Feb  7 11:46:23 2001
From: bernie at pacific.net.hk (Bernie)
Date: Wed, 07 Feb 2001 18:46:23 +0800
Subject: Tkinter: How to remove stacking order of a widget in grid manager
References: <3A80712D.BAF8FC0D@pacific.net.hk> <PH0g6.9255$5H.2228084@typhoon2.ba-dsg.net> <3A81060A.4ED32F2@pacific.net.hk> <3A810C44.EF8E9280@acriter.com>
Message-ID: <3A81277F.B6549DEE@pacific.net.hk>

Hi Johan,

I cannot destory the widget because the cost for the recreation
of the widget is expensive (especially with Pmw widget).  Since
all I needed to do is to reordering the widgets with new stacking
order.

Thanks for the advice!

Bernie


Johan Jonkers wrote:

> I tried the grid_forget on a frame with a single button. It did not
> dissapear. However, if you call the destroy() (or the forget()) method
> on the button, it did dissapear.  Maybe this helps.
>
> Johan
> --
>         -- Programmers don't die; they just GOSUB with no RETURN --



From fredrik at effbot.org  Wed Feb  7 17:52:23 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 07 Feb 2001 16:52:23 GMT
Subject: Tkinter: How to remove stacking order of a widget in grid manager
References: <3A80712D.BAF8FC0D@pacific.net.hk>
Message-ID: <b3fg6.13222$AH6.1995508@newsc.telia.net>

Bernie wrote:
> I have an application which requires reordering the widgets during
> runtime.  I am using the grid manager and found two functions
> grid_remove() and grid_forget().  I thought grid_remove() should
> does the job, it doesn't, neither do grid_forget().
>
> How can that be done?

just calling "grid" again (with or without grid_remove
or grid_forget) works fine for me.  try this:

# a simple regridding test

from Tkinter import *

root = Tk()

def flip():
    buttons.reverse()
    for i in range(10):
        buttons[i].grid(column=0, row=i)

buttons = []
for i in range(10):
    b = Button(root, text=i, width=20, command=flip)
    b.grid(column=0, row=i)
    buttons.append(b)

mainloop()

(just click on any of the buttons to move them around)

Cheers /F




From bernie at pacific.net.hk  Wed Feb  7 18:46:31 2001
From: bernie at pacific.net.hk (Bernie)
Date: Thu, 08 Feb 2001 01:46:31 +0800
Subject: Tkinter: How to remove stacking order of a widget in grid manager
References: <3A80712D.BAF8FC0D@pacific.net.hk> <b3fg6.13222$AH6.1995508@newsc.telia.net>
Message-ID: <3A8189F6.69B6495A@pacific.net.hk>

Hi Fredrik,

Have you try to <tab> around the widget (using the
<tab> key on keybroad).  If you just call grid again,
the tab order (or stacking order) of the widgets will
stays the same as you first created.

I use grid_forget() before grid() because I thought
grid_forget() should help me to solve the stacking
order problem as well.  I don't know if it is a bug
with the TK grid manager or I misunderstood the
documentation.  The bottom line is I am looking
for a solution to reorder the widgets, as well as it's
stacking order.

Any idea how can that be done?

Thanks in advance!


Bernie

Fredrik Lundh wrote:

> Bernie wrote:
> > I have an application which requires reordering the widgets during
> > runtime.  I am using the grid manager and found two functions
> > grid_remove() and grid_forget().  I thought grid_remove() should
> > does the job, it doesn't, neither do grid_forget().
> >
> > How can that be done?
>
> just calling "grid" again (with or without grid_remove
> or grid_forget) works fine for me.  try this:
>
> # a simple regridding test
>
> from Tkinter import *
>
> root = Tk()
>
> def flip():
>     buttons.reverse()
>     for i in range(10):
>         buttons[i].grid(column=0, row=i)
>
> buttons = []
> for i in range(10):
>     b = Button(root, text=i, width=20, command=flip)
>     b.grid(column=0, row=i)
>     buttons.append(b)
>
> mainloop()
>
> (just click on any of the buttons to move them around)
>
> Cheers /F

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------



From btheld2 at my-deja.com  Tue Feb  6 23:49:48 2001
From: btheld2 at my-deja.com (btheld2 at my-deja.com)
Date: Tue, 06 Feb 2001 22:49:48 GMT
Subject: Getting result of dir(object) in C++
Message-ID: <95pv28$1de$1@nnrp1.deja.com>

I need to get a string in C++ that contains all of the functions and
member data of a PyObject.

I know how to do this in raw python (dir(ob.__class__)), but how do I
do this in C++?  Assume I have a PyObject* retreived from
PyEval_EvalCode("objectstring")...

Ideas?


Sent via Deja.com
http://www.deja.com/


From pen_man at my-deja.com  Wed Feb  7 00:07:56 2001
From: pen_man at my-deja.com (penman)
Date: Tue, 06 Feb 2001 23:07:56 GMT
Subject: Getting one character at a time from sockets
Message-ID: <95q044$2f9$1@nnrp1.deja.com>

Using ordinary TCP/IP sockets with the socket module in Python, I need
to get a single character at a time from the client; something
like "talk" in unix. How could I do that? (should be platform
independent if possible)

recv(1) didn't work. It waited till the carriage return came in. (non-
blocking didn't work either)

Thanks in advance.


Sent via Deja.com
http://www.deja.com/


From sholden at holdenweb.com  Wed Feb  7 00:34:55 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 6 Feb 2001 18:34:55 -0500
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com>
Message-ID: <iW%f6.4311$x4.77684@e420r-atl2.usenetserver.com>

"penman" <pen_man at my-deja.com> wrote in message
news:95q044$2f9$1 at nnrp1.deja.com...
> Using ordinary TCP/IP sockets with the socket module in Python, I need
> to get a single character at a time from the client; something
> like "talk" in unix. How could I do that? (should be platform
> independent if possible)
>
> recv(1) didn't work. It waited till the carriage return came in. (non-
> blocking didn't work either)
>
> Thanks in advance.
>
>
> Sent via Deja.com
> http://www.deja.com/
>
I suspect the problem here is the sending end: if you can convince the
transport layer to flush its buffers after getting one character from the
sending application then you should, in theory, be able to receive that
single byte (since the PSH flag should cause the receiving end to debuffer).

This isn't easy without closing the socket, however ... have you considered
revising your application architecture?

The Unix Socket FAQ has the following (and a lot more) to say about this
subject:

11. How can I force a socket to send the data in its buffer?

>From Richard Stevens (rstevens at noao.edu):

You can't force it. Period. TCP makes up its own mind as to when it can send
data. Now, normally when you call write() on a TCP socket, TCP will indeed
send a segment, but there's no guarantee and no way to force this. There are
lots of reasons why TCP will not send a segment: a closed window and the
Nagle algorithm are two things to come immediately to mind.

(Snipped suggestion from Andrew Gierth to use TCP_NODELAY)

Setting this only disables one of the many tests, the Nagle algorithm. But
if the original poster's problem is this, then setting this socket option
will help.

A quick glance at tcp_output() shows around 11 tests TCP has to make as to
whether to send a segment or not.

Now from Dr. Charles E. Campbell Jr. (cec at gryphon.gsfc.nasa.gov):

As you've surmised, I've never had any problem with disabling Nagle's
algorithm. Its basically a buffering method; there's a fixed overhead for
all packets, no matter how small. Hence, Nagle's algorithm collects small
packets together (no more than .2sec delay) and thereby reduces the amount
of overhead bytes being transferred. This approach works well for rcp, for
example: the .2 second delay isn't humanly noticeable, and multiple users
have their small packets more efficiently transferred. Helps in university
settings where most folks using the network are using standard tools such as
rcp and ftp, and programs such as telnet may use it, too.

However, Nagle's algorithm is pure havoc for real-time control and not much
better for keystroke interactive applications (control-C, anyone?). It has
seemed to me that the types of new programs using sockets that people write
usually do have problems with small packet delays. One way to bypass Nagle's
algorithm selectively is to use "out-of-band" messaging, but that is limited
in its content and has other effects (such as a loss of sequentiality) (by
the way, out-of-band is often used for that ctrl-C, too).

More from Vic:

So to sum it all up, if you are having trouble and need to flush the socket,
setting the TCP_NODELAY option will usually solve the problem. If it
doesn't, you will have to use out-of-band messaging, but according to
Andrew, "out-of-band data has its own problems, and I don't think it works
well as a solution to buffering delays (haven't tried it though). It is not
'expedited data' in the sense that exists in some other protocols; it is
transmitted in-stream, but with a pointer to indicate where it is."

etc., etc., ...

regards
 Steve



From pen_man at my-deja.com  Wed Feb  7 00:56:33 2001
From: pen_man at my-deja.com (penman)
Date: Tue, 06 Feb 2001 23:56:33 GMT
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com>
Message-ID: <95q2vd$535$1@nnrp1.deja.com>

In article <95q044$2f9$1 at nnrp1.deja.com>,
  penman <pen_man at my-deja.com> wrote:
> Using ordinary TCP/IP sockets with the socket module in Python, I need
> to get a single character at a time from the client; something
> like "talk" in unix. How could I do that? (should be platform
> independent if possible)
>
> recv(1) didn't work. It waited till the carriage return came in. (non-
> blocking didn't work either)
>
> Thanks in advance.
>

What I am trying to do is a text editor, something like vi, on the
client side, in a BBS for telnet connections. When the client
pushes "i" without CR, the server has to notice it and send some
special characters to move the cursor up one line.


Sent via Deja.com
http://www.deja.com/


From donn at oz.net  Wed Feb  7 04:40:53 2001
From: donn at oz.net (Donn Cave)
Date: 7 Feb 2001 03:40:53 GMT
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com>
Message-ID: <95qg45$5b0$0@216.39.151.169>

Quoth penman <pen_man at my-deja.com>:
| Using ordinary TCP/IP sockets with the socket module in Python, I need
| to get a single character at a time from the client; something
| like "talk" in unix. How could I do that? (should be platform
| independent if possible)
|
| recv(1) didn't work. It waited till the carriage return came in. (non-
| blocking didn't work either)

Don't mess around with non-blocking.  Even recv(8000) will return
more or less immediately if you send any data at all, so it seems
kind of likely you're sending no data until the CR.  Substitute a
test client that has its data built in, and just send('i') to the
server.  I bet recv() will get it right away, Nagel algorithms
notwithstanding.

	Donn Cave, donn at oz.net


From pen_man at my-deja.com  Wed Feb  7 05:24:46 2001
From: pen_man at my-deja.com (penman)
Date: Wed, 07 Feb 2001 04:24:46 GMT
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com> <95qg45$5b0$0@216.39.151.169>
Message-ID: <95qim7$i38$1@nnrp1.deja.com>

In article <95qg45$5b0$0 at 216.39.151.169>,
  Donn Cave <donn at oz.net> wrote:
> Quoth penman <pen_man at my-deja.com>:
> | Using ordinary TCP/IP sockets with the socket module in Python, I
need
> | to get a single character at a time from the client; something
> | like "talk" in unix. How could I do that? (should be platform
> | independent if possible)
> |
> | recv(1) didn't work. It waited till the carriage return came in.
(non-
> | blocking didn't work either)
>
> Don't mess around with non-blocking.  Even recv(8000) will return
> more or less immediately if you send any data at all, so it seems
> kind of likely you're sending no data until the CR.  Substitute a
> test client that has its data built in, and just send('i') to the
> server.  I bet recv() will get it right away, Nagel algorithms
> notwithstanding.
>
> 	Donn Cave, donn at oz.net
>

Thank you, Donn. But the problem is that I have to use ordinary telnet
clients, such as CRT, Netterm, so let the users access to the server
without any specialy programs.

I wonder why nobody ever had the need to recv 1 character "unbuffered"
in Python.


Sent via Deja.com
http://www.deja.com/


From erno-news at erno.iki.fi  Wed Feb  7 07:48:19 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 07 Feb 2001 08:48:19 +0200
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com> <95qg45$5b0$0@216.39.151.169> <95qim7$i38$1@nnrp1.deja.com>
Message-ID: <kulmrj0z0s.fsf@lasipalatsi.fi>

In article <95qim7$i38$1 at nnrp1.deja.com>, penman <pen_man at my-deja.com>
writes:

| In article <95qg45$5b0$0 at 216.39.151.169>,
|   Donn Cave <donn at oz.net> wrote:
|| Don't mess around with non-blocking.  Even recv(8000) will return
|| more or less immediately if you send any data at all, so it seems
|| kind of likely you're sending no data until the CR.  Substitute a
|| test client that has its data built in, and just send('i') to the
|| server.  I bet recv() will get it right away, Nagel algorithms
|| notwithstanding.

| Thank you, Donn. But the problem is that I have to use ordinary telnet
| clients, such as CRT, Netterm, so let the users access to the server
| without any specialy programs.

the telnet protocol provides a mechanism to negotiate options such
as the character-at-a-time mode. the default is line at a time,
i believe. the rfcs have the details...

| I wonder why nobody ever had the need to recv 1 character "unbuffered"
| in Python.

again: there is no buffering on the python side here.

  -- erno


From db3l at fitlinxx.com  Thu Feb  8 04:16:55 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 07 Feb 2001 22:16:55 -0500
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com> <95qg45$5b0$0@216.39.151.169> <95qim7$i38$1@nnrp1.deja.com>
Message-ID: <uy9vhvp7c.fsf@ctwd0143.fitlinxx.com>

penman <pen_man at my-deja.com> writes:

> Thank you, Donn. But the problem is that I have to use ordinary telnet
> clients, such as CRT, Netterm, so let the users access to the server
> without any specialy programs.

It sounds like you aren't performing telnet options negotiation when
you get a connection.  At a minimum, even without that if you support
telnet clients you should process telnet IAC (hex 255) commands that
may show up in the stream from time to time or you may think that
you're getting bad data over the session.

By default, the telnet NVT (network virtual terminal) model is a line
buffered, half-duplex system.  So in the absence of other negotiation,
the clients are going to assume that they should only transmit upon
line termination, and also that they should perform local echoing of
characters being entered.  This works out nicely to permit telnet to
be used as a simple tool to connect to network servers that aren't
expecting it, but it does mean there's no character-at-a-time
transmission by default.

You can be sure you're in this mode if you are seeing what you type in
at your telnet client, but you haven't written code at your server to
explicitly echo received information back.

At a minimum, I would probably suggest having your server negotiate
the suppress-go-ahead and echo options (enabling both).  That means
that you should transmit a "DO SUPPRESS-GO-AHEAD" and a "WILL ECHO"
(unless you really want the client to handle local echo - not typical
in a full-duplex environment) over the stream during the startup of
the session.  The "DO" will request that the client transmit
character-by-character without needing GAs, and the latter will
indicate to the client that you are going to echo back what you
receive so the client will stop local echoing.  In response you should
expect to see a "WILL SUPPRESS-GO-AHEAD" and a "DO ECHO", which you'll
have to absorb and/or track with your options handling.

You might find the telnetlib.py library module to provide some very
basic examples of option processing, although it's geared towards
connecting to a server (not being a server) and doesn't really do full
options processing itself.

The real reference is in the RFCs - I'd probably suggest using:

    RFC854 (TELNET PROTOCOL SPECIFICATION)
    RFC855 (TELNET OPTION SPECIFICATIONS)
    RFC857 (TELNET ECHO OPTION)
    RFC858 (TELNET SUPPRESS GO AHEAD OPTION)

> I wonder why nobody ever had the need to recv 1 character "unbuffered"
> in Python.

I think you're generalizing too much.  Your issue is with the telnet
protocol, and not with raw socket communications.  recv() has no
impact on how a client may or may not be buffering network
communication - and note that it's the client doing the buffering
here, not your code.  And the reason the client is buffering it is
because the client is following the telnet protocol specification.  If
you were to write your own code that read a character from the
keyboard, did a send() on it, you would see the information come
through more immediately.

It is still true, however, that any TCP based system is subject to
various implementation details of TCP that may affect real time
delivery of data - by using TCP you are implicitly accepting that
fact.  Some items (e.g., the nagle algorithm) can impose several
hundred ms of latency even when no other buffering is going on at the
application layer.  It's part and parcel of the protocol - TCP makes
no guarantees about latency of delivery - just that the data will get
there intact and in order.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From rturpin at my-deja.com  Wed Feb  7 00:10:47 2001
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Tue, 06 Feb 2001 23:10:47 GMT
Subject: Scope: Simple and explicit better than nested?
Message-ID: <95q09f$2j1$1@nnrp1.deja.com>

The recent discussions of nested scope have convinced me that
this is the wrong and unpythonic solution to newbies confused
by global, local, and builtin namespaces. Instead of making
name resolution more complex and more implicit, just to give
people something familiar, let's solve the problem by
requiring everything to be explicit. Let's do away with both
global and builtin namespaces. There is only one namespace,
and it is the local namespace. Both "builtin" and the
containing module (if any) are automatically entered into it.
But to get to THEIR objects would require EXPLICIT scoping:

    if (builtin.callable(foo)):
        my_module.doSomething(foo)

The only unscoped names would be local variables and
parameters. Period, full stop. That should be easy to
understand, even for Pascal and SmallTalk programmers.

Along with this change, we should eliminate the "from"
statement. Names inside modules would be explicitly scoped.
Always. Simple and explicit. That is the pythonic way to go.

Hating-name-space-rules-ly yrs,
Russell


Sent via Deja.com
http://www.deja.com/


From dan at eevolved.com  Wed Feb  7 01:31:57 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 07 Feb 2001 00:31:57 GMT
Subject: Scope: Simple and explicit better than nested?
References: <95q09f$2j1$1@nnrp1.deja.com>
Message-ID: <1I0g6.65557$Pm2.1354846@news20.bellglobal.com>

rturpin at my-deja.com wrote:

This idea is good for the ease it provides, but it breaks _A LOT_ of old 
python code... I think that is the prime reason it will never be 
implemented :(

But if that's not an issue, you got my vote. I think for your idea to be a 
serious contender, however, you will have to make a patch that implements 
it :)

Dan


From scarblac at pino.selwerd.nl  Wed Feb  7 18:27:59 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 7 Feb 2001 17:27:59 GMT
Subject: Scope: Simple and explicit better than nested?
References: <95q09f$2j1$1@nnrp1.deja.com>
Message-ID: <slrn9830t1.ueq.scarblac@pino.selwerd.nl>

rturpin at my-deja.com <rturpin at my-deja.com> wrote in comp.lang.python:
> But to get to THEIR objects would require EXPLICIT scoping:
> 
>     if (builtin.callable(foo)):
>         my_module.doSomething(foo)
> 
> The only unscoped names would be local variables and
> parameters. Period, full stop. That should be easy to
> understand, even for Pascal and SmallTalk programmers.

Having to use builtin.range, builtin.int, builtin.open, builtin.repr, etc
etc completely kills the ease of use of Python. And it breaks all existing
code. In short, a very silly idea :-)

-- 
Remco Gerlich


From rturpin at my-deja.com  Thu Feb  8 17:17:53 2001
From: rturpin at my-deja.com (rturpin at my-deja.com)
Date: Thu, 08 Feb 2001 16:17:53 GMT
Subject: Scope: Simple and explicit better than nested?
References: <95q09f$2j1$1@nnrp1.deja.com> <slrn9830t1.ueq.scarblac@pino.selwerd.nl>
Message-ID: <95ugr5$tbc$1@nnrp1.deja.com>

In article <slrn9830t1.ueq.scarblac at pino.selwerd.nl>,
  scarblac at pino.selwerd.nl wrote:
> Having to use builtin.range, builtin.int, builtin.open,
> builtin.repr, etc etc completely kills the ease of use
> of Python. ..

Not at all. There is MUCH more to Python that makes it an
easy and great language than a few dozen builtin functions.
And if typing a few more characters makes it easier for
people corrupted by Pascal to understand what's going on,
well ..

> And it breaks all existing code.

OK, so we leave the local and builtin namespaces, and get
rid only of the global namespace. That breaks code only
for people who use the "from" statement, and we get rid
of that, too. Module names are always explicitly scoped,
as they should be anyway. This should still be simple
enough for Pascal programmers to understand. I bet I can
have a patch tomorrow that eliminates the "from" statement.

Despite that, I suspect we are doomed to nested scope.
For some reason, people prefer the complexity they learned
at their mother's knees to true simplicity. Python seems
destined to inherit the sins of the languages we left
behind. Feh.

Russell



Sent via Deja.com
http://www.deja.com/


From slhath at home.com  Wed Feb  7 00:56:02 2001
From: slhath at home.com (Scott Hathaway)
Date: Tue, 06 Feb 2001 23:56:02 GMT
Subject: python training class in dallas, tx
Message-ID: <ma0g6.54230$bb.951761@news1.rdc1.tx.home.com>

Hello Everyone,

If this is not appropriate for this newsgroup, please let me know which
group would be appropriate.

My company is providing a beginner's python training class April 4-6 in
Dallas, TX.  Anyone interested,
please send a mail to hathaways at infotrain.com and I will send you pricing,
etc.

Thanks,
Scott Hathaway




From fakeaddress at nospammail.com  Wed Feb  7 00:57:08 2001
From: fakeaddress at nospammail.com (Ben Catanzariti)
Date: Wed, 7 Feb 2001 10:57:08 +1100
Subject: Python Users Group Australia
Message-ID: <t43q59.pfd.ln@mail>

Hi,

I am looking for the existence of a Python Users Group in Australia ... does
one exist??

If so how do I get in contact with them:

Elif ... are there other parties interested in forming one?

thanks

Ben




From bbhaydon at my-deja.com  Wed Feb  7 03:53:04 2001
From: bbhaydon at my-deja.com (bbhaydon at my-deja.com)
Date: Wed, 07 Feb 2001 02:53:04 GMT
Subject: Python Users Group Australia
References: <t43q59.pfd.ln@mail>
Message-ID: <95qdad$dns$1@nnrp1.deja.com>

No, but I'd be interested in one...

regards,

Brett

In article <t43q59.pfd.ln at mail>,
  "Ben Catanzariti" <fakeaddress at nospammail.com> wrote:
> Hi,
>
> I am looking for the existence of a Python Users Group in
Australia ... does
> one exist??
>
> If so how do I get in contact with them:
>
> Elif ... are there other parties interested in forming one?
>
> thanks
>
> Ben
>
>


Sent via Deja.com
http://www.deja.com/


From MarkH at ActiveState.com  Wed Feb  7 04:34:20 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 07 Feb 2001 03:34:20 GMT
Subject: Python Users Group Australia
References: <t43q59.pfd.ln@mail>
Message-ID: <3A80C151.3000504@ActiveState.com>

Ben Catanzariti wrote:

> Hi,
> 
> I am looking for the existence of a Python Users Group in Australia ... does
> one exist??

Not as such... But there is a mailing list:

http://starship.python.net/mailman/listinfo/python-au

Mark.



From fakeaddress at nospammail.com  Wed Feb  7 05:14:07 2001
From: fakeaddress at nospammail.com (Ben Catanzariti)
Date: Wed, 7 Feb 2001 15:14:07 +1100
Subject: Python Users Group Australia
References: <t43q59.pfd.ln@mail> <3A80C151.3000504@ActiveState.com>
Message-ID: <p6iq59.lec.ln@mail>

Mark

As you already have a mailing list up and running ... would you like to
expand that to an Australian Python Users Group? or APUG ... (I think I
think I got involved in IT just to use acronymns :)

There is an Australian Users Group conference running this weekend ... ie
(linux, Perl, Cold fusion etc etc no Python however :(  ... I would consider
going to try and gather a base membership ...



"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A80C151.3000504 at ActiveState.com...
> Ben Catanzariti wrote:
>
> > Hi,
> >
> > I am looking for the existence of a Python Users Group in Australia ...
does
> > one exist??
>
> Not as such... But there is a mailing list:
>
> http://starship.python.net/mailman/listinfo/python-au
>
> Mark.
>




From MarkH at ActiveState.com  Wed Feb  7 06:45:07 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 07 Feb 2001 05:45:07 GMT
Subject: Python Users Group Australia
References: <t43q59.pfd.ln@mail> <3A80C151.3000504@ActiveState.com> <p6iq59.lec.ln@mail>
Message-ID: <3A80DFF8.1090308@ActiveState.com>

Ben Catanzariti wrote:


> As you already have a mailing list up and running ... would you like to
> expand that to an Australian Python Users Group? or APUG ... (I think I
> think I got involved in IT just to use acronymns :)

Umm - not really :-)  Although you are obviously free to use this list to organize such a beast.  I would be happy for this list to be the official mailing list of such a UG, but have no time to do anything at all towards organizing such a group.

> There is an Australian Users Group conference running this weekend ... ie
> (linux, Perl, Cold fusion etc etc no Python however :(  ... I would consider
> going to try and gather a base membership ...

That would be good - in Melbourne too :-)  The Python conference is that week, and I may not be back in time.

Mark.



From fakeaddress at nospammail.com  Wed Feb  7 05:21:31 2001
From: fakeaddress at nospammail.com (Ben Catanzariti)
Date: Wed, 7 Feb 2001 15:21:31 +1100
Subject: Python Users Group Australia
References: <t43q59.pfd.ln@mail> <3A80C151.3000504@ActiveState.com>
Message-ID: <lkiq59.b5d.ln@mail>

Actually the Users Group conference is on Saturday 10 March, 2001

This is from the website ...

USER GROUPS OPEN DAY
 Saturday, 10 March, 2001
 10 am-3 pm
VTR Consulting is please to announce that our first IT event in the new year
will be the User Groups Open Day. The event will take place at VTR's new
premises, Roseneath Place, between 10 am and 3 pm on 10 February 2001.
Roseneath Place is located in South Melbourne, off Market Street, between
Clarendon and Cecil Streets. For further information, concerning our
location, please click on the link
The idea behind the organisation of the Open Day is for VTR to facilitate a
meeting between IT professionals and the numerous user groups that together
represent a wide range of IT skills. Groups that will be represented on the
day include: Visual Basic (AVBUG), Delphi (ADUG), Java (AJUG), System
Administrators (SAGE-AU), Linux (LUV), Technical Writers (ASTC), Cold Fusion
(CFUG), etc...The Open Day will represent an excellent opportunity to to
talk to members  about their group activities and the benefits of membership
(a 10% membership discount, in any of the user groups, will be offered on
the day!) in an informal setting.
....




From swun at esec.com.au  Wed Feb  7 06:25:39 2001
From: swun at esec.com.au (Sam Wun)
Date: Wed, 07 Feb 2001 16:25:39 +1100
Subject: IP address and its netmask
Message-ID: <3A80DC53.F3488945@esec.com.au>

Hi,

I want to know whether 203.21.67.89 is in 203.21.67.0/32 network. How do
I do that in Python?

Thanks
sam



From DOUGS at oceanic.com  Wed Feb  7 15:15:15 2001
From: DOUGS at oceanic.com (Doug Stanfield)
Date: Wed, 7 Feb 2001 04:15:15 -1000 
Subject: IP address and its netmask
Message-ID: <8457258D741DD411BD3D0050DA62365907A5E3@huina.oceanic.com>

I found this:

ftp://ftp.cendio.se/pub/playground/python/ipaddr-1.1.tar.gz

By way of the Vaults:

http://www.vex.net/parnassus/

Its a module that has a lot of very useful utility if you're working with IP
numbers and networks.  For example, your problem (changed slightly to make
it a challenge):

[dougs at lawehana dougs]$ python
Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201
(egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import ipaddr
>>> aNet = ipaddr.network('203.21.67.0','255.255.252.0',ipaddr.DEMAND_NONE)
>>> aNet
{network 203.21.67.0/22}
>>> anIP = ipaddr.ipaddr('203.21.67.89')
>>> anIP
ipaddr.ipaddr('203.21.67.89')
>>> aNet.overlaps(anIP)
1
>>> otherIP = ipaddr.ipaddr('204.21.67.89')
>>> aNet.overlaps(otherIP)
0
>>>

-Doug-

> -----Original Message-----
> From: Sam Wun [mailto:swun at esec.com.au]
> Sent: Tuesday, February 06, 2001 7:26 PM
> To: python-list at python.org
> Subject: IP address and its netmask
> 
> 
> Hi,
> 
> I want to know whether 203.21.67.89 is in 203.21.67.0/32 

A /32 is a host address.  Maybe you meant something like /22?
> network. How do
> I do that in Python?
> 
> Thanks
> sam
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 



From nlymbo at my-deja.com  Wed Feb  7 07:02:30 2001
From: nlymbo at my-deja.com (nlymbo at my-deja.com)
Date: Wed, 07 Feb 2001 06:02:30 GMT
Subject: perl regex to python
Message-ID: <95qodk$mh9$1@nnrp1.deja.com>

Hi all, in perl i would do this to pick apart a date string:

if ($date_str =~ m#(\d+)/(\d+)/(\d+)#) {
  print "month: $1, $day: $2, $year: $3\n";
}

How does one save the matches in a reg expression using Python??
Thanks.

sf-l8


Sent via Deja.com
http://www.deja.com/


From dalke at acm.org  Wed Feb  7 07:19:36 2001
From: dalke at acm.org (Andrew Dalke)
Date: Tue, 6 Feb 2001 23:19:36 -0700
Subject: perl regex to python
References: <95qodk$mh9$1@nnrp1.deja.com>
Message-ID: <95qpgv$7fj$1@slb7.atl.mindspring.net>

nlymbo at my-deja.com asked:
>Hi all, in perl i would do this to pick apart a date string:
>
>if ($date_str =~ m#(\d+)/(\d+)/(\d+)#) {
>  print "month: $1, $day: $2, $year: $3\n";
>}

The most obvious Python equivalent is

import string
month, day, year = string.split($date_str, "/")

or (in Python 2.0)

month, day, year = date_str.split("/")

This is not identical to what the Perl snippet does.
Eg, imagine "1/2/3/4/5/6/7/8"

My code will call this an error while your code
reads the first three numbers.  If you want the
less strict behaviour, ask again and someone will
likely point out a pure regex solution.  However,
it's much less common in Python to use regexps as
compared to Perl.

                    Andrew
                    dalke at acm.org





From hamish_lawson at yahoo.co.uk  Wed Feb  7 11:16:41 2001
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: Wed, 07 Feb 2001 10:16:41 GMT
Subject: perl regex to python
References: <95qodk$mh9$1@nnrp1.deja.com>
Message-ID: <95r7a6$21q$1@nnrp1.deja.com>

nlymbo at my-deja.com wrote:

> Hi all, in perl i would do this to pick apart a date string:
>
> if ($date_str =~ m#(\d+)/(\d+)/(\d+)#) {
>   print "month: $1, $day: $2, $year: $3\n";
> }
>
> How does one save the matches in a reg expression using Python??

Python's re module provides the functionality of Perl's regular
expressions. The main difference is that Python doesn't have special
syntax and variables for dealing with REs (m//, $1, etc.), preferring
to do it via functions, methods and objects.

The re module provides a search function which returns a match object
if the search was successful or None if not. The group() method of the
match object returns the specified matched group. (When specifying REs
it's often a good idea to use the r"" raw-string notation to prevent
expansion of backslashed characters by Python itself.)

    import re
    date_str = "Today is 02/07/2001"
    m = re.search(r"(\d+)/(\d+)/(\d+)", date_str)
    if m:
        print "month: %s, day: %s, year: %s" % \
            (m.group(1), m.group(2), m.group(3))

Since the groups in your example are being printed out in order, you
could use the groups() method instead to return a tuple of the matched
groups:

        print "month: %s, day: %s, year: %s" % m.groups()

If you were going to be matching dates in a number of candidate
strings, then you would probably want to compile the regular expression
into a matcher object first and then repeatedly call the matcher
object's own search() method:

    matcher = re.compile(r"(\d+)/(\d+)/(\d+)")
    m = matcher.search(date_str)

This is similar to Perl's 'o' modifier for m//, but demonstrates
Python's preference of having an explicit object to represent the
compiled RE.

Hamish Lawson


Sent via Deja.com
http://www.deja.com/


From scarblac at pino.selwerd.nl  Wed Feb  7 18:54:19 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 7 Feb 2001 17:54:19 GMT
Subject: perl regex to python
References: <95qodk$mh9$1@nnrp1.deja.com>
Message-ID: <slrn9832ec.uii.scarblac@pino.selwerd.nl>

nlymbo at my-deja.com <nlymbo at my-deja.com> wrote in comp.lang.python:
> Hi all, in perl i would do this to pick apart a date string:
> 
> if ($date_str =~ m#(\d+)/(\d+)/(\d+)#) {
>   print "month: $1, $day: $2, $year: $3\n";
> }
> 
> How does one save the matches in a reg expression using Python??
> Thanks.

If you're on Linux (or some other system that has strptime(), ie not Windows),
it'd be much easier to parse dates with t=time.strptime(s,"%m/%d/%Y"), and
similar.

>>> import time
>>> time.asctime(time.strptime("11/05/2001","%m/%d/%Y"))
'Mon Nov  5 00:00:00 2001'

If you need this on Windows, there are pure Python implementations out
there, somewhere.

-- 
Remco Gerlich


From f8dy at my-deja.com  Wed Feb  7 21:05:27 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Wed, 07 Feb 2001 20:05:27 GMT
Subject: perl regex to python
References: <95qodk$mh9$1@nnrp1.deja.com> <slrn9832ec.uii.scarblac@pino.selwerd.nl>
Message-ID: <95s9pv$1h0$1@nnrp1.deja.com>

In article <slrn9832ec.uii.scarblac at pino.selwerd.nl>,
  scarblac at pino.selwerd.nl wrote:
> >>> import time
> >>> time.asctime(time.strptime("11/05/2001","%m/%d/%Y"))
> 'Mon Nov  5 00:00:00 2001'
>
> If you need this on Windows, there are pure Python implementations out
> there, somewhere.

http://www.fukt.hk-r.se/~flognat/hacks/strptime.py

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From rayvd at nospam.firetail.org  Wed Feb  7 07:32:44 2001
From: rayvd at nospam.firetail.org (Ray Van Dolson)
Date: Wed, 07 Feb 2001 06:32:44 GMT
Subject: Radix sort algorithm in Python question
Message-ID: <g_5g6.12851$1%2.627095@sjc-read.news.verio.net>

I'm designing a Radix sort algorithm in Python (for an assignment).  I have 
the algorithm done, but it's not exactly fast.  Since I'm supposed to be 
comparing this to Quicksort (my Python Quicksort implementation is -
infinitely- faster).  Here's the code:

def rSort(a,n,maxLen):
	bins=[[],[],[],[],[],[],[],[],[],[],[]]
	for x in range(0,maxLen):
		if x == 0:
			for y in a:
				bins[y%n].append(y)
		else:
			for y in a:
				origValue=y
				if y < 10**x:
					bins[0].append(y)
				else:
					y=y/(10**x)
					bins[y%n].append(origValue)
		del a
		a=[]
		for section in bins:
			if len(section) > 0:
				for times in range(0,len(section)):
					a.append(section.pop(0))


I'm guessing the problem lies when I am taking the elements out of the 
"bins" and putting them back into the array.  I delete the array and then 
step through the bins popping the numbers out.  When we're talking about 
thousands of elements this takes a while.

Anyone have any suggestions on how I could better streamline this?  I don't 
think radix should be 40 times slower than quicksort, but I could be wrong.

Thanks,
Ray Van Dolson


From fredrik at effbot.org  Wed Feb  7 09:22:50 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 07 Feb 2001 08:22:50 GMT
Subject: Radix sort algorithm in Python question
References: <g_5g6.12851$1%2.627095@sjc-read.news.verio.net>
Message-ID: <uB7g6.13187$AH6.1982516@newsc.telia.net>

Ray Van Dolson wrote:
> del a

you don't have to delete a name that you're about to
reuse on the next line:

(note that variables are name to value mappings,
not values.  deleting a name doesn't always make
the value go away, and binding the name to some-
thing else doesn't affect the original value)

> a=[]
> for section in bins:
>    if len(section) > 0:

if the section is empty, the for loop won't do anything
anyway.  you can get rid of that if statement.

>        for times in range(0,len(section)):
>            a.append(section.pop(0))

the start value to range is optional; range(len(section))
does the same thing.  and since you're looping over the
indices, why not *use* them:

        for index in range(len(section)):
            a.append(section[i])

(pop modifies the section, indexing doesn't)

or better, loop over the items instead:

        for item in section:
            a.append(item)

or better, use the extend method to add the whole
section to the end of the a array:

        a.extend(section)

in 2.0 and later, you can write this as:

        a += section

Cheers /F




From gregj at pobox.com  Wed Feb  7 10:08:00 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 07 Feb 2001 09:08:00 GMT
Subject: Radix sort algorithm in Python question
References: <g_5g6.12851$1%2.627095@sjc-read.news.verio.net>
Message-ID: <95r39d$vaq$1@nnrp1.deja.com>

In article <g_5g6.12851$1%2.627095 at sjc-read.news.verio.net>,
  rayvd at nospam.firetail.org (Ray Van Dolson) wrote:
> I'm designing a Radix sort algorithm in Python (for an assignment).
I have
> the algorithm done, but it's not exactly fast.  Since I'm supposed to
be
> comparing this to Quicksort (my Python Quicksort implementation is -
> infinitely- faster).  Here's the code:

You can improve the function by moving some stuff out of the loops, and
eliminating some special cases. For example, you calculate 10**x twice
each time through your inner loop. Compute it once in the outer loop.
The special case for x==0, tested each time through the outer loop even
though it will only happen once, and the unnecessary test for y < 10**x
each time through the inner loop can be eliminated.

You don't need the parameter n, since the sort radix must be 10 for
this algorithm (which relies on the modulus operator). And the function
can figure out maxLen (number of passes required to sort all of the
values) itself. Finally, the nested loops to copy the values from the
bins back to the list can be turned into a single loop.

Here's my version:

---
def rsort(a):
    "sort list of integers using base-10 radix sort"

    if a:
        bins = [ [],[],[],[],[],[],[],[],[],[] ]

        # get largest element to determine when sort is done
        m = max(a)
        # decimal digit currently partitioning on (1,10,100...)
        r = 1

        while m > r:
            # append each element of a to the end of the bin
            # corresponding to the partition digit
            for e in a:
                bins[(e/r)%10].append(e)

            r = r * 10

            # move values from bins back to a single list
            a = []
            for i in range(10):
                a.extend(bins[i])
                bins[i] = []

    return a
---

I'm sure my implementation can be refined and sped up, but already it
is about 3 times faster than your original. I created a 10,000 element
list of random integers ranging from 1 to 99999. On my system (Python
2.0) my function sorts the list in 0.54 seconds, and yours runs in 1.5
seconds.


>
> def rSort(a,n,maxLen):
> 	bins=[[],[],[],[],[],[],[],[],[],[],[]]
> 	for x in range(0,maxLen):
> 		if x == 0:
> 			for y in a:
> 				bins[y%n].append(y)
> 		else:
> 			for y in a:
> 				origValue=y
> 				if y < 10**x:
> 					bins[0].append(y)
> 				else:
> 					y=y/(10**x)
> 					bins[y%n].append(origValue)
> 		del a
> 		a=[]
> 		for section in bins:
> 			if len(section) > 0:
> 				for times in range(0,len(section)):
> 					a.append(section.pop(0))
>


--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From jpet at eskimo.com  Thu Feb  8 07:49:17 2001
From: jpet at eskimo.com (Jeff Petkau)
Date: Thu, 08 Feb 2001 06:49:17 GMT
Subject: Radix sort algorithm in Python question
References: <g_5g6.12851$1%2.627095@sjc-read.news.verio.net>
Message-ID: <Njrg6.4795$2H3.434880@dfiatx1-snr1.gtei.net>

Ray Van Dolson <rayvd at nospam.firetail.org> wrote in message
news:g_5g6.12851$1%2.627095 at sjc-read.news.verio.net...
> I'm designing a Radix sort algorithm in Python (for an assignment).  I
have
> the algorithm done, but it's not exactly fast.  Since I'm supposed to be
> comparing this to Quicksort (my Python Quicksort implementation is -
> infinitely- faster).  Here's the code:
>
> def rSort(a,n,maxLen):
>     bins=[[],[],[],[],[],[],[],[],[],[],[]]
>     for x in range(0,maxLen):
>         if x == 0:
>             for y in a:
>                 bins[y%n].append(y)
>         else:
>             for y in a:
>                 origValue=y
>                 if y < 10**x:
>                     bins[0].append(y)
>                 else:
>                     y=y/(10**x)
>                     bins[y%n].append(origValue)
>         del a
>         a=[]
>         for section in bins:
>             if len(section) > 0:
>                 for times in range(0,len(section)):
>                     a.append(section.pop(0))
>
>
> I'm guessing the problem lies when I am taking the elements out of the
> "bins" and putting them back into the array.  I delete the array and then
> step through the bins popping the numbers out.  When we're talking about
> thousands of elements this takes a while.
>
> Anyone have any suggestions on how I could better streamline this?  I
don't
> think radix should be 40 times slower than quicksort, but I could be
wrong.
>
> Thanks,
> Ray Van Dolson

First the minor stuff: all of the complicated if/else logic
in the first loop is unnecessary--you end up doing the same
thing in all cases. So you can change it to:

    for x in range(maxLen):
        bins[(y/10**x)%n].append(y)

But the real problem is the 'section.pop(0)' call. Lists in
Python are stored as linear arrays, and pop(0) removes the
*first* element from the array, copying all the other elements
down one position. Since the size of each bin is proportional
to the size of the input array, this changes the algorithm
from O(N) to O(N^2)!

Instead, you could use 'a.extend(section)' to add the
contents of section to a. This doesn't empty out
section, so you have to do that (say with 'del section[:]')
or just recreate the bins array each time through the
loop.

So the faster version is:

def rSort2(a,n,maxLen):
    for x in range(maxLen):
        bins = [[] for i in range(n)]
        for y in a:
            bins[(y/10**x)%n].append(y)
        a=[]
        for section in bins:
            a.extend(section)
    return a

On a 100,000 element list, rSort(x,10,6) takes about 59 seconds
on my machine; rSort2() takes just under 5 seconds.

On a 1,000,000 element list, rSort2(x,10,6) takes 54 seconds;
I expect rSort() to take about an hour and a half, but I'm
tired of waiting for it so I'll just send this message now.

--Jeff





From richard at millinfo.co.za  Wed Feb  7 08:31:00 2001
From: richard at millinfo.co.za (Richard)
Date: Wed, 07 Feb 2001 09:31:00 +0200
Subject: dictionary sub-value lookup
Message-ID: <3A80F9B4.E072037D@millinfo.co.za>

Hi

I am kinda new to this whole object orientation thing, so
there is probably an easy way of doing this, but the
tutorials etc are sparse.

I have an object M which contains N dictionaries. Inside
dictionary N, is an identifier, Y, plus various other
values, eg: Z.... is it possible to find M.Y.Z without
looping through the whole of M, checking Y each time ?

Since I will have to do a LOT of this, and M may be large,
I need a fast method.  

Ideas/comments etc ?

-Bye
-Richard


From sholden at holdenweb.com  Wed Feb  7 17:29:07 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 7 Feb 2001 11:29:07 -0500
Subject: dictionary sub-value lookup
References: <3A80F9B4.E072037D@millinfo.co.za>
Message-ID: <eNeg6.38506$o5.482290@e420r-atl1.usenetserver.com>

"Richard" <richard at millinfo.co.za> wrote in message
news:3A80F9B4.E072037D at millinfo.co.za...
> Hi
>
> I am kinda new to this whole object orientation thing, so
> there is probably an easy way of doing this, but the
> tutorials etc are sparse.
>
It takes a while to get used to the OO way of things, but it will pay
dividends so it's worth the effort.

> I have an object M which contains N dictionaries. Inside
> dictionary N, is an identifier, Y, plus various other
> values, eg: Z.... is it possible to find M.Y.Z without
> looping through the whole of M, checking Y each time ?
>
First obvious question is why are the dictionaries separate?

> Since I will have to do a LOT of this, and M may be large,
> I need a fast method.
>
> Ideas/comments etc ?
>
One simple speed-up, which may or may not help in your case, is to maintain
a global contents dictionary, where each key contains a list of the
dictionaries in which it is defined. That way you'll know exactly which
dictionaries it's worth searching. Of course there's some overhead involved
in this, so you'll have to evaluate the tradeoff.

More detail, or some code, might help in providing more relevant
suggestions.

regards
 Steve




From dan at eevolved.com  Wed Feb  7 18:41:51 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 07 Feb 2001 17:41:51 GMT
Subject: dictionary sub-value lookup
References: <3A80F9B4.E072037D@millinfo.co.za>
Message-ID: <zNfg6.204967$JT5.7467063@news20.bellglobal.com>

Richard wrote:

> Hi
> 
> I am kinda new to this whole object orientation thing, so
> there is probably an easy way of doing this, but the
> tutorials etc are sparse.
> 
> I have an object M which contains N dictionaries. Inside
> dictionary N, is an identifier, Y, plus various other
> values, eg: Z.... is it possible to find M.Y.Z without
> looping through the whole of M, checking Y each time ?
> 
> Since I will have to do a LOT of this, and M may be large,
> I need a fast method.
> 
> Ideas/comments etc ?
> 
> -Bye
> -Richard

What problem are you trying to solve? There may be simpler ways to do it... 
that sounds incredibly complex. You could always have a dictionary that 
contains
{"dictionary_names":dict}
and do
M[N].Y.Z

does this help?
Dan


From scarblac at pino.selwerd.nl  Wed Feb  7 18:47:32 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 7 Feb 2001 17:47:32 GMT
Subject: dictionary sub-value lookup
References: <3A80F9B4.E072037D@millinfo.co.za>
Message-ID: <slrn98321l.uii.scarblac@pino.selwerd.nl>

Richard <richard at millinfo.co.za> wrote in comp.lang.python:
> I am kinda new to this whole object orientation thing, so
> there is probably an easy way of doing this, but the
> tutorials etc are sparse.
> 
> I have an object M which contains N dictionaries. Inside
> dictionary N, is an identifier, Y, plus various other
> values, eg: Z.... is it possible to find M.Y.Z without
> looping through the whole of M, checking Y each time ?
> 
> Since I will have to do a LOT of this, and M may be large,
> I need a fast method.  

Dictionary lookups are fast, but it's rather unclear what you mean
exactly (what's an identifier in a dictionary, with various other values?)

If you have dictionaries of the type {"name1":"value1","name2":"value2" ,...},
and you're searching for a specific *value*, then you have to loop through
the hole thing. Dictionaries are lookup tables, they are one way (try
finding a name in the telephone book, given a phone number).

If you need this, it may be a good idea to keep a second dictionary with
the inverse relationship ({"value1":"name1",...}) but maybe you should first
try to explain yourself a bit better ;-).

-- 
Remco Gerlich


From richard at millinfo.co.za  Thu Feb  8 11:00:14 2001
From: richard at millinfo.co.za (Richard)
Date: Thu, 08 Feb 2001 12:00:14 +0200
Subject: dictionary sub-value lookup
References: <3A80F9B4.E072037D@millinfo.co.za> <slrn98321l.uii.scarblac@pino.selwerd.nl>
Message-ID: <3A826E2E.611F9DBD@millinfo.co.za>

Heya

Okie, <python newbie tries better explanation>:

The object comes from a program.  I have no control over it.
Example of M[0]:
({'Label': 251, 'value': 1275.0})

I just throught there was some easy way of getting M[i] given that
M[i]['Label'] = 251 ... when what I really want is M[i]['value']

ie: given '251', find corrosponding 1275.

Bearing in mind, M is actually HUGE, and M[i] contains about 10
lines of uninteresting data when printed.  There is a LOT of
information here, and I would like the thing to be fast.  I just
thought that maybe there was some funky python or OO way of doing
it (since I am from a Fortran background).

-Bye
-Richard


From mikael at isy.liu.se  Thu Feb  8 13:15:00 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 08 Feb 2001 13:15:00 +0100 (MET)
Subject: dictionary sub-value lookup
In-Reply-To: <3A826E2E.611F9DBD@millinfo.co.za>
Message-ID: <XFMail.010208131500.mikael@isy.liu.se>

On 08-Feb-01 Richard wrote:
 >  The object comes from a program.  I have no control over it.
 >  Example of M[0]:
 >  ({'Label': 251, 'value': 1275.0})
 >  
 >  I just throught there was some easy way of getting M[i] given that
 >  M[i]['Label'] = 251 ... when what I really want is M[i]['value']
 >  
 >  ie: given '251', find corrosponding 1275.
 >  
 >  Bearing in mind, M is actually HUGE, and M[i] contains about 10
 >  lines of uninteresting data when printed.  There is a LOT of
 >  information here, and I would like the thing to be fast.  I just
 >  thought that maybe there was some funky python or OO way of doing
 >  it (since I am from a Fortran background).

Just my two cents:

If you are going to do a lot of lookups of this kind, I think you
should consider preprocessing the data, thus transforming

  {'Label': 251, 'value': 1275.0}

to 

  {251: 1275.0}

and after that do your lookups.

I leave the funky OO-stuff to others.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    08-Feb-01
Time:    13:09:03

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From aahz at panix.com  Thu Feb  8 17:08:20 2001
From: aahz at panix.com (Aahz Maruch)
Date: 8 Feb 2001 08:08:20 -0800
Subject: dictionary sub-value lookup
References: <3A80F9B4.E072037D@millinfo.co.za> <slrn98321l.uii.scarblac@pino.selwerd.nl> <3A826E2E.611F9DBD@millinfo.co.za>
Message-ID: <95ug9k$j72$1@panix6.panix.com>

In article <3A826E2E.611F9DBD at millinfo.co.za>,
Richard  <rhsteven at eng.uct.ac.za> wrote:
>Okie, <python newbie tries better explanation>:
>
>The object comes from a program.  I have no control over it.
>Example of M[0]:
>({'Label': 251, 'value': 1275.0})
>
>I just throught there was some easy way of getting M[i] given that
>M[i]['Label'] = 251 ... when what I really want is M[i]['value']
>
>ie: given '251', find corrosponding 1275.

For starters, you probably want a real database for this kind of thing.
That way, someone else has already done the work.  But if you really
want to do this in Python, you need to create a dict that looks
something like this:

M[251] = (1275.0, 0)

If it's possible that more than one label of 251 will exist, you need to
do this:

M[251] = [(1275.0, 0), ...]
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From SBrunning at trisystems.co.uk  Thu Feb  8 12:40:08 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 8 Feb 2001 11:40:08 -0000 
Subject: dictionary sub-value lookup
Message-ID: <31575A892FF6D1118F5800600846864D5B17CE@intrepid>

> From:	Richard [SMTP:richard at millinfo.co.za]
> The object comes from a program.  I have no control over it.
> Example of M[0]:
> ({'Label': 251, 'value': 1275.0})
 
Ah-ha, I think that I get it. You don't have a big dictionary, you have a
tuple containing lots of small dictionaries. Correct?
 
> I just throught there was some easy way of getting M[i] given that
> M[i]['Label'] = 251 ... when what I really want is M[i]['value']
> 
> ie: given '251', find corrosponding 1275.
 
So, given a value, you which to find the dictionary in your tuple for which
the key 'Label' maps to your value, and return the 'value' entry from that
dictionary. Correct?
 
> Bearing in mind, M is actually HUGE, and M[i] contains about 10
> lines of uninteresting data when printed.  There is a LOT of
> information here, and I would like the thing to be fast.  I just
> thought that maybe there was some funky python or OO way of doing
> it (since I am from a Fortran background).
 
If you want to retrieve just one value, then I think that you'll just want
to loop through the dictionaries looking for the one you want. If you want
to retrieve lots of values, you might want to build another dictionary keyed
by the 'Label' value, containing the 'value' value.

Do all the dictionaries have the same key values? I.e., do they all have
'Label' as a key and 'value' as a key?

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From ruediger.maehl at dzsh.de  Thu Feb  8 12:52:57 2001
From: ruediger.maehl at dzsh.de (Ruediger Maehl)
Date: Thu, 8 Feb 2001 12:52:57 +0100
Subject: dictionary sub-value lookup
References: <3A80F9B4.E072037D@millinfo.co.za>
Message-ID: <newscache$kysf8g$6ua$1@www.dzsh.de>

Richard schrieb in Nachricht <3A80F9B4.E072037D at millinfo.co.za>...
>I have an object M which contains N dictionaries. Inside
>dictionary N, is an identifier, Y, plus various other
>values, eg: Z.... is it possible to find M.Y.Z without
>looping through the whole of M, checking Y each time ?


Hi Richard,

do you mean something similar to the path of a file,
but only for internal objects?

If so, you should perhaps consider using XML for the representation
of your objects/data.
Have a look at: http://www.python.org/sigs/xml-sig/

R?diger






From stksee at yahoo.co.jp  Wed Feb  7 09:28:17 2001
From: stksee at yahoo.co.jp (stksee at yahoo.co.jp)
Date: Wed, 07 Feb 2001 17:28:17 +0900
Subject: How can I change newline code?
Message-ID: <3A810721.68294483@yahoo.co.jp>

It seems "\n" is treated as LF.

Instead, I want to set "\n" as "CRLF" in string object.
What should I do ?








From stephen_purcell at yahoo.com  Wed Feb  7 09:56:56 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 7 Feb 2001 09:56:56 +0100
Subject: How can I change newline code?
In-Reply-To: <3A810721.68294483@yahoo.co.jp>; from stksee@yahoo.co.jp on Wed, Feb 07, 2001 at 05:28:17PM +0900
References: <3A810721.68294483@yahoo.co.jp>
Message-ID: <20010207095656.A6730@freedom.puma-ag.com>

stksee at yahoo.co.jp wrote:
> 
> It seems "\n" is treated as LF.
> 
> Instead, I want to set "\n" as "CRLF" in string object.
> What should I do ?


'\n' will always be LF.  You should use 'os.linesep' to get the right
end-of-line character(s) for the machine running your program:

 >>> import os
 >>> os.linesep
 '\012'


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From dtulpan at free.fr  Wed Feb  7 11:08:44 2001
From: dtulpan at free.fr (kyrsa)
Date: Wed, 07 Feb 2001 10:08:44 GMT
Subject: need some help for sending a msg in python
Message-ID: <M89g6.809$pC5.1923174@nnrp6.proxad.net>

Hi,
I have big problem about this part  of the file :


   temp_mail.write('From: %s\nTo: %s\n' % (fr, values))
    temp_mail.write('MIME-Version: 1.0\n')
    temp_mail.write('Content-Type: text/plain; charset=iso-8859-1\n')
    temp_mail.write('Content-Transfer-Encoding: 8bit\n')
    temp_mail.write("Subject: %s\n" % subject)
    temp_mail.write('\n')

    currentLine = inputFile.readline()
    while (currentLine != ""):
    temp_mail.write(currentLine)
    currentLine = inputFile.readline()

    # Envoi du mail

temp_mail.write('--------------------------------------------------------\n'
)
    temp_mail.write('Pour vous desabonner de la liste de diffusion :\n')

temp_mail.write('http://www.domaine.com/listedif/desabon.php3?email='+values
+'\n')
    temp_mail.close()
    rc = os.system('%s %s < %s' % (sendmail_path, values, temp_mail_name))



when i'm testing it, it's writing :

  File "mail_list.py", line 87

temp_mail.write('http://www.domaine.com/listedif/desabon.php3?email='+values
+'\n')
    ^
SyntaxError: invalid syntax



Can you help  me  ???

Thx

Kyrsa




From claird at starbase.neosoft.com  Wed Feb  7 17:36:45 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 7 Feb 2001 10:36:45 -0600
Subject: need some help for sending a msg in python
References: <M89g6.809$pC5.1923174@nnrp6.proxad.net>
Message-ID: <43C7B0D7025A197A.45FA2931D58DFD5C.57F2A1F0A65DC462@lp.airnews.net>

In article <M89g6.809$pC5.1923174 at nnrp6.proxad.net>,
kyrsa <dtulpan at free.fr> wrote:
>Hi,
>I have big problem about this part  of the file :
			.
			.
			.
>    temp_mail.write('Pour vous desabonner de la liste de diffusion :\n')
>
>temp_mail.write('http://www.domaine.com/listedif/desabon.php3?email='+values
>+'\n')
>    temp_mail.close()
>    rc = os.system('%s %s < %s' % (sendmail_path, values, temp_mail_name))
>
>
>
>when i'm testing it, it's writing :
>
>  File "mail_list.py", line 87
>
>temp_mail.write('http://www.domaine.com/listedif/desabon.php3?email='+values
>+'\n')
>    ^
>SyntaxError: invalid syntax
			.
			.
			.
Is the leading white space on that line not consistent
with what's around it?  The only other explanation con-
sistent with the symptom you describe is that there's
an unprintable character someplace outside the quotes
in that statement.

-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From jmg at ecs.soton.ac.uk  Wed Feb  7 12:44:15 2001
From: jmg at ecs.soton.ac.uk (Jacek Generowicz)
Date: 07 Feb 2001 11:44:15 +0000
Subject: Variable depth of nesting
Message-ID: <g0k872soog.fsf@scumbag.ecs.soton.ac.uk>

Is it somehow possible to write loops with variable nesting depths?

For example, I would like to write something like the function
outlined below without having to use the if clauses. Particularly, I
would like to make it work for a GENERAL length of list_of_lists.

def hmm ( list_of_lists ):
    if len( list_of_lists ) == 1:
        for x in list_of_lists[0]:
            # do something
        return

    if len( list_of_lists ) == 2:
        for x in list_of_lists[0]:
            for y in list_of_lists[1]
            # do somehting
        return

    if len( list_of_lists ) == 3:
        for x . . . :
            for y . . . :
                for z . . . :
                    # do something
        return
                    
    if len( list_of_lists ) == 4:

    #etc.
    #etc.


Alternatively, how about list comprehensions with variable order:


def hmm ( l ):
    if len( l ) == 1:
        return [ x for x in l[0] ]

    if len( l ) == 2:
        return [ (x,y) for x in l[0] for y in l[1] ]

    if len( l ) == 3:
        return [ (x,y,z) for x in l[0] for y in l[1] for z in l[2] ]
                    
    #etc.
    #etc.


Any ideas ?

Thanks,

Jacek


From robin at jessikat.fsnet.co.uk  Wed Feb  7 13:28:35 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 7 Feb 2001 12:28:35 +0000
Subject: Variable depth of nesting
References: <g0k872soog.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <vXBaCIAz9Tg6EwEe@jessikat.fsnet.co.uk>

In article <g0k872soog.fsf at scumbag.ecs.soton.ac.uk>, Jacek Generowicz
<jmg at ecs.soton.ac.uk> writes
for the variable length list comprehension I tried

>>> def vlc(L):
...     v = ''
...     f = ''
...     for i in xrange(len(L)):
...             v += ',x%d'%i
...             f += ' for x%d in L[%d]' % (i,i)
...     return eval('[(%s) %s]' % (v[1:],f))
... 

and it seems to work
-- 
Robin Becker


From aahz at panix.com  Wed Feb  7 14:35:03 2001
From: aahz at panix.com (Aahz Maruch)
Date: 7 Feb 2001 05:35:03 -0800
Subject: Variable depth of nesting
References: <g0k872soog.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <95riu7$6f4$1@panix3.panix.com>

In article <g0k872soog.fsf at scumbag.ecs.soton.ac.uk>,
Jacek Generowicz  <jmg at ecs.soton.ac.uk> wrote:
>
>Is it somehow possible to write loops with variable nesting depths?

Try recursion.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Life as a contractor: when they tell you it's going to rain, sometimes
it drizzles and sometimes it pours.


From joconnor at cybermesa.com  Wed Feb  7 15:52:26 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Wed, 07 Feb 2001 14:52:26 GMT
Subject: Variable depth of nesting
References: <g0k872soog.fsf@scumbag.ecs.soton.ac.uk> <95riu7$6f4$1@panix3.panix.com>
Message-ID: <3a8160c3.25290480@news.cybermesa.com>

On 7 Feb 2001 05:35:03 -0800, aahz at panix.com (Aahz Maruch) wrote:

>In article <g0k872soog.fsf at scumbag.ecs.soton.ac.uk>,
>Jacek Generowicz  <jmg at ecs.soton.ac.uk> wrote:
>>
>>Is it somehow possible to write loops with variable nesting depths?
>
>Try recursion.

Here's an example that will recusively print the elements in a list

>>> import types
>>> def printList(aList):
	for item in aList:
		if type(item) == types.ListType:
			printList(item)
		else:
			print item

>>> printList ([1,2,3])
1
2
3
>>> printList ([1,2,3, [4,5,6]])
1
2
3
4
5
6

Take care,

Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum - http://pub1.ezboard.com/fobjectorienteddevelopmentpython

"God himself plays on the bass strings first, when he tunes the soul"


From andrew at intertrader.com  Wed Feb  7 15:30:42 2001
From: andrew at intertrader.com (Andrew Cooke)
Date: Wed, 07 Feb 2001 14:30:42 +0000
Subject: Variable depth of nesting
References: <g0k872soog.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <3A815C12.A375380F@intertrader.com>

I think this is "Power Loops" - see the first chapter (IIRC) of Finkel
at http://cseng.aw.com/book/related/0,3833,0805311912+20,00.html
(although I may be wrong as they never made much sense to me and my
brand-spanking-new firewall won't let me look at the book that URL
indicates any more because it's FTP).

Andrew

Jacek Generowicz wrote:
> 
> Is it somehow possible to write loops with variable nesting depths?
> 
> For example, I would like to write something like the function
> outlined below without having to use the if clauses. Particularly, I
> would like to make it work for a GENERAL length of list_of_lists.
> 
> def hmm ( list_of_lists ):
>     if len( list_of_lists ) == 1:
>         for x in list_of_lists[0]:
>             # do something
>         return
> 
>     if len( list_of_lists ) == 2:
>         for x in list_of_lists[0]:
>             for y in list_of_lists[1]
>             # do somehting
>         return
> 
>     if len( list_of_lists ) == 3:
>         for x . . . :
>             for y . . . :
>                 for z . . . :
>                     # do something
>         return
> 
>     if len( list_of_lists ) == 4:
> 
>     #etc.
>     #etc.
> 
> Alternatively, how about list comprehensions with variable order:
> 
> def hmm ( l ):
>     if len( l ) == 1:
>         return [ x for x in l[0] ]
> 
>     if len( l ) == 2:
>         return [ (x,y) for x in l[0] for y in l[1] ]
> 
>     if len( l ) == 3:
>         return [ (x,y,z) for x in l[0] for y in l[1] for z in l[2] ]
> 
>     #etc.
>     #etc.
> 
> Any ideas ?
> 
> Thanks,
> 
> Jacek


From erno-news at erno.iki.fi  Mon Feb 12 00:14:13 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 12 Feb 2001 01:14:13 +0200
Subject: Variable depth of nesting
References: <g0k872soog.fsf@scumbag.ecs.soton.ac.uk> <3A815C12.A375380F@intertrader.com>
Message-ID: <kulmrczube.fsf@lasipalatsi.fi>

In article <3A815C12.A375380F at intertrader.com>, Andrew Cooke
<andrew at intertrader.com> writes:

| I think this is "Power Loops" - see the first chapter (IIRC) of Finkel
| at http://cseng.aw.com/book/related/0,3833,0805311912+20,00.html
| (although I may be wrong as they never made much sense to me and my
| brand-spanking-new firewall won't let me look at the book that URL
| indicates any more because it's FTP).

thank you for the pointer! i've had a good time reading the book
(about half way through now).  the power loops section is actually in
chapter 2 (section 4). 

the book says:
  Although power loops are elegant, they are subsumed by recursive
  proce-dures, albeit with a loss of elegance and efficiency.

i'm not sure i agree about the elegance part either...

  -- erno


From mrq at for.mat.bham.ac.uk  Wed Feb  7 14:00:28 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Wed, 7 Feb 2001 13:00:28 +0000
Subject: Some basic questions about Tkinter (probably v easy for experts!)
Message-ID: <Pine.GSO.4.21.0102071252070.15785-100000@babbage>

Apologies if this gets overly long - I'm feeling extremely frustrated in
my efforts to get to grips with Tkinter and want to get a good grip on
what I understand and what I am getting wrong.

(1) If I am going to use Pmw, then I need to have Tkinter installed.  But
do I need to understand how to use Tkinter (at least to a basic
degree) before I can use Pmw?

(2) Should a typical Tkinter program contain the following sort of lines:

from import Tkinter *
aardvark = Tk()
aardvark.mainloop()

Am I right in thinking the last line is what produces the Tk window and
that without it nothing would happen?  Do I then just attach widgets the
aardvark object defined above and further widgets to these ones, etc.?
Alternatively, do I need to attach some sort of frame to aardvark first
before I can add buttons, menus, etc.?

(3) How do I run such a program?  Is the right way to open up a DOS box
(I'm using Windows 95) and type  python aardvark.py  at the prompt?

(4) Will the following correctly add a menu to my aardvark object?

from sys import exit
armadillo = Menu(aardvark)
armadillo.add_cascade(label="File")
armadillo.add_command(lable="Exit", command=exit)

If this isn't right - what is right, and more importantly *why* is it
right?

Thanks in advance for your patience and wisdom...

Martyn

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick



From cerutti at together.net  Wed Feb  7 14:41:56 2001
From: cerutti at together.net (Neil Cerutti)
Date: 7 Feb 2001 13:41:56 GMT
Subject: Some basic questions about Tkinter (probably v easy for experts!)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage>
Message-ID: <slrn982kj2.3k.cerutti@fiad06.norwich.edu>

Martyn Quick posted:
>(1) If I am going to use Pmw, then I need to have Tkinter
>installed.  But do I need to understand how to use Tkinter (at
>least to a basic degree) before I can use Pmw?

Yes, definitely. Pmw adds more widgets to your toolbox, but they
act like Tkinter widgets in every way.

For the rest of your questions, the easiest thing will be to
read:

 http://www.pythonware.com/library/tkinter/introduction/

In particular, read Section 1, "Introducing Tkinter". But don't
only read it; also open a Python interpreter (a simple CLI will
work best), and type in the examples as they appear in the
manual.

-- 
Neil Cerutti <cerutti at together.net>


From mrq at for.mat.bham.ac.uk  Wed Feb  7 15:40:45 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Wed, 7 Feb 2001 14:40:45 +0000
Subject: Some basic questions about Tkinter (probably v easy for experts!)
In-Reply-To: <slrn982kj2.3k.cerutti@fiad06.norwich.edu>
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <slrn982kj2.3k.cerutti@fiad06.norwich.edu>
Message-ID: <Pine.GSO.4.21.0102071438280.23402-100000@babbage>

On 7 Feb 2001, Neil Cerutti wrote:

> For the rest of your questions, the easiest thing will be to
> read:
> 
>  http://www.pythonware.com/library/tkinter/introduction/

As I said in my reply to Fredrik Lundh's answer, I did *try* to read this,
but unfortunately I found it rather difficult.  It doesn't actually seem
to be aimed at beginners who are new to Object Oriented Programming.

> In particular, read Section 1, "Introducing Tkinter". But don't
> only read it; also open a Python interpreter (a simple CLI will
> work best), and type in the examples as they appear in the
> manual.

What's a CLI?

Martyn

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick





From cerutti at together.net  Wed Feb  7 17:13:03 2001
From: cerutti at together.net (Neil Cerutti)
Date: 7 Feb 2001 16:13:03 GMT
Subject: Some basic questions about Tkinter (probably v easy for experts!)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <slrn982kj2.3k.cerutti@fiad06.norwich.edu> <Pine.GSO.4.21.0102071438280.23402-100000@babbage>
Message-ID: <slrn982tec.7d.cerutti@fiad06.norwich.edu>

Martyn Quick posted:
>On 7 Feb 2001, Neil Cerutti wrote:
>
>> For the rest of your questions, the easiest thing will be to
>> read:
>> 
>>  http://www.pythonware.com/library/tkinter/introduction/
>
>As I said in my reply to Fredrik Lundh's answer, I did *try* to
>read this, but unfortunately I found it rather difficult.  It
>doesn't actually seem to be aimed at beginners who are new to
>Object Oriented Programming.

Oh! That'll make it difficult to follow. Sorry. Using Tkinter
isn't a good way to learn OOP.

Luckily, the semantics and syntax of OOP is easy to learn
in Python, if not the practice.

>> In particular, read Section 1, "Introducing Tkinter". But don't
>> only read it; also open a Python interpreter (a simple CLI will
>> work best), and type in the examples as they appear in the
>> manual.
>
>What's a CLI?

It stands for Command Line Interface. I meant to imply that you
should use, e.g., on Windows, python.exe instead of IDLE (the
graphical integrated development environment that comes with some
versions of Python). Tkinter acts funny under IDLE, since IDLE is
written in Tkinter.

Typing along with the demos, seeing what happens as each line
is entered, should be illuminating. It's also kind of fun.

Here's a small demo. I'm not going to use "import *" here, to
hopefully make it a little clearer. You don't need to type in the
comments. ;-)

import Tkinter

# Create the root window. All widgets and other windows will be
# controlled by this window. Tk is a "class" in the Tkinter
# library. root becomes an "instance", or instantiation, of the
# class Tk after this line is run.
root = Tkinter.Tk()

# Use the title method of the window object to set the text in the
# title bar of the window.
root.title("Greetings")

# Create an instance of Tkinter Button class, and call it
# hello_button. The Button class needs some parameters in order
# to be created properly. The first parameter tells the button
# who its "master", or controller, is. The next argument contains
# both a keyword, and its value. "text" is an option of the
# Tkinter button widget; its value becomes the text printed on
# the button. Only the first parameter (root) is required, but
# including keyword arguments is the quickest way to set the
# initial options for new widgets.
hello_button = Tkinter.Button(root, text="Hello World")

# The button has been created, but it doesn't know where to
# appear in its master. For that, you must use a "geometry
# manager". There are several available in Tkinter, but the
# easiest is "pack". "pack" is a powerful but squinchy geometry
# manager that always uses the minimum possible space.
hello_button.pack()

# There will be a button, but it doesn't do anything. You can set
# the "command" option of your button widget instance to tell
# Tkinter to run arbitrary code when the button is pressed. I
# could have set this option in the line above when the button
# was created; but since the button is already in existence, I
# use the "config" method.

# This is just a function.
def change_hello_message():
  hello_button.config(text="Salutations Planet")

hello_button.config(command=change_hello_message)

# The window won't appear and your other widgets won't function
# until:
root.mainloop()


Have fun.

-- 
Neil Cerutti <cerutti at together.net>


From mrq at for.mat.bham.ac.uk  Thu Feb  8 11:28:23 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Thu, 8 Feb 2001 10:28:23 +0000
Subject: Some basic questions about Tkinter (probably v easy for experts!)
In-Reply-To: <slrn982tec.7d.cerutti@fiad06.norwich.edu>
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <slrn982kj2.3k.cerutti@fiad06.norwich.edu> <Pine.GSO.4.21.0102071438280.23402-100000@babbage> <slrn982tec.7d.cerutti@fiad06.norwich.edu>
Message-ID: <Pine.GSO.4.21.0102081019360.20775-100000@babbage>

On 7 Feb 2001, Neil Cerutti wrote:

> >As I said in my reply to Fredrik Lundh's answer, I did *try* to
> >read this, but unfortunately I found it rather difficult.  It
> >doesn't actually seem to be aimed at beginners who are new to
> >Object Oriented Programming.
> 
> Oh! That'll make it difficult to follow. Sorry. Using Tkinter
> isn't a good way to learn OOP.
> 
> Luckily, the semantics and syntax of OOP is easy to learn
> in Python, if not the practice.

I *think* I'm beginning to get the idea of OOP - it just doesn't come as
second nature yet.  Thanks very much for your example below - I have one
question about it (apologies!) but I've managed to make a start on writing
the code I want.  My own code isn't very OOP yet - but I've put lots of
comments in to remind me to adjust it when I've got a better idea of what
will make sensible classes, etc.

> Here's a small demo. I'm not going to use "import *" here, to
> hopefully make it a little clearer. You don't need to type in the
> comments. ;-)
> 
> [snip example]
>
> # The window won't appear and your other widgets won't function
> # until:
> root.mainloop()

OK, I followed that example and it was very helpful - *however* I think I
am misunderstanding your final comment here.  When I typed the example
into python running in a DOSbox window, the window & button did appear and
did do what they were supposed to before I used root.mainloop().  This
seems contrary to what you've written above.  (Perhaps when I run the
thing as a script I need the mainloop() but less so when it's done at
command prompt??).

Once more, thank you very much for your answer Neil, I found it extremely
helpful.

Martyn


--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick





From cerutti at together.net  Thu Feb  8 17:29:58 2001
From: cerutti at together.net (Neil Cerutti)
Date: 8 Feb 2001 16:29:58 GMT
Subject: Some basic questions about Tkinter (probably v easy for experts!)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <slrn982kj2.3k.cerutti@fiad06.norwich.edu> <Pine.GSO.4.21.0102071438280.23402-100000@babbage> <slrn982tec.7d.cerutti@fiad06.norwich.edu> <Pine.GSO.4.21.0102081019360.20775-100000@babbage>
Message-ID: <slrn985iq8.7m.cerutti@fiad06.norwich.edu>

Martyn Quick posted:
>On 7 Feb 2001, Neil Cerutti wrote:
>
>> >As I said in my reply to Fredrik Lundh's answer, I did *try* to
>> >read this, but unfortunately I found it rather difficult.  It
>> >doesn't actually seem to be aimed at beginners who are new to
>> >Object Oriented Programming.
>> 
>> Oh! That'll make it difficult to follow. Sorry. Using Tkinter
>> isn't a good way to learn OOP.
>> 
>> Luckily, the semantics and syntax of OOP is easy to learn
>> in Python, if not the practice.
>
>I *think* I'm beginning to get the idea of OOP - it just doesn't
>come as second nature yet.  Thanks very much for your example
>below - I have one question about it (apologies!) but I've
>managed to make a start on writing the code I want.  My own code
>isn't very OOP yet - but I've put lots of comments in to remind
>me to adjust it when I've got a better idea of what will make
>sensible classes, etc.

Just remember the syntax. And use Fredrik Lundh's docs to find
out what options, methods, and especially programming patterns
are in there for the widgets you use.

>> Here's a small demo. I'm not going to use "import *" here, to
>> hopefully make it a little clearer. You don't need to type in the
>> comments. ;-)
>> 
>> [snip example]
>>
>> # The window won't appear and your other widgets won't function
>> # until:
>> root.mainloop()
>
>OK, I followed that example and it was very helpful - *however*
>I think I am misunderstanding your final comment here.  When I
>typed the example into python running in a DOSbox window, the
>window & button did appear and did do what they were supposed to
>before I used root.mainloop().  This seems contrary to what
>you've written above.  (Perhaps when I run the thing as a script
>I need the mainloop() but less so when it's done at command
>prompt??).

I'm confused about this issue too, I guess.

-- 
Neil Cerutti <cerutti at together.net>


From fredrik at pythonware.com  Thu Feb  8 19:33:03 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 08 Feb 2001 18:33:03 GMT
Subject: Some basic questions about Tkinter (probably v easy for experts!)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <slrn982kj2.3k.cerutti@fiad06.norwich.edu> <Pine.GSO.4.21.0102071438280.23402-100000@babbage> <slrn982tec.7d.cerutti@fiad06.norwich.edu> <Pine.GSO.4.21.0102081019360.20775-100000@babbage> <slrn985iq8.7m.cerutti@fiad06.norwich.edu>
Message-ID: <zDBg6.13591$AH6.2042144@newsc.telia.net>

Martin Quick/Neil Cerutti:
> >OK, I followed that example and it was very helpful - *however*
> >I think I am misunderstanding your final comment here.  When I
> >typed the example into python running in a DOSbox window, the
> >window & button did appear and did do what they were supposed to
> >before I used root.mainloop().  This seems contrary to what
> >you've written above.
>
> I'm confused about this issue too, I guess.

Under some versions/builds/platforms/phases-of-the-moon, Tkinter
uses the PyOS_EventHook mechanism to keep the event loop running
under the interactive prompt.

> >(Perhaps when I run the thing as a script
> >I need the mainloop() but less so when it's done at command
> >prompt??).

Exactly.

Cheers /F




From fredrik at effbot.org  Wed Feb  7 14:42:00 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 07 Feb 2001 13:42:00 GMT
Subject: Some basic questions about Tkinter (probably v easy for experts!)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage>
Message-ID: <Igcg6.13212$AH6.1990695@newsc.telia.net>

Martyn Quick wrote:
> (1) If I am going to use Pmw, then I need to have Tkinter installed.  But
> do I need to understand how to use Tkinter (at least to a basic
> degree) before I can use Pmw?

cannot hurt -- Pmw just adds (a whole bunch of) new
components to Tkinter.

> (2) Should a typical Tkinter program contain the following sort of lines:
>
> from import Tkinter *
> aardvark = Tk()
> aardvark.mainloop()
>
> Am I right in thinking the last line is what produces the Tk window and
> that without it nothing would happen?  Do I then just attach widgets the
> aardvark object defined above and further widgets to these ones, etc.?

have you read the introduction document?  the first chapter
seems to answer your questions:

http://www.pythonware.com/library/tkinter/introduction/hello-tkinter.htm

> (3) How do I run such a program?  Is the right way to open up a DOS box
> (I'm using Windows 95) and type  python aardvark.py  at the prompt?

fwiw, that's what I usually do (when I'm not using PythonWorks).

> (4) Will the following correctly add a menu to my aardvark object?

short answer: almost.

> from sys import exit
> armadillo = Menu(aardvark)

this creates a menu component instance.

> armadillo.add_cascade(label="File")

this adds a submenu to the admadillo menubar.  but you
didn't pass in the submenu instance, so there's nothing to
display when the user clicks on "File"...

try this instead:

    filemenu = Menu(armadillo) # create a submenu
    armadillo.add_cascade(label="File", menu=filemenu)

> armadillo.add_command(lable="Exit", command=exit)

if you fix the typo, this adds a command item to the armadillo
menubar.  when that item is selected, Tkinter calls sys.exit.

if you want to add it to the file menu instead of the armadillo
menubar, try this:

    filemenu.add_command(label="Exit", command=exit)

finally, you need to attach the menu to your window:

    aardvark.config(menu=armadillo)

> If this isn't right - what is right, and more importantly *why* is it
> right?

http://www.pythonware.com/library/tkinter/introduction/x953-menus.htm

hope this helps!

Cheers /F




From mrq at for.mat.bham.ac.uk  Wed Feb  7 15:02:43 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Wed, 7 Feb 2001 14:02:43 +0000
Subject: Some basic questions about Tkinter (probably v easy for experts!)
In-Reply-To: <Igcg6.13212$AH6.1990695@newsc.telia.net>
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net>
Message-ID: <Pine.GSO.4.21.0102071351520.20781-100000@babbage>

On Wed, 7 Feb 2001, Fredrik Lundh wrote:

> have you read the introduction document?  the first chapter
> seems to answer your questions:

I've tried to read the introduction document, but I found it rather
confusing, I'm afraid.  I'm very new to this game and the use of lowercase
and uppercase variants of the same English word to denote almost the same
thing caused me problems.   It is to avoid the confusion that I introduced
aardvark and armadillo as the names for my widgets below.

> > (4) Will the following correctly add a menu to my aardvark object?
> 
> short answer: almost.
> 
> > from sys import exit
> > armadillo = Menu(aardvark)
> 
> this creates a menu component instance.
> 
> > armadillo.add_cascade(label="File")
> 
> this adds a submenu to the admadillo menubar.  but you
> didn't pass in the submenu instance, so there's nothing to
> display when the user clicks on "File"...
> 
> try this instead:
> 
>     filemenu = Menu(armadillo) # create a submenu
>     armadillo.add_cascade(label="File", menu=filemenu)

Hold it - is "filemenu" a user defined name or is it built into Tkinter?
Would the following work equally well?

	dongle = Menu(armadillo)
	armadillo.add_cascade(label="File", menu=dongle)

> > armadillo.add_command(lable="Exit", command=exit)
> 
> if you fix the typo, this adds a command item to the armadillo
> menubar.  when that item is selected, Tkinter calls sys.exit.

Oh good.  (And "Oops!" about the typo.)

> if you want to add it to the file menu instead of the armadillo
> menubar, try this:
> 
>     filemenu.add_command(label="Exit", command=exit)
> 
> finally, you need to attach the menu to your window:
> 
>     aardvark.config(menu=armadillo)

What *exactly* does config do?
The Introduction document (p26) says "We use the config method to attach
it to the root window".  But I couldn't find anywhere telling me what the
config method was.

Furthermore in the example on p26 of the document it has the following
code:

menu = Menu(root)
root.config(menu=menu)

But what happens if I call the menu bar on my root "aardvark"?
Say

aardvark = Menu(root)

Does the next line contain "menu=aardvark", "aardvark=menu", or
"aardvark=aardvark"?

> 
> > If this isn't right - what is right, and more importantly *why* is it
> > right?
> 
> http://www.pythonware.com/library/tkinter/introduction/x953-menus.htm

Ah yes - this is precisely the printed page that I have in front of me
that I'm not really understanding.

A related question is the Second Tkinter Program in that document (Hello,
Again).  The command use to quite is  frame.quit  .  Why is this used
instead of  sys.exit()  ?    Am I right in thinking that  frame  is the
user defined name for the Frame, so if I wrote

aardvark = Frame(master)

I would use  command = aardvark.quit


Thanks for the previous reply!

Martyn


--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick






From fredrik at effbot.org  Wed Feb  7 15:56:03 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 07 Feb 2001 14:56:03 GMT
Subject: Some basic questions about Tkinter (and Python)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102071351520.20781-100000@babbage>
Message-ID: <7mdg6.13217$AH6.1992583@newsc.telia.net>

Martyn Quick wrote:
> I'm very new to this game and the use of lowercase and
> uppercase variants of the same English word to denote
> almost the same thing caused me problems.

Python is case-sensitive.  You better get used to it ;-)

by convention, "foo" is a variable (a named reference to
an object) or a function/method, "Foo" is a class, and
"FOO" is a constant.

(functions, methods, classes and constants are objects
too, of course)

> >     filemenu = Menu(armadillo) # create a submenu
> >     armadillo.add_cascade(label="File", menu=filemenu)
>
> Hold it - is "filemenu" a user defined name or is it built into Tkinter?

"filemenu" is just a variable name.  If you assign a value
to a name (or to be more precise, if you bind a name to
a value), you "own" that name.

(note that the Tkinter introduction assumes that you can
read Python code fluently, and that you have a basic under-
standing of Python's object and class model).

> Would the following work equally well?
>
> dongle = Menu(armadillo)
> armadillo.add_cascade(label="File", menu=dongle)

yes.

> The Introduction document (p26) says "We use the config method to attach
> it to the root window".  But I couldn't find anywhere telling me what the
> config method was.

http://www.pythonware.com/library/tkinter/introduction/widget-configuration.htm

the syntax is:

    widget.config(option=value)

Where "option" is the configuration option you want
to change, and "value" is the value you want to set it
to.  You can also set several options per call:

    widget.config(option1=value1, option2=value2, ...)

Also read up on keyword arguments (a Python feature).

> A related question is the Second Tkinter Program in that document (Hello,
> Again).  The command use to quite is  frame.quit  .  Why is this used
> instead of  sys.exit()  ?

quit causes Tkinter to return from the call to mainloop, while
sys.exit terminates the Python interpreter.  If calling mainloop
is the last thing you do in your program, it doesn't really matter
which one you use...

> Am I right in thinking that  frame  is the user defined name for
> the Frame, so if I wrote

it's a variable name (a named reference to the widget, not
necessarily the name of the Frame instance itself.  see the
section on widget names for more info)

>
> aardvark = Frame(master)
>
> I would use  command = aardvark.quit

yes.

Cheers /F






From claird at starbase.neosoft.com  Wed Feb  7 16:36:58 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 7 Feb 2001 09:36:58 -0600
Subject: Some basic questions about Tkinter (and Python)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102071351520.20781-100000@babbage> <7mdg6.13217$AH6.1992583@newsc.telia.net>
Message-ID: <5A26DA93165A7B85.352CA80CD738363E.9B52D4C723B787A0@lp.airnews.net>

In article <7mdg6.13217$AH6.1992583 at newsc.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
			.
			.
			.
>(note that the Tkinter introduction assumes that you can
>read Python code fluently, and that you have a basic under-
>standing of Python's object and class model).
			.
			.
			.
So I guess the answer is that we don't have a
Tkinter intro for those new to Python ...  Per-
haps it's best to recommend that Mr. Quick
simply buy John Grayson's book without more
delay.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From mrq at for.mat.bham.ac.uk  Thu Feb  8 18:38:19 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Thu, 8 Feb 2001 17:38:19 +0000
Subject: Some basic questions about Tkinter (and Python)
In-Reply-To: <5A26DA93165A7B85.352CA80CD738363E.9B52D4C723B787A0@lp.airnews.net>
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102071351520.20781-100000@babbage> <7mdg6.13217$AH6.1992583@newsc.telia.net> <5A26DA93165A7B85.352CA80CD738363E.9B52D4C723B787A0@lp.airnews.net>
Message-ID: <Pine.GSO.4.21.0102081734460.16415-100000@babbage>

On 7 Feb 2001, Cameron Laird wrote:

> So I guess the answer is that we don't have a
> Tkinter intro for those new to Python ...  Per-
> haps it's best to recommend that Mr. Quick
> simply buy John Grayson's book without more
> delay.

You're not the first person to recommend me this book on this
newsgroup.  Unfortunately, we don't seem to have it in the library at the
university where I work, as otherwise it would be in my hands by now.  I
know you say buy the book, but I can't afford that - perhaps you earn more
than I do?

I have a friend who is a computer programmer (and definitely earns more
than I do ;-)  who recommended that I give python a try.  He is going to
lend me his copy of a book on python on Saturday - Hammond & Robinson,
"Python Programming on Win32".  I don't know whether this has sufficient
info on Tkinter to help, but it's about the only non-online source that I
can get hold of.

Martyn

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick





From claird at starbase.neosoft.com  Fri Feb  9 18:21:28 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 9 Feb 2001 11:21:28 -0600
Subject: Some basic questions about Tkinter (and Python)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <7mdg6.13217$AH6.1992583@newsc.telia.net> <5A26DA93165A7B85.352CA80CD738363E.9B52D4C723B787A0@lp.airnews.net> <Pine.GSO.4.21.0102081734460.16415-100000@babbage>
Message-ID: <29D1A1897E293522.BDB05FE58D09D409.B94B74AE507E733D@lp.airnews.net>

In article <Pine.GSO.4.21.0102081734460.16415-100000 at babbage>,
Martyn Quick  <mrq at for.mat.bham.ac.uk> wrote:
			.
			.
			.
>You're not the first person to recommend me this book on this
>newsgroup.  Unfortunately, we don't seem to have it in the library at the
>university where I work, as otherwise it would be in my hands by now.  I
>know you say buy the book, but I can't afford that - perhaps you earn more
>than I do?
As your countryman David Ricardo so vividly
expressed it, it's the *comparative* advan-
tage, not absolute income, which best explains
our behavior.

Does your University not offer inter-library
loan services?

When I write, "buy", that abbreviates an
entire constellation of assumptions and reali-
zations.  Most bluntly, if it's worth the
expense to the kingdom to keep you titled as
a Research Fellow in Mathematics, then it's
worth it to someone to ensure a book which 
will benefit you so much reaches your hands
safely.

If no one else steps forward, Dr. Quick, I
will; write me privately and I'll arrange for
a copy to reach you.
>
>I have a friend who is a computer programmer (and definitely earns more
>than I do ;-)  who recommended that I give python a try.  He is going to
>lend me his copy of a book on python on Saturday - Hammond & Robinson,
>"Python Programming on Win32".  I don't know whether this has sufficient
>info on Tkinter to help, but it's about the only non-online source that I
>can get hold of.
This book has many virtues.  Its coverage of
Tkinter occupies ten pages, and is slightly
outdated.  It uses those ten pages to good
advantage, though.
			.
			.
			.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From SBrunning at trisystems.co.uk  Fri Feb  9 10:29:25 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 9 Feb 2001 09:29:25 -0000 
Subject: Some basic questions about Tkinter (and Python)
Message-ID: <31575A892FF6D1118F5800600846864D5B17D7@intrepid>

> From:	Martyn Quick [SMTP:mrq at for.mat.bham.ac.uk]
> I have a friend who is a computer programmer (and definitely earns more
> than I do ;-)  who recommended that I give python a try.  He is going to
> lend me his copy of a book on python on Saturday - Hammond & Robinson,
> "Python Programming on Win32".  I don't know whether this has sufficient
> info on Tkinter to help, but it's about the only non-online source that I
> can get hold of.
 
This book doesn't cover Tkinter in any detail, I'm afraid. (5 or 6 pages,
IIRC.) *Very* good book, but it doesn't *intend* to cover everything - it
concentrates on Win 32 stuff.

FWIW, I find Grayson's book *very* tough going. I have *no* experience of
GUI building at all, and there isn't much in the way of gentle introduction.
It's probably a good book for learning Tkinter if you already know Python
pretty well and know how to build a GUI. What *I* need is a gentle
introduction to the art of building an event driver GUI using Tkinter, and I
haven't found one yet.

BTW, I picked up Andre Lessa's 'Python Developers Handbook' at the weekend
(Sams, ?33) - it has a very brief intro to Python, but then moves on to
cover a wide range of advanced topics. Looks good so far. This has a section
on GUI building with Tkinter, which I haven't reached yet. I'll report back
when I've finished, and had a play.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From mrq at for.mat.bham.ac.uk  Thu Feb  8 11:33:45 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Thu, 8 Feb 2001 10:33:45 +0000
Subject: Some basic questions about Tkinter (probably v easy for experts!)
In-Reply-To: <Igcg6.13212$AH6.1990695@newsc.telia.net>
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net>
Message-ID: <Pine.GSO.4.21.0102081029430.20775-100000@babbage>

On Wed, 7 Feb 2001, Fredrik Lundh wrote:
> 
> have you read the introduction document?  the first chapter
> seems to answer your questions:
> 
> http://www.pythonware.com/library/tkinter/introduction/hello-tkinter.htm

Thanks very much for your answer Fredrik - I managed to get some menus
working last night, so I think I'm starting to get the hang of
things.  I'll try having a thorough re-read of the Introduction now that
I'm getting an idea of the syntax.

> > armadillo.add_command(lable="Exit", command=exit)
> 
> if you fix the typo, this adds a command item to the armadillo
> menubar.  when that item is selected, Tkinter calls sys.exit.

I've realized one point that was confusing me.
When I use sys.exit to finish a python script I would type

sys.exit()

but in the above you use  exit  rather than  exit().
Is it generally the case that you don't use the brackets in the
comannd=...  part?  (Sorry if this is easily to be found in the
Introduction document!)  Presumably this means that I can't use a command
which relies on parameters?

Thanks once again...

Martyn

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick





From fredrik at pythonware.com  Thu Feb  8 19:33:04 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 08 Feb 2001 18:33:04 GMT
Subject: Some basic questions about Tkinter (probably v easy for experts!)
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102081029430.20775-100000@babbage>
Message-ID: <ADBg6.13592$AH6.2042423@newsc.telia.net>

Martyn Quick wrote:
> I've realized one point that was confusing me.
> When I use sys.exit to finish a python script I would type
>
> sys.exit()
>
> but in the above you use  exit  rather than  exit().
> Is it generally the case that you don't use the brackets in the
> comannd=...  part?

not unless you want to *call* the function.

> (Sorry if this is easily to be found in the Introduction document!)

it's a Python thing, not a Tkinter thing.  functions (and
methods, classes, modules, etc) are objects -- just like
integers, strings, and other simple types:

to call an object, add parentheses:

>>> import sys
>>> sys # this name points to a module object
<module 'sys' (built-in)>
>>> sys.exit # this name points to a function
<built-in function exit>
>>> x = sys.exit # add another name for the same function
>>> x
<built-in function exit>
>>> sys.exit() # call the function
$

(after the assignment, you could also have typed x())

so in the following statement, we're passing the *function*
itself as the command value:

    widget.config(command=sys.exit)

while in the following statement, we're *calling* the function.
if the function returns (sys.exit() doesn't), the return value is
passed on to the widget:

    widget.config(command=sys.exit())

(functions can return function objects, so this might actually
make sense in some cases.  but not for sys.exit() ;-)

> Presumably this means that I can't use a command which relies on
> parameters?

not directly -- there's something called lambdas (in Python, not
in Tkinter), but it's usually better to use a plain function:

    def myfunc():
        otherfunc(1, 2, 3)

    b = Button(master, command=myfunc)

Cheers /F




From claird at starbase.neosoft.com  Fri Feb  9 18:25:22 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 9 Feb 2001 11:25:22 -0600
Subject: Tkinter idioms (was: Some basic questions about Tkinter (probably v easy for experts!))
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102081029430.20775-100000@babbage> <ADBg6.13592$AH6.2042423@newsc.telia.net>
Message-ID: <50B7C024B0AA609B.8CB8E2F63E2849E7.2783498FE2A00925@lp.airnews.net>

In article <ADBg6.13592$AH6.2042423 at newsc.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
			.
			.
			.
>not directly -- there's something called lambdas (in Python, not
>in Tkinter), but it's usually better to use a plain function:
>
>    def myfunc():
>        otherfunc(1, 2, 3)
>
>    b = Button(master, command=myfunc)
>
>Cheers /F
>
>

Tell me if this is pertinent--I find myself writing a
lot of formulaic lambdas in my Tkinter compositions.
Example:  a Button command has a no-arg signature, but
when I want to gang some other binding to the same com-
mand, I have to write a trivial lambda to ignore the
supplied event.  Is there something I'm missing?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From johngrayson at home.com  Sun Feb 11 01:01:38 2001
From: johngrayson at home.com (John Grayson)
Date: Sun, 11 Feb 2001 00:01:38 GMT
Subject: Tkinter idioms (was: Some basic questions about Tkinter (probably v easy for experts!))
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102081029430.20775-100000@babbage> <ADBg6.13592$AH6.2042423@newsc.telia.net> <50B7C024B0AA609B.8CB8E2F63E2849E7.2783498FE2A00925@lp.airnews.net>
Message-ID: <964kov$tbl$1@nnrp1.deja.com>

In article
<50B7C024B0AA609B.8CB8E2F63E2849E7.2783498FE2A00925 at lp.airnews.net>,
  claird at starbase.neosoft.com (Cameron Laird) wrote:
> In article <ADBg6.13592$AH6.2042423 at newsc.telia.net>,
> Fredrik Lundh <fredrik at effbot.org> wrote:
>
	.
>
	.
>
	.
> >not directly -- there's something called lambdas (in Python, not
> >in Tkinter), but it's usually better to use a plain function:
> >
> >    def myfunc():
> >        otherfunc(1, 2, 3)
> >
> >    b = Button(master, command=myfunc)
> >
> >Cheers /F
> >
> >
>
> Tell me if this is pertinent--I find myself writing a
> lot of formulaic lambdas in my Tkinter compositions.
> Example:  a Button command has a no-arg signature, but
> when I want to gang some other binding to the same com-
> mand, I have to write a trivial lambda to ignore the
> supplied event.  Is there something I'm missing?
> --
>
> Cameron Laird <claird at NeoSoft.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html
>

It is a common practice to put a default arg as the first
argument of a function, then you can call the function
from an event or a command:

def buttonAction(e=None):
    ....

or:

    def buttonAction(self, e=None):
        ....


(Page 106 PTkP)

    John


Sent via Deja.com
http://www.deja.com/


From nvithadt at bellsouth.net  Wed Feb  7 14:48:51 2001
From: nvithadt at bellsouth.net (nvithadt at bellsouth.net)
Date: Wed, 7 Feb 2001 08:48:51 -0500
Subject: using external dll's in Python 2.0
Message-ID: <kncg6.892$EA5.74766@news1.atl>

Hey folk,
    I want to be able to use some 3rd party dll in a python 2.0 program I'm writing. I've searched through the message threads and came across a solution using something called calldll that was available on the parnassus(sp?) site. However, this seems to over useful for python 1.5 . After modifying the makefile script and using vc++ 6.0 I can get a calldll.pyd to be created but whenever I try to import it into my project python dies. I'm fairly new to python as well. I was hoping maybe in the win32 extension libraries there would be some info but alas no. There is LoadLibrary and GetProcAddress but seems to be no way to execute a function pointer. I really would hope I do not have to write a python extension module since there are a lot of functions in the dll's that I would like to use.

TIA, Jason
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010207/1c6cae76/attachment.htm>

From glandrum at my-deja.com  Wed Feb  7 16:02:19 2001
From: glandrum at my-deja.com (Greg Landrum)
Date: Wed, 07 Feb 2001 15:02:19 GMT
Subject: using external dll's in Python 2.0
References: <kncg6.892$EA5.74766@news1.atl>
Message-ID: <95ro1k$fg7$1@nnrp1.deja.com>

In article <kncg6.892$EA5.74766 at news1.atl>,
  <nvithadt at bellsouth.net> wrote:
>     I want to be able to use some 3rd party dll in a python 2.0
program =
> I'm writing. I've searched through the message threads and came
across a =
> solution using something called calldll that was available on the =
> parnassus(sp?) site. However, this seems to over useful for python
1.5 . =

There is a binary build of calldll for python 2.0 available from:
http://www.nightmare.com/software.html

-greg



Sent via Deja.com
http://www.deja.com/


From raikov at cc.gatech.edu  Wed Feb  7 15:09:55 2001
From: raikov at cc.gatech.edu (Ivan Raikov)
Date: 07 Feb 2001 09:09:55 -0500
Subject: Python 1.5.2 and threads under SCO OpenServer 5.0.4
Message-ID: <m3hf26inyk.fsf@fkth15.eastnet.gatech.edu>

Greetings,


        I've been trying to get Python 1.5.2 to compile with thread
support under SCO OpenServer 5.0.4 (i586-pc-sco3.2v5.0.4), and finally
managed to get it (almost) working after I compiled it with GCC 2.95
and linked it against the GNU Pthreads library (pth-1.3.7).

        The problem is, that although most of the threading examples
in the 'Demo/threads' directory appear to be working, the script
called 'telnet.py' is having problems: once it connects, it receives
the login prompt from the server, but after the user name is sent, it
never receives the server's responses -- apparently after the
listening thread is created. 

        I also haven't been able to get Zope to work; Medusa
appears to be running, and it does return an error message when I send
an invalid request, but for some reason I can never get a response
from Zope itself.

        So I was wondering if anyone has had similar difficulties
under this platform, and knows how to go about resolving them. Any
help is much appreciated.


        Thanks,
        Ivan Raikov


-- 
Ivan Raikov 
800 Fowler Street, Atlanta, Georgia 30313-2554
Phone: 404-892-6579 x. 106
Email: raikov at cc.gatech.edu


From michael at stroeder.com  Wed Feb  7 15:41:25 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Wed, 07 Feb 2001 15:41:25 +0100
Subject: python-crypto list hosted at SURFnet.nl
Message-ID: <3A815E95.5399B6C3@stroeder.com>

HI!

I'm glad to announce that SURFnet in the Netherlands agreed on
providing the service to host the mailing list for crypto-related
Python development:

  PYTHON-CRYPTO at NIC.SURFNET.NL

The reason for this new mailing list is that the situation regarding
export regulations for cryptographic code from the U.S. is still not
clear enough. IMHO we can avoid a problems with crypto code posted
to the mailing list by hosting in a country known to be very liberal
in this field.

Please join the list at:

  http://listserv.surfnet.nl/archives/python-crypto.html

Hope to see you there.

Citizens of the U.S. or any other country with export regulations
for cryptographic software have to take care about themselves for
not getting into trouble by posting code to this list.

Ciao, Michael.


From dan at eevolved.com  Wed Feb  7 18:30:55 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 07 Feb 2001 17:30:55 GMT
Subject: python-crypto list hosted at SURFnet.nl
References: <3A815E95.5399B6C3@stroeder.com>
Message-ID: <jDfg6.71100$Pm2.1445232@news20.bellglobal.com>

Michael Str?der wrote:

> HI!
> 
> I'm glad to announce that SURFnet in the Netherlands agreed on
> providing the service to host the mailing list for crypto-related
> Python development:
> 
>   PYTHON-CRYPTO at NIC.SURFNET.NL
> 
> The reason for this new mailing list is that the situation regarding
> export regulations for cryptographic code from the U.S. is still not
> clear enough. IMHO we can avoid a problems with crypto code posted
> to the mailing list by hosting in a country known to be very liberal
> in this field.
> 
> Please join the list at:
> 
>   http://listserv.surfnet.nl/archives/python-crypto.html
> 
> Hope to see you there.
> 
> Citizens of the U.S. or any other country with export regulations
> for cryptographic software have to take care about themselves for
> not getting into trouble by posting code to this list.
> 
> Ciao, Michael.

I know you are not a legal expert, but it seems you have more knowledge 
than me in this area. Does this export control only matter to US citizens? 
What if crypto code written by Canadian citizens is hosted on a US web 
server? (PyAES on sourceforge.net)

thanks,
Dan



From michael at stroeder.com  Wed Feb  7 18:53:42 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Wed, 07 Feb 2001 18:53:42 +0100
Subject: python-crypto list hosted at SURFnet.nl
References: <3A815E95.5399B6C3@stroeder.com> <jDfg6.71100$Pm2.1445232@news20.bellglobal.com>
Message-ID: <3A818BA6.A40A2C27@stroeder.com>

Dan Parisien wrote:
> 
> Michael Str?der wrote:
> 
> > Citizens of the U.S. or any other country with export regulations
> > for cryptographic software have to take care about themselves for
> > not getting into trouble by posting code to this list.
> 
> I know you are not a legal expert, but it seems you have more knowledge
> than me in this area. Does this export control only matter to US 
> citizens?
> What if crypto code written by Canadian citizens is hosted on a US web
> server? (PyAES on sourceforge.net)

Canadian and U.S. law seemed to be pretty similar. At least from a
European point of view. ;-)

Sorry, I can't give any useful information. I'd suggest digging the
openssl-dev mailing list archives with search string "export
restrictions" since this topic was discussed to death there.

Ciao, Michael.


From akuchlin at mems-exchange.org  Wed Feb  7 23:28:54 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 07 Feb 2001 17:28:54 -0500
Subject: python-crypto list hosted at SURFnet.nl
References: <3A815E95.5399B6C3@stroeder.com> <jDfg6.71100$Pm2.1445232@news20.bellglobal.com>
Message-ID: <3dhf26gmah.fsf@ute.cnri.reston.va.us>

Dan Parisien <dan at eevolved.com> writes:
> I know you are not a legal expert, but it seems you have more knowledge 
> than me in this area. Does this export control only matter to US citizens? 

The last time I looked into this, Canada had looser regulations than
the US; if software is freely redistributable, it's considered
equivalent to public domain, and therefore can be exported from Canada
without filling out any forms or doing anything special.  

> What if crypto code written by Canadian citizens is hosted on a US web 
> server? (PyAES on sourceforge.net)

It would be OK to import the code, but the person/entity hosting it
inside the US could potentially run into trouble (though current US
regulations are quite loose for open source software, explicitly
mentioning it as an exception).

--amk


From ryu-jin at bigfoot.com  Wed Feb  7 17:40:30 2001
From: ryu-jin at bigfoot.com (Ryujin)
Date: Wed, 07 Feb 2001 16:40:30 GMT
Subject: SSL Client-Server
Message-ID: <95ru3l$2hk$1@ryu.it>

I have Python2.0 and OpenSSL-0.9.6 installed
Anybody knows where to find a tutorial showing how to create
a client-server ssl socket connection? I'd like to realize a simple
client connecting and sending something to the server using SSL
I could only find out this:                

s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)                
s.connect((HOST, PORT))
ssl = socket.ssl(s, keyfile, certfile)

whatabout the server? and whatbout the documentation?
Any help would be really appreciated.

matte



-- 
---------------------------------------
Memelli Matteo					
Could u imagine a world without Windoz?	 
ryu-jin at bigfoot.com			 
---------------------------------------



From l.szyster at ibm.net  Thu Feb  8 15:31:55 2001
From: l.szyster at ibm.net (Laurent Szyster)
Date: Thu, 08 Feb 2001 15:31:55 +0100
Subject: SSL Client-Server
References: <95ru3l$2hk$1@ryu.it>
Message-ID: <3A82ADDB.F7B6902@ibm.net>

Ryujin wrote:
> 
> I have Python2.0 and OpenSSL-0.9.6 installed
> Anybody knows where to find a tutorial showing how
> to create a client-server ssl socket connection? I'd
> like to realize a simple client connecting and sending
> something to the server using SSL
>
> I could only find out this:
> 
> s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
> s.connect((HOST, PORT))
> ssl = socket.ssl(s, keyfile, certfile)
> 
> whatabout the server? and whatbout the documentation?
> Any help would be really appreciated.

Have a look at the m2crypto package, it has it all:

	http://mars.post1.com/home/ngps/m2/ 

Good luck ...


Laurent Szyster


From Jerry.Spicklemire at IFLYATA.COM  Wed Feb  7 17:43:21 2001
From: Jerry.Spicklemire at IFLYATA.COM (Spicklemire, Jerry)
Date: Wed, 7 Feb 2001 11:43:21 -0500 
Subject: Red Hat Linux and poplib.py
Message-ID: <977A39E65CFCD3119ABF00D0B741D849E0FEB3@innt-73.ata.com>

Hi Python fans,

I'm hoping someone can point me in the right direction. 
When I try to use a contributed Zope module that calls 
poplib.py, Python v.2.5.2, I get a couple of error messages. 
The first hangs on:

>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 132, in
> ListMessages
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 68, in
> UpdateStatus
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 42, in 
> Connect
>     (Object: POPMailAccountBase)
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 183, in
> pass_
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 146, in
> _shortcmd
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 125, in
> _getresp
> error_proto: (see above)

with an error message of:

> > Error Type: error_proto
> > Error Value: -ERR Not implemented

This makes me think that poplib.py is trying to interact 
with some PopMail feature that isn't implemented in the 
Red Hat v.5.2 distribution.

My guess is that the second error is the result of the first, 
but it really is just a guess. The error is:

>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 132, in
> ListMessages
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 68, in
> UpdateStatus
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 42, in 
> Connect
>     (Object: POPMailAccountBase)
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 183, in
> pass_
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 146, in
> _shortcmd
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 125, in
> _getresp
> error_proto: (see above)

which is identical to the one above, but with the message:

> > Error Type: error_proto
> > Error Value: -ERR Mailbox in use

It seems like the popmail server isn't updated when the 
client process drops, and keeps the session open. When that 
happens, I can't even connect to the user mailbox from
Eudora, or even directly from the Python prompt.

BTW, I was able to connect to the mailbox via the Python 
prompt once, after the client process was killed manually.
I used some of the examples in the Beazley book, and 
elswhere, to invoke poplib.py, sign on, etc.

I searched the Web for references to Red Hat and PopMail 
problems, and found a short E-Mail thread about this. 
 
 http://plug.skylab.org/200009/msg00187.html

 http://plug.skylab.org/200009/msg00307.html

The recommended solution was to use another PopMail server, 
such as CUCIPOP. Since I'm not the system admin, I don't 
have much access to the host, and such notions aren't often 
warmly received. It was far from painless just to get Python 
and Zope up and running, though we do have one Linux guru on 
staff.

Any ideas would be greatly appreciated!

Thanks,
Jerry S.




From donn at u.washington.edu  Wed Feb  7 20:20:58 2001
From: donn at u.washington.edu (Donn Cave)
Date: 7 Feb 2001 19:20:58 GMT
Subject: Red Hat Linux and poplib.py
References: <mailman.981564352.10076.python-list@python.org>
Message-ID: <95s76q$nde$1@nntp6.u.washington.edu>

Quoth "Spicklemire, Jerry" <Jerry.Spicklemire at IFLYATA.COM>:
| I'm hoping someone can point me in the right direction. 
| When I try to use a contributed Zope module that calls 
| poplib.py, Python v.2.5.2, I get a couple of error messages. 
| The first hangs on:
|
| >   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
| >   line 132, in
| > ListMessages
| >     (Object: POPMailAccountBase)
| >   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
| >   line 68, in
| > UpdateStatus
| >     (Object: POPMailAccountBase)
| >   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
| >   line 42, in 
| > Connect
| >     (Object: POPMailAccountBase)
| >   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
| >   line 183, in
| > pass_
| >   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
| >   line 146, in
| > _shortcmd
| >   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
| >   line 125, in
| > _getresp
| > error_proto: (see above)
|
| with an error message of:
|
| > > Error Type: error_proto
| > > Error Value: -ERR Not implemented
|
| This makes me think that poplib.py is trying to interact 
| with some PopMail feature that isn't implemented in the 
| Red Hat v.5.2 distribution.

If you look at line 183 in poplib.py, you'll see that it's trying
to supply a password for a clear text login.  According to my
interpretation of "implemented", it would be very odd if that
really wasn't implemented, but there could in theory be a site
policy that refused it for non-SSL-encrypted connections.  Or
the problem could really be the context in which the login was
attempted.

If you're lucky, perhaps someone out there will read this and
remember the same problem with that version of Zope, that version
of Red Hat Linux and that version of ipop3d, and how it was resolved.
Otherwise, you will have to start eliminating the variables yourself.
I am not interested in Zope or Red Hat Linux, but if you can tell
me what poplib is doing and if you can narrow down to a specific
version of ipop3d, I can look into that.  I would think you're running
ipop3d, but I haven't seen any version that has a "Not implemented"
message.  Just telnet to port 110 on the mail host, and copy down
what it says.  While you're there, try "user bogus" and then
"pass bogus" and see what happens.  (Then "quit".)  To get the
story on poplib, maybe you need debugging output in _putcmd() or
_shortcmd().  To a file, probably with a different name every
session.

	Donn Cave, donn at u.washington.edu


From aleaxit at yahoo.com  Wed Feb  7 18:08:58 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 7 Feb 2001 18:08:58 +0100
Subject: ScriptControl+Python+win32all+Procedures -- known bug?
Message-ID: <95rvfg0b3j@news1.newsguy.com>

I'm trying to load Python as ActiveScripting compliant
language into an application that uses Microsoft
ScriptControl, but there seems to be a bug with
(whatever is it in IActiveScript that ScriptControl
uses to implement its Procedures collection...) -- or
is the application perhaps doing something wrong?  I
did manage to reproduce the problem on a toy-scale with
a tiny application that does nothing but:

1 -- instantiate a ScriptControl
2 -- set its Language property
3 -- AddCode to it a script textfile's contents
4 -- get its Procedures collection
5 -- get the Count property of said collection
6 -- get the _NewEnum property of said collection
7 -- enumerate on said _NewEnum

and it all works fine if Language is 'VBScript' (and
the textfile a correct foo.vbs script); however, with
the same application code, if Language is 'Python'
(and the textfile a correct foo.py script), steps 5
and 6 each fail (and thus 7 is not even tried) with an
HRESULT of 0x8002000b (meaning "Invalid index").

VC++6 code for steps 1-5 can be summarized as:

#import "msscript.ocx"
   ...
    MSScriptControl::IScriptControlPtr
pScr("MSScriptControl.ScriptControl");
    pScr->Language = "Python";
    pScr->AddCode("def foo():\n  pass\n");
    MSScriptControl::IScriptProcedureCollectionPtr pProcedures;
    pProcedures = pScr->Procedures;
    /* so far so good, but...: */
    printf("Number of procedures is %d\n", pProcedures->GetCount());
    /* this last line raises a _com_error C++ exception, as above
       detailed, and so does most any other attempt at doing
       anything with pProcedures. */

Is this a known bug of build 135 of ActivePython?  I
cannot find it on the ActiveState site (and thus have
submitted it there, hoping it's not a duplicate!).  Any
help/advice/workaround would be welcome, anyway!


Alex





From MarkH at ActiveState.com  Thu Feb  8 01:29:29 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 08 Feb 2001 00:29:29 GMT
Subject: ScriptControl+Python+win32all+Procedures -- known bug?
References: <95rvfg0b3j@news1.newsguy.com>
Message-ID: <3A81E780.2000804@ActiveState.com>

[Alex]

> Is this a known bug of build 135 of ActivePython? 

It is now
> I cannot find it on the ActiveState site (and thus have
> submitted it there

And that is precisely why it is now known ;-)

> Any help/advice/workaround would be welcome, anyway!

Sorry - I dont know of any.  I mailed you asking for a test case, and I will look into it.  I also have a debugging related ActiveScript bug outstanding, so I will look at them both soon...

Mark.



From jon_warner at my-deja.com  Wed Feb  7 18:20:57 2001
From: jon_warner at my-deja.com (jon_warner at my-deja.com)
Date: Wed, 07 Feb 2001 17:20:57 GMT
Subject: anydbm can't find type of dbm file
Message-ID: <95s05c$njd$1@nnrp1.deja.com>

Hi,

I am a newbie to the Python language, so there may be some errors in the
following source code, but I wonder if you guys could help me work out
what is going on here?

I have written a test program to find out how object persistence works
in Python, and when it is run Python complains that anydbm cannot
determine the dbm file type (i.e. whether it is gdbm orany other type of
dbm file). Here is the code:

#!/usr/bin/python

import sys
import shelve

class link_url:
	def __init__(self, url_link, description, author, type):
		self.url  = url_link
		self.desc = description
		self.auth = author
		self.type = type

def input_url_hand():
	global new_url
	_url = raw_input("What is the name of the URL? ")
	_auth = raw_input("What is your name? ")
	_desc = raw_input("Give a description ")
	_type = raw_input("What type of URL is it? ")
	new_url = link_url(_url, _desc, _auth, _type)

def get_url_link(url_in):
	sys.stdout.write(url_in.url)

def store_link(link_in):
	link_store = shelve.open('my_db')
	link_store[link_in.url]=link_in
	link_store.close

def print_links():
	link_store = shelve.open('my_db')
	for name in link_store.keys():
		print link_store[name]

def test():
	input_url_hand()
	get_url_link(new_url)
	store_link(new_url)
	print_links()

if __name__ == "__main__":
	test()


And when this is run, the output is:

Traceback (innermost last):
  File "link_url.py", line 42, in ?
    test()
  File "link_url.py", line 38, in test
    store_link(new_url)
  File "link_url.py", line 26, in store_link
    link_store = shelve.open('my_db') # open, with (g)dbm filename
  File "/usr/lib/python1.5/shelve.py", line 152, in open
    return DbfilenameShelf(filename, flag)
  File "/usr/lib/python1.5/shelve.py", line 142, in __init__
    Shelf.__init__(self, anydbm.open(filename, flag))
  File "/usr/lib/python1.5/anydbm.py", line 83, in open
    raise error, "db type could not be determined"
anydbm.error: db type could not be determined

Is there anything I can do to correct this? The Python I am using is:

Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  [GCC egcs-2.91.66
19990314/Linux (egcs
- on linux-i386

Thanks in advance,

Jonathan












Sent via Deja.com
http://www.deja.com/


From jafo at tummy.com  Thu Feb  8 07:44:55 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 7 Feb 2001 23:44:55 -0700
Subject: anydbm can't find type of dbm file
In-Reply-To: <95s05c$njd$1@nnrp1.deja.com>; from jon_warner@my-deja.com on Wed, Feb 07, 2001 at 05:20:57PM +0000
References: <95s05c$njd$1@nnrp1.deja.com>
Message-ID: <20010207234455.G505@tummy.com>

>  File "/usr/lib/python1.5/anydbm.py", line 83, in open
>    raise error, "db type could not be determined"
>anydbm.error: db type could not be determined

Please see http://www.tummy.com/radiuscontext/errata.html for a discussion
of this and possible ways to fix it.

Sean
-- 
 The politician was gone but unnoticed, like the period after the Dr on a
 Dr Pepper can.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From jon_warner at my-deja.com  Thu Feb  8 18:39:38 2001
From: jon_warner at my-deja.com (jon_warner at my-deja.com)
Date: Thu, 08 Feb 2001 17:39:38 GMT
Subject: anydbm can't find type of dbm file
References: <95s05c$njd$1@nnrp1.deja.com> <mailman.981614822.29236.python-list@python.org>
Message-ID: <95ulko$2eg$1@nnrp1.deja.com>

In article <mailman.981614822.29236.python-list at python.org>,
  Sean Reifschneider <jafo at tummy.com> wrote:
> >  File "/usr/lib/python1.5/anydbm.py", line 83, in open
> >    raise error, "db type could not be determined"
> >anydbm.error: db type could not be determined
>
> Please see http://www.tummy.com/radiuscontext/errata.html for a
discussion
> of this and possible ways to fix it.

Thanks Sean,

I'll install a newer Python.

Jonathan


Sent via Deja.com
http://www.deja.com/


From marvind at glue.umd.edu  Wed Feb  7 19:30:52 2001
From: marvind at glue.umd.edu (Arvind Mani)
Date: Wed, 7 Feb 2001 13:30:52 -0500
Subject: embed Tkinter in C
Message-ID: <95s3m1$db2$1@hecate.umd.edu>

Hi,

I am planning to build a user interface with Tkinter and integrate it with
an existing C program.

Can someone tell me where to find links to Tkinter C API and a tutorial

Thanks,
marvind




From wware at world.std.com  Wed Feb  7 20:19:41 2001
From: wware at world.std.com (Will Ware)
Date: Wed, 7 Feb 2001 19:19:41 GMT
Subject: embed Tkinter in C
References: <95s3m1$db2$1@hecate.umd.edu>
Message-ID: <G8EJ0t.83J@world.std.com>

Arvind Mani (marvind at glue.umd.edu) wrote:
> I am planning to build a user interface with Tkinter and integrate it with
> an existing C program.
> Can someone tell me where to find links to Tkinter C API and a tutorial

Tkinter is a Python wrapper for the Tk gui toolkit, which is written in
C and runs on Unix and Windows (and the Mac as well, I believe). If your
program is written in C, you might want to connect directly to the Tk
toolkit. Here's some info on Tk's C API:
http://www.bbso.njit.edu/Documentations/TclTkMan/tk4.2b1/tk4.2b1API.html

A possibly better strategy might be to create a Python wrapper for your
C code, and then write Python code that connects your C code to Tkinter.
This will probably be easier and more maintainable. Some resources that
would be useful for this are the Extending-and-Embedding tutorial:
http://www.python.org/doc/current/ext/ext.html
and Swig, an automatic generator for wrapper code:
http://www.swig.org/
http://www.swig.org/Doc1.1/HTML/Python.html

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From fredrik at effbot.org  Wed Feb  7 20:27:27 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Wed, 07 Feb 2001 19:27:27 GMT
Subject: embed Tkinter in C
References: <95s3m1$db2$1@hecate.umd.edu> <G8EJ0t.83J@world.std.com>
Message-ID: <zkhg6.12011$Qb7.1830004@newsb.telia.net>

> Here's some info on Tk's C API:
> http://www.bbso.njit.edu/Documentations/TclTkMan/tk4.2b1/tk4.2b1API.html

for the latest/greatest docs, see:

    http://dev.scriptics.com/

Cheers /F






From jinfengl at uci.edu  Wed Feb  7 19:39:34 2001
From: jinfengl at uci.edu (Jinfeng Liu)
Date: Wed, 7 Feb 2001 10:39:34 -0800
Subject: Synchronized method in Jython
Message-ID: <95s4qd$j1$1@news.service.uci.edu>

Hi,

I want to make a method synchrnized as the following in Java:

public synchronized f(...)

how to specify the keyword 'synchronized' in Jython?

Thanks.





From wilson.austin.aj at bhp.com.au  Wed Feb  7 19:45:47 2001
From: wilson.austin.aj at bhp.com.au (Austin Wilson)
Date: Thu, 8 Feb 2001 05:45:47 +1100
Subject: Problem with py2exe
Message-ID: <95s54j$c72$1@gossamer.itmel.bhp.com.au>

Hi

I have just downloaded py2exe and I am trying to freeze a program.  However,
I am getting the following output.  Can anyone help.

BTW:  I am using Python20 and wxWindows 2.2.2 on Win98.

Thanks
Austin


changing back to 'C:\Program Files\Python20'
creating dist\oww\oww.exe
Using stub 'py2exe\run.exe'
Resolving binary dependencies
Traceback (most recent call last):
  File "setup.py", line 6, in ?
    scripts=["oww.py"],
  File "c:\program files\python20\lib\distutils\core.py", line 138, in setup
    dist.run_commands()
  File "c:\program files\python20\lib\distutils\dist.py", line 829, in
run_commands
    self.run_command(cmd)
  File "c:\program files\python20\lib\distutils\dist.py", line 849, in
run_command
    cmd_obj.run()
  File "py2exe\py2exe.py", line 365, in run
    self.copy_dependend_dlls(final_dir, use_runw, dlls)
  File "py2exe\py2exe.py", line 417, in copy_dependend_dlls
    alldlls.remove(self.get_exe_stub(use_runw))
ValueError: list.remove(x): x not in list





From peter at engcorp.com  Thu Feb  8 06:42:44 2001
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 08 Feb 2001 00:42:44 -0500
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au>
Message-ID: <3A8231D4.1E5951A6@engcorp.com>

Austin Wilson wrote:
> 
> I have just downloaded py2exe and I am trying to freeze a program.  However,
> I am getting the following output.  Can anyone help.
> 
> BTW:  I am using Python20 and wxWindows 2.2.2 on Win98.
> 
> changing back to 'C:\Program Files\Python20'
[snip]
> ValueError: list.remove(x): x not in list

I got the same thing ("list.remove(x) raising ValueError") 
but happily py2exe _still_ managed to generate the 
executable successfully and I was able to run the resulting
program without difficulty!

Not that this helps much, but I have to say I was
pretty pleased that something that came without
guarantees, which I couldn't take the time to
fix when it failed, still managed to do the job
I needed in a pinch.

A big Thank You to the py2exe crew!  :-)


From thomas.heller at ion-tof.com  Thu Feb  8 16:45:04 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 8 Feb 2001 16:45:04 +0100
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au> <3A8231D4.1E5951A6@engcorp.com>
Message-ID: <95ueuo$iseg7$1@ID-59885.news.dfncis.de>

"Peter Hansen" <peter at engcorp.com> wrote in message news:3A8231D4.1E5951A6 at engcorp.com...
> Austin Wilson wrote:
> >
> > I have just downloaded py2exe and I am trying to freeze a program.  However,
> > I am getting the following output.  Can anyone help.
> >
> > BTW:  I am using Python20 and wxWindows 2.2.2 on Win98.
> >
> > changing back to 'C:\Program Files\Python20'
> [snip]
> > ValueError: list.remove(x): x not in list
>
> I got the same thing ("list.remove(x) raising ValueError")
> but happily py2exe _still_ managed to generate the
> executable successfully and I was able to run the resulting
> program without difficulty!

I have just uploaded a new release 0.2.2.

If the problem still occurs with this version,
I would certainly like to find this bug.

Thanks,

Thomas




From wilson.austin.aj at bhp.com.au  Thu Feb  8 19:29:34 2001
From: wilson.austin.aj at bhp.com.au (Austin Wilson)
Date: Fri, 9 Feb 2001 05:29:34 +1100
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au> <3A8231D4.1E5951A6@engcorp.com> <95ueuo$iseg7$1@ID-59885.news.dfncis.de>
Message-ID: <95uoi6$n5$1@gossamer.itmel.bhp.com.au>

I used version 0.2.1.

I have now successfully gotten the program to work but I had to do the
following

1. Copy the exe file generated by py2exe from the dist directory to the
winexe directory
2. Copy wxc.pyd and wx22_2.dll from the wxPython directory into the winexe
directory.

Regards
Austin


"Thomas Heller" <thomas.heller at ion-tof.com> wrote in message
news:95ueuo$iseg7$1 at ID-59885.news.dfncis.de...
>
> "Peter Hansen" <peter at engcorp.com> wrote in message
news:3A8231D4.1E5951A6 at engcorp.com...
> > Austin Wilson wrote:
> > >
> > > I have just downloaded py2exe and I am trying to freeze a program.
However,
> > > I am getting the following output.  Can anyone help.
> > >
> > > BTW:  I am using Python20 and wxWindows 2.2.2 on Win98.
> > >
> > > changing back to 'C:\Program Files\Python20'
> > [snip]
> > > ValueError: list.remove(x): x not in list
> >
> > I got the same thing ("list.remove(x) raising ValueError")
> > but happily py2exe _still_ managed to generate the
> > executable successfully and I was able to run the resulting
> > program without difficulty!
>
> I have just uploaded a new release 0.2.2.
>
> If the problem still occurs with this version,
> I would certainly like to find this bug.
>
> Thanks,
>
> Thomas
>
>




From robin at alldunn.com  Fri Feb  9 02:33:57 2001
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 8 Feb 2001 17:33:57 -0800
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au> <3A8231D4.1E5951A6@engcorp.com> <95ueuo$iseg7$1@ID-59885.news.dfncis.de> <95uoi6$n5$1@gossamer.itmel.bhp.com.au>
Message-ID: <ePHg6.1649$QtQ.68550953@news.randori.com>

"Austin Wilson" <wilson.austin.aj at bhp.com.au> wrote in message
news:95uoi6$n5$1 at gossamer.itmel.bhp.com.au...
> I used version 0.2.1.
>
> I have now successfully gotten the program to work but I had to do the
> following
>
> 1. Copy the exe file generated by py2exe from the dist directory to the
> winexe directory
> 2. Copy wxc.pyd and wx22_2.dll from the wxPython directory into the winexe
> directory.
>

Try either 0.2.0 or 0.2.2.  I've had success with wxPython apps with both os
these versions.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!





From wilson.austin.aj at bhp.com.au  Fri Feb  9 20:40:46 2001
From: wilson.austin.aj at bhp.com.au (Austin Wilson)
Date: Sat, 10 Feb 2001 06:40:46 +1100
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au> <3A8231D4.1E5951A6@engcorp.com> <95ueuo$iseg7$1@ID-59885.news.dfncis.de> <95uoi6$n5$1@gossamer.itmel.bhp.com.au> <ePHg6.1649$QtQ.68550953@news.randori.com>
Message-ID: <961h3k$jrb$1@gossamer.itmel.bhp.com.au>

I have just tried 0.2.2 and it seems to have worked.

Regards
Austin

"Robin Dunn" <robin at alldunn.com> wrote in message
news:ePHg6.1649$QtQ.68550953 at news.randori.com...
>
> "Austin Wilson" <wilson.austin.aj at bhp.com.au> wrote in message
> news:95uoi6$n5$1 at gossamer.itmel.bhp.com.au...
> > I used version 0.2.1.
> >
> > I have now successfully gotten the program to work but I had to do the
> > following
> >
> > 1. Copy the exe file generated by py2exe from the dist directory to the
> > winexe directory
> > 2. Copy wxc.pyd and wx22_2.dll from the wxPython directory into the
winexe
> > directory.
> >
>
> Try either 0.2.0 or 0.2.2.  I've had success with wxPython apps with both
os
> these versions.
>
>
> --
> Robin Dunn
> Software Craftsman
> robin at AllDunn.com       Java give you jitters?
> http://wxPython.org      Relax with wxPython!
>
>
>




From thomas.heller at ion-tof.com  Thu Feb  8 10:48:05 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 8 Feb 2001 10:48:05 +0100
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au>
Message-ID: <95tq0o$ie1b4$1@ID-59885.news.dfncis.de>

"Austin Wilson" <wilson.austin.aj at bhp.com.au> wrote in message news:95s54j$c72$1 at gossamer.itmel.bhp.com.au...
> Hi
>
> I have just downloaded py2exe and I am trying to freeze a program.  However,
> I am getting the following output.  Can anyone help.
>
> BTW:  I am using Python20 and wxWindows 2.2.2 on Win98.
>
> Thanks
> Austin
>
>
> changing back to 'C:\Program Files\Python20'
> creating dist\oww\oww.exe
> Using stub 'py2exe\run.exe'
> Resolving binary dependencies
> Traceback (most recent call last):
>   File "setup.py", line 6, in ?
>     scripts=["oww.py"],
>   File "c:\program files\python20\lib\distutils\core.py", line 138, in setup
>     dist.run_commands()
>   File "c:\program files\python20\lib\distutils\dist.py", line 829, in
> run_commands
>     self.run_command(cmd)
>   File "c:\program files\python20\lib\distutils\dist.py", line 849, in
> run_command
>     cmd_obj.run()
>   File "py2exe\py2exe.py", line 365, in run
>     self.copy_dependend_dlls(final_dir, use_runw, dlls)
>   File "py2exe\py2exe.py", line 417, in copy_dependend_dlls
>     alldlls.remove(self.get_exe_stub(use_runw))
> ValueError: list.remove(x): x not in list
>
>
>
Which py2exe version did you use?
Can you send me your script by private email so that I can take a look?

Thanks for trying it,

Thomas




From moshez at zadka.site.co.il  Wed Feb  7 20:18:36 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: 7 Feb 2001 13:18:36 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Feb  7)
Message-ID: <3E48CCD4391A1F23.06B04BA04C4438B1.98267277E79758A2@lp.airnews.net>

    Jeremy Hylton announces Python 2.1a2
	http://deja.com/=dnc/getdoc.xp?AN=723393159.5
    Sean Reifschneider follows quickly with SRPMS
	http://deja.com/=dnc/getdoc.xp?AN=723773806

    Phil Thompson announces SIP v2.3, a SWIG competitor
	http://deja.com/=dnc/getdoc.xp?AN=723883027

    Fredrik Lundh writes a Python debug function
	http://deja.com/=dnc/getdoc.xp?AN=723981430	
    gives link to an "import" guide
	http://deja.com/=dnc/getdoc.xp?AN=723209491

    Michael Hudson gives Python/C API hints
	http://deja.com/=dnc/getdoc.xp?AN=722933254
    Martin von Loewis gives some more advice
	http://deja.com/=dnc/getdoc.xp?AN=723356637

    Alex Martelli gives advice for Python/C++ integration
	http://deja.com/=dnc/getdoc.xp?AN=722002473	
    Fredrik Lundh gives some more hints
	http://deja.com/=dnc/getdoc.xp?AN=721882739

    Nick Mathewson shows an implementation of VB's "with"
	http://deja.com/=dnc/getdoc.xp?AN=724311465

    Oleg Broytmann and Berthol Hollman manage to port kjbuckets to Python 2.0
	http://deja.com/=dnc/getdoc.xp?AN=723158049
	http://deja.com/=dnc/getdoc.xp?AN=723327091

    Thomas Wouters gives a summary of text-munching methods in Python
	http://deja.com/=dnc/getdoc.xp?AN=721910944

    Andrew Kuchling announces a ZODB guide
	http://deja.com/=dnc/getdoc.xp?AN=724271836

    Christian Tismer announces RPMS for Stackless 2.0 for SuSE
	http://deja.com/=dnc/getdoc.xp?AN=721884526


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org
      Notice especially the master FAQ
          http://www.python.org/doc/FAQ.html

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
          http://deja.com/group/comp.lang.python.announce

      Andrew Kuchling writes marvelous summaries twice a month of
      the action on the python-dev mailing list, where the future
      of Python is truly determined
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collect Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From jmarshal at mathworks.com  Wed Feb  7 20:42:36 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 7 Feb 2001 19:42:36 GMT
Subject: python abstract syntax trees
Message-ID: <95s8fc$59g$1@news.mathworks.com>

In ther parser module, expr and suite return concrete syntax trees.
Does anyone know of any quick routines that build more minimal abstract
syntax trees?


From nas at arctrix.com  Wed Feb  7 23:28:03 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Wed, 7 Feb 2001 14:28:03 -0800
Subject: python abstract syntax trees
In-Reply-To: <95s8fc$59g$1@news.mathworks.com>; from jmarshal@mathworks.com on Wed, Feb 07, 2001 at 07:42:36PM +0000
References: <95s8fc$59g$1@news.mathworks.com>
Message-ID: <20010207142803.A10885@glacier.fnational.com>

On Wed, Feb 07, 2001 at 07:42:36PM +0000, Joshua Marshall wrote:
> In ther parser module, expr and suite return concrete syntax trees.
> Does anyone know of any quick routines that build more minimal abstract
> syntax trees?

Look at Tools/compiler

  Neil



From MarkH at ActiveState.com  Thu Feb  8 01:31:27 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 08 Feb 2001 00:31:27 GMT
Subject: python abstract syntax trees
References: <95s8fc$59g$1@news.mathworks.com>
Message-ID: <3A81E7F6.4080908@ActiveState.com>

Joshua Marshall wrote:

> In ther parser module, expr and suite return concrete syntax trees.
> Does anyone know of any quick routines that build more minimal abstract
> syntax trees?

Check out the "compiler" project - the source code exists in the CVS 
tree, and I believe may be in the normal Python 2.1 distro.  It uses an 
AST flattening module before working with the tree.  The Python for .NET 
compiler uses the exact same code, and it was all based on "p2c"'s effort.

Mark.



From c941520 at alinga.newcastle.edu.au  Wed Feb  7 20:58:07 2001
From: c941520 at alinga.newcastle.edu.au (Ben de Luca)
Date: Wed, 7 Feb 2001 11:58:07 -0800
Subject: im new to python threading question
Message-ID: <3a81470b$0$16373$7f31c96c@news01.syd.optusnet.com.au>

I want to start a new object and pass an object to it how do i do that?

if

def funks(s):
     do stuff with f



how do i call def in a new thread and send it args s




From Bill.Scherer at VerizonWireless.com  Wed Feb  7 14:15:16 2001
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Wed, 07 Feb 2001 08:15:16 -0500
Subject: im new to python threading question
References: <3a81470b$0$16373$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <3A814A64.3970A79A@VerizonWireless.com>

Ben de Luca wrote:
> 
> I want to start a new object and pass an object to it how do i do that?
> 
> if
> 
> def funks(s):
>      do stuff with f
> 
> how do i call def in a new thread and send it args s


import thread

thread.start_new_thread(funk, (s))




-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From fredrik at pythonware.com  Wed Feb  7 20:42:35 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 07 Feb 2001 19:42:35 GMT
Subject: im new to python threading question
References: <3a81470b$0$16373$7f31c96c@news01.syd.optusnet.com.au> <mailman.981551801.983.python-list@python.org>
Message-ID: <Lyhg6.12013$Qb7.1830937@newsb.telia.net>

Bill Scherer wrote:
> > def funks(s):
> >      do stuff with f
> >
> > how do i call def in a new thread and send it args s
>
> import thread
> thread.start_new_thread(funk, (s))

better make that:

    thread.start_new_thread(funks, (s,))

or even better, use the "threading" module instead.

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From **MailFreeAmerica**  Wed Feb  7 21:24:23 2001
From: **MailFreeAmerica** (OpenMind)
Date: Wed, 7 Feb 2001 12:24:23 -0800
Subject: Is Python "Compiled"?
Message-ID: <3a81ad07_3@goliath2.newsfeeds.com>

Recently started reading about Python.  The language looks very promising!

I am confused, however, about the impressions I have received.   If I write
an application in Python, would it be distributed in the form of text
modules?  Or would they be compiled into bytecode like Java and Visual
Basic?

Too, is there a native-code compiler or front end available for Python yet,
something like the GNU C preprocessor I seem to recall?

Thanks

___




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From embed at geocities.com  Wed Feb  7 21:58:00 2001
From: embed at geocities.com (Warren Postma)
Date: Wed, 7 Feb 2001 15:58:00 -0500
Subject: Is Python "Compiled"?
References: <3a81ad07_3@goliath2.newsfeeds.com>
Message-ID: <5Iig6.1295$D3.5361@tor-nn1.netcom.ca>

"OpenMind" <**Mail Free America**> wrote in message
news:3a81ad07_3 at goliath2.newsfeeds.com...
> Recently started reading about Python.  The language looks very promising!
>
> I am confused, however, about the impressions I have received.   If I
write
> an application in Python, would it be distributed in the form of text
> modules?  Or would they be compiled into bytecode like Java and Visual
> Basic?

Yes you can distribute a text script, and yes you could distribute a pyc
file (similar to a java .class).  I think distributing the text script is
highly encouraged.

You can make an EXE also, but the EXE is generally just a container for
python bytecode, similar to a VB4 EXE.


> Too, is there a native-code compiler or front end available for Python
yet,
> something like the GNU C preprocessor I seem to recall?

Various efforts to build a Compiler are in "research stages".  Dynamic
languages don't respond well to attempts to compile them.   .  Nevertheless,
some very smart people seem to be trying.

Warren




From hannah at schlund.de  Thu Feb  8 17:07:07 2001
From: hannah at schlund.de (Hannah Schroeter)
Date: 8 Feb 2001 17:07:07 +0100
Subject: Is Python "Compiled"?
References: <3a81ad07_3@goliath2.newsfeeds.com> <5Iig6.1295$D3.5361@tor-nn1.netcom.ca>
Message-ID: <95ug7b$qp0$1@c3po.schlund.de>

Hello!

In article <5Iig6.1295$D3.5361 at tor-nn1.netcom.ca>,
Warren Postma <embed at geocities.com> wrote:

>[...]

>Various efforts to build a Compiler are in "research stages".  Dynamic
>languages don't respond well to attempts to compile them.   .  Nevertheless,
>some very smart people seem to be trying.

What about Lisp, which is very often compiled to native code?

Besides, JIT technology is also there, already for a long time
(look at Smalltalk implementation techniques and see stuff implemented
a decade ago which they just "discover" for Java).

Bottom line: I don't think that "dynamic languages don't respond well
to attempts to compile them". Techniques are there (again, see Lisp,
see Smalltalk, see Cecil/vortex for examples). I'd believe that
techniques from Smalltalk implementations could also apply to the
Python language.

>Warren

Kind regards,

Hannah.


From sholden at holdenweb.com  Thu Feb  8 18:32:29 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 12:32:29 -0500
Subject: Is Python "Compiled"?
References: <3a81ad07_3@goliath2.newsfeeds.com> <5Iig6.1295$D3.5361@tor-nn1.netcom.ca> <95ug7b$qp0$1@c3po.schlund.de>
Message-ID: <xOAg6.11196$x4.177559@e420r-atl2.usenetserver.com>

"Hannah Schroeter" <hannah at schlund.de> wrote in message
news:95ug7b$qp0$1 at c3po.schlund.de...
> Hello!
>
> In article <5Iig6.1295$D3.5361 at tor-nn1.netcom.ca>,
> Warren Postma <embed at geocities.com> wrote:
>
> >[...]
>
> >Various efforts to build a Compiler are in "research stages".  Dynamic
> >languages don't respond well to attempts to compile them.   .
Nevertheless,
> >some very smart people seem to be trying.
>
> What about Lisp, which is very often compiled to native code?
>
> Besides, JIT technology is also there, already for a long time
> (look at Smalltalk implementation techniques and see stuff implemented
> a decade ago which they just "discover" for Java).
>
> Bottom line: I don't think that "dynamic languages don't respond well
> to attempts to compile them". Techniques are there (again, see Lisp,
> see Smalltalk, see Cecil/vortex for examples). I'd believe that
> techniques from Smalltalk implementations could also apply to the
> Python language.
>
> >Warren
>
> Kind regards,
>
> Hannah.
>
Well, LISP is kind of special: the typical list representation and
car()/cdr() operations are pretty efficient if properly compiled. But most
interpreted languages end up "compiling" down into calls to the interpreter
bytecode routines, which doesn't save much space or time. That was certainly
true for Icon, anyway, and I suspect would be for other similar languages.

JIT might or might not be a win for Python, but programs would still end up
lugging a lot of support code around, and without static typing a lot of
decisions still have to be deferred until run-time.

regards
 Steve




From hannah at schlund.de  Thu Feb  8 19:06:25 2001
From: hannah at schlund.de (Hannah Schroeter)
Date: 8 Feb 2001 19:06:25 +0100
Subject: Is Python "Compiled"?
References: <3a81ad07_3@goliath2.newsfeeds.com> <5Iig6.1295$D3.5361@tor-nn1.netcom.ca> <95ug7b$qp0$1@c3po.schlund.de> <xOAg6.11196$x4.177559@e420r-atl2.usenetserver.com>
Message-ID: <95un71$alp$1@c3po.schlund.de>

Hello!

In article <xOAg6.11196$x4.177559 at e420r-atl2.usenetserver.com>,
Steve Holden <sholden at holdenweb.com> wrote:
>[...]

>Well, LISP is kind of special: the typical list representation and
>car()/cdr() operations are pretty efficient if properly compiled. But most
>interpreted languages

A language can't be classified as interpreted or compiled. Implementations
can be.

Lisp has one implementation (e.g. sbcl) which can both interpret,
compile to bytecode or compile to native code. Another implementation
(e.g. clisp) can only interpret or compile to bytecode. There could
be an implementation that would only do minimal compilation (only that
which is required by the standard, mostly macro expansion) and would
thus be an almost only interpreting implementation. Yet another possibility
would be compiling everything to native, even forms from the interactive
REPL.

>end up "compiling" down into calls to the interpreter
>bytecode routines, which doesn't save much space or time. That was certainly
>true for Icon, anyway, and I suspect would be for other similar languages.

Again, that's a property of implementations, not of languages.
Sure, some languages might be especially difficult to compile to
efficient native code (or efficient bytecodes), others make it easier.

>JIT might or might not be a win for Python, but programs would still end up
>lugging a lot of support code around,

Support code, yes. Many (esp. the free ones) Lisp systems don't create
standalone executables but images to be run with the runtime, which
provides low level functionality like image loading, GC, low level
object representation, predefined functions (as long as they are not
implemented in Lisp themselves). Many Smalltalk implementations do
the same, and still, that's no hinderance in using native processor
instructions in the image file, amounting to more than just calling
one runtime function after the other.

>and without static typing a lot of
>decisions still have to be deferred until run-time.

See again Smalltalk. That's an area where many gains are possible
to make. One is static analysis: if, in a method call x.y(), the
compiler can infer that the type of x is A, the compiler can statically
code a call to A's y() method. If the compiler knows that x can be of
two types B or C, it can code something like
  if (typeof(x) is B)
    call B's y() method
  else
    call C's y() method
which on many processors is even faster than C++'s vtables (one
indirect jump), because for many branch prediction units, indirect
jumps are poison. (The SmallEiffel people do *all* method dispatches
using such a technique and found the break even point [where vtable
would be better] to be over around 50 possible object types).
The other comes from the JIT world. At precompilation time, code
an instrumented method call:
  runtime_call_method(x, "y", (no arguments))
That can gather evidence about the real call patterns (i.e. which
methods get really called for how often), after some time, it will
change that code to something resembling that evidence:
  call B's checked y() method
where
  B's checked y() method is:
    1:
    if (typeof(self) is NOT B)
      jump out to the runtime system indicating the type mismatch, passing
      self to the RTS for re-dispatching
    2:
    here, the real code of the y() method follows.

(i.e. methods are two-entry-point functions, the first is the checked
entry point, the second is unchecked [compiler is statically sure about
the type)

The code after 1 can usually translated to less than a handful of machine
instructions.

>regards
> Steve

Kind regards,

Hannah.


From com-zjensen at msn.com  Sat Feb 10 08:09:32 2001
From: com-zjensen at msn.com (Zac Jensen)
Date: Sat, 10 Feb 2001 00:09:32 -0700
Subject: Is Python "Compiled"?
References: <3a81ad07_3@goliath2.newsfeeds.com> <5Iig6.1295$D3.5361@tor-nn1.netcom.ca> <95ug7b$qp0$1@c3po.schlund.de> <xOAg6.11196$x4.177559@e420r-atl2.usenetserver.com>
Message-ID: <962pff$1jo$1@news.xmission.com>

I'd have to say, use Java... then Use Jython, because, with the combination,
there is extreme power and a nice little cesspool mix of compiled,
interpreted lingo, you've got java compiled to java bytecode which isn't
really compiled but.. is.  Then Python with jython that is interpreted but
compiles somehow into a java class unless modified, and has full dynamic
access to all the java classes at any given time... I am using this nice
combination to make a very VERY mallable "bbs" if any of you rednecks have
heard of one, This will allow extreme "configuration" of the system, and
also provide crossplatform abilities and speed to boot.

the key here is OPTIONS, OPTIONS, OPTIONS!!
!@#!@#!@$%!@$!%^#$%^!@#$!@#%!^@$%&%

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:xOAg6.11196$x4.177559 at e420r-atl2.usenetserver.com...
> "Hannah Schroeter" <hannah at schlund.de> wrote in message
> news:95ug7b$qp0$1 at c3po.schlund.de...
> > Hello!
> >
> > In article <5Iig6.1295$D3.5361 at tor-nn1.netcom.ca>,
> > Warren Postma <embed at geocities.com> wrote:
> >
> > >[...]
> >
> > >Various efforts to build a Compiler are in "research stages".  Dynamic
> > >languages don't respond well to attempts to compile them.   .
> Nevertheless,
> > >some very smart people seem to be trying.
> >
> > What about Lisp, which is very often compiled to native code?
> >
> > Besides, JIT technology is also there, already for a long time
> > (look at Smalltalk implementation techniques and see stuff implemented
> > a decade ago which they just "discover" for Java).
> >
> > Bottom line: I don't think that "dynamic languages don't respond well
> > to attempts to compile them". Techniques are there (again, see Lisp,
> > see Smalltalk, see Cecil/vortex for examples). I'd believe that
> > techniques from Smalltalk implementations could also apply to the
> > Python language.
> >
> > >Warren
> >
> > Kind regards,
> >
> > Hannah.
> >
> Well, LISP is kind of special: the typical list representation and
> car()/cdr() operations are pretty efficient if properly compiled. But most
> interpreted languages end up "compiling" down into calls to the
interpreter
> bytecode routines, which doesn't save much space or time. That was
certainly
> true for Icon, anyway, and I suspect would be for other similar languages.
>
> JIT might or might not be a win for Python, but programs would still end
up
> lugging a lot of support code around, and without static typing a lot of
> decisions still have to be deferred until run-time.
>
> regards
>  Steve
>
>




From sburr at home.com  Fri Feb  9 05:13:10 2001
From: sburr at home.com (Steven Burr)
Date: Fri, 09 Feb 2001 04:13:10 GMT
Subject: Is Python "Compiled"?
References: <3a81ad07_3@goliath2.newsfeeds.com> <5Iig6.1295$D3.5361@tor-nn1.netcom.ca> <95ug7b$qp0$1@c3po.schlund.de>
Message-ID: <sburr-0802012113100001@cx937488-a.chnd1.az.home.com>

In article <95ug7b$qp0$1 at c3po.schlund.de>, hannah at schlund.de (Hannah
Schroeter) wrote:

> Hello!
> 
> In article <5Iig6.1295$D3.5361 at tor-nn1.netcom.ca>,
> Warren Postma <embed at geocities.com> wrote:
> 
> >[...]
> 
> >Various efforts to build a Compiler are in "research stages".  Dynamic
> >languages don't respond well to attempts to compile them.   .  Nevertheless,
> >some very smart people seem to be trying.
> 
> What about Lisp, which is very often compiled to native code?
> 
> Besides, JIT technology is also there, already for a long time
> (look at Smalltalk implementation techniques and see stuff implemented
> a decade ago which they just "discover" for Java).
> 
[snip]

According to Apple's documentation, Objective C, the programming language
for the Mac OSX Cocoa environment, also supports dynamic typing and
dynamic binding.

See http://developer.apple.com/techpubs/macosx/Cocoa/ObjectiveC/ObjC.pdf

Since I'm at best a programming diletante, this may be entirely beside the
point, and if so forgive me.  I thought it might be useful information for
those attempting to build a Python compiler.


From embed at geocities.com  Wed Feb  7 21:52:24 2001
From: embed at geocities.com (Warren Postma)
Date: Wed, 7 Feb 2001 15:52:24 -0500
Subject: portable INI-file read/write extension for Python - any interest?
Message-ID: <RCig6.1291$D3.5109@tor-nn1.netcom.ca>

I happened to need a fast way to read and modify lot of ini files, and I
happened to not be on Windows, and so I adapted an existing a set of native
C routines to do INI files that I had lying around since the good old MS-DOS
days. (Thank you Charlie Beerman, wherever you are.)  I decided to gussy
them up, and wrap them up nicely for Python.  They are thread safe so if
you're writing an app that's half C and half Python, it's ideal, since you
wouldn't want to use two different buffered INI file libraries, one for your
C applications and another for your Python app (imagine the "cache
conflicts" you'd run into).

Now, for anyone already using Windows only, the win32api module has a simple
way to do this in native python, but then, it's not as fast.  Also, I might
write a 100% python version, so I can use the same ini file API when I want
to use Python on a platform where no C compiler is available.  I use INI
files a lot. Can you tell?

Here's a little code sample that I believe illustrates a "friendlier"
approach to INI files, more Python-esque that win32api:

import ini
filename='c:\\winnt\\system.ini'
for sec in ini.sections(filename):
    print '['+sec+']'
    for key,value in ini.getsection(filename,sec):
        print key,'=',value

I have compiled and tested it and have a pyd file for Windows available. It
is admittedly already superceded by win32api module's functions
"GetProfileVal",
and "SetProfileVal" however ugly those functions are ugly bare win32
wrappers. Also, I'm still not quite sure how to list all sections in a file,
given only the win32api module.

The only notable thing about my wrapper, and the INI functions I used are
that they only use stdio.h and string.h, and is portable anywhere you've got
a C compiler.

Anybody interested in it? I'll be putting it up at starship, once I get it
compiled on Linux, but I just thought someone else might have a use for it.

Here are some details on the API:

------------------------ ini.flush ------------------------
flush()
   -> Writes current INI file from memory to disk.
      Happens automatically when you read the next INI file.
  Example
     ini.flush()



------------------------ ini.get ------------------------
get(filename,section,item,default=None)
   -> Returns value, typically a string, unless an integer
      default value is provided.  If no default is specified,
      and no value is found, then None is returned.

  Sample "MYFILE.INI":
    [mainsection]
    item1=This is a string.
    item2=3

  Example:
    x = ini.get('myfile.ini', 'mainsection', 'item1')
    #x == 'This is a string.'
    x = ini.get('myfile.ini', 'mainsection', 'item2', -1)
    #x ==  3
    x = ini.get('myfile.ini', 'mainsection', 'item3' )
    #x ==   None

----------------------- ini.getsection ------------------------
getsection(filename,section)
   -> Returns entire section of an INI file as a list of
      (name,value) tuples

  Sample "MYFILE.INI":
    [mainsection]
    item1=This is a string.
    item2=3

  Example:
    pprint.pprint ( ini.getsection('myfile.ini', 'mainsection') )
    [ ('item1','This is a string'),
      ('item2','3') ]

----------------------- ini.sections ------------------------
sections(filename)
   -> Returns list of sections in the INI file as a list of
      strings.

  Sample "MYFILE.INI":
    [mainsection]
    item=X
    [anothersection]
    item=Y

  Example:
    print `ini.sections('myfile.ini')`
    [ 'mainsection', 'anothersection' ]

-----------------------  ini.set ------------------------

set(filename,section,item,value)
   -> Sets an ini file value in memory. Returns None.
      Raises exceptions on failure.
   -> if value is None, the item is DELETED from the ini file
   -> 'value' must be convertable to string via str(value)
   -> You must call ini.flush() to write all the changes at once,
      or changes will not be flushed until a different INI
      file is read.
  Example
     ini.set('myfile.ini', 'newsection', 'item3', 999 )

----------------------- (end) -----------------------


Will send a pre-release-zip of this to you on request. Will release and
announce once I compile Linux binaries. There's no use calling it portable
when I haven't written a gcc makefile yet! :-)


Warren Postma




From robin at alldunn.com  Wed Feb  7 23:16:46 2001
From: robin at alldunn.com (Robin Dunn)
Date: Wed, 7 Feb 2001 14:16:46 -0800
Subject: portable INI-file read/write extension for Python - any interest?
References: <RCig6.1291$D3.5109@tor-nn1.netcom.ca>
Message-ID: <9Qjg6.1142$q_C.86507560@news.randori.com>

>
> Will send a pre-release-zip of this to you on request. Will release and
> announce once I compile Linux binaries. There's no use calling it portable
> when I haven't written a gcc makefile yet! :-)
>

Why not use Distutils?  Then you only have to do it once and it works most
other places.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!





From ruediger.maehl at dzsh.de  Thu Feb  8 13:45:33 2001
From: ruediger.maehl at dzsh.de (Ruediger Maehl)
Date: Thu, 8 Feb 2001 13:45:33 +0100
Subject: portable INI-file read/write extension for Python - any interest?
References: <RCig6.1291$D3.5109@tor-nn1.netcom.ca>
Message-ID: <newscache$cevf8g$f5b$1@www.dzsh.de>

Warren Postma schrieb in Nachricht ...
> [snip]

>Here's a little code sample that I believe illustrates a "friendlier"
>approach to INI files, more Python-esque that win32api:
>
>import ini
>filename='c:\\winnt\\system.ini'
>for sec in ini.sections(filename):
>    print '['+sec+']'
>    for key,value in ini.getsection(filename,sec):
>        print key,'=',value
>


Looks like you do not read the ini-file into memory?

> [snip]
>Will send a pre-release-zip of this to you on request. Will release and
>announce once I compile Linux binaries. There's no use calling it portable
>when I haven't written a gcc makefile yet! :-)


I am interested!

R?diger
ruediger.maehl at web.de




From jcollin at exis.net  Wed Feb  7 21:58:56 2001
From: jcollin at exis.net (Jay Collins)
Date: Wed, 7 Feb 2001 15:58:56 -0500
Subject: UID/GID in python
Message-ID: <Pine.LNX.4.21.0102071557300.552-100000@tarpon.exis.net>

What is the best way to get the uid.gid of a file in python?

is there someting that when you do like info(filename) returns the
gid,uid, file size times etc?




From matt at mondoinfo.com  Wed Feb  7 22:11:05 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Wed, 07 Feb 2001 21:11:05 GMT
Subject: UID/GID in python
References: <Pine.LNX.4.21.0102071557300.552-100000@tarpon.exis.net>
Message-ID: <slrn983ef9.12k.matt@happy-hour.mondoinfo.com>

On Wed, 7 Feb 2001 15:58:56 -0500, Jay Collins <jcollin at exis.net>
wrote:

>What is the best way to get the uid.gid of a file in python?

os.stat("foo")[stat.ST_UID]

and similarly with stat.ST_GID

>is there someting that when you do like info(filename) returns the
>gid,uid, file size times etc?

os.stat() returns the data as a tuple and the stat module has symbolic
names for the items in the tuple.

Regards,
Matt


From jh at web.de  Wed Feb  7 22:32:55 2001
From: jh at web.de (Jürgen Hermann)
Date: Wed, 7 Feb 2001 22:32:55 +0100
Subject: Trace function with automatic source code location reporting
Message-ID: <95sffm$i49$05$1@news.t-online.com>

Hi!

The following is the core of a tracing function, is there any less obscure
way to get the location of the caller?

def trace(*args):
    try:
        raise RuntimeError
    except RuntimeError:
        import sys, traceback, string
        caller = sys.exc_info()[2].tb_frame.f_back
        print "%s [%d]: %s" % (caller.f_code.co_filename, caller.f_lineno,
string.join(map(str, args)))
        del caller

trace("first call", 1)

trace("second call", 2.0)

def func():
    trace("third call", 1+2)

func()


Bye, J?rgen




From Uwe.Hoffmann at bc.boehringer-ingelheim.com  Thu Feb  8 09:03:24 2001
From: Uwe.Hoffmann at bc.boehringer-ingelheim.com (Uwe.Hoffmann at bc.boehringer-ingelheim.com)
Date: Thu, 8 Feb 2001 09:03:24 +0100 
Subject: Trace function with automatic source code location reporting
Message-ID: <14074F0BA53BD411803000A0C9DD65111F0EF1@bc0exch7.bc.de.bic>

Hi!

J?rgen Hermann wrote:

>>The following is the core of a tracing function, is there any less obscure
>>                 way to get the location of the caller?


Maybe you are looking for :            
	tb = traceback.extract_stack()[-2]
            print "%s,%i,%s: " %(tb[0],tb[1],tb[2],) ,

Regards
	Uwe



From rhyde99 at email.msn.com  Wed Feb  7 22:35:11 2001
From: rhyde99 at email.msn.com (Roger H)
Date: Wed, 7 Feb 2001 13:35:11 -0800
Subject: Python wierdness
Message-ID: <OeV743UkAHA.278@cpmsnbbsa07>

In my Python musings I came across a bit of seeming wierdness.
I threw some code at the Python 1.5.2 command line to demonstrate:

    >>> # This function prints "bar" if the input
    ... # variable modulo 0.1 is equal to 0.1, and
    ... # "no bar" if it is not equal to 0.1:
    ...
    >>> def foo(n):
    ...     if n % 0.1 == 0.1:
    ...             print "bar"
    ...     else:
    ...             print "no bar"
    ...
    >>> # Here I create a variable x with the value 0.3:
    ...
    >>> x = 0.3
    >>>
    >>> # Let's check whether x modulo 0.1 is equal to 0.1:
    ...
    >>> x % 0.1
    0.1
    >>>
    >>> # Obviously, x modulo 0.1 is indeed equal to 0.1.
    ... # If we pass it to foo(), we should see "bar", right?
    ...
    >>> foo(x)
    no  bar
    >>>
    >>> # Hmmm... "no bar."  So foo() doesn't agree that
    ... # x modulo 0.1 is equal to 0.1, even though the
    ... # command line shows that it is.  Why not?

What's going on here, and how could one get around it?

~Roger the Shrubber~




From mwh21 at cam.ac.uk  Wed Feb  7 22:44:34 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Feb 2001 21:44:34 +0000
Subject: Python wierdness
References: <OeV743UkAHA.278@cpmsnbbsa07>
Message-ID: <m3d7cu2mnx.fsf@atrus.jesus.cam.ac.uk>

"Roger H" <rhyde99 at email.msn.com> writes:
[snip]
>     >>> x = 0.3
>     >>>
>     >>> # Let's check whether x modulo 0.1 is equal to 0.1:
>     ...
>     >>> x % 0.1
>     0.1
>     >>>
>     >>> # Obviously, x modulo 0.1 is indeed equal to 0.1.
>     ... # If we pass it to foo(), we should see "bar", right?
[snip]
> What's going on here, and how could one get around it?

Floating point is going on here.  Python 2.0 is more helpful in this
regard:

>>> 0.3 % 0.1
0.099999999999999978
>>> 0.1
0.10000000000000001
>>> 0.3 % 0.1 - 0.1
-2.7755575615628914e-17

Comparing floats for equality is always fraught.  Enjoy :-) (I may be
a mathematician, but you aren't going to catch me doing numerical
analysis...)

Cheers,
M.

-- 
  I'd certainly be shocked to discover a consensus.  ;-)
                                      -- Aahz Maruch, comp.lang.python


From mjackson at wc.eso.mc.xerox.com  Wed Feb  7 22:52:28 2001
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 7 Feb 2001 21:52:28 GMT
Subject: Python wierdness
References: <OeV743UkAHA.278@cpmsnbbsa07>
Message-ID: <95sg2s$bjs$1@news.wrc.xerox.com>

"Roger H" <rhyde99 at email.msn.com> writes:

> What's going on here, and how could one get around it?

Display rounding of finite-precision floating point arithmetic:

yngvi> python
Python 1.5.2 (#7, May  6 1999, 14:39:45)  [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 0.3 % 0.1 - 0.1
-2.77555756156e-17

In other words, 0.3 % 0.1 == 0.1 is indeed false.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Having your book made into a movie is like having
	your ox made into a bouillon cube.
				- Bill Neely




From fredrik at pythonware.com  Wed Feb  7 23:06:18 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 07 Feb 2001 22:06:18 GMT
Subject: Python wierdness
References: <OeV743UkAHA.278@cpmsnbbsa07>
Message-ID: <uFjg6.12051$Qb7.1838781@newsb.telia.net>

Roger wrote:
> In my Python musings I came across a bit of seeming wierdness

not Python's fault, really.

>     >>> x % 0.1
>     0.1
>     >>>
>     >>> # Obviously, x modulo 0.1 is indeed equal to 0.1.

obviously, x modulo 0.1 *prints* as 0.1 under 1.5.2.
doesn't mean that it contains exactly the same bits,
does it?  let's see:

>>> (0.3 % 0.1) - 0.1
-2.77555756156e-017

nope.

>     ... # If we pass it to foo(), we should see "bar", right?

obviously not.

>     >>> # Hmmm... "no bar."  So foo() doesn't agree that
>     ... # x modulo 0.1 is equal to 0.1, even though the
>     ... # command line shows that it is.  Why not?

full story here:
http://www.python.org/cgi-bin/moinmoin/RepresentationError

also note that Python 2.0 shows more precision:

    >>> x = 0.3
    >>> x % 0.1
    0.099999999999999978

> What's going on here, and how could one get around it?

unless you know exactly what you're doing, don't use "=="
on floating point values.  alternatives:

    if abs(a - b) < some_small_value:
        ...

or even

    if str(a) == str(b):
        ...

Cheers /F




From samschul at pacbell.net  Wed Feb  7 23:42:45 2001
From: samschul at pacbell.net (Sam Schulenburg)
Date: Wed, 07 Feb 2001 22:42:45 GMT
Subject: Python wierdness
References: <OeV743UkAHA.278@cpmsnbbsa07> <uFjg6.12051$Qb7.1838781@newsb.telia.net>
Message-ID: <95sj13$aq0$1@nnrp1.deja.com>

This advice also applies to other languages.
Sam Schulenburg

> > What's going on here, and how could one get around it?
>
> unless you know exactly what you're doing, don't use "=="
> on floating point values.  alternatives:
>
>     if abs(a - b) < some_small_value:
>         ...
>
> or even
>
>     if str(a) == str(b):
>         ...
>
> Cheers /F
>
>


Sent via Deja.com
http://www.deja.com/


From donn at u.washington.edu  Wed Feb  7 23:15:21 2001
From: donn at u.washington.edu (Donn Cave)
Date: 7 Feb 2001 22:15:21 GMT
Subject: Python wierdness
References: <OeV743UkAHA.278@cpmsnbbsa07>
Message-ID: <95shdp$ose$1@nntp6.u.washington.edu>

Quoth "Roger H" <rhyde99 at email.msn.com>:
| In my Python musings I came across a bit of seeming wierdness.
| I threw some code at the Python 1.5.2 command line to demonstrate:
...
|     >>> # Here I create a variable x with the value 0.3:
|     ...
|     >>> x = 0.3
|     >>>
|     >>> # Let's check whether x modulo 0.1 is equal to 0.1:
|     ...
|     >>> x % 0.1
|     0.1
|     >>>
|     >>> # Obviously, x modulo 0.1 is indeed equal to 0.1.

It wouldn't be as obvious in Python 2.0.  Try this -

  >>> y = x % 0.1
  >>> '%.18f' % y
  '0.099999999999999978'

It isn't Python's fault, computers just don't have very good
heads for this sort of math.

	Donn Cave, donn at u.washington.edu


From teyc at bigfoot.com  Wed Feb  7 22:42:34 2001
From: teyc at bigfoot.com (Chui Tey)
Date: Thu, 8 Feb 2001 07:42:34 +1000
Subject: Touch in python?
Message-ID: <95sfla$adf$1@bunyip.cc.uq.edu.au>

Hi,

Is there an equivalent of 'touch' in python? I would like to change the file
modified time.

Chui




From jh at web.de  Wed Feb  7 22:56:50 2001
From: jh at web.de (Jürgen Hermann)
Date: Wed, 7 Feb 2001 22:56:50 +0100
Subject: Touch in python?
References: <95sfla$adf$1@bunyip.cc.uq.edu.au>
Message-ID: <95sg8e$mod$00$1@news.t-online.com>

"Chui Tey" <teyc at bigfoot.com> schrieb im Newsbeitrag
news:95sfla$adf$1 at bunyip.cc.uq.edu.au...
> Is there an equivalent of 'touch' in python? I would like to change the
file
> modified time.

os.utime()




From gangli at msn.com  Thu Feb  8 20:55:15 2001
From: gangli at msn.com (gangli at msn.com)
Date: Thu, 08 Feb 2001 19:55:15 GMT
Subject: Touch in python?
References: <95sfla$adf$1@bunyip.cc.uq.edu.au> <95sg8e$mod$00$1@news.t-online.com>
Message-ID: <95utiv$a55$1@nnrp1.deja.com>

In article <95sg8e$mod$00$1 at news.t-online.com>,
  "J?rgen Hermann" <jh at web.de> wrote:
>
> "Chui Tey" <teyc at bigfoot.com> schrieb im Newsbeitrag
> news:95sfla$adf$1 at bunyip.cc.uq.edu.au...
> > Is there an equivalent of 'touch' in python? I would like to change
the
> file
> > modified time.
>
> os.utime()
>
>
os.utime won't work against directory on Windows


Sent via Deja.com
http://www.deja.com/


From hwanjoyu at uiuc.edu  Wed Feb  7 23:45:38 2001
From: hwanjoyu at uiuc.edu (Hwanjo Yu)
Date: Wed, 7 Feb 2001 16:45:38 -0600
Subject: Q: urlopen and timeout
Message-ID: <rekg6.522$853.6664@vixen.cso.uiuc.edu>

Hi,
It seems there is no way for us to set up the timeout explicitly in the
urlopen function.
Although it may have the default timeout in the function, it seems to wait
forever sometimes because of the unexpected web site's behavior.  Is there
any way to set the timeout explicitly ?
And,
The socket function also doesn't have timeout parameter, so I don't have any
idea when the socket connection function is gonna return.
How to set timeout for these cases ?
Thanks.




From sholden at holdenweb.com  Thu Feb  8 00:10:32 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 7 Feb 2001 18:10:32 -0500
Subject: urlopen and timeout
References: <rekg6.522$853.6664@vixen.cso.uiuc.edu>
Message-ID: <yFkg6.39714$o5.496051@e420r-atl1.usenetserver.com>

A case for Tom O'Malley's timeoutsocket module. You can set a default
timeout on sockets before you try to connect on them, which even works when
the socket connection is made inside a library routine.

Take a look at

    http://www.timo-tasi.org/python/timeoutsocket.py

regards
 Steve
--
Tools, training and technology to help you meet your information needs



"Hwanjo Yu" <hwanjoyu at uiuc.edu> wrote in message
news:rekg6.522$853.6664 at vixen.cso.uiuc.edu...
> Hi,
> It seems there is no way for us to set up the timeout explicitly in the
> urlopen function.
> Although it may have the default timeout in the function, it seems to wait
> forever sometimes because of the unexpected web site's behavior.  Is there
> any way to set the timeout explicitly ?
> And,
> The socket function also doesn't have timeout parameter, so I don't have
any
> idea when the socket connection function is gonna return.
> How to set timeout for these cases ?
> Thanks.
>
>
>



From phd at phd.pp.ru  Thu Feb  8 13:30:14 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 8 Feb 2001 15:30:14 +0300 (MSK)
Subject: Q: urlopen and timeout
In-Reply-To: <rekg6.522$853.6664@vixen.cso.uiuc.edu>
Message-ID: <Pine.LNX.4.30.0102081529260.20064-100000@fep132.fep.ru>

On Wed, 7 Feb 2001, Hwanjo Yu wrote:
> Although it may have the default timeout in the function, it seems to wait
> forever sometimes because of the unexpected web site's behavior.  Is there
> any way to set the timeout explicitly ?
> And,
> The socket function also doesn't have timeout parameter, so I don't have any
> idea when the socket connection function is gonna return.
> How to set timeout for these cases ?

   timeoutsocket.py at http://www.timo-tasi.org/python/

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From m.bless at gmx.de  Wed Feb  7 23:55:49 2001
From: m.bless at gmx.de (Martin Bless)
Date: Wed, 07 Feb 2001 22:55:49 GMT
Subject: String formatting characters - looking for details
Message-ID: <3a81cfa2.1708851@news.muenster.de>

(1) Where can I find some more details about the string formatting
characters?
(2) How  I print hex values with zeros padded left, like 0F or 0A?
print '%X' % 15  # two digits, wanted

(3) This is from the Python docs - I'm looking for more information on
how to use the formatting characters.

Thanks,
Martin

------------------- from the Python docs --------
2.1.5.2 String Formatting Operations 

String objects have one unique built-in operation: the % operator
(modulo) with a string left argument interprets this string as a C
sprintf() format string to be applied to the right argument, and
returns the string resulting from this formatting operation. 

The right argument should be a tuple with one item for each argument
required by the format string; if the string requires a single
argument, the right argument may also be a single non-tuple object.2.5
The following format characters are understood: %, c, s, i, d, u, o,
x, X, e, E, f, g, G. Width and precision may be a * to specify that an
integer argument specifies the actual width or precision. The flag
characters -, +, blank, # and 0 are understood. The size specifiers h,
l or L may be present but are ignored.
---------------------



From sholden at holdenweb.com  Thu Feb  8 00:17:07 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 7 Feb 2001 18:17:07 -0500
Subject: String formatting characters - looking for details
References: <3a81cfa2.1708851@news.muenster.de>
Message-ID: <6Pkg6.39716$o5.497716@e420r-atl1.usenetserver.com>

"Martin Bless" <m.bless at gmx.de> wrote in message
news:3a81cfa2.1708851 at news.muenster.de...
> (1) Where can I find some more details about the string formatting
> characters?
> (2) How  I print hex values with zeros padded left, like 0F or 0A?
> print '%X' % 15  # two digits, wanted
>
> (3) This is from the Python docs - I'm looking for more information on
> how to use the formatting characters.
>
> Thanks,
> Martin
>
[ doc snipped]

1. If you have a UNIX system handy, the man page on printf(3) will tell you
most of what you need to know. I've just loaded Cygwin on my laptop, and
that also has the right stuff in it. Or a "C" library reference book, if you
have one.

2. Leading zeroes can be obtained with a "0" flag, and ".2" specifies a
field width of two characters. The following example should help.

>>> "%0.2X" % 2
'02'

3. See 1.

regards
 Steve




From f8dy at my-deja.com  Thu Feb  8 03:28:15 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Thu, 08 Feb 2001 02:28:15 GMT
Subject: String formatting characters - looking for details
References: <3a81cfa2.1708851@news.muenster.de> <6Pkg6.39716$o5.497716@e420r-atl1.usenetserver.com>
Message-ID: <95t07p$m5g$1@nnrp1.deja.com>

In article <6Pkg6.39716$o5.497716 at e420r-atl1.usenetserver.com>,
  "Steve Holden" <sholden at holdenweb.com> wrote:
> "Martin Bless" <m.bless at gmx.de> wrote in message
> news:3a81cfa2.1708851 at news.muenster.de...
> > (1) Where can I find some more details about the string formatting
> > characters?
> > (2) How  I print hex values with zeros padded left, like 0F or 0A?
> > print '%X' % 15  # two digits, wanted
>
> 1. If you have a UNIX system handy, the man page on printf(3) will tell you

And if you don't, it's Google to the rescue!  Search for "printf format
identifiers".  I found lots of very useful pages, like this one: 
http://www.cs.huji.ac.il/support/docs/reference/C/FUNCTIONS/format.html

-M
--
You're smart; why haven't you learned Python yet?  http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From jafo at tummy.com  Thu Feb  8 07:40:51 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 7 Feb 2001 23:40:51 -0700
Subject: String formatting characters - looking for details
In-Reply-To: <95t07p$m5g$1@nnrp1.deja.com>; from f8dy@my-deja.com on Thu, Feb 08, 2001 at 02:28:15AM +0000
References: <3a81cfa2.1708851@news.muenster.de> <6Pkg6.39716$o5.497716@e420r-atl1.usenetserver.com> <95t07p$m5g$1@nnrp1.deja.com>
Message-ID: <20010207234051.E505@tummy.com>

On Thu, Feb 08, 2001 at 02:28:15AM +0000, Mark Pilgrim wrote:
>And if you don't, it's Google to the rescue!  Search for "printf format
>identifiers".  I found lots of very useful pages, like this one: 
>http://www.cs.huji.ac.il/support/docs/reference/C/FUNCTIONS/format.html

Alternatively, check the latest CVS of the documents.  I finally got
Evelyn to submit some patches to the docs which should provide complete
documentation of the % formatting options.

The unfortunate thing is that this patch sat here internally for months
since I wrote the content, hoping to get marked up...

Sean
-- 
 Q:  What kind of dog goes "BOFH!  BOFH!"?
 A:  A rootweiler
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From fredrik at pythonware.com  Thu Feb  8 07:40:21 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 08 Feb 2001 06:40:21 GMT
Subject: String formatting characters - looking for details
References: <3a81cfa2.1708851@news.muenster.de>
Message-ID: <pbrg6.13514$AH6.2023410@newsc.telia.net>

Martin Bless wrote:
> (1) Where can I find some more details about the string formatting
> characters?

your nearest C/Unix manual.  or the Python 2.1 docs:

http://python.sourceforge.net/devel-docs/lib/typesseq-strings.html

> (2) How  I print hex values with zeros padded left, like 0F or 0A?
> print '%X' % 15  # two digits, wanted

print "%02X" % 15

> (3) This is from the Python docs - I'm looking for more information on
> how to use the formatting characters.

searching the net for "sprintf" brings up an infinite number
of manual pages.  (fwiw, the perl manual page says these
formatting codes are "universally known"...)

Cheers /F




From scarblac at pino.selwerd.nl  Thu Feb  8 12:36:07 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 8 Feb 2001 11:36:07 GMT
Subject: String formatting characters - looking for details
References: <3a81cfa2.1708851@news.muenster.de>
Message-ID: <slrn9850l7.vp5.scarblac@pino.selwerd.nl>

Martin Bless <m.bless at gmx.de> wrote in comp.lang.python:
> (1) Where can I find some more details about the string formatting
> characters?
> (2) How  I print hex values with zeros padded left, like 0F or 0A?
> print '%X' % 15  # two digits, wanted

'%02X'. '%2X' pads with spaces so it is width 2, '%02X' pads with zeroes.

> (3) This is from the Python docs - I'm looking for more information on
> how to use the formatting characters.

Well, the Python docs say what you can use, not what it does... The best
option is probably to look at a Unix man page for printf. Type 'man printf'
at a Unix/Linux prompt or into Google.

> ------------------- from the Python docs --------
> 2.1.5.2 String Formatting Operations 
> 
> String objects have one unique built-in operation: the % operator
> (modulo) with a string left argument interprets this string as a C
> sprintf() format string to be applied to the right argument, and
> returns the string resulting from this formatting operation. 
> 
> The right argument should be a tuple with one item for each argument
> required by the format string; if the string requires a single
> argument, the right argument may also be a single non-tuple object.2.5
> The following format characters are understood: %, c, s, i, d, u, o,
> x, X, e, E, f, g, G. Width and precision may be a * to specify that an
> integer argument specifies the actual width or precision. The flag
> characters -, +, blank, # and 0 are understood. The size specifiers h,
> l or L may be present but are ignored.
> ---------------------
 
-- 
Remco Gerlich


From m.bless at gmx.de  Thu Feb  8 20:02:50 2001
From: m.bless at gmx.de (Martin Bless)
Date: Thu, 08 Feb 2001 19:02:50 GMT
Subject: String formatting characters - looking for details
References: <3a81cfa2.1708851@news.muenster.de>
Message-ID: <3a82eb27.915141@news.muenster.de>

Thanks to all of you.

Yes, I'm learning Python, I'm using Python, and I love it.
It's probably hard to imagine, but I don't have a  Linux/Unix machine
at hand. That was part of the problem. And I've done some languages
but never really C.  And I did some searching but obviously knew to
little to find "the needle in that heap of hay" (Brutal attempt of
translating a german saying ...

So, thanks again,  you really helped me.

Martin



From btheld2 at my-deja.com  Thu Feb  8 00:06:26 2001
From: btheld2 at my-deja.com (btheld2 at my-deja.com)
Date: Wed, 07 Feb 2001 23:06:26 GMT
Subject: SWIG and __doc__ strings
Message-ID: <95skdb$c19$1@nnrp1.deja.com>

Is there a way for the automatically generated .py file created by SWIG
to include function and class documentation strings (for the __doc__
attribute)?  Examples?

Ben


Sent via Deja.com
http://www.deja.com/


From burton at tx420.org  Thu Feb  8 00:20:48 2001
From: burton at tx420.org (JackBurton)
Date: Wed, 7 Feb 2001 17:20:48 -0600
Subject: Error during make?
Message-ID: <3a81d7bd$0$65278$38c90ffe@news1.twtelecom.net>

Hi, I was installing Python on Slackware Linux 2.2.6 kernel, and  am getting
the error message:
 make[1]: *** [selectmodule.o] Error 1
make[1]: Leaving directory `/home/jay/Python/Python-2.0/Modules'
make: *** [Modules] Error 2


I was wondering if anyone could help me with this?

advthanxance
Jack




From thomas at xs4all.net  Mon Feb 12 18:49:14 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 12 Feb 2001 18:49:14 +0100
Subject: Error during make?
In-Reply-To: <3a81d7bd$0$65278$38c90ffe@news1.twtelecom.net>; from burton@tx420.org on Wed, Feb 07, 2001 at 05:20:48PM -0600
References: <3a81d7bd$0$65278$38c90ffe@news1.twtelecom.net>
Message-ID: <20010212184914.C4924@xs4all.nl>

On Wed, Feb 07, 2001 at 05:20:48PM -0600, JackBurton wrote:
> Hi, I was installing Python on Slackware Linux 2.2.6 kernel, and  am getting
> the error message:
>  make[1]: *** [selectmodule.o] Error 1
> make[1]: Leaving directory `/home/jay/Python/Python-2.0/Modules'
> make: *** [Modules] Error 2

> I was wondering if anyone could help me with this?

The real error is somewhere above the 'make: Error 1' line. You are probably
trying to compile a module for which you lack the right include files, but
it's impossible to say without seeing more of the make error messages.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From dsavitsk at e-coli.net  Thu Feb  8 01:02:29 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Wed, 7 Feb 2001 18:02:29 -0600
Subject: sorting question
Message-ID: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net>

i would like to iterate through all possible combinations of the characters
in a string, and i am not thinking very clearly about how to approach this.
that is, if the initial string is 'bar', i want to print
['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
any hints?

thanks,
ds




From sholden at holdenweb.com  Thu Feb  8 01:36:15 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 7 Feb 2001 19:36:15 -0500
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net>
Message-ID: <YVlg6.39726$o5.503818@e420r-atl1.usenetserver.com>

"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:tmlg6.61202$Ch.11526305 at newsrump.sjc.telocity.net...
> i would like to iterate through all possible combinations of the
characters
> in a string, and i am not thinking very clearly about how to approach
this.
> that is, if the initial string is 'bar', i want to print
> ['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
> any hints?
>
> thanks,
> ds
>
Tim Peters was working on something a while ago, see

    http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=558836665

but I'm not sure where it eventually went to live

regards
 Steve
--
Tools, training and technology to help you meet your information needs







From gregj at pobox.com  Thu Feb  8 10:39:56 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 08 Feb 2001 09:39:56 GMT
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net>
Message-ID: <95tphb$a05$1@nnrp1.deja.com>

In article <tmlg6.61202$Ch.11526305 at newsrump.sjc.telocity.net>,
  "dsavitsk" <dsavitsk at e-coli.net> wrote:
> i would like to iterate through all possible combinations of the
characters
> in a string, and i am not thinking very clearly about how to approach
this.
> that is, if the initial string is 'bar', i want to print
> ['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
> any hints?

I adapted Sedgewick's permutation function (from Algorithms in C, pp
628-629). My solution (below) first generates a list of all
permutations of integers in the range (1..n), where n is the length of
the string to be permuted. It then translates the original string into
a list of permutations using the list of integers. Because generating
the list of permuted integers is expensive, you could cache the lists
so you wouldn't have to do it for every 3- or 4-character string.

Note that this process gets very slow and the lists get very long for
even short strings... after five characters you can see the hit.

`````````````````````````````````````
def visit(k, val, id, v, perms):
    "permutation by exhaustive search, from sedgewick"
    id += 1
    val[k] = id
    if id == v:
        perms.append(val[:])
    else:
        for i in range(v):
            if val[i] == 0:
                visit(i, val, id, v, perms)
    id -= 1
    val[k] = 0

def permutations(n):
    "generate list of all permutations of the integers 1..n"
    perms = []
    val = [0] * n
    visit(0, val, -1, n, perms)
    return perms

def permutestring(s):
    "generate list of all permutations of a string"
    perms = permutations(len(s))
    result = []
    for p in perms:
        t = ""
        for i in p:
            t += s[i-1]
        result.append(t)
    return result

#print permutations(4)
print permutestring("bar")
print permutestring("abcd")

`````````````````````````````````````


--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From stephen_purcell at yahoo.com  Thu Feb  8 12:20:45 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Thu, 8 Feb 2001 12:20:45 +0100
Subject: sorting question
In-Reply-To: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net>; from dsavitsk@e-coli.net on Wed, Feb 07, 2001 at 06:02:29PM -0600
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net>
Message-ID: <20010208122045.B7052@freedom.puma-ag.com>

dsavitsk wrote:
> i would like to iterate through all possible combinations of the characters
> in a string, and i am not thinking very clearly about how to approach this.
> that is, if the initial string is 'bar', i want to print
> ['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
> any hints?

The following works for me:

def permute(choices, fixed=''):
    bag = []                   
    numchoices = len(choices)
    for pos in range(numchoices):
        start = choices[pos]
        rest = choices[:pos] + choices[pos+1:]
        bag.extend(permute(rest, fixed + start))
        if numchoices == 1: bag.append(fixed + start + rest)
    return bag

print permute('bar')
print permute('1234')

It produces the following output:

['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
['1234', '1243', '1324', '1342', '1423', '1432', '2134', '2143', '2314', '2341 '2413', '2431', '3124', '3142', '3214', '3241', '3412', '3421', '4123', '4132 '4213', '4231', '4312', '4321']


With longer strings, you probably don't want to store the results in memory,
so you can use a visitor function instead of a list:

def permute2(choices, visitfunc, fixed=''):
    numchoices = len(choices)              
    for pos in range(numchoices):
        start = choices[pos]
        rest = choices[:pos] + choices[pos+1:]
        permute2(rest, visitfunc, fixed + start)
        if numchoices == 1: visitfunc(fixed + start + rest)

permutations = []
permute2('1234567', permutations.append)  # pretty silly example of a visitor
print permutations

If your strings have repeated characters and you want only unique combinations:

permutations = {}
def visitor(v):
    permutations[v] = None
permute2('ababcbed', visitor)
print permutations.keys()


Hope that helps,

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From aleaxit at yahoo.com  Thu Feb  8 14:11:07 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 8 Feb 2001 14:11:07 +0100
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net>
Message-ID: <95u5tc010vo@news2.newsguy.com>

"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:tmlg6.61202$Ch.11526305 at newsrump.sjc.telocity.net...
> i would like to iterate through all possible combinations of the
characters
> in a string, and i am not thinking very clearly about how to approach
this.
> that is, if the initial string is 'bar', i want to print
> ['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
> any hints?

Perhaps not very fast, but, IMHO, pretty cool:

class Perm:
    def __init__(self, seq):
        self._seq = list(seq)
        self._len = len(self._seq)
    def __getitem__(self, x):
        index = x
        indices = [-1]*(self._len-1)
        for n in range(self._len,1,-1):
            index,indices[self._len-n] = divmod(index,n)
        if index: raise IndexError,x
        result = self._seq[:]
        for i in range(len(indices)):
            index = i+indices[i]
            result[i],result[index] = result[index],result[i]
        return result

and now

for p in Perm('bar'):
    print ''.join(p)

>>> for p in Perm('bar'):
...     print ''.join(p)
...
bar
abr
rab
bra
arb
rba
>>>

the idea is to associate to each number between 0 and
one less than the factorial of N (where N is the length
of the sequence being permuted) a specific permutation,
which is a list of N integers -- the first in
range(0,N), the second in range(0,N-1), ... the last
always 0 (so, no need to record or compute it).  As
usual for enumeration tasks, it looks a bit like a
problem of 'counting in some base', except that, here,
it's several bases -- N for the first digit, N-1 for
the second one, and so on.  Unfortunately I don't think
there is a close-form way to get the list of N integers
from the number it represents, whence the loop with
the divmod in it.

The second half of the __getitem__ method then gets
from the list of numbers to an actual permutation, by
a way that will remind one of 'Knuth shuffling' if one
has ever programmed the latter:-).  It can be done
in other good ways, though.


Alex





From alistair.watts at businesscollaborator.com  Thu Feb  8 16:17:16 2001
From: alistair.watts at businesscollaborator.com (Alistair Watts)
Date: Thu, 08 Feb 2001 15:17:16 +0000
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net> <95u5tc010vo@news2.newsguy.com>
Message-ID: <3A82B87C.2B48AC9D@businesscollaborator.com>

  A little recursive solution to the problem.

  The idea here is firstly consider the base case of a string of length
one.  Easy, it's just that string.  

  Secondly, for a string of length n, where n>1 choose each character in
turn for the start of the result and use the characters left over to
construct the tail.

  The tail is all the combinations of the remaining letters.  Which is
the same problem again except the string is now of length n-1.  So
recurse until we reach the base case.

  Not sure if this is faster than Alex's code, but it's a few line less.

  Regards...

Alistair

-----------------------------------

text = 'bar'

def f(s):
  if len(s) == 1:
    return s[0]
  r = []
  for i in range(len(s)):
    for rest in f(s[:i] + s[i+1:]):
      r.append(s[i] + rest)
  return r

>>> f(text)
['bar', 'bra', 'abr', 'arb', 'rba', 'rab']

-----------------------------------

Alex Martelli wrote:
> 
> "dsavitsk" <dsavitsk at e-coli.net> wrote in message
> news:tmlg6.61202$Ch.11526305 at newsrump.sjc.telocity.net...
> > i would like to iterate through all possible combinations of the
> characters
> > in a string, and i am not thinking very clearly about how to approach
> this.
> > that is, if the initial string is 'bar', i want to print
> > ['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
> > any hints?
> 
> Perhaps not very fast, but, IMHO, pretty cool:
> 
> class Perm:
>     def __init__(self, seq):
>         self._seq = list(seq)
>         self._len = len(self._seq)
>     def __getitem__(self, x):
>         index = x
>         indices = [-1]*(self._len-1)
>         for n in range(self._len,1,-1):
>             index,indices[self._len-n] = divmod(index,n)
>         if index: raise IndexError,x
>         result = self._seq[:]
>         for i in range(len(indices)):
>             index = i+indices[i]
>             result[i],result[index] = result[index],result[i]
>         return result
> 
> and now
> 
> for p in Perm('bar'):
>     print ''.join(p)
> 
> >>> for p in Perm('bar'):
> ...     print ''.join(p)
> ...
> bar
> abr
> rab
> bra
> arb
> rba
> >>>
> 
> the idea is to associate to each number between 0 and
> one less than the factorial of N (where N is the length
> of the sequence being permuted) a specific permutation,
> which is a list of N integers -- the first in
> range(0,N), the second in range(0,N-1), ... the last
> always 0 (so, no need to record or compute it).  As
> usual for enumeration tasks, it looks a bit like a
> problem of 'counting in some base', except that, here,
> it's several bases -- N for the first digit, N-1 for
> the second one, and so on.  Unfortunately I don't think
> there is a close-form way to get the list of N integers
> from the number it represents, whence the loop with
> the divmod in it.
> 
> The second half of the __getitem__ method then gets
> from the list of numbers to an actual permutation, by
> a way that will remind one of 'Knuth shuffling' if one
> has ever programmed the latter:-).  It can be done
> in other good ways, though.
> 
> Alex
> 
> --
> http://mail.python.org/mailman/listinfo/python-list

-- 
-----------------------------------
Business Collaborator Team
Enviros Software Solutions
http://www.businesscollaborator.com
-----------------------------------



From dsavitsk at e-coli.net  Thu Feb  8 17:30:04 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 8 Feb 2001 10:30:04 -0600
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net>
Message-ID: <gQzg6.67860$Ch.12128127@newsrump.sjc.telocity.net>

Thank you all for the suggestions.

i am trying to unscramble scrambled words (which i am frustratingly bad at)
so i send all of the permutations to Word's spell checker and see what it
thinks are words.  I'll post the code this afternoon when i'm done in case
anyone wants it.

doug

"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:tmlg6.61202$Ch.11526305 at newsrump.sjc.telocity.net...
> i would like to iterate through all possible combinations of the
characters
> in a string, and i am not thinking very clearly about how to approach
this.
> that is, if the initial string is 'bar', i want to print
> ['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
> any hints?
>
> thanks,
> ds
>
>




From andrew_dot_henshaw_at_earthling_dot_net  Fri Feb  9 03:57:31 2001
From: andrew_dot_henshaw_at_earthling_dot_net (Andrew Henshaw)
Date: Thu, 8 Feb 2001 21:57:31 -0500
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net> <gQzg6.67860$Ch.12128127@newsrump.sjc.telocity.net>
Message-ID: <t86nh2l57s7j43@corp.supernews.com>

Depending upon your constraints, another way to approach this problem may be
better.  If you can get your word (or Word) dictionary, then you can build a
compiled form that collapses your search.  The compiled form of a word takes
the letters of a word and sorts them.   Use the compiled form as a
dictionary key to the original word (or words).  For example:
'bar' and 'bra' both compile to 'abr'.  Thus

words = {'arb': ('bar', bra')}

When you want to check a word, compile it by sorting its letters and test
its key against your words dictionary.

    a=list('bar')
    a.sort()
    compiled_word = string.join(a, '')
    if words.has_key(compiled_word):
        print words[compiled_word]

This would be blazingly fast, but would depend upon compiling your word
list.  There are lots of pretty complete text lists of English words
available on the web.

Andrew Henshaw



"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:gQzg6.67860$Ch.12128127 at newsrump.sjc.telocity.net...
> Thank you all for the suggestions.
>
> i am trying to unscramble scrambled words (which i am frustratingly bad
at)
> so i send all of the permutations to Word's spell checker and see what it
> thinks are words.  I'll post the code this afternoon when i'm done in case
> anyone wants it.
>
> doug
>
> "dsavitsk" <dsavitsk at e-coli.net> wrote in message
> news:tmlg6.61202$Ch.11526305 at newsrump.sjc.telocity.net...
> > i would like to iterate through all possible combinations of the
> characters
> > in a string, and i am not thinking very clearly about how to approach
> this.
> > that is, if the initial string is 'bar', i want to print
> > ['bar', 'bra', 'abr', 'arb', 'rba', 'rab']
> > any hints?
> >
> > thanks,
> > ds
> >
> >
>
>




From andrew_dot_henshaw_at_earthling_dot_net  Fri Feb  9 04:46:10 2001
From: andrew_dot_henshaw_at_earthling_dot_net (Andrew Henshaw)
Date: Thu, 8 Feb 2001 22:46:10 -0500
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net> <gQzg6.67860$Ch.12128127@newsrump.sjc.telocity.net> <t86nh2l57s7j43@corp.supernews.com>
Message-ID: <t86qc9j0iith02@corp.supernews.com>

"Andrew Henshaw" <andrew_dot_henshaw_at_earthling_dot_net> wrote in message
news:t86nh2l57s7j43 at corp.supernews.com...
> Depending upon your constraints, another way to approach this problem may
be
> better.  If you can get your word (or Word) dictionary, then you can build
a
> compiled form that collapses your search.  The compiled form of a word
takes
> the letters of a word and sorts them.   Use the compiled form as a
> dictionary key to the original word (or words).  For example:
> 'bar' and 'bra' both compile to 'abr'.  Thus
...snip..
> This would be blazingly fast, but would depend upon compiling your word
> list.  There are lots of pretty complete text lists of English words
> available on the web.
...snip...

There is a word list available at
http://www-personal.umich.edu/~jlawler/wordlist.html
using this list and the code at the end of this note, I get the following
output:

>>> import Anagram
>>> a = Anagram.Anagram('c:/temp/wordlist.txt')
>>> a.unjumble('mujleb')
['jumble']
>>> a.unjumble('redaps')
['spared', 'spread']
>>>

(Interesting. It missed 'parsed', 'drapes', and 'rasped'.  You might look
for a better list!)
On my machine, it took eight seconds to load and compile the list.

AH

#################################
import string

class Anagram:
    def __init__(self, words_file_name):
        self.word_dict = {}
        for word in open(words_file_name).readlines():
            word = string.strip(word)
            if word:
                compiled = self.compile(word)
                if self.word_dict.has_key(compiled):
                    self.word_dict[compiled].append(word)
                else:
                    self.word_dict[compiled] = [word]

    def compile(self, word):
        l = list(word)
        l.sort()
        return string.join(l, '')

    def unjumble(self, word):
        compiled = self.compile(word)
        if self.word_dict.has_key(compiled):
            print self.word_dict[compiled]
        else:
            print '**not found**'





From ten.llun at ttam  Fri Feb  9 09:34:11 2001
From: ten.llun at ttam (Matthew Schinckel)
Date: Fri, 09 Feb 2001 19:04:11 +1030
Subject: sorting question
References: <tmlg6.61202$Ch.11526305@newsrump.sjc.telocity.net> <gQzg6.67860$Ch.12128127@newsrump.sjc.telocity.net>
Message-ID: <3a83ac08_7@news.chariot.net.au>

In message <gQzg6.67860$Ch.12128127 at newsrump.sjc.telocity.net>, "dsavitsk"
wrote:
> Thank you all for the suggestions.
> 
> i am trying to unscramble scrambled words (which i am frustratingly bad at)
> so i send all of the permutations to Word's spell checker and see what it
> thinks are words.  I'll post the code this afternoon when i'm done in case
> anyone wants it.
> 

Hmm, sort of sounds similar to what I did the year before last.

The local 'Newspaper' (I give it more credit than it is due :-), the Adelaide
Advertiser, has a 9-letter word puzzle, where 9 letters are given, and the aim
is to make as many words as possible that are 4 letters or more, and contain
the magic 'middle' letter.

For instance:

C   A   R
D   A   B
R   O   O

would give many words, with cardboard as the 9-letter word.

My school had a competition where each homegroup had to try and find as many
words, and the results were tallied over the whole term.

Naturally, I wrote a python program, and took all of the fun out of it.

If anyone is interested, you might find it at:

http://www.chariot.net.au/~jaq/matt/NineLetter.zip

(It's almost 2Mb, and includes a large, almost[?] comprehensive word list.)
-- 
Matthew Schinckel
matt at null dot net



From pinard at iro.umontreal.ca  Thu Feb  8 01:06:40 2001
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=e7ois?= Pinard)
Date: 07 Feb 2001 19:06:40 -0500
Subject: Martin Loewis? (sorry)
Message-ID: <oqk872t4vj.fsf@lin2.sram.qc.ca>

Hi, Martin.  (Sorry for the noise to the list.)  I cannot reach you, getting:

   <martin at mira.cs.tu-berlin.de>... Deferred: Connection refused by mira.cs.tu-berlin.de.
   Message could not be delivered for 1 week
   Message will be deleted from queue
   Reporting-MTA: dns; mail.cs.tu-berlin.de
   Arrival-Date: Wed, 31 Jan 2001 15:46:09 +0100 (MET)

This was about the Translation Project, Python, and other fun things...
Do you have a better address?

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From nospam at nospam.com  Thu Feb  8 01:23:19 2001
From: nospam at nospam.com (MrBill)
Date: Wed, 7 Feb 2001 19:23:19 -0500
Subject: Image Capture And Processing
Message-ID: <95sok6$gvt$1@bob.news.rcn.net>

I would like to capture and process images from my USB connected Logitech
digicam.

Does anyone know where I would start to acquire images from the camera and
control it from python (or any other programming environment).  I'm running
on windoze 2000.

Thanks,
--
Bill Rodgers




From bfunk33 at qwest.NOSPAM.net.invalid.net.com.gov  Thu Feb  8 02:22:16 2001
From: bfunk33 at qwest.NOSPAM.net.invalid.net.com.gov (Bill Funk)
Date: Thu, 08 Feb 2001 01:22:16 GMT
Subject: Image Capture And Processing
References: <95sok6$gvt$1@bob.news.rcn.net>
Message-ID: <3a81f3d1.25710123@news.qwest.net>

On Wed, 7 Feb 2001 19:23:19 -0500, "MrBill" <nospam at nospam.com> wrote:

>I would like to capture and process images from my USB connected Logitech
>digicam.
>
>Does anyone know where I would start to acquire images from the camera and
>control it from python (or any other programming environment).  I'm running
>on windoze 2000.
>
>Thanks,
>--
>Bill Rodgers
>
>


While I don't have that particular camera, a good place to start would
be with the bundled software.
Did you get any?
If so, there should be help files or a manual.
If not, you will need to find a way to *see* the camera, and to do
this with USB usually requires the right drivers for both the camera
and the OS. Often, these can be downloaded from the camera maker's
site on the Web.

Or, if the camera has removable media, some sort of reader for the
media wil make things immensly easier. For most of us, a USB reader is
going to be the fastest/easiest. These make the memory card look like
a removable drive (like a floppy), and you can use Windows Explorer to
copy the image files to your hard drive.
>From there, any graphics program can access the files, and you are
set!

================
Bill Funk

bfunk1 on Ebay


From tjg at exceptionalminds.com  Thu Feb  8 02:17:47 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Wed, 7 Feb 2001 17:17:47 -0800
Subject: ARP Anyone?
Message-ID: <20010207171747.E27007@trufflehunter.avalongroup.net>

Hi all,

Has anyone ever done anything with ARP and Python? I'm looking for a way to
track new workstations on a network (before they ever get an IP address)
using python.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  22 days  5:29 hours ago<<



From db3l at fitlinxx.com  Thu Feb  8 04:43:32 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 07 Feb 2001 22:43:32 -0500
Subject: ARP Anyone?
References: <mailman.981595013.18967.python-list@python.org>
Message-ID: <usnlpvnyz.fsf@ctwd0143.fitlinxx.com>

Timothy Grant <tjg at exceptionalminds.com> writes:

> Has anyone ever done anything with ARP and Python? I'm looking for a way to
> track new workstations on a network (before they ever get an IP address)
> using python.

I just use the system "arp" command with os.popen and parse the output :-)

While there may be some slight differences in format or command line
options across platforms, it's probably the most portable way to
handle it since the underlying access to the ARP tables differs far
more (especially between Unix and Windows systems).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From jafo at tummy.com  Thu Feb  8 07:37:41 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 7 Feb 2001 23:37:41 -0700
Subject: ARP Anyone?
In-Reply-To: <usnlpvnyz.fsf@ctwd0143.fitlinxx.com>; from db3l@fitlinxx.com on Wed, Feb 07, 2001 at 10:43:32PM -0500
References: <mailman.981595013.18967.python-list@python.org> <usnlpvnyz.fsf@ctwd0143.fitlinxx.com>
Message-ID: <20010207233741.D505@tummy.com>

On Wed, Feb 07, 2001 at 10:43:32PM -0500, David Bolen wrote:
>I just use the system "arp" command with os.popen and parse the output :-)

I'd consider parsing the output of tcpdump set up to only dump ARP
packets, or seeing what I could do about talking to libpcap or
something (isn't there a module for libpcap?).

Sean
-- 
 Let's just say that your monkeys aren't quite typing Shakespeare.
   -- Sean Reifschneider, speaking about Quicken support, 2001
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From dlaroche at dsuper.net  Thu Feb  8 04:02:00 2001
From: dlaroche at dsuper.net (Denis Laroche)
Date: Thu, 08 Feb 2001 03:02:00 GMT
Subject: newbie question
Message-ID: <3A820E6B.CB29AFBC@dsuper.net>

Hello Python enthusiasts,

I just started to learn the language this week.

Could someone tell me how to recreate a list or a dictionary object from
its string representation obtained with repr().

Thanks in advance for any help.

--
Denis Laroche
Pratt & Whitney Canada
Otterburn Park, Qu?bec, Canada, J3H 5P4
dlaroche at dsuper.net





From sholden at holdenweb.com  Thu Feb  8 04:03:02 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 7 Feb 2001 22:03:02 -0500
Subject: newbie question
References: <3A820E6B.CB29AFBC@dsuper.net>
Message-ID: <y3og6.40178$o5.513584@e420r-atl1.usenetserver.com>

"Denis Laroche" <dlaroche at dsuper.net> wrote in message
news:3A820E6B.CB29AFBC at dsuper.net...
> Hello Python enthusiasts,
>
> I just started to learn the language this week.
>
Good stuff.

> Could someone tell me how to recreate a list or a dictionary object from
> its string representation obtained with repr().
>
> Thanks in advance for any help.
>

Just use the eval() built-in:

>>> d = {1:"one", 'two':2}
>>> s = repr(d)
>>> print s
{'two': 2, 1: 'one'}
>>> d1 = eval(s)
>>> d1
{'two': 2, 1: 'one'}
>>>
regards
 Steve
--
Tools, training and technology to help you meet your information needs







From erno-news at erno.iki.fi  Thu Feb  8 05:15:21 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 08 Feb 2001 06:15:21 +0200
Subject: newbie question
References: <3A820E6B.CB29AFBC@dsuper.net>
Message-ID: <kuhf2524km.fsf@lasipalatsi.fi>

In article <3A820E6B.CB29AFBC at dsuper.net>, Denis Laroche
<dlaroche at dsuper.net> writes:

| Hello Python enthusiasts,
| I just started to learn the language this week.

| Could someone tell me how to recreate a list or a dictionary object from
| its string representation obtained with repr().

you can't do that in general, but if you only put objects that
have a eval()-able string representation inside your list/dictionary,
you can use eval().

>>> l = [1, 2]
>>> eval(repr(l)) == l
1
>>> l = [sys.stdout, l.append]
>>> eval(repr(l)) == l
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<string>", line 1

    [<open file '<stdout>', mode 'w' at 804ab68>, <built-in method append of list object at 8068e38>]
     ^
SyntaxError: invalid syntax

if you want to serialize stuff, the pickle module is more suited
to that (although it can't handle files and such either).

  -- erno


From scarblac at pino.selwerd.nl  Thu Feb  8 12:30:35 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 8 Feb 2001 11:30:35 GMT
Subject: newbie question
References: <3A820E6B.CB29AFBC@dsuper.net>
Message-ID: <slrn9850aq.vp5.scarblac@pino.selwerd.nl>

Denis Laroche <dlaroche at dsuper.net> wrote in comp.lang.python:
> Hello Python enthusiasts,
> 
> I just started to learn the language this week.
> 
> Could someone tell me how to recreate a list or a dictionary object from
> its string representation obtained with repr().
> 
> Thanks in advance for any help.

Often, eval() can restore the value that you got with repr():
>>> eval(repr(["a",3,(1,2,3)]))
["a",3,(1,2,3)]

Of course, this only works if the repr() string has enough information and
was designed for this use (that is, numbers, strings, list etc work, but
once you start using class instances etc it probably won't).

-- 
Remco Gerlich


From matt at netasol.com  Thu Feb  8 04:31:07 2001
From: matt at netasol.com (Matt Fair)
Date: Wed, 07 Feb 2001 21:31:07 -0600
Subject: Reportlab fonts
Message-ID: <3A8212FB.5F15BF60@netasol.com>

Hello,
Does anyone know if it is possible yet to expand the standard 14 fonts
when generating pdfs with reportlab?  In the user manual it says that
they are working on embedding fonts into the pdf.  What are the
limitations to pdfs?  Are the 14 standard fonts on the clients computer
with Acrobat?  Would there be a way to expand the fonts even though you
could embed the fonts into the pdf (the user would require to have the
fonts on the computer).  How accurate can you make these pdf for
printing?  I need to make printable files that are very accurate to
where the text is and the size and font.
Thanks
Matt




From robin at jessikat.fsnet.co.uk  Thu Feb  8 11:08:51 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 8 Feb 2001 10:08:51 +0000
Subject: Reportlab fonts
References: <mailman.981603842.18070.python-list@python.org>
Message-ID: <QAlIJvAzAng6Ewjp@jessikat.fsnet.co.uk>

In article <mailman.981603842.18070.python-list at python.org>, Matt Fair
<matt at netasol.com> writes
>Hello,
>Does anyone know if it is possible yet to expand the standard 14 fonts
>when generating pdfs with reportlab?  In the user manual it says that
>they are working on embedding fonts into the pdf.  What are the
>limitations to pdfs?  Are the 14 standard fonts on the clients computer
>with Acrobat?  Would there be a way to expand the fonts even though you
>could embed the fonts into the pdf (the user would require to have the
>fonts on the computer).  How accurate can you make these pdf for
>printing?  I need to make printable files that are very accurate to
>where the text is and the size and font.
>Thanks
>Matt
>
>
so far as I know standard reportlab only allows the 14 standard fonts. A
colleague reports an extended version with japanese fonts. I'll repost
this into the reportlab-users mail list.
-- 
Robin Becker


From ekibismuth at privacyx.com  Thu Feb  8 05:20:08 2001
From: ekibismuth at privacyx.com (eki)
Date: Thu, 08 Feb 2001 04:20:08 GMT
Subject: Newbie Query
Message-ID: <Y7pg6.9822$En4.180927@typhoon.jacksonville.mediaone.net>

Hello all,

I am just getting restarted in programming after learning a bit of BASIC
back in the seventies. I would like to take a course online in programming
in python.

Does anyone know of any beginners classes which are being offered online?

Thanks,

Eki




From sholden at holdenweb.com  Thu Feb  8 13:53:44 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 07:53:44 -0500
Subject: Newbie Query
References: <Y7pg6.9822$En4.180927@typhoon.jacksonville.mediaone.net>
Message-ID: <kJwg6.42991$o5.533528@e420r-atl1.usenetserver.com>


--
--
Tools, training and technology to help you meet your information needs



"eki" <ekibismuth at privacyx.com> wrote in message
news:Y7pg6.9822$En4.180927 at typhoon.jacksonville.mediaone.net...
> Hello all,
>
> I am just getting restarted in programming after learning a bit of BASIC
> back in the seventies. I would like to take a course online in programming
> in python.
>
> Does anyone know of any beginners classes which are being offered online?
>
> Thanks,
>
> Eki
>
>
>



From sholden at holdenweb.com  Thu Feb  8 13:54:31 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 07:54:31 -0500
Subject: Newbie Query
References: <Y7pg6.9822$En4.180927@typhoon.jacksonville.mediaone.net>
Message-ID: <UJwg6.10923$x4.162687@e420r-atl2.usenetserver.com>

"eki" <ekibismuth at privacyx.com> wrote in message
news:Y7pg6.9822$En4.180927 at typhoon.jacksonville.mediaone.net...
> Hello all,
>
> I am just getting restarted in programming after learning a bit of BASIC
> back in the seventies. I would like to take a course online in programming
> in python.
>
> Does anyone know of any beginners classes which are being offered online?
>
> Thanks,
>
> Eki
>
Go to www.python.org and look for the tutorial under "Documentation". There
are also links to lots of other good stuff there.

regards
 Steve




From smnordby at yahoo.com  Thu Feb  8 13:56:09 2001
From: smnordby at yahoo.com (Steve)
Date: Thu, 08 Feb 2001 04:56:09 -0800
Subject: Newbie Query
References: <Y7pg6.9822$En4.180927@typhoon.jacksonville.mediaone.net>
Message-ID: <3A829769.141DE1A4@yahoo.com>

You may find Alan Gauld's online tutorial very helpful since he gives
examples in both BASIC and Python
http://www.crosswinds.net/~agauld/

-Steve-

eki wrote:
> 
> Hello all,
> 
> I am just getting restarted in programming after learning a bit of BASIC
> back in the seventies. I would like to take a course online in programming
> in python.
> 
> Does anyone know of any beginners classes which are being offered online?
> 
> Thanks,
> 
> Eki


From -$Paul$- at verence.demon.co.uk  Thu Feb  8 19:49:09 2001
From: -$Paul$- at verence.demon.co.uk (Paul Wright)
Date: 8 Feb 2001 18:49:09 -0000
Subject: Newbie Query
References: <Y7pg6.9822$En4.180927@typhoon.jacksonville.mediaone.net>
Message-ID: <95upn5$avv$1@verence.demon.co.uk>

In article <Y7pg6.9822$En4.180927 at typhoon.jacksonville.mediaone.net>,
eki <ekibismuth at privacyx.com> wrote:
>I am just getting restarted in programming after learning a bit of BASIC
>back in the seventies. I would like to take a course online in programming
>in python.
>
>Does anyone know of any beginners classes which are being offered online?

http://www.livewires.org.uk/python/

Declaration of interests: I'm the maintainer for this at the moment.

-- 
----- Paul Wright ------| Shin: a device for finding furniture in the dark.
-paul.wright at pobox.com--| 
http://pobox.com/~pw201 | 


From lluang at northwestern.edu  Thu Feb  8 20:27:35 2001
From: lluang at northwestern.edu (Louis Luangkesorn)
Date: Thu, 08 Feb 2001 13:27:35 -0600
Subject: Newbie Query
References: <Y7pg6.9822$En4.180927@typhoon.jacksonville.mediaone.net>
Message-ID: <3A82F327.71D7A668@northwestern.edu>

The following article, Python Books Online by Stephen Figgens is about online
books that are aimed at people learning how to program using Python as a
language of instruction (as opposed to people who know how to program and are
trying to learn Python)  It seems to be directed directly at you.  It points
to Alan Gauld's text which Steve referred to as well as  How to Think Like a
Computer Scientist by Allen B. Downey and Jeffrey Elkner, which has been used
as a text for High School Computer science courses (and also has C++ and Java
counterparts).

http://www.oreillynet.com/pub/a/python/2001/02/07/pythonnews.html

eki wrote:

> Hello all,
>
> I am just getting restarted in programming after learning a bit of BASIC
> back in the seventies. I would like to take a course online in programming
> in python.
>
> Does anyone know of any beginners classes which are being offered online?
>
> Thanks,
>
> Eki

--
Louis Luangkesorn
lluang at northwestern.edu   http://pubweb.nwu.edu/~kll560
Who am I? They mock me, these lonely questions of mine. Whoever I am, Thou
knowest, O God, I am thine.
        - Dietrich Bonhoeffer "Who am I?"




From danielk at aracnet.com  Thu Feb  8 06:02:51 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Wed, 07 Feb 2001 21:02:51 -0800
Subject: from/import and reload()
Message-ID: <en948tkqvcd14vqu78ojt7c59jl8h2rob3@4ax.com>

Is there a way to reload a module in interactive mode if the module was
originally loaded using

	from module import *

I suppose I could delete all of the namespace elements but for larger modules
this would be impractical.

Daniel Klein

"One's name is the most important possession one has."
				-- Anonymous

[ please also respond directly to me as my news server does not pick up all
news messages. Thanks. ]


From fredrik at pythonware.com  Thu Feb  8 07:42:42 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 08 Feb 2001 06:42:42 GMT
Subject: from/import and reload()
References: <en948tkqvcd14vqu78ojt7c59jl8h2rob3@4ax.com>
Message-ID: <Cdrg6.13515$AH6.2023467@newsc.telia.net>

Daniel Klein wrote:
> Is there a way to reload a module in interactive mode if the module was
> originally loaded using
>
> from module import *

import module # to get a module object
reload(module) # to reload the module
from module import * # to reimport all public names

Cheers /F




From gs234 at cam.ac.uk  Thu Feb  8 06:28:16 2001
From: gs234 at cam.ac.uk (Gaute B Strokkenes)
Date: 08 Feb 2001 05:28:16 +0000
Subject: Sorting strings.
Message-ID: <4a1yt93frj.fsf@kern.srcf.societies.cam.ac.uk>

Hello!

One of my current pet projects is a python script that parses and
sorts a library catalogue.  The catalogue is stored as a plain text
file, and all that needs doing is to read in the author, title,
etc. of each book, with some cleverness to detect multi-line entries,
comments etc.

After all this is done, it is time to sort the entries in appropriate
order before we print them out again.  Currently I do this:

def entry_cmp(a, b):
    if a.author < b.author:
        return -1
    if a.author > b.author:
        return 1
    if a.title < b.title:
        return -1
    if a.title > b.title:
        return 1
    if a.prefix < b.prefix:
        return -1
    if a.prefix > b.prefix:
        return 1
    if a.code < b.code:
        return -1
    if a.code > b.code:
        return 1
    return 0

...

ll.sort(entry_cmp)

The problem with this approach is that string comparison doesn't quite
have the behaviour that I would like.  For instance, I'd like numbers
to come after letters, and I'd like to be case insensitive.  Is there
a reasonably clean way to do this?

-- 
Big Gaute                               http://www.srcf.ucam.org/~gs234/
I'm having a BIG BANG THEORY!!


From gregj at pobox.com  Thu Feb  8 08:54:15 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 08 Feb 2001 07:54:15 GMT
Subject: Sorting strings.
References: <4a1yt93frj.fsf@kern.srcf.societies.cam.ac.uk>
Message-ID: <95tjb7$4ic$1@nnrp1.deja.com>

In article <4a1yt93frj.fsf at kern.srcf.societies.cam.ac.uk>,
  Gaute B Strokkenes <gs234 at cam.ac.uk> wrote:
> ...
> The problem with this approach is that string comparison doesn't quite
> have the behaviour that I would like.  For instance, I'd like numbers
> to come after letters, and I'd like to be case insensitive.  Is there
> a reasonably clean way to do this?

You can write a function to compare strings according to your rules:
ignore case, make the digits greater than letters, and (maybe) ignore
punctuation and anything else not alphanumeric. Then use that function
to compare author, title, etc. rather than <, ==, >.

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From tismer at tismer.com  Thu Feb  8 12:43:48 2001
From: tismer at tismer.com (Christian Tismer)
Date: Thu, 08 Feb 2001 12:43:48 +0100
Subject: Sorting strings.
References: <4a1yt93frj.fsf@kern.srcf.societies.cam.ac.uk>
Message-ID: <3A828674.177DFEAE@tismer.com>


Gaute B Strokkenes wrote:
> 
> Hello!
> 
> One of my current pet projects is a python script that parses and
> sorts a library catalogue.  The catalogue is stored as a plain text
> file, and all that needs doing is to read in the author, title,
> etc. of each book, with some cleverness to detect multi-line entries,
> comments etc.
> 
> After all this is done, it is time to sort the entries in appropriate
> order before we print them out again.  Currently I do this:
> 
> def entry_cmp(a, b):
>     if a.author < b.author:
>         return -1
>     if a.author > b.author:
>         return 1
>     if a.title < b.title:
>         return -1
>     if a.title > b.title:
>         return 1
>     if a.prefix < b.prefix:
>         return -1
>     if a.prefix > b.prefix:
>         return 1
>     if a.code < b.code:
>         return -1
>     if a.code > b.code:
>         return 1
>     return 0
> 
> ...
> 
> ll.sort(entry_cmp)
> 
> The problem with this approach is that string comparison doesn't quite
> have the behaviour that I would like.  For instance, I'd like numbers
> to come after letters, and I'd like to be case insensitive.  Is there
> a reasonably clean way to do this?

There are several ways. As someone else pointed out, you can of
course make your cmp function as sophisticated as necessary.

Slightly less powerful, but reasonably faster is to calculate
a suitable key field (which might be a tuple) and do the
sorting on that, without supplying an extra cmp function.

Whatever you choose, you need to process your data a little,
either during comparison, or in preparation of the whole
sort.

The following seems quite natural to me:
For case insensitive sorting, use string.upper() to build
a case insensitive key. For proper sorting by numbers,
parse the numbers off from your string, and turn them
into real integer numbers.

You want numbers to come after letters.
This is a little uncomfortable, since comparison by
default assumes that strings are bigger than numbers.
In a cmp() implementation, you can address this directly.
But if you use the key pre-generation method, you need
to modify your key strings a little to make that work as
expected.

Here a function that parses one of your strings and splits
it into a (string, number, string) tuple:

import string, re

rex = re.compile("([^0-9]*)([0-9]*)(.*)")

def create_key(s):
  s = string.upper(s)
  s1, num, s2 = rex.match(s).groups()
  try:
    num = int(num)
    s1 = s1 + "x"
  except:
    num = -1
  return s1, num, s2

This function takes a string and creates a tuple for you
which contains a number matched inside your string, or
-1 if there is none. The sorting order is case independent,
and numbers go after strings. The latter is accomplished
by appending "x" to s1, if we found a number.

Example:

>>> x=map(create_key, ("ab", "ab1", "abc"))
>>> x.sort()
>>> x
[('AB', -1, ''), ('ABC', -1, ''), ('ABx', 1, '')]

Some mapping like this should be done for all of your
keys to sort, in the right order. Then you put all these
keys together to one tuple, and you attach your real
data to the end.

Example, just for data like (author, title):

def sortit(data_list):
  sorter = []
  for author, title in data_list:
  sorter.append( (create_key(author), create_key(title), (author, title)) )
  #                     # here a key     # there a key   # and here the data
  sorter.sort()
  result = []
  for record in sorter:
    result.append(record[-1]) # just the data
  return result

hope this helps - ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com



From gtalvola at nameconnector.com  Thu Feb  8 15:58:05 2001
From: gtalvola at nameconnector.com (Geoff Talvola)
Date: Thu, 08 Feb 2001 09:58:05 -0500
Subject: Sorting strings.
References: <4a1yt93frj.fsf@kern.srcf.societies.cam.ac.uk> <3A828674.177DFEAE@tismer.com>
Message-ID: <3A82B3FC.42BD0D68@NameConnector.com>

Christian Tismer wrote:

> Gaute B Strokkenes wrote:
>
> > The problem with this approach is that string comparison doesn't quite
> > have the behaviour that I would like.  For instance, I'd like numbers
> > to come after letters, and I'd like to be case insensitive.  Is there
> > a reasonably clean way to do this?
>
> There are several ways. As someone else pointed out, you can of
> course make your cmp function as sophisticated as necessary.
>
> Slightly less powerful, but reasonably faster is to calculate
> a suitable key field (which might be a tuple) and do the
> sorting on that, without supplying an extra cmp function.
>

Here's a utility function that makes it easy to sort by providing
a function that constructs a sort key:

def sortUsingKeyFunc(l, keyfunc):
    l = [(keyfunc(x), x) for x in l]
    l.sort()
    return [y for x,y in l]

To use it, simply provide a "keyfunc" that constructs a sort key
for the items in the list.  For example, to sort a list of strings
alphabetically:

>>> l = ['E', 'd', 'A', 'b', 'C']
>>> import string
>>> sortUsingKeyFunc(l, string.upper)
['A', 'b', 'C', 'd', 'E']

Or to sort a list of dictionaries on "name" case-insensitively,
and secondarily on "value":

>>> l = [{'name':'AAA', 'value':5},
...      {'name':'aaa', 'value':4},
...      {'name':'AAA', 'value':3},
...      {'name':'bbb', 'value':2}]
>>> def keyfunc(d):
...     return d['name'].upper(), d['value']
...
>>> from pprint import pprint
>>> pprint(sortUsingKeyFunc(l, keyfunc))
[{'value': 3, 'name': 'AAA'},
 {'value': 4, 'name': 'aaa'},
 {'value': 5, 'name': 'AAA'},
 {'value': 2, 'name': 'bbb'}]

Note that sortUsingKeyFunc returns a new list -- the original list is unmodified.

--


- Geoff Talvola
  Parlance Corporation
  gtalvola at NameConnector.com




From aleaxit at yahoo.com  Thu Feb  8 13:29:21 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 8 Feb 2001 13:29:21 +0100
Subject: Sorting strings.
References: <4a1yt93frj.fsf@kern.srcf.societies.cam.ac.uk>
Message-ID: <95u3f20utd@news2.newsguy.com>

"Gaute B Strokkenes" <gs234 at cam.ac.uk> wrote in message
news:4a1yt93frj.fsf at kern.srcf.societies.cam.ac.uk...
    [snip]
> After all this is done, it is time to sort the entries in appropriate
> order before we print them out again.  Currently I do this:
>
> def entry_cmp(a, b):
>     if a.author < b.author:
>         return -1
>     if a.author > b.author:
>         return 1
>     if a.title < b.title:
>         return -1
>     if a.title > b.title:
>         return 1
>     if a.prefix < b.prefix:
>         return -1
>     if a.prefix > b.prefix:
>         return 1
>     if a.code < b.code:
>         return -1
>     if a.code > b.code:
>         return 1
>     return 0
>
> ...
>
> ll.sort(entry_cmp)
>
> The problem with this approach is that string comparison doesn't quite
> have the behaviour that I would like.  For instance, I'd like numbers
> to come after letters, and I'd like to be case insensitive.  Is there
> a reasonably clean way to do this?

Not only there is, but it may make your whole program faster.  The
key idea is to apply sort, *WITHOUT* an argument, to an _auxiliary_
list, whose items are tuples -- each corresponding to a "suitably
transformed version" of the corresponding item you want to sort,
followed by said 'corresponding item'.  After the sort, you pick
the 'corresponding item's only, discarding the auxiliary fields
that you computed once just for sorting purposes.  (I'm used to
calling this design pattern a 'Schwartzian transform' but every time
I do so somebody nitpicks about it:-).

To ensure comparison is case-insensitive you need to map all
letters to (e.g.) uppercase, and to ensure digits come after
letters you must map them to characters that follow the letters
in the alphabet (if you map letters to uppercase, you may at
the same time map digits 0=9 to e.g. letters a-j, which do
come after the uppercase ones).

For these transformations, you use two functions in the
string module -- maketrans prepares a transformation matrix,
translate applies it.  You'll have problems in locales where
the numbers of uppercase and lowercase letters differ, but
that is a complication best met separately.  In normal locales:

import string
assert(len(string.lowercase) == len(string.uppercase))
from =  string.lowercase + string.digits
onto = (string.uppercase + string.lowercase)[:len(from)]
transform = string.maketrans(from, onto)

and now you can call string.translate(x, transform) to
get the transformed version of any string x.

So, your sort task can now become:

def mysort(alist):
    def ok(x):
        return string.translate(x, transform)
    temp = [(ok(x.author), ok(x.title),
        ok(x.prefix), ok(x.code), x) for x in alist]
    temp.sort()
    return [ x[-1] for x in temp ]

and you'll call it as

li = mysort(li)


The order of items with identical (transformed) author,
title, prefix and code will vary between the two versions,
but since .sort is not a stable sort you may not care all
that much anyway.  It's easy to make THIS sort stable,
by the way -- just consider "index" as the penultimate
entry in each tuple, e.g.:

    temp = [(ok(alist[i].author), ok(alist[i].title),
        ok(alist[i].prefix), ok(alist[i].code), i)
            for i in range(len(alist))]
    temp.sort()
    return [ alist[x[-1]] for x in temp ]


I would consider this 'reasonably clean' or better... it
makes runtime generally faster, and source code more compact,
while adding more flexibility to how sorting is performed...
few program-design 'trade-offs' are as one-sided as this!-)


Alex





From apederse at my-deja.com  Thu Feb  8 08:17:30 2001
From: apederse at my-deja.com (apederse at my-deja.com)
Date: Thu, 08 Feb 2001 07:17:30 GMT
Subject: urllib.open('http://www.redherring.com/')
Message-ID: <95th67$36j$1@nnrp1.deja.com>

I have some problems with urllib. Any URL works
just fine except http://www.redherring.com/ which
returns nothing, no headers, no nothing. (Red
Herring works fine from any browser). I
have tried to rule out any possible causes, such
as cache and problems regarding headerfields by
emulating a real browser request from netscape and
konqueror (using fancyURLopener) and I have
monitored network traffic using tcpdump.  Using
tcpdump it looks like there is a request sendt to
vip.redherring.com and a response coming back.
Could anyone look into this problem and see if
this is related my particular computer and network
configurations or if it is a problem lying within
urllib or python.

best regards,
/Asle


Sent via Deja.com
http://www.deja.com/


From gregj at pobox.com  Thu Feb  8 08:49:55 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 08 Feb 2001 07:49:55 GMT
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com>
Message-ID: <95tj33$4g1$1@nnrp1.deja.com>

In article <95th67$36j$1 at nnrp1.deja.com>,
  apederse at my-deja.com wrote:
> I have some problems with urllib. Any URL works
> just fine except http://www.redherring.com/ which
> returns nothing, no headers, no nothing

It's broken for me, too. I can get to the site with my browser, but
urllib.urlopen() hangs for a long time, then aborts with a "connection
reset by peer" error. The same thing happens with httplib.

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From andrew at intertrader.com  Thu Feb  8 09:35:52 2001
From: andrew at intertrader.com (Andrew Cooke)
Date: Thu, 08 Feb 2001 08:35:52 +0000
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com>
Message-ID: <3A825A68.B1023097@intertrader.com>

Maybe they're checking an HTTP header line and only responding to known
browsers?

Andrew

Greg Jorgensen wrote:
> 
> In article <95th67$36j$1 at nnrp1.deja.com>,
>   apederse at my-deja.com wrote:
> > I have some problems with urllib. Any URL works
> > just fine except http://www.redherring.com/ which
> > returns nothing, no headers, no nothing
> 
> It's broken for me, too. I can get to the site with my browser, but
> urllib.urlopen() hangs for a long time, then aborts with a "connection
> reset by peer" error. The same thing happens with httplib.
> 
> --
> Greg Jorgensen
> Portland, Oregon, USA
> gregj at pobox.com
> 
> Sent via Deja.com
> http://www.deja.com/


From gregj at pobox.com  Thu Feb  8 10:53:00 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Thu, 08 Feb 2001 09:53:00 GMT
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com>
Message-ID: <95tq9r$afs$1@nnrp1.deja.com>

In article <3A825A68.B1023097 at intertrader.com>,
  Andrew Cooke <andrew at intertrader.com> wrote:
>
> Maybe they're checking an HTTP header line and only responding to
known
> browsers?

I tried sending some standard headers, but it still hung and then came
back with "connection reset by peer."

``````````````````````````````````
import httplib

h = httplib.HTTP("redherring.com")
h.connect()
h.putrequest("GET","/")
h.putheader("Accept", "text/html")
h.putheader("Accept", "text/plain")
h.putheader("User-Agent", "Mozilla/4.08 [en] (Windows NT 5.0)")
h.putheader("Accept", "image/gif, image/jpeg")
h.putheader("Accept-Language", "en")
h.putheader("Accept-Charset", "iso-8859-1,*,utf-8")
h.endheaders()

errcode, errmsg, headers = h.getreply()
f = h.getfile()
s = f.read()
f.close()

print h
print "content length: %d" % len(s)
``````````````````````````````````````````


--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From apederse at my-deja.com  Fri Feb  9 08:48:21 2001
From: apederse at my-deja.com (apederse at my-deja.com)
Date: Fri, 09 Feb 2001 07:48:21 GMT
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com>
Message-ID: <9607c5$bkj$1@nnrp1.deja.com>

In article <95tq9r$afs$1 at nnrp1.deja.com>,
  Greg Jorgensen <gregj at pobox.com> wrote:
> In article <3A825A68.B1023097 at intertrader.com>,
>   Andrew Cooke <andrew at intertrader.com> wrote:
> >
> > Maybe they're checking an HTTP header line and only responding to
> known
> > browsers?
>
> I tried sending some standard headers, but it still hung and then came
> back with "connection reset by peer."
>
> ``````````````````````````````````
> import httplib
>
> h = httplib.HTTP("redherring.com")
> h.connect()
> h.putrequest("GET","/")
> h.putheader("Accept", "text/html")
> h.putheader("Accept", "text/plain")
> h.putheader("User-Agent", "Mozilla/4.08 [en] (Windows NT 5.0)")
> h.putheader("Accept", "image/gif, image/jpeg")
> h.putheader("Accept-Language", "en")
> h.putheader("Accept-Charset", "iso-8859-1,*,utf-8")
> h.endheaders()
>
> errcode, errmsg, headers = h.getreply()
> f = h.getfile()
> s = f.read()
> f.close()
>
> print h
> print "content length: %d" % len(s)
> ``````````````````````````````````````````
>
> --
> Greg Jorgensen
> Portland, Oregon, USA
> gregj at pobox.com
>
> Sent via Deja.com
> http://www.deja.com/
>

to me it looks like the problem lies whithin python itself? is there any
point in reporting such bugs when everyone will soon be using python 2.0
anyway?

/Asle


Sent via Deja.com
http://www.deja.com/


From stephen_purcell at yahoo.com  Fri Feb  9 09:59:58 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 9 Feb 2001 09:59:58 +0100
Subject: Everyone will soon be using 2.0? (was: urllib.open('http://www.redherring.com/'))
In-Reply-To: <9607c5$bkj$1@nnrp1.deja.com>; from apederse@my-deja.com on Fri, Feb 09, 2001 at 07:48:21AM +0000
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com>
Message-ID: <20010209095958.A15906@freedom.puma-ag.com>

apederse at my-deja.com wrote:
> 
> to me it looks like the problem lies whithin python itself? is there any
> point in reporting such bugs when everyone will soon be using python 2.0
> anyway?
> 

Will they? There's a whole slew of people who will still be with 1.5.2 until
the likes of Redhat & Debian release Python 2 packages with an official
release. Some of the changes from 1.5.2 to 2.0 are very significant
(byte-code and c-extension incompatibility) and I can't imagine that the
major distributions will be able to standardise on 2.0 until all the
python add-ons have been re-built, tested and packaged too.

I for one will be sticking with 1.5.2 until there are 2.0 versions of every
add-on I use (pygresql, PIL and others) available or easily installable on
the majority of Linux boxes in my organisation and beyond. That will be at
least 6 months, in my estimation. (For Windows users, this time will be shorter
because they are used to gathering EXEs from random locations on the internet.)

Further, I think that people will be starting new projects with 1.5.2 for some
time yet, in the same way that there are Java projects starting off now with
JDK 1.1.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From apederse at my-deja.com  Fri Feb  9 13:00:45 2001
From: apederse at my-deja.com (apederse at my-deja.com)
Date: Fri, 09 Feb 2001 12:00:45 GMT
Subject: Everyone will soon be using 2.0? (was: urllib.open('http://www.redherring.com/'))
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com> <mailman.981709353.11863.python-list@python.org>
Message-ID: <960m5a$n5i$1@nnrp1.deja.com>

In article <mailman.981709353.11863.python-list at python.org>,
  Steve Purcell <stephen_purcell at yahoo.com> wrote:
> apederse at my-deja.com wrote:
> >
> > to me it looks like the problem lies whithin python itself? is there
any
> > point in reporting such bugs when everyone will soon be using python
2.0
> > anyway?
> >
>
> Will they? There's a whole slew of people who will still be with 1.5.2
until
> the likes of Redhat & Debian release Python 2 packages with an
official
> release. Some of the changes from 1.5.2 to 2.0 are very significant
> (byte-code and c-extension incompatibility) and I can't imagine that
the
> major distributions will be able to standardise on 2.0 until all the
> python add-ons have been re-built, tested and packaged too.
>
> I for one will be sticking with 1.5.2 until there are 2.0 versions of
every
> add-on I use (pygresql, PIL and others) available or easily
installable on
> the majority of Linux boxes in my organisation and beyond. That will
be at
> least 6 months, in my estimation. (For Windows users, this time will
be shorter
> because they are used to gathering EXEs from random locations on the
internet.)
>
> Further, I think that people will be starting new projects with 1.5.2
for some
> time yet, in the same way that there are Java projects starting off
now with
> JDK 1.1.
>
> -Steve
>
> --
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Available for consulting and training.
> "Even snakes are afraid of snakes." -- Steven Wright
>
>

So, what is the procedure like when reporting bugs? There should
probably be more people looking into the problem first? ..And I do agree
with you on sticking to 1.5.2 for a while still. Myself, I will wait for
the next release of Redhat before I start porting to Python 2.0.

/Asle


Sent via Deja.com
http://www.deja.com/


From akuchlin at mems-exchange.org  Fri Feb  9 16:07:10 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 09 Feb 2001 10:07:10 -0500
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com>
Message-ID: <3ditmj99pd.fsf@ute.cnri.reston.va.us>

apederse at my-deja.com writes:
> to me it looks like the problem lies whithin python itself? is there any
> point in reporting such bugs when everyone will soon be using python 2.0
> anyway?

I don't believe urllib or httplib were massively rewritten for 2.0
(nor will they be in 2.1), so if it's a bug, it's probably still
there.  On the other hand, if you try it with 2.0 and it works fine,
then there won't be much interest; a 1.5.3 release is not going to
happen.

--amk



From apederse at my-deja.com  Fri Feb  9 16:53:08 2001
From: apederse at my-deja.com (apederse at my-deja.com)
Date: Fri, 09 Feb 2001 15:53:08 GMT
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com> <3ditmj99pd.fsf@ute.cnri.reston.va.us>
Message-ID: <9613ov$367$1@nnrp1.deja.com>

In article <3ditmj99pd.fsf at ute.cnri.reston.va.us>,
  Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> apederse at my-deja.com writes:
> > to me it looks like the problem lies whithin python itself? is there
any
> > point in reporting such bugs when everyone will soon be using python
2.0
> > anyway?
>
> I don't believe urllib or httplib were massively rewritten for 2.0
> (nor will they be in 2.1), so if it's a bug, it's probably still
> there.  On the other hand, if you try it with 2.0 and it works fine,
> then there won't be much interest; a 1.5.3 release is not going to
> happen.
>
> --amk
>
>

a 1.5.3 release is a bit drastic ...a simple bugfix will do!

/Asle


Sent via Deja.com
http://www.deja.com/


From fredrik at pythonware.com  Fri Feb  9 17:22:02 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 09 Feb 2001 16:22:02 GMT
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com> <3ditmj99pd.fsf@ute.cnri.reston.va.us>
Message-ID: <KOUg6.14337$AH6.2082159@newsc.telia.net>

Andrew Kuchling wrote:
> I don't believe urllib or httplib were massively rewritten for 2.0
> (nor will they be in 2.1), so if it's a bug, it's probably still there.

on my box, it sometimes work under 1.5.2, and never
works under an unmodified 2.0 installation.

however, the following hack works for both versions
(note the explicit socket shutdown).

    import httplib

    h = httplib.HTTP("www.redherring.com")
    h.set_debuglevel(1)
    h.putrequest("GET","/")
    h.endheaders()

    try:
        h.sock.shutdown(1) # 1.5.2
    except AttributeError:
        h._conn.sock.shutdown(1) # 2.0

    errcode, errmsg, headers = h.getreply()
    f = h.getfile()
    s = f.read()
    f.close()

    print "content length: %d" % len(s)

this prints:

    connect: (www.redherring.com, 80)
    send: 'GET / HTTP/1.0\015\012'
    send: '\015\012'
    reply: 'HTTP/1.1 200 OK\015\012'
    header: Server: Netscape-Enterprise/4.0
    header: Date: Fri, 09 Feb 2001 16:15:36 GMT
    header: Set-cookie: AnalysisUserId=28177981735337; path=/;
        expires=Friday, 31 Dec-2010 23:59:59 GMT
    header: AnalysisUserId: 28177981735337
    header: Content-type: text/html
    header: Connection: close

Cheers /F




From andrew at intertrader.com  Fri Feb  9 20:14:46 2001
From: andrew at intertrader.com (Andrew Cooke)
Date: Fri, 09 Feb 2001 19:14:46 +0000
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com> <3ditmj99pd.fsf@ute.cnri.reston.va.us> <KOUg6.14337$AH6.2082159@newsc.telia.net>
Message-ID: <3A8441A6.BF6333FF@intertrader.com>

Just a thought (I'm sure you've checked), but if it needs the socket
closed what's happening at your end about persistence and/or content
length?  If it happened on every site it would be pretty convincing
evidence of content-length in the request being too large (IMHO).  Maybe
their server is broken wrt content length in some way?

Andrew

Fredrik Lundh wrote:
> 
> Andrew Kuchling wrote:
> > I don't believe urllib or httplib were massively rewritten for 2.0
> > (nor will they be in 2.1), so if it's a bug, it's probably still there.
> 
> on my box, it sometimes work under 1.5.2, and never
> works under an unmodified 2.0 installation.
> 
> however, the following hack works for both versions
> (note the explicit socket shutdown).
> 
>     import httplib
> 
>     h = httplib.HTTP("www.redherring.com")
>     h.set_debuglevel(1)
>     h.putrequest("GET","/")
>     h.endheaders()
> 
>     try:
>         h.sock.shutdown(1) # 1.5.2
>     except AttributeError:
>         h._conn.sock.shutdown(1) # 2.0
> 
>     errcode, errmsg, headers = h.getreply()
>     f = h.getfile()
>     s = f.read()
>     f.close()
> 
>     print "content length: %d" % len(s)
> 
> this prints:
> 
>     connect: (www.redherring.com, 80)
>     send: 'GET / HTTP/1.0\015\012'
>     send: '\015\012'
>     reply: 'HTTP/1.1 200 OK\015\012'
>     header: Server: Netscape-Enterprise/4.0
>     header: Date: Fri, 09 Feb 2001 16:15:36 GMT
>     header: Set-cookie: AnalysisUserId=28177981735337; path=/;
>         expires=Friday, 31 Dec-2010 23:59:59 GMT
>     header: AnalysisUserId: 28177981735337
>     header: Content-type: text/html
>     header: Connection: close
> 
> Cheers /F


From apederse at my-deja.com  Mon Feb 12 11:42:48 2001
From: apederse at my-deja.com (apederse at my-deja.com)
Date: Mon, 12 Feb 2001 10:42:48 GMT
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com> <3ditmj99pd.fsf@ute.cnri.reston.va.us> <KOUg6.14337$AH6.2082159@newsc.telia.net>
Message-ID: <968en7$nhn$1@nnrp1.deja.com>

In article <KOUg6.14337$AH6.2082159 at newsc.telia.net>,
  "Fredrik Lundh" <fredrik at effbot.org> wrote:
> Andrew Kuchling wrote:
> > I don't believe urllib or httplib were massively rewritten for 2.0
> > (nor will they be in 2.1), so if it's a bug, it's probably still there.
>
> on my box, it sometimes work under 1.5.2, and never
> works under an unmodified 2.0 installation.
>
> however, the following hack works for both versions
> (note the explicit socket shutdown).
>

is it possible to do a similar fix to urllib? (For convenience reasons I
always use urllib in my applications...)

btw. I'm not very familiar with how the Python community works, but I'm
thinking at this point, shouldn't whoever that has been involved in the
programming of httplib and urllib be informed about the problems?

/Asle


Sent via Deja.com
http://www.deja.com/


From jeremy at alum.mit.edu  Mon Feb 12 16:00:03 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Mon, 12 Feb 2001 10:00:03 -0500 (EST)
Subject: urllib.open('http://www.redherring.com/')
In-Reply-To: <968en7$nhn$1@nnrp1.deja.com>
References: <95th67$36j$1@nnrp1.deja.com>
	<95tj33$4g1$1@nnrp1.deja.com>
	<3A825A68.B1023097@intertrader.com>
	<95tq9r$afs$1@nnrp1.deja.com>
	<9607c5$bkj$1@nnrp1.deja.com>
	<3ditmj99pd.fsf@ute.cnri.reston.va.us>
	<KOUg6.14337$AH6.2082159@newsc.telia.net>
	<968en7$nhn$1@nnrp1.deja.com>
Message-ID: <14983.64115.291676.385646@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "AP" == apederse  <apederse at my-deja.com> writes:

  AP> In article <KOUg6.14337$AH6.2082159 at newsc.telia.net>,
  AP>   "Fredrik Lundh" <fredrik at effbot.org> wrote:
  >> Andrew Kuchling wrote:
  >> > I don't believe urllib or httplib were massively rewritten for
  >> > 2.0 (nor will they be in 2.1), so if it's a bug, it's probably
  >> > still there.
  >>
  >> on my box, it sometimes work under 1.5.2, and never works under
  >> an unmodified 2.0 installation.
  >>
  >> however, the following hack works for both versions (note the
  >> explicit socket shutdown).

  AP> is it possible to do a similar fix to urllib? (For convenience
  AP> reasons I always use urllib in my applications...)

  AP> btw. I'm not very familiar with how the Python community works,
  AP> but I'm thinking at this point, shouldn't whoever that has been
  AP> involved in the programming of httplib and urllib be informed
  AP> about the problems?

The preferred mechanism for reporting bugs is the SourceForge Bug
Tracker: http://sourceforge.net/bugs/?group_id=5470

Jeremy



From apederse at my-deja.com  Mon Feb 12 11:43:28 2001
From: apederse at my-deja.com (apederse at my-deja.com)
Date: Mon, 12 Feb 2001 10:43:28 GMT
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <95tj33$4g1$1@nnrp1.deja.com> <3A825A68.B1023097@intertrader.com> <95tq9r$afs$1@nnrp1.deja.com> <9607c5$bkj$1@nnrp1.deja.com> <3ditmj99pd.fsf@ute.cnri.reston.va.us> <KOUg6.14337$AH6.2082159@newsc.telia.net>
Message-ID: <968eog$ni4$1@nnrp1.deja.com>

In article <KOUg6.14337$AH6.2082159 at newsc.telia.net>,
  "Fredrik Lundh" <fredrik at effbot.org> wrote:
> Andrew Kuchling wrote:
> > I don't believe urllib or httplib were massively rewritten for 2.0
> > (nor will they be in 2.1), so if it's a bug, it's probably still there.
>
> on my box, it sometimes work under 1.5.2, and never
> works under an unmodified 2.0 installation.
>
> however, the following hack works for both versions
> (note the explicit socket shutdown).
>

is it possible to do a similar fix to urllib? (For convenience reasons I
always use urllib in my applications...)

btw. I'm not very familiar with how the Python community works, but I'm
thinking at this point, shouldn't whoever that has been involved in the
programming of httplib and urllib be informed about the problems?

/Asle


Sent via Deja.com
http://www.deja.com/


From duncan at rcp.co.uk  Thu Feb  8 11:11:58 2001
From: duncan at rcp.co.uk (Duncan Booth)
Date: Thu, 8 Feb 2001 10:11:58 +0000 (UTC)
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com>
Message-ID: <Xns9042677DFF798duncanrcpcouk@127.0.0.1>

apederse at my-deja.com wrote in <95th67$36j$1 at nnrp1.deja.com>:

>I have some problems with urllib. Any URL works
>just fine except http://www.redherring.com/ which
>returns nothing, no headers, no nothing. (Red
>Herring works fine from any browser).

It works fine when I try it (using Python 2.0). Changing things like the 
accept header which might be expected to make a difference don't seem to 
have any effect.


From aahz at panix.com  Fri Feb  9 16:36:49 2001
From: aahz at panix.com (Aahz Maruch)
Date: 9 Feb 2001 07:36:49 -0800
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com>
Message-ID: <9612qh$kmd$1@panix2.panix.com>

In article <95th67$36j$1 at nnrp1.deja.com>,  <apederse at my-deja.com> wrote:
>
>I have some problems with urllib. Any URL works just fine except
>http://www.redherring.com/ which returns nothing, no headers, no
>nothing. (Red Herring works fine from any browser). I have tried to
>rule out any possible causes, such as cache and problems regarding
>headerfields by emulating a real browser request from netscape and
>konqueror (using fancyURLopener) and I have monitored network traffic
>using tcpdump.  Using tcpdump it looks like there is a request sendt
>to vip.redherring.com and a response coming back.  Could anyone look
>into this problem and see if this is related my particular computer and
>network configurations or if it is a problem lying within urllib or
>python.

I've just tried with a straight telnet session, and that also fails.
So, it's not just Python.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From nas at arctrix.com  Fri Feb  9 17:28:36 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Fri, 9 Feb 2001 08:28:36 -0800
Subject: urllib.open('http://www.redherring.com/')
In-Reply-To: <9612qh$kmd$1@panix2.panix.com>; from aahz@panix.com on Fri, Feb 09, 2001 at 07:36:49AM -0800
References: <95th67$36j$1@nnrp1.deja.com> <9612qh$kmd$1@panix2.panix.com>
Message-ID: <20010209082836.B15525@glacier.fnational.com>

Aahz Maruch on www.redherring.com:
> I've just tried with a straight telnet session, and that also fails.
> So, it's not just Python.

It seems to require proper (ie "\r\n") line endings.

  Neil



From andrew at intertrader.com  Fri Feb  9 20:16:43 2001
From: andrew at intertrader.com (Andrew Cooke)
Date: Fri, 09 Feb 2001 19:16:43 +0000
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com> <9612qh$kmd$1@panix2.panix.com> <mailman.981736207.23829.python-list@python.org>
Message-ID: <3A84421B.A10AA8B0@intertrader.com>

Ah!  I've had this problem with other sites (not using Python).  There
is an obsscure server out there (can't remember the name of the company)
that is sensitive to this.

Andrew

Neil Schemenauer wrote:
> 
> Aahz Maruch on www.redherring.com:
> > I've just tried with a straight telnet session, and that also fails.
> > So, it's not just Python.
> 
> It seems to require proper (ie "\r\n") line endings.
> 
>   Neil


From chakie at infa.abo.fi  Thu Feb  8 08:20:56 2001
From: chakie at infa.abo.fi (Jan Ekholm)
Date: 8 Feb 2001 09:20:56 +0200
Subject: Geometric classes (lines, planes etc)?
Message-ID: <3a8248d8_1@newsflash.abo.fi>

Hi all,

For a project we would need to use some classes for manipulating geometric
entities. We need something like planes, lines, points etc, and
possibilities to manipulate and calculate with them. We do however not
want to draw them in any way, so I'm not looking for anything graphical
here, just something that allows me to represent simple geometric shapes
in space.

Is there anything available that we could use, or do we need to build our
own? A Google search turned up a *lot* about Tk (which is not what we
want) and about using Python with Qt and Gnome.

I apologise if this is a RTFM or FAQ. In that case please point me in the
right direction.

Kind regards,
	Chakie

-- 
--------------------+--------------------------------------------------------
Jan 'Chakie' Ekholm |     Balrog New Media    http://www.balrog.fi/
   Linux Inside     | I'm the blue screen of death, nobody hears your screams


From pearu at cens.ioc.ee  Thu Feb  8 10:48:36 2001
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Thu, 8 Feb 2001 11:48:36 +0200
Subject: Geometric classes (lines, planes etc)?
In-Reply-To: <3a8248d8_1@newsflash.abo.fi>
References: <3a8248d8_1@newsflash.abo.fi>
Message-ID: <Pine.LNX.4.21.0102081143220.27799-100000@kev.ioc.ee>

I have written a Python interface to cddlib, a library for computational
geometry. See

http://cens.ioc.ee/projects/polyhedron/

It does not define planes, lines, points, etc but they can be defined in
the terms of computational geometry. Depends what do you want to do with
planes, lines, etc....

Pearu

On 8 Feb 2001, Jan Ekholm wrote:

> For a project we would need to use some classes for manipulating geometric
> entities. We need something like planes, lines, points etc, and
> possibilities to manipulate and calculate with them. We do however not
> want to draw them in any way, so I'm not looking for anything graphical
> here, just something that allows me to represent simple geometric shapes
> in space.
> 
> Is there anything available that we could use, or do we need to build our
> own? A Google search turned up a *lot* about Tk (which is not what we
> want) and about using Python with Qt and Gnome.



From ten.llun at ttam  Fri Feb  9 09:39:25 2001
From: ten.llun at ttam (Matthew Schinckel)
Date: Fri, 09 Feb 2001 19:09:25 +1030
Subject: Geometric classes (lines, planes etc)?
References: <3a8248d8_1@newsflash.abo.fi>
Message-ID: <3a83ac09_7@news.chariot.net.au>

In message <3a8248d8_1 at newsflash.abo.fi>, Jan Ekholm wrote:
> 
> Hi all,
> 
> For a project we would need to use some classes for manipulating geometric
> entities. We need something like planes, lines, points etc, and
> possibilities to manipulate and calculate with them. We do however not
> want to draw them in any way, so I'm not looking for anything graphical
> here, just something that allows me to represent simple geometric shapes
> in space.
> 
> Is there anything available that we could use, or do we need to build our
> own? A Google search turned up a *lot* about Tk (which is not what we
> want) and about using Python with Qt and Gnome.

You can have a look at my rudimentary CAD framework - at this stage it does
only what you are after. (And is incomplete :-)

http://www.chariot.net.au/~jaq/matt/


-- 
Matthew Schinckel
matt at null dot net



From fgeiger at datec.at  Thu Feb  8 08:34:31 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Thu, 8 Feb 2001 08:34:31 +0100
Subject: ZODB: What does change if classes get persistent?
Message-ID: <95ti61$2p$1@newsreaderm1.core.theplanet.net>

Hello all,

1) "How do you lock your web app data" was a question I asked about a week
ago (many thanks to all who contributed!!!). Alex Martelli pointed me to try
working with a database. So I started to do so. A few days later I came
across A.M. Kuchling's contribution "Alpha release of ZODB programming
guide". As I plan to "upgrade" my cgi lessons to Zope, I thought it'd be a
good idea to try ZODB which is part of Zope. I downloaded the package from
AMK's site and gave it a try, because it indeed sounds rather promising,
what I read in the docs.

So I played around but soon ran into troubles:

I have a class

class TCarts(TCollection, Persistence.Persistent):
   def __init__(self):
      TCollection.__init__(self)

which inherits from TCollection. TCollection wraps a dictionary.

After mixin' in Persistence.Persistent I get the following error

TCollection.__init__(self, self)
TypeError: unbound method must be called with class instance 1st argument

upon creation.

Does anybody know what's going on here behind the scenes?


2) A few contributors of "How do you lock your web app data" stated that
using a lock file to protect data against multiple access is somewhat
problematic especially in case of app crahes. This is the very problem I
encountered with my ferw lines of ZODB app! This astonished me a bit. ZODB
provides such sophisticated functionality to a user's app but runs into
troubles because of a simple lock file? BTW, the error message is

File "C:\Program Files\Python20\ZODB\lock_file.py", line 115, in lock_file
raise error, (StorageSystemError: Could not lock the database file.
There must be another process that has opened the file.


Any suggestions how to proceed? Did I miss something substantial?

Many thanks in advance and best regards
Franz GEIGER






From akuchlin at mems-exchange.org  Thu Feb  8 16:35:23 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 08 Feb 2001 10:35:23 -0500
Subject: ZODB: What does change if classes get persistent?
References: <95ti61$2p$1@newsreaderm1.core.theplanet.net>
Message-ID: <3d1yt96vd0.fsf@ute.cnri.reston.va.us>

"Franz GEIGER" <fgeiger at datec.at> writes:
> class TCarts(TCollection, Persistence.Persistent):
>    def __init__(self):
>       TCollection.__init__(self)
> 
> TCollection.__init__(self, self)
> TypeError: unbound method must be called with class instance 1st argument
> 
> upon creation.
> 
> Does anybody know what's going on here behind the scenes?

Ooh, something I forgot to cover.  TCarts is no longer a regular
Python class, but an ExtensionClass; this breaks the usual convention
for calling superclass constructors. (Does Python 2.1 try to fix this?
Anyone from DC know?)

Anyway, the fix would be to use this instead:
     TCarts.inheritedAttribute('__init__')(self)

See http://www.digicool.com/releases/ExtensionClass/ for a lengthier
discussion (search for __init__).

> File "C:\Program Files\Python20\ZODB\lock_file.py", line 115, in lock_file
> raise error, (StorageSystemError: Could not lock the database file.
> There must be another process that has opened the file.

Use ZEO if you need to have multiple processes accessing data at the
same time.  The lowest-level class, the Storage class, isn't written
to handle concurrent access to the same files.  You probably couldn't
run two completely different instances of Oracle and point them at the
same disk partitions, either, and the ZODB's storages impose similar
limitations.

--amk


From fgeiger at datec.at  Fri Feb  9 08:01:01 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Fri, 9 Feb 2001 08:01:01 +0100
Subject: ZODB: What does change if classes get persistent?
References: <95ti61$2p$1@newsreaderm1.core.theplanet.net> <3d1yt96vd0.fsf@ute.cnri.reston.va.us>
Message-ID: <9604jg$76v$1@newsreaderm1.core.theplanet.net>

Thank you, Andrew,

that did the trick!

Best regards
Franz


"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3d1yt96vd0.fsf at ute.cnri.reston.va.us...
> "Franz GEIGER" <fgeiger at datec.at> writes:
> > class TCarts(TCollection, Persistence.Persistent):
> >    def __init__(self):
> >       TCollection.__init__(self)
> >
> > TCollection.__init__(self, self)
> > TypeError: unbound method must be called with class instance 1st
argument
> >
> > upon creation.
> >
> > Does anybody know what's going on here behind the scenes?
>
> Ooh, something I forgot to cover.  TCarts is no longer a regular
> Python class, but an ExtensionClass; this breaks the usual convention
> for calling superclass constructors. (Does Python 2.1 try to fix this?
> Anyone from DC know?)
>
> Anyway, the fix would be to use this instead:
>      TCarts.inheritedAttribute('__init__')(self)
>
> See http://www.digicool.com/releases/ExtensionClass/ for a lengthier
> discussion (search for __init__).
>
> > File "C:\Program Files\Python20\ZODB\lock_file.py", line 115, in
lock_file
> > raise error, (StorageSystemError: Could not lock the database file.
> > There must be another process that has opened the file.
>
> Use ZEO if you need to have multiple processes accessing data at the
> same time.  The lowest-level class, the Storage class, isn't written
> to handle concurrent access to the same files.  You probably couldn't
> run two completely different instances of Oracle and point them at the
> same disk partitions, either, and the ZODB's storages impose similar
> limitations.
>
> --amk




From mix77 at usa.net  Thu Feb  8 10:26:38 2001
From: mix77 at usa.net (Mix)
Date: Thu, 8 Feb 2001 11:26:38 +0200
Subject: CGI-Python Image viewing
Message-ID: <95tp4p$bev$1@ctb-nnrp2.saix.net>

I am new to CGI scripting AND Python. What I am trying to do is,
display an image on an html page that has been retrieved from a Postgre
database
using Python (cgi). My problem is : that image that not show up on the page,
regardless of where it is sitting on my html or cgi directory!
(I am running Apache Web server under Redhat 7.0 with the default linux
kernel).
But, I CAN display the image on a plain html page (i.e without the
scripting).

Please help!
Thanks




From anders.eriksson at morateknikutveckling.se  Thu Feb  8 13:42:38 2001
From: anders.eriksson at morateknikutveckling.se (Anders M Eriksson)
Date: Thu, 08 Feb 2001 13:42:38 +0100
Subject: CGI-Python Image viewing
References: <95tp4p$bev$1@ctb-nnrp2.saix.net>
Message-ID: <se458t834sp9dokn5ibkuudpauovc6abbl@4ax.com>

On Thu, 8 Feb 2001 11:26:38 +0200, "Mix" <mix77 at usa.net> wrote:

>I am new to CGI scripting AND Python. What I am trying to do is,
>display an image on an html page that has been retrieved from a Postgre
>database
>using Python (cgi). My problem is : that image that not show up on the page,
>regardless of where it is sitting on my html or cgi directory!

There are several things that must work for you to be able to show an
image.

1.	You need to create a python script ;-) At
http://starship.python.net/crew/davem/cgifaq/faqw.cgi?req=show&file=faq02.010.htp
you can see how it's done...

2.	You need to make the script executable!
3.	<img src="myscript.py">


If I have to guess then your problem is probably the Content-Type!


// Anders
    


From sholden at holdenweb.com  Thu Feb  8 15:57:42 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 09:57:42 -0500
Subject: CGI-Python Image viewing
References: <95tp4p$bev$1@ctb-nnrp2.saix.net>
Message-ID: <pxyg6.10982$x4.169982@e420r-atl2.usenetserver.com>

Mix" <mix77 at usa.net> wrote in message
news:95tp4p$bev$1 at ctb-nnrp2.saix.net...
> I am new to CGI scripting AND Python. What I am trying to do is,
> display an image on an html page that has been retrieved from a Postgre
> database
> using Python (cgi). My problem is : that image that not show up on the
page,
> regardless of where it is sitting on my html or cgi directory!
> (I am running Apache Web server under Redhat 7.0 with the default linux
> kernel).
> But, I CAN display the image on a plain html page (i.e without the
> scripting).
>
> Please help!
> Thanks
>
>
Perhaps you should give us a little more information: for example, is it the
HTML you are retrieving from Postgres, or the image as well? An example of
the generated / retrieved HTML would be useful.

Generally speaking you should be sending HTML to the client with an <img
src="..."> tag in it to specify the location of the graphic. If the value of
the "src" attribute is absolute (begins with a /) then the browser will
request by sending a path from the server's root. If, on the other hand,
it's relative (does not begin with a /) then the browser will request it by
generating an absolute path based on the URL which was used to access the
HTML.

All this assumes you haven't put <BASE> tagging in your HTML.

Does this help?

regards
 Steve




From tonyc at bellsouth.net  Thu Feb  8 11:00:58 2001
From: tonyc at bellsouth.net (Tony Corrente)
Date: Thu, 08 Feb 2001 10:00:58 GMT
Subject: Accessing Widget With GTK 
Message-ID: <3A826E68.E77DB89@bellsouth.net>

Hello.

I am very new to Python so please excuse me if this question is stupid
but I can't find an example anywhere.  I've created a form using Glade.
The form has a entry field name pName and a button.  I've run glc.py and
created the Python code. Everything worked..  glc created 2 files. the
main program and the window handler file for the button.  What I want to
know is... How do I update pName when the button is clicked? I've tried
just using
def on_bNext_clicked(widget, mainObj):
    pname.set_text("This is a test")
when I run the program and click the button I get the following error
Traceback (innermost last):
  File "/usr/lib/python1.5/site-packages/gtk.py", line 125, in __call__
    ret = apply(self.func, a)
  File "window1Handlers.py", line 15, in on_bNext_clicked
    pName.set_text("This is a test")
NameError: pName

How do I access pName ( or any other widget in the window ) ???
Could anyone please guide me in the right direction ???

Thanks in advance...










From snake-or-bear at worldofbeverage.net  Wed Feb 14 10:28:31 2001
From: snake-or-bear at worldofbeverage.net (snake-or-bear at worldofbeverage.net)
Date: Wed, 14 Feb 2001 14:28:31 +0500
Subject: Accessing Widget With GTK
References: <3A826E68.E77DB89@bellsouth.net>
Message-ID: <A%Ai6.3981$Jp2.115389@news6.giganews.com>

This is just a guess,  but could it be a scoping problem?  Does your
event handler know who pName is?  If it's a class you might want to set
it up as self.pName...
Could you post more code?

In article <3A826E68.E77DB89 at bellsouth.net>, "Tony Corrente"
<tonyc at bellsouth.net> wrote:

> Hello.
> 
> I am very new to Python so please excuse me if this question is stupid
> but I can't find an example anywhere.  I've created a form using Glade.
> The form has a entry field name pName and a button.  I've run glc.py and
> created the Python code. Everything worked..  glc created 2 files. the
> main program and the window handler file for the button.  What I want to
> know is... How do I update pName when the button is clicked? I've tried
> just using def on_bNext_clicked(widget, mainObj):
>     pname.set_text("This is a test")
> when I run the program and click the button I get the following error
> Traceback (innermost last):
>   File "/usr/lib/python1.5/site-packages/gtk.py", line 125, in __call__
>     ret = apply(self.func, a)
>   File "window1Handlers.py", line 15, in on_bNext_clicked
>     pName.set_text("This is a test")
> NameError: pName
> 
> How do I access pName ( or any other widget in the window ) ??? Could
> anyone please guide me in the right direction ???
> 
> Thanks in advance...
> 
> 
> 
> 
> 
> 
> 
>


From rickp at telocity.com  Thu Feb 15 00:10:37 2001
From: rickp at telocity.com (Rick Pasotto)
Date: Wed, 14 Feb 2001 23:10:37 GMT
Subject: Accessing Widget With GTK 
References: <3A826E68.E77DB89@bellsouth.net>
Message-ID: <slrn98m43d.lic.rickp@tc.niof.net>

On Thu, 08 Feb 2001 10:00:58 GMT in comp.lang.python, Tony Corrente wrote:
> Hello.
> 
> I am very new to Python so please excuse me if this question is stupid
> but I can't find an example anywhere.  I've created a form using Glade.
> The form has a entry field name pName and a button.  I've run glc.py and
> created the Python code.

Where do I find this? (glc.py)

-- 
"Freedom is just chaos with better lighting."
		-- Alan Dean Foster
		   Rick Pasotto email: rickp at telocity.com


From rtrocca at libero.it  Thu Feb 15 08:26:27 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Thu, 15 Feb 2001 07:26:27 GMT
Subject: Accessing Widget With GTK
References: <3A826E68.E77DB89@bellsouth.net>
Message-ID: <1eoua2y.f9os6sgr2ddyN%rtrocca@libero.it>

Well, usually, in order to avoid that I use classes so it is easy to
access a self.pName from a class method.
I'm writing an application in python using a glade interface. I
preferred to load the xml file generated by glade by hand and connect
signals to class methods. Ok, this is just a personal opinion.
Anyway probably it is just a name scope problem.
try with glob(al)
 Riccardo 

Tony Corrente <tonyc at bellsouth.net> wrote:

> Hello.
> 
> I am very new to Python so please excuse me if this question is stupid
> but I can't find an example anywhere.  I've created a form using Glade.
> The form has a entry field name pName and a button.  I've run glc.py and
> created the Python code. Everything worked..  glc created 2 files. the
> main program and the window handler file for the button.  What I want to
> know is... How do I update pName when the button is clicked? I've tried
> just using
> def on_bNext_clicked(widget, mainObj):
>     pname.set_text("This is a test")
> when I run the program and click the button I get the following error
> Traceback (innermost last):
>   File "/usr/lib/python1.5/site-packages/gtk.py", line 125, in __call__
>     ret = apply(self.func, a)
>   File "window1Handlers.py", line 15, in on_bNext_clicked
>     pName.set_text("This is a test")
> NameError: pName
> 
> How do I access pName ( or any other widget in the window ) ???
> Could anyone please guide me in the right direction ???
> 
> Thanks in advance...


From pierre.puiseux at univ-pau.fr  Thu Feb  8 11:02:27 2001
From: pierre.puiseux at univ-pau.fr (Pierre Puiseux)
Date: Thu, 08 Feb 2001 11:02:27 +0100
Subject: SWIG and C++
Message-ID: <3A826EB3.B25975B9@univ-pau.fr>

> 
> Hello,
> 
> is it possible with SWIG, to import in python a C++ class with 2 methods
> WITH THE SAME NAME but with different arguments.For instance :
> //file truc.h
> class truc
> {...
>         void mymethod(int, int);
>         void mymethod(float);
> }
> //
> 
> //File : truc.i
> %module truc
> 
> %{
> #include "truc.h"
> %}
> 
> %include truc.h
> //
> 
> //here is the swig command and the error message :
> >swig -c++ -shadow -python truc.i
> >Generating wrappers for Python truc.h : Line 17. Function truc_mymethod (member mymethod) multiply defined (2nd definition ignored).
> 
> Thanks
> 
> --
> Pierre PUISEUX
> Universit? de Pau et des Pays de l'Adour
> Laboratoire de  Math?matiques Appliqu?es
> T?l: (+) 33-5-59-92-31-67
> http://www.univ-pau.fr/~puiseux/home.htm


From tjs at mail.csse.monash.edu.au  Thu Feb  8 12:31:29 2001
From: tjs at mail.csse.monash.edu.au (Toby J Sargeant)
Date: Thu, 8 Feb 2001 22:31:29 +1100
Subject: SWIG and C++
In-Reply-To: <3A826EB3.B25975B9@univ-pau.fr>; from pierre.puiseux@univ-pau.fr on Thu, Feb 08, 2001 at 11:02:27AM +0100
References: <3A826EB3.B25975B9@univ-pau.fr>
Message-ID: <20010208223129.A4148@fangorn.csse.monash.edu.au>

On Thu, Feb 08, 2001 at 11:02:27AM +0100, Pierre Puiseux wrote:
> 
> Hello,
> 
> is it possible with SWIG, to import in python a C++ class with 2 methods
> WITH THE SAME NAME but with different arguments.For instance :
> //file truc.h
> class truc
> {...
>         void mymethod(int, int);
>         void mymethod(float);
> }
> //
> 
> //File : truc.i
> %module truc
> 
> %{
> #include "truc.h"
> %}
> 
> %include truc.h
> //
> 
> //here is the swig command and the error message :
> >swig -c++ -shadow -python truc.i
> >Generating wrappers for Python truc.h : Line 17. Function truc_mymethod (member mymethod) multiply defined (2nd definition ignored).
> 
> Thanks

>From memory, the answer is yes, you can. You might like to look at the sip
files for something like PyQT to find out how, though. I don't know why
what you're doing isn't working.

One caveat to this that affected me when I was playing with PyQT
that may still be true is that if you have two methods that have the same
number of arguments, and all the arguments of the second method are
castable to the types of the arguments in the first, then the sip wrapper
will blindly do the cast, and end up calling the first method, even though
the types of the arguments you pass match the signature of the second.

The particular example I'm referring to is

QWMatrix::map(int,int) vs. QWMatrix::map(float,float)

and the net result was that any floating point values passed to map() were
implicitly cast to ints, rather than using the floating point version.

Toby.



From kdahlhaus at yahoo.com  Sun Feb 11 05:53:18 2001
From: kdahlhaus at yahoo.com (Kevin Dahlhausen)
Date: Sun, 11 Feb 2001 04:53:18 GMT
Subject: SWIG and C++
References: <3A826EB3.B25975B9@univ-pau.fr> <mailman.981632045.28779.python-list@python.org>
Message-ID: <2Voh6.1930$ws2.121282@ord-read.news.verio.net>

Yes, but it's kind of hacky at the moment.  You have to remap each of the
overidden functions to a different exported name.  Then create a python
function of the correct name that looks at the input paramters and
dispatches the current function.  I didn't think of this, I just use it.
The FLTK wrappers on Sourceforge (pyfltk.sourceforge.net) use this
method.

I'm looking into the Boost Python Library, which supports this and also
allows subclassing of C++ classes in Python.


In article <mailman.981632045.28779.python-list at python.org>, "Toby J
Sargeant" <tjs at mail.csse.monash.edu.au> wrote:

> On Thu, Feb 08, 2001 at 11:02:27AM +0100, Pierre Puiseux wrote:
>> 
>> Hello,
>> 
>> is it possible with SWIG, to import in python a C++ class with 2
>> methods WITH THE SAME NAME but with different arguments.For instance :
>> //file truc.h
>> class truc
>> {...
>>         void mymethod(int, int); void mymethod(float);
>> }
>> //
>> 
>> //File : truc.i
>> %module truc
>> 
>> %{
>> #include "truc.h"
>> %}
>> 
>> %include truc.h
>> //
>> 
>> //here is the swig command and the error message :
>> >swig -c++ -shadow -python truc.i Generating wrappers for Python truc.h
>> >: Line 17. Function truc_mymethod (member mymethod) multiply defined
>> >(2nd definition ignored).
>> 
>> Thanks
> 
> From memory, the answer is yes, you can. You might like to look at the
> sip files for something like PyQT to find out how, though. I don't know
> why what you're doing isn't working.
> 
> One caveat to this that affected me when I was playing with PyQT that
> may still be true is that if you have two methods that have the same
> number of arguments, and all the arguments of the second method are
> castable to the types of the arguments in the first, then the sip
> wrapper will blindly do the cast, and end up calling the first method,
> even though the types of the arguments you pass match the signature of
> the second.
> 
> The particular example I'm referring to is
> 
> QWMatrix::map(int,int) vs. QWMatrix::map(float,float)
> 
> and the net result was that any floating point values passed to map()
> were implicitly cast to ints, rather than using the floating point
> version.
> 
> Toby.
>


From greg at perceval.be  Thu Feb  8 12:54:00 2001
From: greg at perceval.be (Gregoire Welraeds)
Date: Thu, 8 Feb 2001 12:54:00 +0100 (CET)
Subject: None assigment
Message-ID: <Pine.BSF.4.21.0102081248050.4685-100000@localhost>

While playing a bit with python 2.0, I found that I can assign some value
to None

EG: None= 2

which I found to be very pernicious because then the following, for
example, won't work any more:

>>> b= filter(None, [1,2,'',3])

Bug or feature ?

If it is a feature (I suppose it is not a bug :), what is the interest of
this ?

Gregoire Welraeds
Perceval Development team
-------------------------------------------------------------------------------
Perceval Technologies sa/nv	Tel: +32-2-6409194		
Rue Tenbosch, 9			Fax: +32-2-6403154		
B-1000 Brussels			general information:   info at perceval.net
BELGIUM				technical information: helpdesk at perceval.net
URL: http://www.perceval.be/
-------------------------------------------------------------------------------




From SBrunning at trisystems.co.uk  Thu Feb  8 13:20:27 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 8 Feb 2001 12:20:27 -0000 
Subject: None assigment
Message-ID: <31575A892FF6D1118F5800600846864D5B17D1@intrepid>

> From:	Gregoire Welraeds [SMTP:greg at perceval.be]
> While playing a bit with python 2.0, I found that I can assign some value
> to None
(snip)
> If it is a feature (I suppose it is not a bug :), what is the interest of
> this ?
 
Feature. 'None' is just a label, and just like any other label, you can
always assign to it. I'm not aware of any way of preventing a label from
being assigned to.

Doesn't make it a good idea, though...

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk





-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From aahz at panix.com  Thu Feb  8 17:10:12 2001
From: aahz at panix.com (Aahz Maruch)
Date: 8 Feb 2001 08:10:12 -0800
Subject: None assigment
References: <mailman.981633367.1329.python-list@python.org>
Message-ID: <95ugd4$jka$1@panix6.panix.com>

In article <mailman.981633367.1329.python-list at python.org>,
Gregoire Welraeds  <greg at perceval.be> wrote:
>
>While playing a bit with python 2.0, I found that I can assign some value
>to None
>
>EG: None= 2
>
>which I found to be very pernicious because then the following, for
>example, won't work any more:
>
>>>> b= filter(None, [1,2,'',3])
>
>Bug or feature ?

Neither, really; it's Just The Way Things Work.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From f8dy at my-deja.com  Thu Feb  8 19:16:59 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Thu, 08 Feb 2001 18:16:59 GMT
Subject: None assigment
References: <mailman.981633367.1329.python-list@python.org> <95ugd4$jka$1@panix6.panix.com>
Message-ID: <95unqf$4j3$1@nnrp1.deja.com>

In article <95ugd4$jka$1 at panix6.panix.com>,
  aahz at panix.com (Aahz Maruch) wrote:
> In article <mailman.981633367.1329.python-list at python.org>,
> Gregoire Welraeds  <greg at perceval.be> wrote:
> >
> >While playing a bit with python 2.0, I found that I can assign some
value
> >to None
> >
> >EG: None= 2
> >
> >which I found to be very pernicious because then the following, for
> >example, won't work any more:
> >
> >>>> b= filter(None, [1,2,'',3])
> >
> >Bug or feature ?
>
> Neither, really; it's Just The Way Things Work.

Indeed.  I once had a similar experience; while playing around in the
Python IDE, I defined a variable 'dir' to hold some directory path.
Then days later (having never shut down Python), I wondered why the
built-in 'dir' function didn't work.

Which leads to the following questions:
- Are there any true constants in Python?  (I don't think so, but
correct me if I'm wrong.)
- Once you've redefined a built-in function like 'dir', is there any
way to get it back without restarting Python?

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From aahz at panix.com  Thu Feb  8 20:06:45 2001
From: aahz at panix.com (Aahz Maruch)
Date: 8 Feb 2001 11:06:45 -0800
Subject: None assigment
References: <mailman.981633367.1329.python-list@python.org> <95ugd4$jka$1@panix6.panix.com> <95unqf$4j3$1@nnrp1.deja.com>
Message-ID: <95uqo5$ek9$1@panix6.panix.com>

In article <95unqf$4j3$1 at nnrp1.deja.com>,
Mark Pilgrim  <f8dy at my-deja.com> wrote:
>
>- Once you've redefined a built-in function like 'dir', is there any
>way to get it back without restarting Python?

del dir
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From amused at webamused.com  Thu Feb  8 20:26:25 2001
From: amused at webamused.com (amused at webamused.com)
Date: Thu, 08 Feb 2001 19:26:25 GMT
Subject: None assigment
References: <mailman.981633367.1329.python-list@python.org> <95ugd4$jka$1@panix6.panix.com> <95unqf$4j3$1@nnrp1.deja.com>
Message-ID: <95ursl$8j4$1@nnrp1.deja.com>

In article <95unqf$4j3$1 at nnrp1.deja.com>,
  Mark Pilgrim <f8dy at my-deja.com> wrote:
>
> Indeed.  I once had a similar experience; while playing around in the
> Python IDE, I defined a variable 'dir' to hold some directory path.
> Then days later (having never shut down Python), I wondered why the
> built-in 'dir' function didn't work.
>
> Which leads to the following questions:
> - Are there any true constants in Python?  (I don't think so, but
> correct me if I'm wrong.)
> - Once you've redefined a built-in function like 'dir', is there any
> way to get it back without restarting Python?
>
> -M
> --

You can delete it from the local namespace with del:

Python 2.1a1 (#9, Jan 22 2001, 20:58:30) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> dir
<built-in function dir>
>>> dir = "Spam"
>>> dir
'Spam'
>>> del dir
>>> dir
<built-in function dir>
>>>

  Joshua



Sent via Deja.com
http://www.deja.com/


From greg at perceval.be  Thu Feb  8 18:12:27 2001
From: greg at perceval.be (Gregoire Welraeds)
Date: Thu, 8 Feb 2001 18:12:27 +0100 (CET)
Subject: None assigment
In-Reply-To: <m24ry51euy.fsf@mycroft.actrix.gen.nz>
Message-ID: <Pine.BSF.4.21.0102081633310.4685-100000@localhost>

> There's nothing special about the name "None"; it's just a variable,
> like any other. 
<snip>
> After doing "None = 2", you can "del None" to get the default value 
> back.

If we follow that logic, I could use any non assigned variable to have the
following working:

>>> a= [1,'',3]
>>> filter(b,a)

but this won't work as the interpreter complains that there is no variable
named b.

Always according to this logic, I have tried to run the following:

>>> a= [1,'',3]
>>> b= 0 # b could be anything
>>> del b
>>> filter(b,a)

but this won't work either as stated in the documentation. Referencing a
"deleted" variable is an error at least until another value is assigned to
it (see del statement section in the tutorial).

It seems that (the so called variable) None is neither a variable like any
other, nor is it a label as stated by Simon Brunning in another post.

Furthermore, I found that the documentation is rather confusing about this
question. Here's what I found while browsing:

>From Python Reference Manual:

-	6.1 Expression statements 	
	...
	in Python, procedures return the value None
(http://www.python.org/doc/2.0b1/ref/exprstmts.html#l2h-334)

-	3.2 The standard type hierarchy 
	...
	This type has a single value. There is a single object with this
	value. This object is accessed through the built-in name None. It 
	is used to signify the absence of a value in many situations,
	e.g., it is returned from functions that don't explicitly return
	anything. Its truth value is false. 	
(http://www.python.org/doc/2.0b1/ref/types.html#l2h-70)

>From Python Library Reference:

-	3.4 types -- Names for all built-in types
	NoneType 
       		The type of None.
	... 
(http://www.python.org/doc/2.0b1/lib/module-types.html#l2h-279)

- 	2.1.1 Truth Value Testing 
	Any object can be tested for truth value, for use in an if or while
	condition or as operand of the Boolean operations below. The following
	values are considered false: 
		None  
		...
(http://www.python.org/doc/2.0b1/lib/truth.html#l2h-5)

>From Python/C API Reference Manual

-	7.1.2 The None Object
	...
(http://www.python.org/doc/2.0b1/api/noneObject.html#l2h-246)
	
And finally, to add to the confusion, I have try the following:

>>> a= [1,'',3]
>>> filter(0,a)

for what the system complains that 0 is a non function type ! 

Now If we resume:
None is a value
None is a built in type
None is an object
None is an alias (as I understand it) for lambda x: x in the case of
filter(None, some_list)

Now that I look at this problem, I'm totally confused about None. Any
explanaition is welcome. Anyway, I don't understand that one can override
None. Could you give me at least one single good reason to do that.

Gregoire & Xavier
>From Perceval Development team, (python division :)

On 9 Feb 2001, Paul Foley wrote:

~# Date: 09 Feb 2001 02:30:45 +1300
~# From: Paul Foley <mycroft at actrix.gen.nz>
~# To: Gregoire Welraeds <greg at perceval.be>
~# Subject: Re: None assigment
~# 
~# On Thu, 8 Feb 2001 12:54:00 +0100 (CET), Gregoire Welraeds wrote:
~# 
~# > While playing a bit with python 2.0, I found that I can assign some value
~# > to None
~# 
~# > EG: None= 2
~# 
~# > which I found to be very pernicious because then the following, for
~# > example, won't work any more:
~# 
~# >>>> b= filter(None, [1,2,'',3])
~# 
~# > Bug or feature ?
~# 
~# Neither, really.
~# 
~# > If it is a feature (I suppose it is not a bug :), what is the interest of
~# > this ?
~# 
~# There's nothing special about the name "None"; it's just a variable,
~# like any other.  Any other behaviour would be inconsistent.  None is
~# usually bound to the special "None" value (note: values have nothing
~# to do with names; the None value is not really related to the name
~# "None"), that's all.
~# 
~# After doing "None = 2", you can "del None" to get the default value
~# back.
~# 
~# -- 
~# Qui beneficium dedit, taceat; narret qui accepit                 -- Seneca
~# 
~# (setq reply-to
~#   (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))
~# 







From dsh8290 at rit.edu  Thu Feb  8 18:36:00 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 8 Feb 2001 12:36:00 -0500
Subject: None assigment
In-Reply-To: <Pine.BSF.4.21.0102081633310.4685-100000@localhost>; from greg@perceval.be on Thu, Feb 08, 2001 at 06:12:27PM +0100
References: <m24ry51euy.fsf@mycroft.actrix.gen.nz> <Pine.BSF.4.21.0102081633310.4685-100000@localhost>
Message-ID: <20010208123600.A12837@harmony.cs.rit.edu>

On Thu, Feb 08, 2001 at 06:12:27PM +0100, Gregoire Welraeds wrote:
| 
| > There's nothing special about the name "None"; it's just a variable,
| > like any other. 
| <snip>
| > After doing "None = 2", you can "del None" to get the default value 
| > back.
| 
| If we follow that logic, I could use any non assigned variable to have the
| following working:
| 
| >>> a= [1,'',3]
| >>> filter(b,a)
| 
| but this won't work as the interpreter complains that there is no variable
| named b.

This is because the interpreter tries to evaluate all expressions
before the actual call to the functions.  When the interpreter tries
to evaluate the expression  b  it can't.  When it evaluates the
expression  None  it can.  If you were to "del None" , then try and
pass it to a function it would fail the same way.  (Or instead of
using the del statement, just remove the name None from the
dictionary)

| 
| 	
| And finally, to add to the confusion, I have try the following:
| 
| >>> a= [1,'',3]
| >>> filter(0,a)
| 
| for what the system complains that 0 is a non function type ! 

Correct.

| 
| Now If we resume:
| None is a value
| None is a built in type
| None is an object
| None is an alias (as I understand it) for lambda x: x in the case of
| filter(None, some_list)
| 
| Now that I look at this problem, I'm totally confused about None. Any

"None" is a name/binding.  By default it refers to an object whose
type is called "None".  This object is a Singleton, and is built in.

Objects can have many properties, one of them is "callable".  If you
were to make a replacement class for None, you could do something like:

class MyNone :
	def __call__( self , arg ) :
		return arg


If I try this :

n = MyNone()
filter( n , [ 1 , 2 , 3 ] )


it would be the same as using the builtin None object.  (Unless I
missed some sublties)

None is simply an object, provided at interpreter startup, that has
some (special) properties.  (All classes define new objects, and each
has "special" properties)

| explanaition is welcome. Anyway, I don't understand that one can override
| None. Could you give me at least one single good reason to do that.

Being able to "override" None is simply how the language works.
Python allows you to assign an object (everything in python is an
object!) to any name.  "None" isn't a special case.  I can't think of
any good reason to assign a new object to "None", but that doesn't
mean the language disallows it.  (Ab)using None in this manner would
certainly obfuscate your code.

I also don't think it is really necessary for the interpreter to
disallow assignment to None since it adds complexity to the
interpreter's workings without providing any real advantage.  Instead,
programmer discipline to use None appropriately works just fine IMO. :-)

-D




From SBrunning at trisystems.co.uk  Thu Feb  8 18:34:37 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 8 Feb 2001 17:34:37 -0000 
Subject: None assigment
Message-ID: <31575A892FF6D1118F5800600846864D5B17D5@intrepid>

> From:	Gregoire Welraeds [SMTP:greg at perceval.be]
> > After doing "None = 2", you can "del None" to get the default value 
> > back.
 
Hmm. This is true - I didn't know that.
 
> If we follow that logic, I could use any non assigned variable to have the
> following working:
> 
> >>> a= [1,'',3]
> >>> filter(b,a)
> 
> but this won't work as the interpreter complains that there is no variable
> named b.
 
You can't do this with *ordinary* labels - by del'ing them, you are
un-assigning them. But the 'None' label is *extraordinary*, obviously.
 
> It seems that (the so called variable) None is neither a variable like any
> other, nor is it a label as stated by Simon Brunning in another post.
 
It *is* a label - try this:

>>>print None
None
>>> None = 'spam'
>>> print None
spam

So, clearly, None is a label.
 
	 Now that I look at this problem, I'm totally confused about None.
Any
> explanaition is welcome. Anyway, I don't understand that one can override
> None. Could you give me at least one single good reason to do that.
 
I can give you a good reason *not* to - it will confuse the hell out of
people. ;-)

I'm sure one of the Python gurus will give you a bit more background, but I
think that the short answer is this - leave 'None' alone to get on with its
job, and you won't have any problems.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From dsh8290 at rit.edu  Thu Feb  8 19:15:23 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 8 Feb 2001 13:15:23 -0500
Subject: None assigment
In-Reply-To: <31575A892FF6D1118F5800600846864D5B17D5@intrepid>; from SBrunning@trisystems.co.uk on Thu, Feb 08, 2001 at 05:34:37PM +0000
References: <31575A892FF6D1118F5800600846864D5B17D5@intrepid>
Message-ID: <20010208131523.A13104@harmony.cs.rit.edu>

On Thu, Feb 08, 2001 at 05:34:37PM +0000, Simon Brunning wrote:
| > From:	Gregoire Welraeds [SMTP:greg at perceval.be]
| > > After doing "None = 2", you can "del None" to get the default value 
| > > back.
|  
| Hmm. This is true - I didn't know that.

I didn't realize this, and had never tried. 

|  
| > If we follow that logic, I could use any non assigned variable to have the
| > following working:
| > 
| > >>> a= [1,'',3]
| > >>> filter(b,a)
| > 
| > but this won't work as the interpreter complains that there is no variable
| > named b.
|  
| You can't do this with *ordinary* labels - by del'ing them, you are
| un-assigning them. But the 'None' label is *extraordinary*, obviously.

I didn't realize this either, and had never tried.  Check this out :

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> del None
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  NameError: There is no variable named 'None'
>>>
>>> None = 1
>>> del None
>>> del None
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  NameError: There is no variable named 'None'
>>>


Interesting.  So normally "None" is a special keyword (that is a
reference to an object) ,  but that doesn't prevent one from creating
a local variable named "None" that shadows the keyword.

:-)
-D




From nas at arctrix.com  Thu Feb  8 19:55:40 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Thu, 8 Feb 2001 10:55:40 -0800
Subject: None assigment
In-Reply-To: <20010208131523.A13104@harmony.cs.rit.edu>; from dsh8290@rit.edu on Thu, Feb 08, 2001 at 01:15:23PM -0500
References: <31575A892FF6D1118F5800600846864D5B17D5@intrepid> <20010208131523.A13104@harmony.cs.rit.edu>
Message-ID: <20010208105540.A13016@glacier.fnational.com>

On Thu, Feb 08, 2001 at 01:15:23PM -0500, D-Man wrote:
> Interesting.  So normally "None" is a special keyword (that is a
> reference to an object) ,  but that doesn't prevent one from creating
> a local variable named "None" that shadows the keyword.

Sounds like something another P language would do.  Things are
simpler.  None is a builtin.  It exists in the __builtin__
module.  The buildin module is searched for names after the
global namespace is searched.

    >>> print None # find the binding for None, it is in __builtin__
    None
    >>> None = "spam" # make a global binding for None
    >>> print None
    spam
    >>> import __builtin__
    >>> print __builtin__.None # __builtin__ binding is still there
    None
    >>> del None # remove global binding
    >>> print None
    None
    >>> __builtin__.None = "spam" # change __builtin__ binding (evil)
    >>> print None
    spam
 
Cheers,

  Neil



From dsh8290 at rit.edu  Thu Feb  8 20:05:48 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 8 Feb 2001 14:05:48 -0500
Subject: None assigment
In-Reply-To: <20010208105540.A13016@glacier.fnational.com>; from nas@arctrix.com on Thu, Feb 08, 2001 at 10:55:40AM -0800
References: <31575A892FF6D1118F5800600846864D5B17D5@intrepid> <20010208131523.A13104@harmony.cs.rit.edu> <"from dsh8290"@rit.edu> <20010208105540.A13016@glacier.fnational.com>
Message-ID: <20010208140547.A13209@harmony.cs.rit.edu>

On Thu, Feb 08, 2001 at 10:55:40AM -0800, Neil Schemenauer wrote:
| On Thu, Feb 08, 2001 at 01:15:23PM -0500, D-Man wrote:
| > Interesting.  So normally "None" is a special keyword (that is a
| > reference to an object) ,  but that doesn't prevent one from creating
| > a local variable named "None" that shadows the keyword.
| 
| Sounds like something another P language would do.  Things are
| simpler.  None is a builtin.  It exists in the __builtin__
| module.  The buildin module is searched for names after the
| global namespace is searched.

So why can't I 'del' it then?  It is the response from del that made
me think it was a special keyword.

-D




From sholden at holdenweb.com  Thu Feb  8 20:17:14 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 14:17:14 -0500
Subject: None assigment
References: <31575A892FF6D1118F5800600846864D5B17D5@intrepid> <20010208131523.A13104@harmony.cs.rit.edu> <"from dsh8290"@rit.edu> <20010208105540.A13016@glacier.fnational.com> <mailman.981659244.16551.python-list@python.org>
Message-ID: <VkCg6.43229$o5.547288@e420r-atl1.usenetserver.com>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.981659244.16551.python-list at python.org...
> On Thu, Feb 08, 2001 at 10:55:40AM -0800, Neil Schemenauer wrote:
> | On Thu, Feb 08, 2001 at 01:15:23PM -0500, D-Man wrote:
> | > Interesting.  So normally "None" is a special keyword (that is a
> | > reference to an object) ,  but that doesn't prevent one from creating
> | > a local variable named "None" that shadows the keyword.
> |
> | Sounds like something another P language would do.  Things are
> | simpler.  None is a builtin.  It exists in the __builtin__
> | module.  The buildin module is searched for names after the
> | global namespace is searched.
>
> So why can't I 'del' it then?  It is the response from del that made
> me think it was a special keyword.
>
try

del __builtins__.None

regards
 Steve




From tanzer at swing.co.at  Fri Feb  9 07:19:04 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Fri, 09 Feb 2001 07:19:04 +0100
Subject: None assigment 
In-Reply-To: Your message of "Thu, 08 Feb 2001 13:15:23 EST."
             <20010208131523.A13104@harmony.cs.rit.edu> 
Message-ID: <m14R6tc-000wcEC@swing.co.at>

D-Man <dsh8290 at rit.edu> wrote:

> | You can't do this with *ordinary* labels - by del'ing them, you are
> | un-assigning them. But the 'None' label is *extraordinary*, obviously.
> 
> I didn't realize this either, and had never tried.  Check this out :
> 
> Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> >>> del None
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   NameError: There is no variable named 'None'
> >>>
> >>> None = 1
> >>> del None
> >>> del None
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   NameError: There is no variable named 'None'
> >>>
> 
> 
> Interesting.  So normally "None" is a special keyword (that is a
> reference to an object) ,  but that doesn't prevent one from creating
> a local variable named "None" that shadows the keyword.

None is not a keyword. It's a binding living in the __builtin__ scope.
Therefore a plain `del None' doesn't work.

Try this:

tanzer [lib] 2 $ python2.0 
Python 2.0 (#5, Jan 30 2001, 11:08:38) 
[GCC 2.7.2.1] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import __builtin__
>>> del __builtin__.None 
>>> None
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: There is no variable named 'None'
>>> 

As others repeatedly pointed out, it's not a good idea to do such
things in real code.

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From sholden at holdenweb.com  Thu Feb  8 19:52:31 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 13:52:31 -0500
Subject: None assigment
References: <31575A892FF6D1118F5800600846864D5B17D5@intrepid> <mailman.981656286.3737.python-list@python.org>
Message-ID: <KZBg6.43227$o5.546119@e420r-atl1.usenetserver.com>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.981656286.3737.python-list at python.org...
[ ... ]
>
> Interesting.  So normally "None" is a special keyword (that is a
> reference to an object) ,  but that doesn't prevent one from creating
> a local variable named "None" that shadows the keyword.
>
Nope, it's just in __builtins__, and remains there whether you bind the same
name in the module scope or not. You can delete it from the module scope,
thereby making the one in __builtins__ visible once more.  And, as I pointed
out in an earlier posting, you can even bind __builtins__.None to a new
value if you want, to the confusion of all.

regards
 Steve




From mwh21 at cam.ac.uk  Thu Feb  8 19:53:30 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 08 Feb 2001 18:53:30 +0000
Subject: None assigment
References: <31575A892FF6D1118F5800600846864D5B17D5@intrepid> <mailman.981656286.3737.python-list@python.org>
Message-ID: <m34ry52ehh.fsf@atrus.jesus.cam.ac.uk>

D-Man <dsh8290 at rit.edu> writes:

> Interesting.  So normally "None" is a special keyword (that is a
> reference to an object) , but that doesn't prevent one from creating
> a local variable named "None" that shadows the keyword.

Not really.  It's a builtin.  Ever used a variable called list?

Cheers,
M.

-- 
  The "of course, while I have no problem with this at all, it's
  surely too much for a lesser being" flavor of argument always 
  rings hollow to me.                       -- Tim Peters, 29 Apr 1998


From sholden at holdenweb.com  Thu Feb  8 18:49:13 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 12:49:13 -0500
Subject: None assigment
References: <mailman.981652444.20002.python-list@python.org>
Message-ID: <V3Bg6.43113$o5.544085@e420r-atl1.usenetserver.com>

"Gregoire Welraeds" <greg at perceval.be> wrote in message
news:mailman.981652444.20002.python-list at python.org...
>
[ ... ]
>
> Now that I look at this problem, I'm totally confused about None. Any
> explanaition is welcome. Anyway, I don't understand that one can override
> None. Could you give me at least one single good reason to do that.
>
The built in name None is defined as the value None, which is the only value
of type None.

When you assign to None, you create a new name in your module (or function)
namespace, and bind it to whatever value is assigned.  This new name will be
found before the built in name, since the built in namespace is the last to
be searched. When you delete None you delete it from your module or function
namespace, and the built in name again becomes available.

To quote Mark Lutz' excellent advice from "Learning Python": don't do that!
There is no "single good reason" to do it. This is Python, however, and just
because there's no good reason to do something doesn't mean the implementers
will implement a prohibition. You are free to shoot yourself in the foot if
you want...

There really is no advantage, except perhaps in obfuscated Python contests,
to redefining None. It's just an implementation artefact that you can do it.
Try to forget it's possible and I'm sure you'll feel better!

> Gregoire & Xavier
> From Perceval Development team, (python division :)
>
regards
 Steve



From aahz at panix.com  Thu Feb  8 18:58:42 2001
From: aahz at panix.com (Aahz Maruch)
Date: 8 Feb 2001 09:58:42 -0800
Subject: None assigment
References: <mailman.981652444.20002.python-list@python.org>
Message-ID: <95umoi$is5$1@panix2.panix.com>

In article <mailman.981652444.20002.python-list at python.org>,
Gregoire Welraeds  <greg at perceval.be> wrote:
>
>Always according to this logic, I have tried to run the following:
>
>>>> a= [1,'',3]
>>>> b= 0 # b could be anything
>>>> del b
>>>> filter(b,a)
>
>but this won't work either as stated in the documentation. Referencing a
>"deleted" variable is an error at least until another value is assigned to
>it (see del statement section in the tutorial).
>
>It seems that (the so called variable) None is neither a variable like any
>other, nor is it a label as stated by Simon Brunning in another post.

What you're missing is the issue of Python scopes.  Try this:

b=2
def foo():
    b=4
    print b
    del b
    global b
    print b
foo()

None has a bit of magic associated with it, but not much.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From root at rainerdeyke.com  Fri Feb  9 02:20:58 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 09 Feb 2001 01:20:58 GMT
Subject: None assigment
References: <mailman.981652444.20002.python-list@python.org> <95umoi$is5$1@panix2.panix.com>
Message-ID: <_BHg6.195181$ge4.67634054@news2.rdc2.tx.home.com>

"Aahz Maruch" <aahz at panix.com> wrote in message
news:95umoi$is5$1 at panix2.panix.com...
> What you're missing is the issue of Python scopes.  Try this:
>
> b=2
> def foo():
>     b=4
>     print b
>     del b
>     global b
>     print b
> foo()

And what *you*'re missing is that whether a variable is local or global is
determined at compile time (in the absensce of 'exec' or 'from ... import
*').  The above does not work.  However, Whether a global variable is taken
from the module dictionary or '__builtins__' is a run-time decision.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From aahz at panix.com  Fri Feb  9 05:49:44 2001
From: aahz at panix.com (Aahz Maruch)
Date: 8 Feb 2001 20:49:44 -0800
Subject: None assigment
References: <mailman.981652444.20002.python-list@python.org> <95umoi$is5$1@panix2.panix.com> <_BHg6.195181$ge4.67634054@news2.rdc2.tx.home.com>
Message-ID: <95vst8$1na$1@panix2.panix.com>

In article <_BHg6.195181$ge4.67634054 at news2.rdc2.tx.home.com>,
Rainer Deyke <root at rainerdeyke.com> wrote:
>"Aahz Maruch" <aahz at panix.com> wrote in message
>news:95umoi$is5$1 at panix2.panix.com...
>>
>> What you're missing is the issue of Python scopes.  Try this:
>>
>> b=2
>> def foo():
>>     b=4
>>     print b
>>     del b
>>     global b
>>     print b
>> foo()
>
>And what *you*'re missing is that whether a variable is local or global is
>determined at compile time (in the absensce of 'exec' or 'from ... import
>*').  The above does not work.  

It doesn't, eh?  Did you try running it?  (After a few blunders of my
own, I rarely post code to c.l.py without testing it -- just to make
sure.)  Note that this is not the same as the issue with None, but it's
still a scoping issue.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From root at rainerdeyke.com  Fri Feb  9 07:00:00 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 09 Feb 2001 06:00:00 GMT
Subject: global, was Re: None assigment
References: <mailman.981652444.20002.python-list@python.org> <95umoi$is5$1@panix2.panix.com> <_BHg6.195181$ge4.67634054@news2.rdc2.tx.home.com> <95vst8$1na$1@panix2.panix.com>
Message-ID: <AHLg6.197584$ge4.68265785@news2.rdc2.tx.home.com>

"Aahz Maruch" <aahz at panix.com> wrote in message
news:95vst8$1na$1 at panix2.panix.com...
> In article <_BHg6.195181$ge4.67634054 at news2.rdc2.tx.home.com>,
> Rainer Deyke <root at rainerdeyke.com> wrote:
> >"Aahz Maruch" <aahz at panix.com> wrote in message
> >news:95umoi$is5$1 at panix2.panix.com...
> >>
> >> What you're missing is the issue of Python scopes.  Try this:
> >>
> >> b=2
> >> def foo():
> >>     b=4
> >>     print b
> >>     del b
> >>     global b
> >>     print b
> >> foo()
> >
> >And what *you*'re missing is that whether a variable is local or global
is
> >determined at compile time (in the absensce of 'exec' or 'from ... import
> >*').  The above does not work.
>
> It doesn't, eh?  Did you try running it?  (After a few blunders of my
> own, I rarely post code to c.l.py without testing it -- just to make
> sure.)  Note that this is not the same as the issue with None, but it's
> still a scoping issue.

Strangely enough, it does work.  I'm really confused now.  My understanding
had been that 'global' is not a normal statement executed at run-time, but
something which modifies the meaning of a name for the whole function.  It
seems I was right about the former but wrong about the latter.  Consider the
following:

b = 6

def foo():
  b = 4
  if 0:
    global b
  print b

def bar():
  b = 4
  if 1:
    global b
  print b

def baz(n):
  b = 4
  if n:
    global b
  print b

foo() # Prints '4'.
bar() # Prints '6'.
baz(0) # Prints '6', not '4'!
baz(1) # Print '6'.

(None of them modified the global variable 'b'.)


>From this I conclude that the behavior of 'global' is totally unpredictable
except when it is placed above all non-'global' statements in a function.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From mwh21 at cam.ac.uk  Fri Feb  9 09:18:32 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 09 Feb 2001 08:18:32 +0000
Subject: global, was Re: None assigment
References: <mailman.981652444.20002.python-list@python.org> <95umoi$is5$1@panix2.panix.com> <_BHg6.195181$ge4.67634054@news2.rdc2.tx.home.com> <95vst8$1na$1@panix2.panix.com> <AHLg6.197584$ge4.68265785@news2.rdc2.tx.home.com>
Message-ID: <m3y9vg1d7r.fsf@atrus.jesus.cam.ac.uk>

"Rainer Deyke" <root at rainerdeyke.com> writes:

> Strangely enough, it does work.  I'm really confused now.  My understanding
> had been that 'global' is not a normal statement executed at run-time, but
> something which modifies the meaning of a name for the whole function.  It
> seems I was right about the former but wrong about the latter.  Consider the
> following:
> 
> b = 6
> 
> def foo():
>   b = 4
>   if 0:
>     global b
>   print b
> 
> def bar():
>   b = 4
>   if 1:
>     global b
>   print b
> 
> def baz(n):
>   b = 4
>   if n:
>     global b
>   print b
> 
> foo() # Prints '4'.
> bar() # Prints '6'.
> baz(0) # Prints '6', not '4'!
> baz(1) # Print '6'.
> 
> (None of them modified the global variable 'b'.)
> 
> 
> From this I conclude that the behavior of 'global' is totally
> unpredictable except when it is placed above all non-'global'
> statements in a function.

Which version are you using?  They all print 4 for me, and all but
foo() do modify the value of `b'.  foo() doesn't because what follows
an

if 0:

isn't compiled.

Cheers,
M.

-- 
  Premature optimization is the root of all evil in programming.  
                                                       -- C.A.R. Hoare


From mwh21 at cam.ac.uk  Fri Feb  9 09:19:57 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 09 Feb 2001 08:19:57 +0000
Subject: global, was Re: None assigment
References: <mailman.981652444.20002.python-list@python.org> <95umoi$is5$1@panix2.panix.com> <_BHg6.195181$ge4.67634054@news2.rdc2.tx.home.com> <95vst8$1na$1@panix2.panix.com> <AHLg6.197584$ge4.68265785@news2.rdc2.tx.home.com> <m3y9vg1d7r.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <m3vgqk1d5e.fsf@atrus.jesus.cam.ac.uk>

Michael Hudson <mwh21 at cam.ac.uk> writes:

> Which version are you using?  They all print 4 for me, and all but
> foo() do modify the value of `b'.  foo() doesn't because what follows
> an

OTOH, that was a version that had the PEP 227 code checked into it.
With an older version, I see what you see.  Wierd.

Cheers,
M.

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme


From jmarshal at mathworks.com  Fri Feb  9 16:51:51 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 9 Feb 2001 15:51:51 GMT
Subject: global, was Re: None assigment
References: <mailman.981652444.20002.python-list@python.org> <95umoi$is5$1@panix2.panix.com> <_BHg6.195181$ge4.67634054@news2.rdc2.tx.home.com> <95vst8$1na$1@panix2.panix.com> <AHLg6.197584$ge4.68265785@news2.rdc2.tx.home.com> <m3y9vg1d7r.fsf@atrus.jesus.cam.ac.uk> <m3vgqk1d5e.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <9613mn$9v$1@news.mathworks.com>

Michael Hudson <mwh21 at cam.ac.uk> wrote:
> Michael Hudson <mwh21 at cam.ac.uk> writes:

>> Which version are you using?  They all print 4 for me, and all but
>> foo() do modify the value of `b'.  foo() doesn't because what follows
>> an

> OTOH, that was a version that had the PEP 227 code checked into it.
> With an older version, I see what you see.  Wierd.

Sounds like a bug fix.


From jeremy at alum.mit.edu  Fri Feb  9 20:13:14 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 09 Feb 2001 19:13:14 GMT
Subject: global, was Re: None assigment
References: <mailman.981652444.20002.python-list@python.org> <95umoi$is5$1@panix2.panix.com> <_BHg6.195181$ge4.67634054@news2.rdc2.tx.home.com> <95vst8$1na$1@panix2.panix.com> <AHLg6.197584$ge4.68265785@news2.rdc2.tx.home.com> <m3y9vg1d7r.fsf@atrus.jesus.cam.ac.uk> <m3vgqk1d5e.fsf@atrus.jesus.cam.ac.uk> <9613mn$9v$1@news.mathworks.com>
Message-ID: <961fg0$etd$1@nnrp1.deja.com>

In article <9613mn$9v$1 at news.mathworks.com>,
  Joshua Marshall <jmarshal at mathworks.com> wrote:
> Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > Michael Hudson <mwh21 at cam.ac.uk> writes:
>
> >> Which version are you using?  They all print 4 for me, and all but
> >> foo() do modify the value of `b'.  foo() doesn't because what
follows
> >> an
>
> > OTOH, that was a version that had the PEP 227 code checked into it.
> > With an older version, I see what you see.  Wierd.
>
> Sounds like a bug fix.
>

As Tim will explain in a post that hasn't made it to DejaNews yet,
earlier versions of Python did not define the behavior of assignment
before a global statement.

The recent round of compiler changes uses separate passes to determine a
name's scope and to generate code for loads and stores.  As a result,
the global statement affects all assignments.

It's unclear what we should happen in this case.  It could be an error,
since it's dodgy and the behavior will change with 2.1.

--
-- Jeremy Hylton, <http://www.python.org/~jeremy/>


Sent via Deja.com
http://www.deja.com/


From tim.one at home.com  Fri Feb  9 21:04:15 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 9 Feb 2001 15:04:15 -0500
Subject: global, was Re: None assigment
In-Reply-To: <961fg0$etd$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEHCIOAA.tim.one@home.com>

[Jeremy Hylton]
> As Tim will explain in a post that hasn't made it to DejaNews yet,
> earlier versions of Python did not define

Neither does 2.1:  changing the implementation didn't change the Ref Man,
and the Ref Man still declines to define the semantics or promise that the
behavior today will persist tomorrow.

> the behavior of assignment

Or any other reference.

> before a global statement.
> ...
> It's unclear what we should happen in this case. It could be an error,
> since it's dodgy and the behavior will change with 2.1.

"Undefined behavior" is unPythonic and should be wiped out whenever
possible.  That these things were dodgy was known from the start, but when
the language was just getting off the ground there were far more important
things to do than generate errors for every conceivable abuse of the
language.  Now that the language is still getting off the ground <wink>,
that's still true.  But changes in the meantime have made it much easier to
identify some of these cases; like:

> The recent round of compiler changes uses separate passes to determine a
> name's scope and to generate code for loads and stores.

The behavior of "global x" after a reference to x has never been defined,
but it's never been reasonably easy to identify and complain about it.  Now
that name classification is done by design instead of by an afterthought
"optimization pass", it should be much easier to gripe.  +1 on making this
an error now.  And if 2.1 is relaxed to again allow "import *" at function
scope in some cases, either that should at least raise a warning, or the Ref
Man should be changed to say that's a defined use of the language.

ambiguity-sucks-ly y'rs  - tim




From tim.one at home.com  Fri Feb  9 08:22:10 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 9 Feb 2001 02:22:10 -0500
Subject: None assigment
In-Reply-To: <95vst8$1na$1@panix2.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEEGIOAA.tim.one@home.com>

[Aahz and Rainer Deyke, volley over a stupid-ass program too long
 without changing topics <wink>]

> b=2
> def foo():
>     b=4
>     print b
>     del b
>     global b
>     print b
> foo()

Just noting that the program's behavior is undefined, so "works" vs "doesn't
work" is a bogus argument (Ref Man, "global" stmt):

    Names listed in a global statement must not be used in the
    same code block textually preceding that global statement.

    ...

    (The current implementation does not enforce the latter two
    restrictions, but programs should not abuse this freedom, as
    future implementations may enforce them or silently change
    the meaning of the program.)

Which is a lot like saying we left bear traps set all over the room, and you
shouldn't abuse your freedom by stepping into one lest you lose a foot or
two <0.9 wink>.

So why doesn't the compiler check for that and gripe?  Don't know.  A good
guess is because Guido figured he had more important things to do at the
time.

and-if-we-knew-what-they-were-i-bet-we'd-agree-ly y'rs  - tim




From daniel.dittmar at sap.com  Thu Feb  8 19:05:52 2001
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Thu, 8 Feb 2001 19:05:52 +0100
Subject: None assigment
References: <mailman.981652444.20002.python-list@python.org>
Message-ID: <95un5v$k6g$1@news1.wdf.sap-ag.de>

None is defined in __builtins__ (type dir (__builtins__)), so it is not a
special name, it's just a name defined in the outer most namespace.

You can locally redefine 'dir' or 'type' as well - and how often did I fall
for that one.

You can even redefine it globally: __builtins__.None = 'Zilch' - not that
you should.

Daniel

--
Daniel Dittmar
daniel.dittmar at sap.com
SAP DB, SAP Labs Berlin
http://www.sapdb.org/





From loewis at informatik.hu-berlin.de  Fri Feb  9 19:02:24 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Feb 2001 19:02:24 +0100
Subject: None assigment
References: <mailman.981652444.20002.python-list@python.org>
Message-ID: <p6qitmj3fbj.fsf@informatik.hu-berlin.de>

Gregoire Welraeds <greg at perceval.be> writes:

> > There's nothing special about the name "None"; it's just a variable,
> > like any other. 
> <snip>
> > After doing "None = 2", you can "del None" to get the default value 
> > back.
> 
> If we follow that logic, I could use any non assigned variable to have the
> following working:

What is the logic here? None is a builtin name (i.e. living in the
__builtins__ dictionary/module), like str, id, and many others. If you
assign to None, then you get another binding in the global or local
namespace, which hides the builtin name. If you del None from the
global namespace, you see the builtin again.

Regards,
Martin


From sholden at holdenweb.com  Thu Feb  8 19:50:00 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 13:50:00 -0500
Subject: None assigment
References: <mailman.981653763.25307.python-list@python.org>
Message-ID: <mXBg6.43226$o5.546257@e420r-atl1.usenetserver.com>

Simon Brunning" <SBrunning at trisystems.co.uk> wrote in message
news:mailman.981653763.25307.python-list at python.org...
> > From: Gregoire Welraeds [SMTP:greg at perceval.be]
> > > After doing "None = 2", you can "del None" to get the default value
> > > back.
>
> Hmm. This is true - I didn't know that.
>
> > If we follow that logic, I could use any non assigned variable to have
the
> > following working:
> >
> > >>> a= [1,'',3]
> > >>> filter(b,a)
> >
> > but this won't work as the interpreter complains that there is no
variable
> > named b.
>
> You can't do this with *ordinary* labels - by del'ing them, you are
> un-assigning them. But the 'None' label is *extraordinary*, obviously.
>
Oh, no it isn't. It's just defined in the __builtins__ namespace. You can't
delete this built in version, but you can, with enough effort, redefine it:

>>> None = 22
>>> print None
22
>>> del None
>>> print None
None
>>> __builtins__.None = 42
>>> print None
42
>>> del None
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: There is no variable named 'None'
>>>


> > It seems that (the so called variable) None is neither a variable like
any
> > other, nor is it a label as stated by Simon Brunning in another post.
>
> It *is* a label - try this:
>
> >>>print None
> None
> >>> None = 'spam'
> >>> print None
> spam
>
> So, clearly, None is a label.
>
Yes, but here you've added another definition of None in the module
namespace, not redefined the one you had in __builtins__ to start with.

> Now that I look at this problem, I'm totally confused about None.
> Any
> > explanaition is welcome. Anyway, I don't understand that one can
override
> > None. Could you give me at least one single good reason to do that.
>
> I can give you a good reason *not* to - it will confuse the hell out of
> people. ;-)
>
> I'm sure one of the Python gurus will give you a bit more background, but
I
> think that the short answer is this - leave 'None' alone to get on with
its
> job, and you won't have any problems.
>
On this we can, I hope, all agree.

regards
 Steve




From sholden at holdenweb.com  Thu Feb  8 20:16:11 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 14:16:11 -0500
Subject: None assigment
References: <mailman.981653763.25307.python-list@python.org> <mXBg6.43226$o5.546257@e420r-atl1.usenetserver.com>
Message-ID: <WjCg6.43228$o5.547199@e420r-atl1.usenetserver.com>

One final little wrinkle:

>>> del __builtins__.None
>>> print None
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: There is no variable named 'None'
>>>

Sorry, I said you coudn't delete the built in version, but you clearly can!

regards
 Steve




From root at rainerdeyke.com  Thu Feb  8 20:45:30 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Thu, 08 Feb 2001 19:45:30 GMT
Subject: None assigment
References: <mailman.981653763.25307.python-list@python.org>
Message-ID: <uHCg6.193739$ge4.67142359@news2.rdc2.tx.home.com>

"Simon Brunning" <SBrunning at trisystems.co.uk> wrote in message
news:mailman.981653763.25307.python-list at python.org...
> > From: Gregoire Welraeds [SMTP:greg at perceval.be]
> > > After doing "None = 2", you can "del None" to get the default value
> > > back.
>
> Hmm. This is true - I didn't know that.
>
> > If we follow that logic, I could use any non assigned variable to have
the
> > following working:
> >
> > >>> a= [1,'',3]
> > >>> filter(b,a)
> >
> > but this won't work as the interpreter complains that there is no
variable
> > named b.
>
> You can't do this with *ordinary* labels - by del'ing them, you are
> un-assigning them. But the 'None' label is *extraordinary*, obviously.

Actually, it's just in '__builtins__'.  When you assign to 'None', you
create a variable in the module namespace which hides '__builtins__.None'.
When you delete it, the 'None' in '__builtins__' is revealed.  This applies
to all labels:

__builtins__.foo = 1
print foo # Prints 1.
foo = 5
print foo # Prints 5.
del foo
print foo # Prints 1 again.  Neat, huh?


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From mwh21 at cam.ac.uk  Thu Feb  8 21:05:48 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 08 Feb 2001 20:05:48 +0000
Subject: None assigment
References: <mailman.981653763.25307.python-list@python.org> <uHCg6.193739$ge4.67142359@news2.rdc2.tx.home.com>
Message-ID: <m31yt92b4z.fsf@atrus.jesus.cam.ac.uk>

"Rainer Deyke" <root at rainerdeyke.com> writes:

> "Simon Brunning" <SBrunning at trisystems.co.uk> wrote in message
> news:mailman.981653763.25307.python-list at python.org...
> > > From: Gregoire Welraeds [SMTP:greg at perceval.be]
> > > > After doing "None = 2", you can "del None" to get the default value
> > > > back.
> >
> > Hmm. This is true - I didn't know that.
> >
> > > If we follow that logic, I could use any non assigned variable to have
> the
> > > following working:
> > >
> > > >>> a= [1,'',3]
> > > >>> filter(b,a)
> > >
> > > but this won't work as the interpreter complains that there is no
> variable
> > > named b.
> >
> > You can't do this with *ordinary* labels - by del'ing them, you are
> > un-assigning them. But the 'None' label is *extraordinary*, obviously.
> 
> Actually, it's just in '__builtins__'.  When you assign to 'None', you
> create a variable in the module namespace which hides '__builtins__.None'.
> When you delete it, the 'None' in '__builtins__' is revealed.  This applies
> to all labels:

It's not very often I come up against something I don't understand
in Python, but try this:

>>> del __builtins__ 

This makes Python, erm, interesting, to use:

>>> import sys
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: __import__ not found

but yet:

>>> None 
>>> 

i.e. None still gets found.

What's going on?  frameobject.c:172 is what's going on:

	if (builtins == NULL) {
		/* No builtins!  Make up a minimal one. */
		builtins = PyDict_New();
		if (builtins == NULL || /* Give them 'None', at least. */
		    PyDict_SetItemString(builtins, "None", Py_None) < 0) {
			Py_DECREF(f);
			return NULL;
		}
	}

Well, I though I'd share my little traipse through Python's innards
with you all...

Cheers,
M.

-- 
  Clue: You've got the appropriate amount of hostility for the
  Monastery, however you are metaphorically getting out of the 
  safari jeep and kicking the lions.                         -- coonec
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From greg at perceval.be  Fri Feb  9 09:38:42 2001
From: greg at perceval.be (Gregoire Welraeds)
Date: Fri, 9 Feb 2001 09:38:42 +0100 (CET)
Subject: None assigment
In-Reply-To: <m31yt92b4z.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <Pine.BSF.4.21.0102090929160.10564-100000@localhost>

In reply to the message of Michael Hudson sent on Feb 8 (see below) :

~# What's going on?  frameobject.c:172 is what's going on:
~# 
~# 	if (builtins == NULL) {
~# 		/* No builtins!  Make up a minimal one. */
~# 		builtins = PyDict_New();
~# 		if (builtins == NULL || /* Give them 'None', at least. */
~# 		    PyDict_SetItemString(builtins, "None", Py_None) < 0) {
~# 			Py_DECREF(f);
~# 			return NULL;
~# 		}
~# 	}

So we guess that the None "object" (label, value, and so on :) is always
to a minimal builtin namespace because it is required to create a
"void" function (function that return nothing). And maybe there is a other
place in the python C code where None have to be defined. 

Maybe None have been added to builtin to avoid special cases. But it
becomes clear, as we understand it, that None is now a special case of
builtin. 

So, following that thought, we think that None should be a reserved word. 

Furthermore, None is some kind of magic value which have a special
conceptual meaning such as nil in other languages or the magic /dev/null
...

And we all agree that programmers should not touch to the value of
__builtin__.None. We think that the ability to override None because None
is a builtin is a conceptual problem.

Gregoire and Xavier
Perceval Development team, python division (and some other more :)
-------------------------------------------------------------------------------
Perceval Technologies sa/nv	Tel: +32-2-6409194		
Rue Tenbosch, 9			Fax: +32-2-6403154		
B-1000 Brussels			general information:   info at perceval.net
BELGIUM				technical information: helpdesk at perceval.net
URL: http://www.perceval.be/
-------------------------------------------------------------------------------

On 8 Feb 2001, Michael Hudson wrote:

~# Date: 08 Feb 2001 20:05:48 +0000
~# From: Michael Hudson <mwh21 at cam.ac.uk>
~# To: python-list at python.org
~# Newsgroups: comp.lang.python
~# Subject: Re: None assigment
~# 
~# "Rainer Deyke" <root at rainerdeyke.com> writes:
~# 
~# > "Simon Brunning" <SBrunning at trisystems.co.uk> wrote in message
~# > news:mailman.981653763.25307.python-list at python.org...
~# > > > From: Gregoire Welraeds [SMTP:greg at perceval.be]
~# > > > > After doing "None = 2", you can "del None" to get the default value
~# > > > > back.
~# > >
~# > > Hmm. This is true - I didn't know that.
~# > >
~# > > > If we follow that logic, I could use any non assigned variable to have
~# > the
~# > > > following working:
~# > > >
~# > > > >>> a= [1,'',3]
~# > > > >>> filter(b,a)
~# > > >
~# > > > but this won't work as the interpreter complains that there is no
~# > variable
~# > > > named b.
~# > >
~# > > You can't do this with *ordinary* labels - by del'ing them, you are
~# > > un-assigning them. But the 'None' label is *extraordinary*, obviously.
~# > 
~# > Actually, it's just in '__builtins__'.  When you assign to 'None', you
~# > create a variable in the module namespace which hides '__builtins__.None'.
~# > When you delete it, the 'None' in '__builtins__' is revealed.  This applies
~# > to all labels:
~# 
~# It's not very often I come up against something I don't understand
~# in Python, but try this:
~# 
~# >>> del __builtins__ 
~# 
~# This makes Python, erm, interesting, to use:
~# 
~# >>> import sys
~# Traceback (most recent call last):
~#   File "<stdin>", line 1, in ?
~# ImportError: __import__ not found
~# 
~# but yet:
~# 
~# >>> None 
~# >>> 
~# 
~# i.e. None still gets found.
~# 
~# What's going on?  frameobject.c:172 is what's going on:
~# 
~# 	if (builtins == NULL) {
~# 		/* No builtins!  Make up a minimal one. */
~# 		builtins = PyDict_New();
~# 		if (builtins == NULL || /* Give them 'None', at least. */
~# 		    PyDict_SetItemString(builtins, "None", Py_None) < 0) {
~# 			Py_DECREF(f);
~# 			return NULL;
~# 		}
~# 	}
~# 
~# Well, I though I'd share my little traipse through Python's innards
~# with you all...
~# 
~# Cheers,
~# M.
~# 
~# -- 
~#   Clue: You've got the appropriate amount of hostility for the
~#   Monastery, however you are metaphorically getting out of the 
~#   safari jeep and kicking the lions.                         -- coonec
~#                -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html
~# -- 
~# http://mail.python.org/mailman/listinfo/python-list
~# 




From SBrunning at trisystems.co.uk  Fri Feb  9 10:08:52 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 9 Feb 2001 09:08:52 -0000 
Subject: None assigment
Message-ID: <31575A892FF6D1118F5800600846864D5B17D6@intrepid>

> From:	Michael Hudson [SMTP:mwh21 at cam.ac.uk]
> >>> del __builtins__ 
> 
> This makes Python, erm, interesting, to use:
> 
> >>> import sys
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: __import__ not found
 
Eeek!
 
> but yet:
> 
> >>> None 
> >>> 
> 
> i.e. None still gets found.
 
Always something new to learn on c.l.py. *Totally* useless of course, but
fascinating nevertheless. ;-)

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk
There is nothing so absurd that it has not been said by philosophers. -
Cicero




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From peter at engcorp.com  Sat Feb 10 18:46:56 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 10 Feb 2001 12:46:56 -0500
Subject: None assigment
References: <mailman.981708020.6548.python-list@python.org>
Message-ID: <3A857E90.18EE0634@engcorp.com>

Gregoire Welraeds wrote:
> 
> Maybe None have been added to builtin to avoid special cases. But it
> becomes clear, as we understand it, that None is now a special case of
> builtin.
> 
> So, following that thought, we think that None should be a reserved word.
> 
[snip]
> 
> And we all agree that programmers should not touch to the value of
> __builtin__.None. We think that the ability to override None because None
> is a builtin is a conceptual problem.

No!  Please, why muck with things that work perfectly well?

I can think of at least one situation in which the ability
to do this could be of some _real_ value.  Making None
a reserved word would prevent me from doing that.  It would
gain you nothing except a stricter language which would
be getting ever closer to those "other" languages which
are so restrictive that they place limits on your imagination
and ability to innovate.

Why, I wonder, do some people think that just because
something "shouldn't" happen, it must therefore 
never be allowed to happen?  If "we all agree that 
programmers should not touch None", can't we just
stop there and leave those mature people to do what
they wish?  (People "shouldn't speed", either, but
I don't think many want governors on their car engine 
actually to _prevent_ it.)

Remember, Python is a dynamic language.  The more
special-case restictions you place on it, the less
like Python it's going to look.

likely-starting-a-flame-war-ly yr's  -Peter

P.S. Identifying useful things that could be done with
assigning to None is left as an exercise to the reader.
I wouldn't want to stifle creativity by mentioning my
idea too quickly. :-)


From clarence at netlojix.com  Sun Feb 11 03:09:18 2001
From: clarence at netlojix.com (Clarence Gardner)
Date: Sat, 10 Feb 2001 18:09:18 -0800
Subject: None assigment
References: <mailman.981708020.6548.python-list@python.org> <3A857E90.18EE0634@engcorp.com>
Message-ID: <981857533.2047761672@news.silcom.com>

On Sat, 10 Feb 2001, Peter Hansen wrote:
>Why, I wonder, do some people think that just because
>something "shouldn't" happen, it must therefore 
>never be allowed to happen?  If "we all agree that 
>
I think, if there is a reason, it is probably that they want
the language to prevent the programmer from *accidentally*
doing bad things that happen to be possible.  But nobody
types
	None = 3
by accident.  And since, unlike some other languages, that
is the only way to do it, there is no purpose in prohibiting it.

$.02 poorer,
Clarence


From tim_one at email.msn.com  Sun Feb 11 07:22:10 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 11 Feb 2001 01:22:10 -0500
Subject: None assigment
In-Reply-To: <981857533.2047761672@news.silcom.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEKLIOAA.tim_one@email.msn.com>

[Peter Hansen]
> ...
> Why, I wonder, do some people think that just because
> something "shouldn't" happen, it must therefore
> never be allowed to happen?  If "we all agree that

[Clarence Gardner]
> I think, if there is a reason, it is probably that they want
> the language to prevent the programmer from *accidentally*
> doing bad things that happen to be possible.  But nobody
> types
> 	None = 3
> by accident.  And since, unlike some other languages, that
> is the only way to do it, there is no purpose in prohibiting it.

Actually, some people do write stuff like:

    for x, y, None in list_of_threetuples:
        print f(x, y)

when they don't care about the third tuple elements.  I know this because I
was one of them <wink>.  Got away with it for months, and just *assumed* the
language supported "None" as a "don't care" assignment target.  This is easy
to believe if you're coming from, say, Perl.

There's a bit of Deep Magic that makes this illusion hard to shake if you
fall under it:

>>> def f():
...     None = 42
>>> None = 66
>>> print f()
None
>>> type(f())
<type 'None'>
>>>

That is, rebinding None has no effect on that functions that fall off the
end still return "the old" value of None; the "intended" value of None is
hardcoded into the fall-off-the-end function return sequence.

In that respect, None is unique among non-keyword names, and for that reason
alone it should be a keyword instead of a global.  I expect that will happen
someday, too.  But it's a race between that and hell freezing over <wink>.

hard-to-get-excited-about-it-even-when-i-was-young-ly y'rs  - tim




From qrczak at knm.org.pl  Sun Feb 18 18:02:10 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 18 Feb 2001 17:02:10 GMT
Subject: None assigment
References: <mailman.981708020.6548.python-list@python.org> <3A857E90.18EE0634@engcorp.com> <981857533.2047761672@news.silcom.com>
Message-ID: <slrn99000i.nse.qrczak@qrnik.zagroda>

Sat, 10 Feb 2001 18:09:18 -0800, Clarence Gardner <clarence at netlojix.com> pisze:

> I think, if there is a reason, it is probably that they want the
> language to prevent the programmer from *accidentally* doing bad
> things that happen to be possible.  But nobody types
> 	None = 3
> by accident.

It's harmless anyway: it rebinds a local name, not the builtin one.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From thomas at xs4all.net  Sun Feb 18 22:10:53 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 18 Feb 2001 22:10:53 +0100
Subject: None assigment
In-Reply-To: <slrn99000i.nse.qrczak@qrnik.zagroda>; from qrczak@knm.org.pl on Sun, Feb 18, 2001 at 05:02:10PM +0000
References: <mailman.981708020.6548.python-list@python.org> <3A857E90.18EE0634@engcorp.com> <981857533.2047761672@news.silcom.com> <slrn99000i.nse.qrczak@qrnik.zagroda>
Message-ID: <20010218221053.B14292@xs4all.nl>

On Sun, Feb 18, 2001 at 05:02:10PM +0000, Marcin 'Qrczak' Kowalczyk wrote:
> Sat, 10 Feb 2001 18:09:18 -0800, Clarence Gardner <clarence at netlojix.com> pisze:

> > I think, if there is a reason, it is probably that they want the
> > language to prevent the programmer from *accidentally* doing bad
> > things that happen to be possible.  But nobody types
> > 	None = 3
> > by accident.

> It's harmless anyway: it rebinds a local name, not the builtin one.

It's *mostly* harmless. Consider

  spam, None, eggs, None = string.split("this is a string")
  viking = filter(None, (spam, eggs))

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From fredrik at pythonware.com  Sun Feb 18 22:35:37 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 21:35:37 GMT
Subject: None assigment
References: <mailman.981708020.6548.python-list@python.org> <3A857E90.18EE0634@engcorp.com> <981857533.2047761672@news.silcom.com> <slrn99000i.nse.qrczak@qrnik.zagroda> <mailman.982530686.24078.python-list@python.org>
Message-ID: <JeXj6.18187$AH6.2495470@newsc.telia.net>

Thomas Wouters wrote:
> It's *mostly* harmless. Consider
>
>   spam, None, eggs, None = string.split("this is a string")
>   viking = filter(None, (spam, eggs))

now put them both in the same function, with the filter
statement on line 5 and the split on line 65.

trust me, you can waste lots of time trying to figure out
what causes that "NameError: None" message on line 5...

Cheers /F




From dsh8290 at rit.edu  Mon Feb 19 03:56:45 2001
From: dsh8290 at rit.edu (D-Man)
Date: Sun, 18 Feb 2001 21:56:45 -0500
Subject: None assigment
In-Reply-To: <JeXj6.18187$AH6.2495470@newsc.telia.net>; from fredrik@pythonware.com on Sun, Feb 18, 2001 at 09:35:37PM +0000
References: <mailman.981708020.6548.python-list@python.org> <3A857E90.18EE0634@engcorp.com> <981857533.2047761672@news.silcom.com> <slrn99000i.nse.qrczak@qrnik.zagroda> <mailman.982530686.24078.python-list@python.org> <JeXj6.18187$AH6.2495470@newsc.telia.net>
Message-ID: <20010218215645.A12311@harmony.cs.rit.edu>

On Sun, Feb 18, 2001 at 09:35:37PM +0000, Fredrik Lundh wrote:
| Thomas Wouters wrote:
| > It's *mostly* harmless. Consider
| >
| >   spam, None, eggs, None = string.split("this is a string")
| >   viking = filter(None, (spam, eggs))
| 
| now put them both in the same function, with the filter
| statement on line 5 and the split on line 65.
| 
| trust me, you can waste lots of time trying to figure out
| what causes that "NameError: None" message on line 5...

True (been there, done that,  with other guns), but,  my terminal is
only 24 lines long.  Functions that don't fit on the screen (usually)
need to be split up better.

:-)

-D




From tonyc at bellsouth.net  Thu Feb  8 13:40:55 2001
From: tonyc at bellsouth.net (Tony Corrente)
Date: Thu, 08 Feb 2001 12:40:55 GMT
Subject: Newbie ?? GUI 
Message-ID: <3A8293E4.2482AEE8@bellsouth.net>

Hello.

I am very new to Python so please excuse me if this question is stupid
but I can't find an example anywhere.  I've created a form using Glade.
The form has a entry field name pName and a button.  I've run glc.py and

created the Python code. Everything worked..  glc created 2 files. the
main program and the window handler file for the button which is
imported into the main program.  What I want to know is... How do I
update pName when the button is clicked? I've tried just using

def on_bNext_clicked(widget, mainObj):
    pname.set_text("This is a test")

when I run the program and click the button I get the following error
Traceback (innermost last):
  File "/usr/lib/python1.5/site-packages/gtk.py", line 125, in __call__
    ret = apply(self.func, a)
  File "window1Handlers.py", line 15, in on_bNext_clicked
    pName.set_text("This is a test")
NameError: pName

How do I access pName ( or any other widget in the window ) ???
Could anyone please guide me in the right direction ???

Thanks in advance...






From paul.robinson at businesscollaborator.com  Thu Feb  8 14:13:56 2001
From: paul.robinson at businesscollaborator.com (Paul Robinson)
Date: Thu, 08 Feb 2001 13:13:56 +0000
Subject: Forking and Threads
Message-ID: <3A829B94.9AF5A831@businesscollaborator.com>


OK, before I start, I realise that threading and forking are not
necessarily well known for playing well with each other. But given that,
and assuming that this is the road I have chosen to go down can anyone
explain why the attached program (also available at:
http://cobweb.quantisci.co.uk/pub/english.cgi/d10030658/forkingtest2.py
) will run happily for several minutes (upto about 10mins so far) and
then just sit there with the main process running with as much CPU time
as the OS will give it!

This is the "distilled down" version of the problem that I'm seeing in a
much larger application.

I've only tested this on:
Python 1.5.2
Red Hat Linux release 6.2 (Zoot)
Kernel 2.2.14-5.0smp on a 2-processor i686

and would very interested in how (and if) it performs on other operating
systems - obviously only those with both threading and forking need
apply.

I find that "watch -n 1 "pstree -c -p -a <user name> | grep python |
grep -v grep"" gives a overview of the running of the program as you can
see the child processes appearing and disappearing.

A large number of child processes and a short run time for the children
seems to cause the crash most quickly. These parameters can be set with
-MC (MAX_CHILDREN) and -MS (MAX_SECONDS the child will wait for).

forkingtest2.py -MC 20 -MS 3 crashes after a while on my dual 350MHz
PII.

Any insight into this behaviour would be gratefully appreciated.

Paul Robinson
Business Collaborator Development Manager
Enviros Software Solutions
WWW: http://www.businesscollaborator.com
-------------- next part --------------
#!/usr/bin/python

import os, sys
import marshal

from threading import Thread, currentThread
from time import sleep, time
from random import choice, seed
from string import atoi

CHILD = 0
MAX_CHILDREN = 5
MAX_SLEEP = 5

def log(*args):
	print '\t'*CHILD, os.getpid(),' : ', currentThread(), \
		' : ', reduce(lambda x, y: x + ' '+ y, map(str,args))

def processLoop():
	children = []
	while 1:

		while len(children) < MAX_CHILDREN:
			pid = os.fork()
			if pid:
				#parent
				children.append(pid)
			else:
				#child
				try:
					global CHILD
					CHILD = CHILD + 1
					#Do stuff
					seed(time())
					pause = choice(range(1,MAX_SLEEP))
					log('Sleeping', pause)
					sleep(pause)
					log('Exitting')
				finally:
					os._exit(0)
		else:
			log('Too many children')

		sleep(3)
		# Wait on children
		for i in range(len(children)-1, -1, -1):
			# Call waitpid (os.WNOHANG - non-blocking)
			status = os.waitpid(children[i], os.WNOHANG)
			if status[0] > 0:
				# Child has been "deaded"
				del children[i]

if __name__ == '__main__':
	try:
		i = sys.argv.index('-MC')
		global MAX_CHILDREN
		MAX_CHILDREN = atoi(sys.argv[i+1])
	except:
		pass

	print 'Set MAX_CHILDREN to', MAX_CHILDREN, ' - use "-MC n" to change'

	try:
		i = sys.argv.index('-MS')
		global MAX_SLEEP
		MAX_SLEEP = atoi(sys.argv[i+1])
		MAX_SLEEP = ((MAX_SLEEP >= 2) and MAX_SLEEP) or 2
	except:
		pass
	print 'Set MAX_SLEEP to', MAX_SLEEP, ' - use "-MS m" to change (where m >= 2)'

	pl = Thread(target=processLoop, name='processLoop')
	pl.start()
	currentThread().setName('Main Loop')
	f = open('/tmp/temp','a')
	while 1:
		log('Looping')
		sleep(5)




From paul.robinson at businesscollaborator.com  Thu Feb  8 15:06:13 2001
From: paul.robinson at businesscollaborator.com (Paul Robinson)
Date: Thu, 08 Feb 2001 14:06:13 +0000
Subject: Forking and Threads
References: <3A829B94.9AF5A831@businesscollaborator.com>
Message-ID: <3A82A7D5.5C88E195@businesscollaborator.com>

Paul Robinson wrote:
> 
> will run happily for several minutes (upto about 10mins so far) and
> then just sit there with the main process running with as much CPU time
> as the OS will give it!

Update:
My last run ( forkingtest2.py -MC 50 -MS 3 ) ran for 24 minutes before
running at a steady 99% CPU!

Paul Robinson
Business Collaborator Development Manager
Enviros Software Solutions
WWW: http://www.businesscollaborator.com



From db3l at fitlinxx.com  Thu Feb  8 23:45:22 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 08 Feb 2001 17:45:22 -0500
Subject: Forking and Threads
References: <3A829B94.9AF5A831@businesscollaborator.com> <mailman.981641418.2058.python-list@python.org>
Message-ID: <uae7wyet9.fsf@ctwd0143.fitlinxx.com>

Paul Robinson <paul.robinson at businesscollaborator.com> writes:

> Update:
> My last run ( forkingtest2.py -MC 50 -MS 3 ) ran for 24 minutes before
> running at a steady 99% CPU!

What do you see in your log file leading up to the hang?  Does there
seem to be continued logging during the CPU burn or is it quiet?  I'd
probably try to figure out just what portion of the code the main
process was in while burning CPU (it sort of looks like it could
really only be in your waitpid loop, but a few additional log messages
might be useful).

Also, you might try exiting the child process more nicely (e.g., with
a sys.exit rather than os._exit) in case it's a question of resource
cleanup of some sort.

BTW, I can't actually run it here (Windows) but do you ever see a log
message from a child other than "1" - it looks like you only increment
CHILD inside the child process itself, so wouldn't all the children
get a value of 1?

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From paul.robinson at businesscollaborator.com  Fri Feb  9 10:26:06 2001
From: paul.robinson at businesscollaborator.com (Paul Robinson)
Date: Fri, 09 Feb 2001 09:26:06 +0000
Subject: Forking and Threads
References: <3A829B94.9AF5A831@businesscollaborator.com> <mailman.981641418.2058.python-list@python.org> <uae7wyet9.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A83B7AE.5D312C6@businesscollaborator.com>

David Bolen wrote:
> 
> Paul Robinson <paul.robinson at businesscollaborator.com> writes:
> 
> > Update:
> > My last run ( forkingtest2.py -MC 50 -MS 3 ) ran for 24 minutes before
> > running at a steady 99% CPU!
> 
> What do you see in your log file leading up to the hang?  Does there
> seem to be continued logging during the CPU burn or is it quiet?

Normal upto the the CPU burn and then nothing.

> I'd
> probably try to figure out just what portion of the code the main
> process was in while burning CPU (it sort of looks like it could
> really only be in your waitpid loop, but a few additional log messages
> might be useful).

Yes, I agree. The original case I was working on (in the larger program)
was instrumented more heavily and I came to the conclusion that it was
getting stuck at a random point! There were sockets and select involved
with that as well so it was slightly more complicated. 90% of the time
it would crash while doing the select - but I guess that was just
because 90% of the program time was spent doing the select!
 
> BTW, I can't actually run it here (Windows) but do you ever see a log
> message from a child other than "1" - it looks like you only increment
> CHILD inside the child process itself, so wouldn't all the children
> get a value of 1?

The value of CHILD is just there in case children are spawning children
(they were in previous versions...)

It ran successfully overnight using Python 2.0!

Paul Robinson
Business Collaborator Development Manager
Enviros Software Solutions
WWW: http://www.businesscollaborator.com



From tim.one at home.com  Fri Feb  9 05:17:43 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 8 Feb 2001 23:17:43 -0500
Subject: Forking and Threads
In-Reply-To: <3A829B94.9AF5A831@businesscollaborator.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEEAIOAA.tim.one@home.com>

[posted & mailed]

[Paul Robinson]
> ...
> http://cobweb.quantisci.co.uk/pub/english.cgi/d10030658/forkingtest2.py
> ) will run happily for several minutes (upto about 10mins so far) and
> then just sit there with the main process running with as much CPU time
> as the OS will give it!
>
> ...
>
> I've only tested this on:
> Python 1.5.2
> Red Hat Linux release 6.2 (Zoot)
> Kernel 2.2.14-5.0smp on a 2-processor i686
>
> and would very interested in how (and if) it performs on other operating
> systems - obviously only those with both threading and forking need
> apply.

Try it under Python 2.0.  Similar problems were reported earlier, but only
(IIRC) under Linux.  A patch went into 2.0 that at least allowed Python's
own Lib/test/test_fork1.py to complete normally.  However, the behavior was
never fully understood; the best guess we were left with is that, across a
fork, the Linux flavor of mutexes and/or condition variables were mistakenly
still sharing some little of piece of internal state (Python's own basic
flavor of lock is implemented under pthreads via a mutex + condvar combo,
and after a fork the observed behavior of Python's internal "global
interpreter lock" didn't make any sense).  2.0 "fixed" that via brute force
panic:  after a fork, the child allocates a brand new interpreter lock for
its own use; it leaks the one inherited from the parent process.  Nobody was
able to do substantially better than this despite countless hours of trying,
so, under the assumption that it was a Linux bug that would eventually get
fixed, we all forgot about it.

blissfully-too<wink>-ly y'rs  - tim




From paul.robinson at businesscollaborator.com  Fri Feb  9 10:50:39 2001
From: paul.robinson at businesscollaborator.com (Paul Robinson)
Date: Fri, 09 Feb 2001 09:50:39 +0000
Subject: Forking and Threads
References: <LNBBLJKPBEHFEDALKOLCMEEAIOAA.tim.one@home.com>
Message-ID: <3A83BD6F.42D236C1@businesscollaborator.com>

Tim Peters wrote:
> 
> [posted & mailed]
> 
> [Paul Robinson]
> > ...
> > http://cobweb.quantisci.co.uk/pub/english.cgi/d10030658/forkingtest2.py
> > ) will run happily for several minutes (upto about 10mins so far) and
> > then just sit there with the main process running with as much CPU time
> > as the OS will give it!

> Try it under Python 2.0.

That was one of my thoughts as well. It ran successfully overnight using
Python 2.0 on the same system.

> 2.0 "fixed" that via brute force
> panic:  after a fork, the child allocates a brand new interpreter lock for
> its own use; it leaks the one inherited from the parent process.

I didn't check the memory usage before I terminated the run this
morning. Should I be expecting significant resource leakage with a long
(and I mean looong) running process?
If the leak is in the forked child then that shouldn't be a problem as
they are short lived.

> Nobody was
> able to do substantially better than this despite countless hours of trying,
> so, under the assumption that it was a Linux bug that would eventually get
> fixed, we all forgot about it.

Hmm. OK. That just leaves me in a slightly tricky situation of verifying
that the other few hundred thousand LOC that make up the software
product also work with Python 2.0! - we had that scheduled for April but
this set back might require being slightly more flexible.

Is there any suggestion of how I could do something similar (or work
around the problem) using 1.5.2?

Thanks for the advice.

Paul Robinson
Business Collaborator Development Manager
Enviros Software Solutions
WWW: http://www.businesscollaborator.com



From maxwell at ebioinformatics.com  Fri Feb  9 12:48:15 2001
From: maxwell at ebioinformatics.com (Peter Maxwell)
Date: Fri, 09 Feb 2001 22:48:15 +1100
Subject: Forking and Threads
References: <LNBBLJKPBEHFEDALKOLCMEEAIOAA.tim.one@home.com> <mailman.981712512.24831.python-list@python.org>
Message-ID: <3A83D8FF.A90F4E4A@ebioinformatics.com>

Paul Robinson wrote:
> Is there any suggestion of how I could do something similar (or work
> around the problem) using 1.5.2?

I had a different forking-and-locking problem which might share a
similar solution: A multithreaded python program running on solaris
which would lock up when running on a multiprocessor machine.  The
program didn't use os.fork but it did use os.popen, os.system and
commands.getstatusoutput and of course at some level those all fork. 
Simultaneous forking by two threads seemed to be the problem so I
replaced os.popen with:

	AT_FORK = thread.allocate_lock()
	def popen(cmd, mode="r"):		
		AT_FORK.acquire()	
		try:
			pipe = os.popen(cmd, mode)
		finally:
			AT_FORK.release()
		return pipe

and implemented replacements for os.system and commands.getstatusoutput
using this version of popen. 

 -- Peter Maxwell


From db3l at fitlinxx.com  Fri Feb  9 19:59:16 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 09 Feb 2001 13:59:16 -0500
Subject: Forking and Threads
References: <LNBBLJKPBEHFEDALKOLCMEEAIOAA.tim.one@home.com> <mailman.981712512.24831.python-list@python.org>
Message-ID: <ur917wum3.fsf@ctwd0143.fitlinxx.com>

Paul Robinson <paul.robinson at businesscollaborator.com> writes:

> Is there any suggestion of how I could do something similar (or work
> around the problem) using 1.5.2?

I'm not familiar with this piece of code directly, but one option
would always be to peek at the 2.0 source in CVS across this change,
and see how easy it would be to transport the delta to a 1.5.2 source
tree and then re-compile your own executable (providing you have
control of the version of python itself with your product).

That way the only thing changing would be the lock behavior across the
fork so the rest of your 1.5.2 testing would still be applicable.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From wonkowatson at yahoo.com  Thu Feb  8 14:46:54 2001
From: wonkowatson at yahoo.com (John Sands)
Date: Thu, 08 Feb 2001 13:46:54 -0000
Subject: Python and JabberCOM
Message-ID: <95u80e+g1f3@eGroups.com>

I'm trying to use JabberCOM from Python and it isn't working. Here's 
some code:

import win32com.client
import pythoncom

class JabberEvents:
    def OnAuthError(self, ErrorText):
        print "OnAuthError", ErrorText
    def OnCommError(self, ErrorText):
        print "OnCommError", ErrorText
    def OnConnect(self):
        print "OnConnect"
def test1():
    jabber = win32com.client.DispatchWithEvents
("JabberCOM.JabberSession", JabberEvents)
    jabber.Server = "jabber.org"
    jabber.Username = "WonkoWatson"
    jabber.Password = "pass"
    jabber.Resource = "sync"
    print jabber.Active
    jabber.DoConnect(0 , 0)
    pythoncom.PumpWaitingMessages()
    win32api.Sleep(10000)
    print jabber.Active

test1()
print "end"

I get no events firing, and the jabber.Active shows 0 both times.
If I comment out any of the Server,Username,Password or Resource 
properties then I get an OnAuthError, so I know the event hook is 
working. I can execute exactly the same code from VB and I get the 
OnConnect event in a few seconds:

    Public WithEvents JSession As JabberSession
    ...
    Set JSession = New JabberSession
    JSession.server = server
    JSession.username = username
    JSession.password = password
    JSession.Resource = "VBJC_Tester"
    JSession.DoConnect False, jatPlainText

Thanks for any advice.





From MarkH at ActiveState.com  Thu Feb  8 22:41:35 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 08 Feb 2001 21:41:35 GMT
Subject: Python and JabberCOM
References: <mailman.981640803.32201.python-list@python.org>
Message-ID: <3A8311A1.3000502@ActiveState.com>

John Sands wrote:


>     jabber.DoConnect(0 , 0)
>     pythoncom.PumpWaitingMessages()
>     win32api.Sleep(10000)
>     print jabber.Active

I am guessing this is your problem - you probably need a "real" message loop.

Try changing it to something like:

jabber.DoConnect(0 , 0)
while not done:
   pythoncom.PumpWaitingMessages()
   win32api.Sleep(0)

print jabber.Active

And have your event handler set "done" to 1 when something interesting happens.  If this works, mail me and we can work out a more appropriate way than a hard-loop pumping these messages.

Mark.



From chris.gonnerman at usa.net  Thu Feb  8 15:13:42 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 8 Feb 2001 08:13:42 -0600
Subject: ANNOUNCE:  new gdmodule
Message-ID: <009601c091d9$5bc45c60$a100000a@local>

With Richard Jones' blessing I have taken over support for his Python GD
module, and issued a
new version.  Visit:

    http://newcenturycomputers.net/projects/gdmodule

if you are interested.

The new version includes support for most GD 1.8.3 functionality including
TrueType font
support.





From nospam at our.site  Thu Feb  8 15:30:24 2001
From: nospam at our.site (nospam at our.site)
Date: Thu, 08 Feb 2001 14:30:24 +0000
Subject: catching all exceptions
Message-ID: <95uati$cm4$1@icrf.news>

This message has been posted by:  Arne Mueller <a.mueller at REMOVE-THIS-TO-SENDicrf.icnet.uk>

Hi All,

I'm (still) using python 1.5.2 and I'd like to catch all exceptions in a
try block then find out what kind of exception excatly was risen and
save the output that would have been produced if I hadn't cought the
exception (or the stack-trace up to this position). Something like this:

try:
    do_something()
except (IOError, ...), value:
    this_exception = Exception_class
    ...
    write_Exception_to_database(this_exception, value)
    raise this_exception, value

The thing is that my program may raise all kinds of exceptions (not only
user defined exceptions), and I'd like to write the exception class (not
the Exceptions base class), the value, and the stack-trace up to this
point as one big string to a database before terminating the program.

currently I'm doing:

try:
  do_something()
except Exceptions, value:
  write_Exception_to_database('Python-Error: ' + repr(value))
  sys.exit(-1)  

with this I think I can catch all predefined exception classes, because
Exceptions is the base class for all other exceptions, but I realy want
to know which derived exception class was raised (and also the
stack-trace).

Is there (simple) way to do that in python (1.5.2)?

	thanks alot for help,

	Arne

-- 
Arne Mueller
Biomolecular Modelling Laboratory
Imperial Cancer Research Fund
44 Lincoln's Inn Fields
London WC2A 3PX, U.K.
phone : +44-(0)207 2693405      | fax :+44-(0)207-269-3534
email : a.mueller at icrf.icnet.uk | http://www.bmm.icnet.uk


From sholden at holdenweb.com  Thu Feb  8 16:11:50 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 10:11:50 -0500
Subject: catching all exceptions
References: <95uati$cm4$1@icrf.news>
Message-ID: <6Lyg6.43015$o5.540092@e420r-atl1.usenetserver.com>

<nospam at our.site> wrote in message news:95uati$cm4$1 at icrf.news...
> This message has been posted by:  Arne Mueller
<a.mueller at REMOVE-THIS-TO-SENDicrf.icnet.uk>
>
> Hi All,
>
> I'm (still) using python 1.5.2 and I'd like to catch all exceptions in a
> try block then find out what kind of exception excatly was risen and
> save the output that would have been produced if I hadn't cought the
> exception (or the stack-trace up to this position). Something like this:
>
> try:
>     do_something()
> except (IOError, ...), value:
>     this_exception = Exception_class
>     ...
>     write_Exception_to_database(this_exception, value)
>     raise this_exception, value
>
In general, of course, you can't be sure whether there is additional
information in the exception, since that depends entirely on how it was
raised (and I don't mean were its parents kind to it).

You presumably realisethat you can catch all exceptions with a simple
"except:" clause?

> The thing is that my program may raise all kinds of exceptions (not only
> user defined exceptions), and I'd like to write the exception class (not
> the Exceptions base class), the value, and the stack-trace up to this
> point as one big string to a database before terminating the program.
>
> currently I'm doing:
>
> try:
>   do_something()
> except Exceptions, value:
>   write_Exception_to_database('Python-Error: ' + repr(value))
>   sys.exit(-1)
>
> with this I think I can catch all predefined exception classes, because
> Exceptions is the base class for all other exceptions, but I realy want
> to know which derived exception class was raised (and also the
> stack-trace).
>
Exceptions defined by library modules *should* define their exceptions as
Exception subclasses, but there is no guarantee they will. Hence the value
of the "except:" clause, which guarantees to catch everything (including
SystemExit and KeyboardInterrupt, so you have to be careful...).

> Is there (simple) way to do that in python (1.5.2)?
>
try:
    some extremely hairy code
except:
    report as effectively as possible on the exception
    raise # re-raise it for outer except clauses to deal with

> thanks alot for help,
>
> Arne
>
The "raise" statement with no exception specifications simply passes the
same exception up to whatever is waiting "above" to deal with it, ultimately
resulting in an interpreter error message if nothing else is available.

regards
 Steve



From bsb at winnegan.de  Thu Feb  8 16:13:39 2001
From: bsb at winnegan.de (Siggy Brentrup)
Date: 08 Feb 2001 16:13:39 +0100
Subject: catching all exceptions
In-Reply-To: <95uati$cm4$1@icrf.news>
References: <95uati$cm4$1@icrf.news>
Message-ID: <87y9vhp5r0.fsf@winnegan.de>

nospam at our.site writes:


[...]

> currently I'm doing:
> 
> try:
>   do_something()
> except Exceptions, value:
>   write_Exception_to_database('Python-Error: ' + repr(value))
>   sys.exit(-1)  
> 
> with this I think I can catch all predefined exception classes, because
> Exceptions is the base class for all other exceptions, but I realy want
> to know which derived exception class was raised (and also the
> stack-trace).
> 
> Is there (simple) way to do that in python (1.5.2)?

import traceback, cStringIO

f = cStringIO.StringIO
traceback.print_exc(None, f)

write_to_database(f.getvalue())

HTH
  Siggy




From bsb at winnegan.de  Thu Feb  8 16:23:16 2001
From: bsb at winnegan.de (Siggy Brentrup)
Date: 08 Feb 2001 16:23:16 +0100
Subject: catching all exceptions
In-Reply-To: <87y9vhp5r0.fsf@winnegan.de>
References: <95uati$cm4$1@icrf.news> <87y9vhp5r0.fsf@winnegan.de>
Message-ID: <87snlpp5az.fsf@winnegan.de>

Sorry for the typo :-(

Siggy Brentrup <bsb at winnegan.de> writes:

> import traceback, cStringIO
> 
> f = cStringIO.StringIO
  f = cStringIO.StringIO()
> traceback.print_exc(None, f)
> 
> write_to_database(f.getvalue())
> 
> HTH
>   Siggy





From a.mueller at icrf.icnet.uk  Thu Feb  8 17:34:00 2001
From: a.mueller at icrf.icnet.uk (Arne Mueller)
Date: Thu, 08 Feb 2001 16:34:00 +0000
Subject: catching all exceptions
References: <95uati$cm4$1@icrf.news> <87y9vhp5r0.fsf@winnegan.de> <mailman.981645915.21597.python-list@python.org>
Message-ID: <3A82CA78.7D9B12F0@icrf.icnet.uk>

Siggy Brentrup wrote:
> 
> Sorry for the typo :-(
> 
> Siggy Brentrup <bsb at winnegan.de> writes:
> 
> > import traceback, cStringIO
> >
> > f = cStringIO.StringIO
>   f = cStringIO.StringIO()
> > traceback.print_exc(None, f)
> >
> > write_to_database(f.getvalue())
> >
> > HTH
> >   Siggy

Thanks alot for the code above, this is what I was looking for. Also
sorry, I should have known about the traceback module - obviously didn't
read the python reference manual carefully anough ;-(

here's some code to do a test:

import sys, traceback, StringIO
x = 5

def boo(n):
    x.append(n) # raises AttributeError

try:
    boo(6)
except:
    f = StringIO.StringIO()
    print "Start"
    traceback.print_exc(None, f)
    #s = f.getvalue() # this works
    s = f.readlines() # empty array
    print s
    f.close()
    print "End"

f.readline() returns an empty array, but shouldn't have 
traceback.print_exc(None, f) filled f, so that 'readlines' can access
it?

	thanks very much,

	Arne



From stephen_purcell at yahoo.com  Thu Feb  8 16:24:36 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Thu, 8 Feb 2001 16:24:36 +0100
Subject: catching all exceptions
In-Reply-To: <95uati$cm4$1@icrf.news>; from nospam@our.site on Thu, Feb 08, 2001 at 02:30:24PM +0000
References: <95uati$cm4$1@icrf.news>
Message-ID: <20010208162436.A8411@freedom.puma-ag.com>

Arne wrote:
> 
> with this I think I can catch all predefined exception classes, because
> Exceptions is the base class for all other exceptions, but I realy want
> to know which derived exception class was raised (and also the
> stack-trace).
> 
> Is there (simple) way to do that in python (1.5.2)?

Yes.

try:
   something
except:
   exc_type, exc_value, exc_traceback = sys.exc_info()
   # Then write to database.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From syver at NOSPAMcyberwatcher.com  Thu Feb  8 15:39:26 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Thu, 8 Feb 2001 15:39:26 +0100
Subject: Python Unicode strangeness.
Message-ID: <Dcyg6.276$xT3.3885@news1.oke.nextra.no>

'%s, %s' % ('??asfd', u'a')
this throws unicode error.

'%s, %s' % (u'a?asfg', u'a')
this doesn't

'%s, %s' % ('a?asfg', 'a')
and this doesn't

'%s, %s' % (u'a?asfg', 'a')
and this doesn't

Is this by design?




From akuchlin at mems-exchange.org  Thu Feb  8 16:42:06 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 08 Feb 2001 10:42:06 -0500
Subject: Python Unicode strangeness.
References: <Dcyg6.276$xT3.3885@news1.oke.nextra.no>
Message-ID: <3dvgql5ghd.fsf@ute.cnri.reston.va.us>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:
> '%s, %s' % ('??asfd', u'a')
> this throws unicode error.
> Is this by design?

I believe so, though either MAL or /F are more likely to give a
definitive reply.  Because one of the arguments to % is a Unicode
string, it will coerce everything to Unicode.  The encoding used when
coercing a regular string into a Unicode string is 7-bit ASCII, so
only characters <128 are handled.  The first string in the tuple has
characters >128, so a UnicodeError is raised.  If you make it a
Unicode string, there's no problem since it doesn't need to be
coerced; if you make 'a' a regular string, no coercion is attempted.
In either case, the exception isn't raised. 

--amk




From loewis at informatik.hu-berlin.de  Fri Feb  9 19:07:57 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 09 Feb 2001 19:07:57 +0100
Subject: Python Unicode strangeness.
References: <Dcyg6.276$xT3.3885@news1.oke.nextra.no>
Message-ID: <p6qg0hn3f2a.fsf@informatik.hu-berlin.de>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:

> '%s, %s' % ('??asfd', u'a')
> this throws unicode error.
> 
> Is this by design?

I think so, you should not combine strings with characters > 128 and
Unicode objects: the result should be Unicode, but you can't be
certain in what encoding the string was. To get this to work, do

'%s, %s' % (unicode('??asfd','latin-1'), u'a')

(if Latin-1 is the encoding that you had intended for that string), or

'%s, %s' % ('??asfd', u'a'.encode('latin-1'))

Regards,
Martin


From cook at maya.com  Thu Feb  8 16:00:49 2001
From: cook at maya.com (Nicole Cook)
Date: Thu, 8 Feb 2001 10:00:49 -0500
Subject: mxODBC problems
Message-ID: <000901c091df$f673e9a0$21fe46c0@maya.com>

I've installed mxODBC on Windows, and I'm still trying to get it to the
point where I can use it.  I know it's all there, and it's not colliding
with win32all.  Is there something wrong with my python path?   I don't know
what to try next.

Nicole


Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [
Copyright 1991-1995 Stichting Mathematisch
>>> import ODBC
>>> import ODBC.Windows
Traceback (innermost last):
  File "<stdin>", line 1, in ?
ImportError: No module named Windows
>>> from ODBC import Windows
Traceback (innermost last):
  File "<stdin>", line 1, in ?
ImportError: cannot import name Windows
>>> from ODBC import *
>>> db = ODBC.Windows.connect
Traceback (innermost last):
  File "<stdin>", line 1, in ?
AttributeError: Windows
>>> from ODBC.Windows import *
Traceback (innermost last):
  File "<stdin>", line 1, in ?
ImportError: No module named Windows
>>>





From db3l at fitlinxx.com  Thu Feb  8 23:53:44 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 08 Feb 2001 17:53:44 -0500
Subject: mxODBC problems
References: <mailman.981644643.15712.python-list@python.org>
Message-ID: <u7l30yefb.fsf@ctwd0143.fitlinxx.com>

"Nicole Cook" <cook at maya.com> writes:

> I've installed mxODBC on Windows, and I'm still trying to get it to the
> point where I can use it.  I know it's all there, and it's not colliding
> with win32all.  Is there something wrong with my python path?   I don't know
> what to try next.

Are you sure you're getting mxODBC and not some other ODBC module?  It
sort of looks like you may still not be getting the module you think
you are?

> Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [
> Copyright 1991-1995 Stichting Mathematisch
> >>> import ODBC

What happens if you "print ODBC.__file__" - you should see it
reference the __init__.py file from your mxODBC installation (normally
beneath the lib\ODBC directory of your Python installation).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From benpark at my-deja.com  Thu Feb  8 16:19:33 2001
From: benpark at my-deja.com (Ben Park)
Date: Thu, 08 Feb 2001 15:19:33 GMT
Subject: i/o for NumPy
Message-ID: <95uddp$q35$1@nnrp1.deja.com>


What are the unformatted i/o functions in NumPy? I see fromstring, but
you have to read the binary data into a string first and then transfer
to an array.

What I am looking for is something like fread/fwrite in Matlab, or
readu/writeu in IDL.


Sent via Deja.com
http://www.deja.com/


From akuchlin at mems-exchange.org  Thu Feb  8 16:21:55 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 08 Feb 2001 10:21:55 -0500
Subject: python-dev summary, Jan. 16-31
Message-ID: <3d4ry56vzg.fsf@ute.cnri.reston.va.us>

Python-dev summary, January 16-31, 2001
=======================================

To comment on material in this python-dev summary, you can simply post
to comp.lang.python / <python-list at python.org>.  These summaries are
archived at <URL:http://www.amk.ca/python/dev/>.

This two week period saw the release of the first alpha of Python 2.1,
and the second alpha was released a few days into February.  There's a
lot of bugfixing going on, much of which isn't of interest here, but
there were a few general discussions worth noting.


2.1alpha1
=========

The first alpha of Python 2.1 was released on January 22, and is the
debut of Python 2.1.  It didn't compile cleanly on Cygwin, and the new
autodetecting build system needs some more work to support more
platforms correctly, but no major problems or brown paper bag bugs
showed up.  

2.1alpha2 was subsequently released on February 2, slightly out of the
bounds of this summary.  2.1a2 incorporated the nested scopes from PEP
227 and the weak reference type from PEP 205.

A draft of "What's New in Python 2.1" is available, and will be continuously
updated until 2.1final is shipped:
	http://www.amk.ca/python/2.1/


unistr()
========

M.-A. Lemburg checked in a unistr() built-in, but later backed out the
change after subsequent discussion.  str() takes any Python object and
converts it to a string, and unistr() would take any object and return
a Unicode string.  ?!ng wondered why the existing unicode() built-in
shouldn't do this -- currently it only accepts a string and an optional
encoding -- or should this functionality be added to str()?

This also raised the question of how a type could return a Unicode
representation of itself.  Should the tp_str slot in the Py_TypeObject
struct be permitted to return either an 8-bit or a Unicode string, or
should a tp_unicode slot be added?  Some discussion followed, but no
clear answer was apparent, so unistr() was removed.  This issue will
probably have to wait for a PEP to be written, which means that it'll
have to wait until 2.2.


Sets
====

Eric S. Raymond posted the docs for a set module he's written with an
eye toward getting it into the standard library.  "The \module{set}
module defines functions for treating lists and other sequences as
mathematical sets, and defines a set class that uses these operations
natively and overloads Python's standard operator set."
	http://mail.python.org/pipermail/python-dev/2001-January/012005.html

But there's already a PEP for a set type, #218 by Greg Wilson:
	http://python.sourceforge.net/peps/pep-0218.html

Reconciling the two proposals doesn't seem difficult, but opinion was
divided: are sets so commonly used that they should be in the standard
library, particularly if dictionaries can be used to implement many
common set operations?  

?!ng pointed out that simply allowing "if key in dict" would make
using dictionaries as sets a bit more natural.  /F agreed, and GvR
added "You know, I've long resisted this, but I agree now -- this is
the right thing."  Thomas Wouters submitted a patch, but it hasn't yet
been checked in.  

A subthread wondered about the related question of 'for X in
dictionary:' Should this do anything, and if so, what?  X could be
"key" or "key, value", but the best choice isn't obvious.  Another
subthread drifted into the question of the Batteries Included
distribution proposed by Moshe Zadka in PEP 206, and yet another one
proposed dictionaries that can be marked as unmodified ("frozen" or
"locked").  No clear conclusions were made, and the problems seem
messy enough that a PEP will be required, so this is another thing
likely to be postponed until 2.2.


Envoi
=====

This will be the last python-dev summary I write, accounting for its
rushed and sketchy quality.

The goal of these summaries has been to make the discussions on
python-dev more visible to the community, letting people offer timely
comments while a thread is still reasonably current and fresh in
memory.  Unfortunately that doesn't seem to happen, and no messages to
python-dev are spurred by them.  Ideally people would come forward to
say "The last summary mentioned you were talking about X. I use X a
lot, and here's what I think: ...", but that isn't happening.

The release of 2.1 offers a second calibration on their effectiveness.
2.1 is the first Python release to have been carried out using PEPs as
the mechanism, so there are no sizable changes in 2.1 that don't have
a corresponding PEP.  Yet many people were *surprised* by some of the
changes in Python 2.1 such as function attributes and nested scopes,
even though PEPs were written and discussed, often in lengthy threads
months ago.  

To me, this makes it crystal clear that the summaries aren't achieving
their goal of making the development process more transparent to the
community.  Perhaps giving the PEPs higher visibility -- posting
announcements of status changes and new PEPs to
comp.lang.python.announce, for example -- would do a better job, and
it would definitely require less effort.  In any case, I feel there's
no point in continuing to write them.

(Should anyone want to volunteer to continue writing them, please do
so; contact me if you want copies of the previous summaries for a
complete archive.)

My thanks to the various people who've offered comments and
encouragement along the way; you've all kept me going over the last 7
months...


Related Links
=============

Python-dev archives:
        http://www.python.org/pipermail/python-dev/

Python project page on SourceForge:
        http://sourceforge.net/projects/python

Python Enhancement Proposals (PEPs):
        http://python.sourceforge.net/peps/

What's New in Python 2.1
        http://www.amk.ca/python/2.1/



From hinsen at cnrs-orleans.fr  Thu Feb  8 17:38:08 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 08 Feb 2001 17:38:08 +0100
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us>
Message-ID: <m3ofwdp1u7.fsf@chinon.cnrs-orleans.fr>

Andrew Kuchling <akuchlin at mems-exchange.org> writes:

> To me, this makes it crystal clear that the summaries aren't achieving
> their goal of making the development process more transparent to the
> community.  Perhaps giving the PEPs higher visibility -- posting

The fact that there are people who read neither the PEPs nor the
summaries doesn't prove that the summaries aren't useful. I have
always read them with much interest, since I don't have the time at
the moment to follow the development in detail. And I would have
shouted out loud if I had noticed a tendency that looks dangerous to
me - it just hasn't happened.

So personally I hope that someone will take up the job of providing
development summaries in the future. And I'd like to thank Andrew
for having done so for many months already.

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From jjlucsy at concentric.net  Thu Feb  8 18:55:42 2001
From: jjlucsy at concentric.net (Joel Lucsy)
Date: Thu, 8 Feb 2001 12:55:42 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <m3ofwdp1u7.fsf@chinon.cnrs-orleans.fr>
Message-ID: <H4Bg6.43114$o5.544557@e420r-atl1.usenetserver.com>

"Konrad Hinsen" <hinsen at cnrs-orleans.fr> wrote in message
news:m3ofwdp1u7.fsf at chinon.cnrs-orleans.fr...
> Andrew Kuchling <akuchlin at mems-exchange.org> writes:
>
> > To me, this makes it crystal clear that the summaries aren't achieving
> > their goal of making the development process more transparent to the
> > community.  Perhaps giving the PEPs higher visibility -- posting
>
> The fact that there are people who read neither the PEPs nor the
> summaries doesn't prove that the summaries aren't useful. I have
> always read them with much interest, since I don't have the time at
> the moment to follow the development in detail. And I would have
> shouted out loud if I had noticed a tendency that looks dangerous to
> me - it just hasn't happened.

Same here.

--
Joel Lucsy (jjlucsy at concentric.net)




From michael at stroeder.com  Thu Feb  8 19:08:33 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 08 Feb 2001 19:08:33 +0100
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <m3ofwdp1u7.fsf@chinon.cnrs-orleans.fr> <H4Bg6.43114$o5.544557@e420r-atl1.usenetserver.com>
Message-ID: <3A82E0A1.9F5BD7EC@stroeder.com>

Joel Lucsy wrote:
> 
> "Konrad Hinsen" <hinsen at cnrs-orleans.fr> wrote in message
> news:m3ofwdp1u7.fsf at chinon.cnrs-orleans.fr...
> > Andrew Kuchling <akuchlin at mems-exchange.org> writes:
> >
> > > To me, this makes it crystal clear that the summaries aren't achieving
> > > their goal of making the development process more transparent to the
> > > community.  Perhaps giving the PEPs higher visibility -- posting
> >
> > The fact that there are people who read neither the PEPs nor the
> > summaries doesn't prove that the summaries aren't useful. I have
> > always read them with much interest,
> 
> Same here.

+1

Ciao, Michael.


From bkc at Murkworks.com  Thu Feb  8 22:02:43 2001
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 8 Feb 2001 16:02:43 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <m3ofwdp1u7.fsf@chinon.cnrs-orleans.fr> <H4Bg6.43114$o5.544557@e420r-atl1.usenetserver.com> <3A82E0A1.9F5BD7EC@stroeder.com>
Message-ID: <nKDg6.3$W4.931@news.clarkson.edu>

And me too. I always read them with interest, specifically because I don't
have the time to follow all the threads personally. I think python-dev
summary is very useful. I'm sorry I can't personally contribute.

--
Brad Clements,
bkc at murkworks.com
"Michael Str?der" <michael at stroeder.com> wrote in message
news:3A82E0A1.9F5BD7EC at stroeder.com...
> Joel Lucsy wrote:
> >
> > "Konrad Hinsen" <hinsen at cnrs-orleans.fr> wrote in message
> > news:m3ofwdp1u7.fsf at chinon.cnrs-orleans.fr...
> > > Andrew Kuchling <akuchlin at mems-exchange.org> writes:
> > >
> > > > To me, this makes it crystal clear that the summaries aren't
achieving
> > > > their goal of making the development process more transparent to the
> > > > community.  Perhaps giving the PEPs higher visibility -- posting
> > >
> > > The fact that there are people who read neither the PEPs nor the
> > > summaries doesn't prove that the summaries aren't useful. I have
> > > always read them with much interest,
> >
> > Same here.
>
> +1
>
> Ciao, Michael.




From andymac at bullseye.apana.org.au  Fri Feb  9 03:11:26 2001
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Fri, 9 Feb 2001 13:11:26 +1100 (EDT)
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <m3ofwdp1u7.fsf@chinon.cnrs-orleans.fr>
Message-ID: <Pine.OS2.4.30.0102091310510.765-100000@CENTRAL>

On 8 Feb 2001, Konrad Hinsen wrote:

> So personally I hope that someone will take up the job of providing
> development summaries in the future. And I'd like to thank Andrew
> for having done so for many months already.

Ditto on both counts.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From dan at sol.control.com  Fri Feb  9 19:57:42 2001
From: dan at sol.control.com (Dan L. Pierson)
Date: 09 Feb 2001 13:57:42 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <m3ofwdp1u7.fsf@chinon.cnrs-orleans.fr>
Message-ID: <m34ry3lm55.fsf@daystar.control.com>

Konrad Hinsen <hinsen at cnrs-orleans.fr> writes:

> The fact that there are people who read neither the PEPs nor the
> summaries doesn't prove that the summaries aren't useful. I have
> always read them with much interest, since I don't have the time at
> the moment to follow the development in detail. And I would have
> shouted out loud if I had noticed a tendency that looks dangerous to
> me - it just hasn't happened.
> 
> So personally I hope that someone will take up the job of providing
> development summaries in the future. And I'd like to thank Andrew
> for having done so for many months already.

I agree.  Thank you, Andrew,  for all of the work you've put into
creating on the consistent high spots of comp.lang.python for the
last serveral months.

One of the reasons that I haven't followed up on the python-dev stuff
as much as I should is that I'm almost exclusively working with Zope
these days and it's still stuck in 1.5.2 land.  Looking through your
window at 2.0 and 2.1 one makes me drool a lot, but I haven't got any
experience with it, just opinions.

Thanks again,

Dan Pierson


From dalke at acm.org  Thu Feb  8 22:21:52 2001
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 8 Feb 2001 14:21:52 -0700
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us>
Message-ID: <95v2lq$hjl$1@slb7.atl.mindspring.net>

amk:
>The release of 2.1 offers a second calibration on their effectiveness.
>2.1 is the first Python release to have been carried out using PEPs as
>the mechanism, so there are no sizable changes in 2.1 that don't have
>a corresponding PEP.  Yet many people were *surprised* by some of the
>changes in Python 2.1 such as function attributes and nested scopes,
>even though PEPs were written and discussed, often in lengthy threads
>months ago.

I for one am not on the python-dev mailing list.  It's too
much mail traffic for me to follow.  There were about 1,000
messages in January and while the signal level is high,
the content is usually beyond my ken.

(Following the newsgroup is easier because I can defer reading
the newsgroup then kill threads I don't want to follow.  I
suppose I can do that with email as well, by redirecting
incoming messages to a special mailbox, but that's a hassle
and I would have to see if it's possible and learn how to
do it.  Barrier to entry.)

Speaking of the archive, it's always been hard for me to
remember where the archive is located.  There's nothing on
the python.org page pointing to it.  I know it's at sourceforge,
but python.sourceforge.com doesn't work nor does .org.  I
finally figured out it was python.sourceforge.net but I
don't see a link to the back archive.  (Sorry, but my ability
to remember names isn't perfect - that's one reason I like
Python :)

There are also no links from the python.org page to the PEPs
although python.sourceforge.net points directory to it.
More importantly, the PEPs are rarely announced to c.l.py,
which means I don't know what's going on.  When a PEP *is*
announced on c.l.py there are discussions about it, which
implies that others are like me in being interested in
giving input but don't have the time to follow all of the
technical details of python-dev.

I used DejaNews to search for threads named "PEP".  The ones
which came up were:
  PEP 4
  PEP 201
  PEP 203
  PEP 214
  PEP 222 (because of you)
  PEP 223
  PEP 224

Given that there are some 40 PEPs, this implies that only 20%
of them were announced to c.l.py, which greatly reduces the
dissemination of their ideas to the rest of the Python world.

Missing Finished PEPs are: 207, 208, 217, 221, 240 and 232.

I'm looking at PEP 0001, "PEP Purpose and Guidlines". It says:

] PEP authors are responsible for collecting community feedback on a
] PEP before submitting it for review.  A PEP that has not been
] discussed on python-list and python-dev will not be accepted for
] review.  However, wherever possible, long open-ended discussions
] on public mailing lists should be avoided.  A better strategy is
] to encourage public feedback directly to the PEP author, who
] collects and integrates the comments back into the PEP.

This is obviously not being done.  I searched Deja for "PEP 217",
which I first knew about from your python-dev summary.  There
was only a single reference from Tim Peters about it.  Take
PEP 232, which is the one on function attributes.  You said

> Yet many people were *surprised* by some of the
> changes in Python 2.1 such as function attributes and nested scopes,
> even though PEPs were written and discussed, often in lengthy threads
> months ago.

The first mention on c.l.py on function attributes was from Jan. 24
 http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=719772745&fmt=text
even though, yes, the PEP was created on Dec. 2.

Jeremy Hylton first mentioned PEP 227 ("Statically Nested
Scopes") on Nov. 19 in the thread "Ruby and Python"
 http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=695547859&fmt=text
then was mentioned again in passing by Tim on Dec. 20.  The PEP
was created on Nov. 1.

No wonder there have been complaints -  how is anyone not
interested in following python-dev supposed to know what is
going on with Python development?

>To me, this makes it crystal clear that the summaries aren't achieving
>their goal of making the development process more transparent to the
>community.

The problem is the summaries are just that - summaries.  I
*really* enjoy reading them because there are a concise description
of what occured and the only way I have these days to know
what is going on with development.

However, consider this latest summary.  I received it on
the 7th, which is 1 to 3 weeks after the discussions it talked
about.  To contribute to a thread, I would need to catch up
on the mailing list to see what arguments have been presented,
then bring up my questions or objections.  But it would be
about two weeks late, which means people may have already
decided on what to do, and making changes is more likely to
be frowned upon.

(Not saying that that will happen, but that's my perception,
which makes it harder for me to decide to put in the effort
to make a useful comment.)


> Perhaps giving the PEPs higher visibility -- posting
>announcements of status changes and new PEPs to
>comp.lang.python.announce, for example -- would do a better job

Just mentioning them on c.l.py as per PEP 0001 would help!

I *really* miss the pre-python-dev newsgroup days.  Spinning
the core technical discussions off has not only reduced my
awareness of Python's direction, but it has almost eliminated
my education in CS topics like continuations.  (Which you
clarified for me at the Houston conference :)  And there
are far fewer posts from Tim Peters than the schriebfest
days.

So I really place the problem on the main python-dev people
who don't bother to disseminate their ideas to the broader
audience of people who develop a lot of Python code but
aren't language designers.

>I feel there's no point in continuing to write them.

I agree with the others who have replied - I will miss the
summaries greatly.  They were nearly the only way I had to
know what was going on with Python development.

>(Should anyone want to volunteer to continue writing them, please do
>so; contact me if you want copies of the previous summaries for a
>complete archive.)

I can't as I don't have the time to follow python-dev, much less
comment anywhere near as intellegently about it as you have.

                    Andrew
                    dalke at acm.org





From dalke at acm.org  Thu Feb  8 22:38:55 2001
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 8 Feb 2001 14:38:55 -0700
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net>
Message-ID: <95v3mt$lnh$1@nntp9.atl.mindspring.net>

Oh yeah, I forgot:
  http://mail.python.org/mailman/listinfo/python-dev

says in bold that

  Subscription is by invitation only.

and adds

  This is a closed list

So just how am I supposed to follow what is going on
in Python development and how do I make comments in a
timely manner?

Feeling snubbed,
                    Andrew
                    dalke at acm.org





From aahz at panix.com  Thu Feb  8 23:17:07 2001
From: aahz at panix.com (Aahz Maruch)
Date: 8 Feb 2001 14:17:07 -0800
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net> <95v3mt$lnh$1@nntp9.atl.mindspring.net>
Message-ID: <95v5t3$d3q$1@panix3.panix.com>

[cc'd to Andrew Kuchling]

In article <95v3mt$lnh$1 at nntp9.atl.mindspring.net>,
Andrew Dalke <dalke at acm.org> wrote:
>
> [...]

I'm adding a strong "Me, too!" to everything Andrew D. said.

The idea doesn't thrill me, but one possible way to handle this would be
to gateway python-dev to a moderated newsgroup.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From n8gray at caltech.edu.is.my.email.address  Sat Feb 10 06:17:49 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Fri, 9 Feb 2001 21:17:49 -0800
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net> <95v3mt$lnh$1@nntp9.atl.mindspring.net> <95v5t3$d3q$1@panix3.panix.com>
Message-ID: <962j8r$i9i@gap.cco.caltech.edu>

Aahz Maruch wrote:

> [cc'd to Andrew Kuchling]
> 
> In article <95v3mt$lnh$1 at nntp9.atl.mindspring.net>,
> Andrew Dalke <dalke at acm.org> wrote:
> >
> > [...]
> 
> I'm adding a strong "Me, too!" to everything Andrew D. said.

Me too, too.

> The idea doesn't thrill me, but one possible way to handle this would be
> to gateway python-dev to a moderated newsgroup.

It sounds like a good idea to me.  

I'm a bit puzzled by AMK's tone of disappointment with the fact that his 
summaries haven't generated messages to python-dev.  It's never been clear 
to me that comments from the peanut gallery were appreciated on python-dev. 
 When subscription to a mailing list is advertised as "by invitiation only" 
and the location of its archives is intentionally obscure, it sends a 
strong message that outsiders are not welcome.

A moderated newsgroup seems like a much better vehicle if you intend for 
the public to participate in the discussion of Python development without 
sending the SNR plummetting.  I imagine most people would hesitate to post 
to a moderated newsgroup without having given their message a modicum of 
thought.  (Unlike *most* newsgroups ;^)

Nonetheless, a hearty "thank you" to Andrew for taking the time to keep us 
up to date for so long.

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From peter at engcorp.com  Sat Feb 10 18:55:50 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 10 Feb 2001 12:55:50 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net> <95v3mt$lnh$1@nntp9.atl.mindspring.net> <95v5t3$d3q$1@panix3.panix.com> <962j8r$i9i@gap.cco.caltech.edu>
Message-ID: <3A8580A6.7210BA53@engcorp.com>

Nathaniel Gray wrote:
> 
> I'm a bit puzzled by AMK's tone of disappointment with the fact that his
> summaries haven't generated messages to python-dev.  It's never been clear
> to me that comments from the peanut gallery were appreciated on python-dev.
>  When subscription to a mailing list is advertised as "by invitiation only"
> and the location of its archives is intentionally obscure, it sends a
> strong message that outsiders are not welcome.

Hear hear!  

When I first heard about python-dev and tried to find it,
but saw the "by invitation only" note, I thought to myself 
"Dang... oh well, at least that guy who posts summaries 
of ongoing discussions from python-dev is willing to post 
those summaries of ongoing discussions from python-dev."

AMK, I appreciate the effort required for such "volunteer" 
work, and the way limited feedback makes it seem like a 
pointless endeavour.  I think there might actually be a _huge_,
but silent (and somewhat silenced), audience for those summaries
and I hate to see them go.  

Rest assured that had I really seen something objectionable, 
and had it not already been available as a thread in c.l.p, 
I would have inquired in c.l.p myself.  The fact that that 
hasn't happened (much?) with me or some others should 
perhaps be taken more as a sign of how well the caretakers
of Python are doing their (volunteer) job, and not as 
a sign of disinterest on our part.

In any case, I too thank you for your past efforts.


From aahz at panix.com  Mon Feb 12 23:21:03 2001
From: aahz at panix.com (Aahz Maruch)
Date: 12 Feb 2001 14:21:03 -0800
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v3mt$lnh$1@nntp9.atl.mindspring.net> <95v5t3$d3q$1@panix3.panix.com> <962j8r$i9i@gap.cco.caltech.edu>
Message-ID: <969nkf$5nd$1@panix3.panix.com>

In article <962j8r$i9i at gap.cco.caltech.edu>,
Nathaniel Gray  <n8gray at caltech.edu.is.my.email.address> wrote:
>
>A moderated newsgroup seems like a much better vehicle if you intend for 
>the public to participate in the discussion of Python development without 
>sending the SNR plummetting.  I imagine most people would hesitate to post 
>to a moderated newsgroup without having given their message a modicum of 
>thought.  (Unlike *most* newsgroups ;^)

Actually, I was thinking that this would be a one-directional gateway
from mailing list to newsgroup, no posting allowed.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

'Gender' isn't a definition, it's a heuristic.  --Aahz


From n8gray at caltech.edu.is.my.email.address  Fri Feb 16 07:31:09 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Thu, 15 Feb 2001 22:31:09 -0800
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v3mt$lnh$1@nntp9.atl.mindspring.net> <95v5t3$d3q$1@panix3.panix.com> <962j8r$i9i@gap.cco.caltech.edu> <969nkf$5nd$1@panix3.panix.com>
Message-ID: <96ihqb$dou@gap.cco.caltech.edu>

Aahz Maruch wrote:
> 
> Actually, I was thinking that this would be a one-directional gateway
> from mailing list to newsgroup, no posting allowed.

True, this would help keep the common folk up to date with Python-dev, but 
it sounded to me like the dev crew was disappointed that there hadn't been 
any feedback generated by the summaries.  If feedback is actually what they 
want (maybe it isn't, and maybe it shouldn't be), I don't see this idea 
solving the problem.

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From aahz at panix.com  Fri Feb 16 17:27:01 2001
From: aahz at panix.com (Aahz Maruch)
Date: 16 Feb 2001 08:27:01 -0800
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <962j8r$i9i@gap.cco.caltech.edu> <969nkf$5nd$1@panix3.panix.com> <96ihqb$dou@gap.cco.caltech.edu>
Message-ID: <96jkcl$f4k$1@panix2.panix.com>

In article <96ihqb$dou at gap.cco.caltech.edu>,
Nathaniel Gray  <n8gray at caltech.edu.is.my.email.address> wrote:
>Aahz Maruch wrote:
>> 
>> Actually, I was thinking that this would be a one-directional gateway
>> from mailing list to newsgroup, no posting allowed.
>
>True, this would help keep the common folk up to date with Python-dev, but 
>it sounded to me like the dev crew was disappointed that there hadn't been 
>any feedback generated by the summaries.  If feedback is actually what they 
>want (maybe it isn't, and maybe it shouldn't be), I don't see this idea 
>solving the problem.

As I said in another post, if Followup-To goes to c.l.py, everything is
hunky-dory.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From db3l at fitlinxx.com  Fri Feb  9 00:03:38 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 08 Feb 2001 18:03:38 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net> <95v3mt$lnh$1@nntp9.atl.mindspring.net>
Message-ID: <u4ry4ydyt.fsf@ctwd0143.fitlinxx.com>

"Andrew Dalke" <dalke at acm.org> writes:

> So just how am I supposed to follow what is going on
> in Python development and how do I make comments in a
> timely manner?

Well, it's not quite as closed as it may sound - the membership on the
list itself is closed to those participating in the core development,
but anyone can send e-mail to the list, and all mail is archived and
can be followed through the archive.

(Although as suggested elsewhere, a gateway to a newsgroup (even if
one-way) would be cool as I find news the most efficient vehicle to
follow active lists.)

I'm not totally clear on the rationale for the closed status (and I'm
not currently on the list), but I'm guessing there was an expectation
that an open list would get lots of lurking subscriptions which might
really ratchet up the load tremendously for spurious discussions.

This way, those directly on the list are those writing the code.  In
effect, it really is a core developers list, but it's been opened up
enough so that anyone can peruse it - or contribute individual mail.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From paul.moore at uk.origin-it.com  Mon Feb 12 17:49:04 2001
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Mon, 12 Feb 2001 17:49:04 +0100
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net> <95v3mt$lnh$1@nntp9.atl.mindspring.net>
Message-ID: <PxOIOsbCfkh0LHOAPQWUK6DwdDmL@4ax.com>

On Thu, 8 Feb 2001 22:38:55 +0100 , "Andrew Dalke" <dalke at acm.org>
wrote:

>Oh yeah, I forgot:
>  http://mail.python.org/mailman/listinfo/python-dev
>
>says in bold that
>
>  Subscription is by invitation only.
>
>and adds
>
>  This is a closed list
>
>So just how am I supposed to follow what is going on
>in Python development and how do I make comments in a
>timely manner?
>
>Feeling snubbed,

I agree entirely. I follow python-dev avidly, but I hate the web
interface for doing so. The tone of the comments on the subscription
page seem to make it vey clear that "casual observers" are not
welcome, and that is all I would be likely to be.

I may well be glad to make the odd comment, but I cannot commit myself
to any significant level of participation, so I am stuck with only the
web interface.

I'd argue for, at the least, making python-dev available freely on a
"read only" basis, and preferably allowing open access, with a strong
policy of keeping postings on-topic (I don't really have a feeling
that the Python community would abuse write access to python-dev, but
it is a potential issue, witness the perl5 porters list...)

But nevertheless, thanks Andrew for the work you have done.

Paul Moore.



From dalke at acm.org  Fri Feb 16 10:09:17 2001
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 16 Feb 2001 02:09:17 -0700
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net> <95v3mt$lnh$1@nntp9.atl.mindspring.net>
Message-ID: <96iqru$h36$1@slb2.atl.mindspring.net>

I wrote:
>  http://mail.python.org/mailman/listinfo/python-dev
>
>says in bold that
>
>  Subscription is by invitation only.
>
>and adds
>
>  This is a closed list

I was rooting around for the iterator thread.  I found

http://mail.python.org/pipermail/python-dev/2001-February/012828.html
which says

> > p.s. I have set up http://groups.yahoo.com/group/python-iter and
> > http://groups.yahoo.com/group/python-sets; Guido, would you prefer
> > discussion of sets and iterators to be moved to these lists, or to
> > stay on python-dev?
>
> Let's move these to egroups for now.
>
> --Guido van Rossum (home page: http://www.python.org/~guido/)

I assume "egroups" is "groups.yahoo.com".  I went to the
list to see the current discussion status.  I could not
because

    Options
    ? Listed in directory
    ? Open membership
    ? Unmoderated
    ? All members may post
>>  ? Archives for members only
    ? Email attachments are permitted

Needless to say, I am not a member so I can't even see the
archives of this thread.  And there is no way I'm going to
give Yahoo and their mutable terms of services my email
address and other personal data.  I'ld rather just complain
annoyingly to c.l.py :)  Besides, it looks like there's only
a single message in the archive.

                    Andrew Dalke
                    dalke at acm.org





From tim.one at home.com  Fri Feb 16 10:41:53 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 04:41:53 -0500
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <96iqru$h36$1@slb2.atl.mindspring.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEMHIPAA.tim.one@home.com>

[Andrew Dalke]
> ...
> Needless to say, I am not a member so I can't even see the
> archives of this thread.  And there is no way I'm going to
> give Yahoo and their mutable terms of services my email
> address and other personal data.  I'ld rather just complain
> annoyingly to c.l.py :)

I liked life better when language changes were slugged out on c.l.py.  There
are soooooo many archives and semi-private mailing lists and utterly pvt
emails now-- and all of them, like this msg, with useless Subject lines --I
just don't know where to look anymore either.  And I spend hours on this
stuff every day of my miserable life.

But, there's so much traffic on Python-Dev alone that I doubt Guido can
afford to spend significant time on c.l.py anymore, and he's been pushing to
move discussions to venues like egroups for at least a year.

If I ever get time to write a PEP, I'll go back to slugging it out on
c.l.py.  But people will complain about that too -- they did before, and I
doubt that human nature is changing as quickly as Python <0.9 wink>.

That said, giving Yahoo an email address isn't going to ruin your credit
rating, and you can lie about everything else.

no-payback-in-being-unique<wink>-ly y'rs  - tim




From akuchlin at mems-exchange.org  Thu Feb  8 23:17:41 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 08 Feb 2001 17:17:41 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net>
Message-ID: <3dpugs6cqi.fsf@ute.cnri.reston.va.us>

"Andrew Dalke" <dalke at acm.org> writes:
> The first mention on c.l.py on function attributes was from Jan. 24
>  http://www.deja.com/[ST_rn=ps]/getdoc.xp?AN=719772745&fmt=text
> even though, yes, the PEP was created on Dec. 2.

I meant python-dev discussion here; clearly the PEPs aren't visible
enough in c.l.p.

> about.  To contribute to a thread, I would need to catch up
> on the mailing list to see what arguments have been presented,
> then bring up my questions or objections.  But it would be
> about two weeks late, which means people may have already
> decided on what to do, and making changes is more likely to
> be frowned upon.

Two weeks late is a bit annoying, but it's much less annoying than
only finding out about objections when the first alpha ships, months
later!  Note that you *can* send messages to python-dev; only
subscription is limited.

> So I really place the problem on the main python-dev people
> who don't bother to disseminate their ideas to the broader
> audience of people who develop a lot of Python code but
> aren't language designers.

Indeed.  I'm increasingly worried that python-dev is becoming too
insular, working on things while the community is unaware of what's
coming.  The fixes for this seem to be:

* Announce PEP drafts and status changes to c.l.py.announce
* Write a set of introductory pages for Python development (this has been 
  brought up before).
* A read-only alias for python-dev would be useful for several people, but 
  may require serious Mailman hackery (I don't know).

Any other suggestions?

--amk


From mjackson at wc.eso.mc.xerox.com  Fri Feb  9 01:59:29 2001
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 9 Feb 2001 00:59:29 GMT
Subject: python-dev summary, Jan. 16-31
References: <3dpugs6cqi.fsf@ute.cnri.reston.va.us>
Message-ID: <95vfdh$3k4$1@news.wrc.xerox.com>

Andrew Kuchling <akuchlin at mems-exchange.org> writes:
> "Andrew Dalke" <dalke at acm.org> writes:

[a whole bunch of stuff to which I add a firm "me, too."]

> > about.  To contribute to a thread, I would need to catch up
> > on the mailing list to see what arguments have been presented,
> > then bring up my questions or objections.  But it would be
> > about two weeks late, which means people may have already
> > decided on what to do, and making changes is more likely to
> > be frowned upon.
> 
> Two weeks late is a bit annoying, but it's much less annoying than
> only finding out about objections when the first alpha ships, months
> later!  Note that you *can* send messages to python-dev; only
> subscription is limited.

Hm.  That wasn't clear to me either, and I *do* keep the python-dev
page bookmarked, and check it out from time to time.  Had I ever felt
well-informed enough to want to speak up I wouldn't have known I could
address python-dev directly.

> > So I really place the problem on the main python-dev people
> > who don't bother to disseminate their ideas to the broader
> > audience of people who develop a lot of Python code but
> > aren't language designers.
> 
> Indeed.  I'm increasingly worried that python-dev is becoming too
> insular, working on things while the community is unaware of what's
> coming.

Risk:  the language moves in the direction of being easier for wizards
to write, at the (probable) expense of becoming less easy for
nonwizards to read.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Having your book made into a movie is like having
	your ox made into a bouillon cube.
				- Bill Neely




From gustav at morpheus.demon.co.uk  Mon Feb 12 22:56:52 2001
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Mon, 12 Feb 2001 21:56:52 +0000
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v2lq$hjl$1@slb7.atl.mindspring.net> <3dpugs6cqi.fsf@ute.cnri.reston.va.us>
Message-ID: <jrmg8t0ae2qm755aucm2c1tep73if7fc3o@4ax.com>

On 08 Feb 2001 17:17:41 -0500, Andrew Kuchling
<akuchlin at mems-exchange.org> wrote:

>Note that you *can* send messages to python-dev; only
>subscription is limited.

Um. So let's just get this straight. Anyone can post to python-dev,
anyone can read it via the web interface, so the only purpose of the
restricted access is to stop people being able to read the content via
a mailreader....???

This is seriously dumb. The subscription process does nothing
practical apart from making it *look like* the Python development team
are insular and unwilling to accept input from the community?

Please, someone. Straighten this out, or explain the reasons behind
the current setup!

Paul.


From tim.one at home.com  Tue Feb 13 00:21:19 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 12 Feb 2001 18:21:19 -0500
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <jrmg8t0ae2qm755aucm2c1tep73if7fc3o@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEOPIOAA.tim.one@home.com>

[Paul Moore]
> Um. So let's just get this straight. Anyone can post to python-dev,
> anyone can read it via the web interface, so the only purpose of the
> restricted access is to stop people being able to read the content via
> a mailreader....???

The purpose is to discourage people from posting.  There's enough traffic on
Python-Dev from the blessed invitees without the flamewar-attracting ease of
hitting "reply" in a mail reader.  If someone wants to post something to
Python-Dev, it takes them some effort to do so, and-- while I don't
particularly care for that myself either --it has been effective in reducing
noise.

You might think that a moderated list would do just as well, but it
wouldn't:  at least half the Python-Dev traffic is about fixing bugs or
build problems in the latest CVS snapshot, and that goes on around the clock
(it's an international crew).  The Python-Dev people are *developing* Python
(writing code, writing docs, writing PEPs, running tests, porting to new
platforms, ...), and need to communicate as quickly as possible about
current development issues.  A moderator would get in the way of that; and
nobody has the spare bandwidth to play moderator anyway.

BTW, back when this kind of thing was discussed on c.l.py, people complained
about *that*.  Business as usual ...

If you want to argue for a language change, you have to get Guido's ear, and
the only sure way to do that (whether or not you're on Python-Dev!) is to
write a PEP.  Some of the Python-Dev discussions are interesting, but I bet
far fewer than you imagine.

then-again-there's-a-sure-way-for-you-to-find-out-ly y'rs  - tim




From paul.moore at uk.origin-it.com  Tue Feb 13 11:17:00 2001
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Tue, 13 Feb 2001 11:17:00 +0100
Subject: python-dev summary, Jan. 16-31
References: <jrmg8t0ae2qm755aucm2c1tep73if7fc3o@4ax.com> <mailman.982020135.1863.python-list@python.org>
Message-ID: <GQmJOiw7EeQ4ouaJEFrC52lYxz1n@4ax.com>

On Tue, 13 Feb 2001 00:21:19 +0100, "Tim Peters" <tim.one at home.com>
wrote:

>The purpose is to discourage people from posting.  There's enough traffic on
>Python-Dev from the blessed invitees without the flamewar-attracting ease of
>hitting "reply" in a mail reader.

Good point. That seems fair.

>If someone wants to post something to Python-Dev, it takes them some effort
>to do so, and-- while I don't particularly care for that myself either --
>it has been effective in reducing noise.

OK, that makes sense. And experience is the best way to know, so if it
has worked, then I can't really argue...

>If you want to argue for a language change, you have to get Guido's ear, and
>the only sure way to do that (whether or not you're on Python-Dev!) is to
>write a PEP.  Some of the Python-Dev discussions are interesting, but I bet
>far fewer than you imagine.

Actually, I'm sad enough that I enjoy watching the process of language
design "in action", and I find python-dev a very interesting list to
read. I know - I've been watching you :-)

>then-again-there's-a-sure-way-for-you-to-find-out-ly y'rs  - tim

only-via-the-web-ly y'rs - Paul. :-)



From tim.one at home.com  Wed Feb 14 09:39:52 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 14 Feb 2001 03:39:52 -0500
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <GQmJOiw7EeQ4ouaJEFrC52lYxz1n@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEELIPAA.tim.one@home.com>

[Paul Moore, among others, wants easier access to Python-Dev]

Just a note to let people know that schemes are afoot to make this easier.
Guido does not want to open the list up, but is quite in favor of making it
easier to follow.  Whether or not you're all being ignored, he certainly
doesn't like it to *appear* that you're being ignored <wink>. Can't yet say
exactly how, though.

"Trust Guido".

that-is-all-ye-know-on-earth-and-all-ye-need-to-know-ly y'rs  - tim




From cfelling at iae.nl  Thu Feb  8 23:44:32 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 8 Feb 2001 23:44:32 +0100
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us>
Message-ID: <95v7gg$8ne$1@animus.fel.iae.nl>

Andrew Kuchling <akuchlin at mems-exchange.org> wrote:

> Envoi
> =====

> This will be the last python-dev summary I write, accounting for its
> rushed and sketchy quality.

This is really, really, really sad news!
You know, by reading your nice wrapups on the discussions in python-dev
I better understood why things changed as they did.  And more over it
gave me a good community feeling, it opened up the development process
to us mere mortals:)

> The goal of these summaries has been to make the discussions on
> python-dev more visible to the community, letting people offer timely
> comments while a thread is still reasonably current and fresh in
> memory.  Unfortunately that doesn't seem to happen, and no messages to

I guess that you did to good a job, your wrapups informed me that new
proposals were thouroughly discussed, and mostly the objections I had
I shared with someone more knowledgable *and* involved in the development
so no need to react.

> The release of 2.1 offers a second calibration on their effectiveness.
> 2.1 is the first Python release to have been carried out using PEPs as
> the mechanism, so there are no sizable changes in 2.1 that don't have
> a corresponding PEP.  Yet many people were *surprised* by some of the
> changes in Python 2.1 such as function attributes and nested scopes,
> even though PEPs were written and discussed, often in lengthy threads
> months ago.  

To me that's because the status of the  PEPS isn't cristal clear.
You have to frequent the PEPS status page and keep an eye on changes
there.  It would be much easier (for me atleast:) if someone (you again:)
would inform us that the status of a PEP changed.


I know all the needed info is availeble right now, but I've to go out
(on the net) and go over lots of messages just to get the info you
summed up so nicely for me all those months.  Prior to your wrapups
I really had no idea were Python development was heading, so I really
liked it.
-- 
groetjes, carel


From embed at geocities.com  Fri Feb  9 15:22:49 2001
From: embed at geocities.com (Warren Postma)
Date: Fri, 9 Feb 2001 09:22:49 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v7gg$8ne$1@animus.fel.iae.nl>
Message-ID: <c4Tg6.1909$D3.8369@tor-nn1.netcom.ca>

"Carel Fellinger" <cfelling at iae.nl> wrote in message
news:95v7gg$8ne$1 at animus.fel.iae.nl...
> Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
>
> > Envoi
> > =====
>
> > This will be the last python-dev summary I write, accounting for its
> > rushed and sketchy quality.
>
> This is really, really, really sad news!

Py_INCREF('news:95v7gg$8ne$1 at animus.fel.iae.nl...')





From tony at lsl.co.uk  Fri Feb  9 11:56:14 2001
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Fri, 9 Feb 2001 10:56:14 -0000
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <3d4ry56vzg.fsf@ute.cnri.reston.va.us>
Message-ID: <002c01c09286$ebe15520$f05aa8c0@lslp7o.int.lsl.co.uk>

I too am very sad that you're deciding to stop writing the python-dev
summaries (sometimes a resounding silence is appreciation, not
disdain!). The summaries have been on my "looking forwards to that" list
for a good while now. Many thanks for what you've done (any chance of
reconsidering now you know how much we *really* love you?).

As with others, I don't have time to follow python-dev itself (via
http://mail.python.org/pipermail/python-dev/ of course - as indicated at
the back of the summaries), but I *do* want to know what is going on
"behind the scenes". The summaries provide(d) a decent intermediate.

If AMK cannot continue with this, is there any scope for the python-dev
crowd getting one of the "core" group to write a summary as (erm) part
of their "paid" work? (ducks brickbats - I don't want to know what hours
you're *actually* working!)

Tibs ("spare time" - I'm sure I've *heard* of it - what did it mean,
now?)

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
Which is safer, driving or cycling?
Cycling - it's harder to kill people with a bike...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)





From phd at phd.pp.ru  Fri Feb  9 12:06:03 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 9 Feb 2001 14:06:03 +0300 (MSK)
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <002c01c09286$ebe15520$f05aa8c0@lslp7o.int.lsl.co.uk>
Message-ID: <Pine.LNX.4.30.0102091404240.25391-100000@fep132.fep.ru>

Hello! Me too! :)

On Fri, 9 Feb 2001, Tony J Ibbs (Tibs) wrote:
> I too am very sad that you're deciding to stop writing the python-dev
> summaries (sometimes a resounding silence is appreciation, not
> disdain!). The summaries have been on my "looking forwards to that" list

   Exactly!

> for a good while now. Many thanks for what you've done (any chance of

   Many thinks. We really appreciate your good job!

> reconsidering now you know how much we *really* love you?).

   Any chance you'll continue the summaries?

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From tim.one at home.com  Fri Feb  9 23:41:29 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 9 Feb 2001 17:41:29 -0500
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <002c01c09286$ebe15520$f05aa8c0@lslp7o.int.lsl.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEHLIOAA.tim.one@home.com>

[Tony J Ibbs (Tibs)]
> ...
> If AMK cannot continue with this, is there any scope for the python-dev
> crowd getting one of the "core" group to write a summary as (erm) part
> of their "paid" work?

Not unless they stopped doing Python development work; this is going to be a
volunteer activity no matter who does it.  Just between us, there were even
some weeks where Andrew's summaries were my best clue about what was
happening on Python-Dev (I read c.l.py first, because, well, *somebody*
ought to <wink>).




From andymac at bullseye.apana.org.au  Sat Feb 10 09:35:29 2001
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Sat, 10 Feb 2001 19:35:29 +1100 (EDT)
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <LNBBLJKPBEHFEDALKOLCOEHLIOAA.tim.one@home.com>
Message-ID: <Pine.OS2.4.30.0102101923330.56-100000@CENTRAL>

On Fri, 9 Feb 2001, Tim Peters wrote:

> I read c.l.py first, because, well, *somebody* ought to <wink>.

And I am most appreciative of your presence and insight.

I have to say that it disappoints me that Guido has found it necessary to
retreat from this community into the monastery of python-dev.  I fear that
Python's priesthood and citizenry will lose contact with each other as a
result.  It won't happen overnight, but incrementally.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From emile at fenx.com  Fri Feb  9 13:10:30 2001
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 9 Feb 2001 04:10:30 -0800
Subject: python-dev summary, Jan. 16-31
References: <mailman.981716953.9412.python-list@python.org>
Message-ID: <960mqj$itm0s$1@ID-11957.news.dfncis.de>

I thought earlier this week that I had missed the recent
summary, and then was pleased to see it show up, and now
saddened to see it pass on.

Thanks for your efforts in the past on this.  It has been
much appreciated.

--

Emile van Sebille
emile at fenx.com
-------------------


"Oleg Broytmann" <phd at phd.pp.ru> wrote in message
news:mailman.981716953.9412.python-list at python.org...
> Hello! Me too! :)
>
> On Fri, 9 Feb 2001, Tony J Ibbs (Tibs) wrote:
> > I too am very sad that you're deciding to stop writing
the python-dev
> > summaries (sometimes a resounding silence is
appreciation, not
> > disdain!). The summaries have been on my "looking
forwards to that" list
>
>    Exactly!
>
> > for a good while now. Many thanks for what you've done
(any chance of
>
>    Many thinks. We really appreciate your good job!
>
> > reconsidering now you know how much we *really* love
you?).
>
>    Any chance you'll continue the summaries?
>
> Oleg.
> ----
>      Oleg Broytmann            http://phd.pp.ru/
phd at phd.pp.ru
>            Programmers don't die, they just GOSUB without
RETURN.
>
>




From mmillikan at vfa.com  Fri Feb  9 15:04:41 2001
From: mmillikan at vfa.com (mmillikan at vfa.com)
Date: 09 Feb 2001 09:04:41 -0500
Subject: python-dev summary, Jan. 16-31
References: <mailman.981716953.9412.python-list@python.org> <960mqj$itm0s$1@ID-11957.news.dfncis.de>
Message-ID: <ud7csrlza.fsf@vfa.com>

+1

Mark millikan


From sandipan at vsnl.com  Fri Feb  9 16:37:51 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Fri, 9 Feb 2001 21:07:51 +0530
Subject: python-dev summary, Jan. 16-31
References: <mailman.981716953.9412.python-list@python.org> <960mqj$itm0s$1@ID-11957.news.dfncis.de> <ud7csrlza.fsf@vfa.com>
Message-ID: <00f101c092ae$43c32080$020000c3@node02>

+3. Saving electrons here ;-)

Though my colleagues and I use a number of other languages, python is
central in the architectures we work with. We strive to code in other
languages to match upto the python style and philosophy. We believe in it.
Watching it grow and evolve (via the dev summary) is, in my opinion, an
important process in our own development as programmers.

Regards,

Sandipan

----- Original Message -----
From: <mmillikan at vfa.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Friday, February 09, 2001 7:34 PM
Subject: Re: python-dev summary, Jan. 16-31


> +1
>
> Mark millikan
> --
> http://mail.python.org/mailman/listinfo/python-list
>




From akuchlin at mems-exchange.org  Fri Feb  9 15:55:52 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 09 Feb 2001 09:55:52 -0500
Subject: python-dev summary, Jan. 16-31
References: <mailman.981716953.9412.python-list@python.org>
Message-ID: <3dofwbhpmv.fsf@ute.cnri.reston.va.us>

Oleg Broytmann <phd at phd.pp.ru> writes:
>    Any chance you'll continue the summaries?

Not really.  All of this is sparked by my wanting to reduce the load,
really; I recently realized that I haven't hacked on *any* of my
projects (Oedipus, amkCrypto, etc.) lately.  Therefore I've dropped
out of my book club (one night a week), dropped the summaries, (two
nights every two weeks, or 15% of my evenings), and after 2.1final is
released I'll probably unsubscribe from python-dev, too.

So, to rescuing the summaries... For the information of potential
volunteers, I'll summarize the process.  Writing them isn't difficult
and can easily be done while watching TV -- I usually pull the mail
archive on my laptop and do just that.  They're just time-consuming,
taking up two evenings at a time.

1) Grab a copy of the mailbox archive for the month.

2) Load it up into mutt (or other MUA of choice), and delete messages 
   outside of the two-week period being summarized.

3) Sort the remaining messages by thread, and go through finding the 
   interesting threads.  What makes a thread interesting?

        * Threads discussing how to fix a particular bug,
          or tracing down a bug's root cause, can be quite lengthy but 
          usually aren't interesting. 

        * Minute discussions of language syntax are rarely interesting,
          and I think even Guido tunes them out after a while.

        * New proposals that spark off a discussion are interesting.
        * Discussions surrounding a PEP are interesting.

4) <time consuming but fun step> Read through each interesting thread,
   pulling out interesting quotes and summarizing the flow of the
   argument.  This is fun because you can introduce your own biases; I
   think this is why people become journalists.

5) <time consuming and tedious step> For each quote, find the URL for
   that message in the archives on mail.python.org, and link to it.
   This is painful; recently I've started spidering the archives with wget
   and grepping them in order to avoid ferreting through the indexes 
   over my modem connection.

   Thinking about it now, it would have been much easier on me to just
   do what Linux Weekly News does, and make a local copy of each
   message.  That means you wouldn't have to worry about links
   breaking (moving Mailman to mail.python.org required regenerating
   the archives, and the links in the October & November summaries are
   now all broken), and the author wouldn't have to hunt for the right
   message URL.  You'd lose the easy ability to follow the thread, of
   course.

6) The summary is written as plain ASCII, and it gets posted, and
   mailed to LWN and LinuxToday.  The HTML version is then just a
   matter of escaping <&>, turning the URLs into <A> tags with an Emacs
   macro, and putting <pre> tags around it; no fancy XML DTD...

--amk



From phd at phd.pp.ru  Fri Feb  9 16:39:44 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 9 Feb 2001 18:39:44 +0300 (MSK)
Subject: python-dev summary, Jan. 16-31
In-Reply-To: <3dofwbhpmv.fsf@ute.cnri.reston.va.us>
Message-ID: <Pine.LNX.4.30.0102091839010.27625-100000@fep132.fep.ru>

On 9 Feb 2001, Andrew Kuchling wrote:
> Oleg Broytmann <phd at phd.pp.ru> writes:
> >    Any chance you'll continue the summaries?
>
> Not really.  All of this is sparked by my wanting to reduce the load,
> really; I recently realized that I haven't hacked on *any* of my
> projects (Oedipus, amkCrypto, etc.) lately.  Therefore I've dropped

   Ah, ok, ok, that's right. Code is even more important! Good luck!

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From SBrunning at trisystems.co.uk  Fri Feb  9 17:11:08 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 9 Feb 2001 16:11:08 -0000 
Subject: python-dev summary, Jan. 16-31
Message-ID: <31575A892FF6D1118F5800600846864D5B17E3@intrepid>

> From:	Andrew Kuchling [SMTP:akuchlin at mems-exchange.org]
> * Announce PEP drafts and status changes to c.l.py.announce
> * Write a set of introductory pages for Python development (this has been 
>   brought up before).
> * A read-only alias for python-dev would be useful for several people, but
> 
>   may require serious Mailman hackery (I don't know).
 
Sounds to me like if these three could be achieved, 99.9% of c.l.py would be
more than happy. I know that *I* would.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From donn at oz.net  Fri Feb  9 18:18:47 2001
From: donn at oz.net (Donn Cave)
Date: 9 Feb 2001 17:18:47 GMT
Subject: python-dev summary, Jan. 16-31
References: <mailman.981735155.18442.python-list@python.org>
Message-ID: <9618pn$mfv$0@216.39.151.169>

Quoth Simon Brunning <SBrunning at trisystems.co.uk>:
|> From:	Andrew Kuchling [SMTP:akuchlin at mems-exchange.org]
|> * Announce PEP drafts and status changes to c.l.py.announce
|> * Write a set of introductory pages for Python development (this has been 
|>   brought up before).
|> * A read-only alias for python-dev would be useful for several people, but
|> 
|>   may require serious Mailman hackery (I don't know).
|  
| Sounds to me like if these three could be achieved, 99.9% of c.l.py would be
| more than happy. I know that *I* would.

Since you presume to speak for 99.9% of c.l.py, it's fair for me
to suggest you don't know what you're talking about.  I doubt
even the present membership of python-dev really has time to read
it, and very few of us with only an observer's interest are going
to be able to keep up.  I'm not making this up, see comments in
this thread.  Most of us don't have even the knowledge to always
follow what's going on.  Which has been part of the value of the
summaries, digested by someone who does.

I would be more than happy if I didn't have any reason to read
python-dev.

	Donn Cave, donn at oz.net


From NoSpam at NoSpam.com  Sun Feb 11 06:33:47 2001
From: NoSpam at NoSpam.com (Tom)
Date: Sun, 11 Feb 2001 05:33:47 GMT
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us>
Message-ID: <%uph6.125320$V22.25862028@news4.rdc1.on.home.com>

I read and appreciate all your 'python-dev summaries'.  I find them
interesting and informative, but more importantly, they make Python 'feel'
more open and inclusive.

Tom.

"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3d4ry56vzg.fsf at ute.cnri.reston.va.us...
> Python-dev summary, January 16-31, 2001
> =======================================
>
> To comment on material in this python-dev summary, you can simply post
> to comp.lang.python / <python-list at python.org>.  These summaries are
> archived at <URL:http://www.amk.ca/python/dev/>.
>
> This two week period saw the release of the first alpha of Python 2.1,
> and the second alpha was released a few days into February.  There's a
> lot of bugfixing going on, much of which isn't of interest here, but
> there were a few general discussions worth noting.
>
>
> 2.1alpha1
> =========
>
> The first alpha of Python 2.1 was released on January 22, and is the
> debut of Python 2.1.  It didn't compile cleanly on Cygwin, and the new
> autodetecting build system needs some more work to support more
> platforms correctly, but no major problems or brown paper bag bugs
> showed up.
>
> 2.1alpha2 was subsequently released on February 2, slightly out of the
> bounds of this summary.  2.1a2 incorporated the nested scopes from PEP
> 227 and the weak reference type from PEP 205.
>
> A draft of "What's New in Python 2.1" is available, and will be
continuously
> updated until 2.1final is shipped:
> http://www.amk.ca/python/2.1/
>
>
> unistr()
> ========
>
> M.-A. Lemburg checked in a unistr() built-in, but later backed out the
> change after subsequent discussion.  str() takes any Python object and
> converts it to a string, and unistr() would take any object and return
> a Unicode string.  ?!ng wondered why the existing unicode() built-in
> shouldn't do this -- currently it only accepts a string and an optional
> encoding -- or should this functionality be added to str()?
>
> This also raised the question of how a type could return a Unicode
> representation of itself.  Should the tp_str slot in the Py_TypeObject
> struct be permitted to return either an 8-bit or a Unicode string, or
> should a tp_unicode slot be added?  Some discussion followed, but no
> clear answer was apparent, so unistr() was removed.  This issue will
> probably have to wait for a PEP to be written, which means that it'll
> have to wait until 2.2.
>
>
> Sets
> ====
>
> Eric S. Raymond posted the docs for a set module he's written with an
> eye toward getting it into the standard library.  "The \module{set}
> module defines functions for treating lists and other sequences as
> mathematical sets, and defines a set class that uses these operations
> natively and overloads Python's standard operator set."
> http://mail.python.org/pipermail/python-dev/2001-January/012005.html
>
> But there's already a PEP for a set type, #218 by Greg Wilson:
> http://python.sourceforge.net/peps/pep-0218.html
>
> Reconciling the two proposals doesn't seem difficult, but opinion was
> divided: are sets so commonly used that they should be in the standard
> library, particularly if dictionaries can be used to implement many
> common set operations?
>
> ?!ng pointed out that simply allowing "if key in dict" would make
> using dictionaries as sets a bit more natural.  /F agreed, and GvR
> added "You know, I've long resisted this, but I agree now -- this is
> the right thing."  Thomas Wouters submitted a patch, but it hasn't yet
> been checked in.
>
> A subthread wondered about the related question of 'for X in
> dictionary:' Should this do anything, and if so, what?  X could be
> "key" or "key, value", but the best choice isn't obvious.  Another
> subthread drifted into the question of the Batteries Included
> distribution proposed by Moshe Zadka in PEP 206, and yet another one
> proposed dictionaries that can be marked as unmodified ("frozen" or
> "locked").  No clear conclusions were made, and the problems seem
> messy enough that a PEP will be required, so this is another thing
> likely to be postponed until 2.2.
>
>
> Envoi
> =====
>
> This will be the last python-dev summary I write, accounting for its
> rushed and sketchy quality.
>
> The goal of these summaries has been to make the discussions on
> python-dev more visible to the community, letting people offer timely
> comments while a thread is still reasonably current and fresh in
> memory.  Unfortunately that doesn't seem to happen, and no messages to
> python-dev are spurred by them.  Ideally people would come forward to
> say "The last summary mentioned you were talking about X. I use X a
> lot, and here's what I think: ...", but that isn't happening.
>
> The release of 2.1 offers a second calibration on their effectiveness.
> 2.1 is the first Python release to have been carried out using PEPs as
> the mechanism, so there are no sizable changes in 2.1 that don't have
> a corresponding PEP.  Yet many people were *surprised* by some of the
> changes in Python 2.1 such as function attributes and nested scopes,
> even though PEPs were written and discussed, often in lengthy threads
> months ago.
>
> To me, this makes it crystal clear that the summaries aren't achieving
> their goal of making the development process more transparent to the
> community.  Perhaps giving the PEPs higher visibility -- posting
> announcements of status changes and new PEPs to
> comp.lang.python.announce, for example -- would do a better job, and
> it would definitely require less effort.  In any case, I feel there's
> no point in continuing to write them.
>
> (Should anyone want to volunteer to continue writing them, please do
> so; contact me if you want copies of the previous summaries for a
> complete archive.)
>
> My thanks to the various people who've offered comments and
> encouragement along the way; you've all kept me going over the last 7
> months...
>
>
> Related Links
> =============
>
> Python-dev archives:
>         http://www.python.org/pipermail/python-dev/
>
> Python project page on SourceForge:
>         http://sourceforge.net/projects/python
>
> Python Enhancement Proposals (PEPs):
>         http://python.sourceforge.net/peps/
>
> What's New in Python 2.1
>         http://www.amk.ca/python/2.1/
>




From jurgen.defurne at philips.com  Thu Feb  8 16:23:55 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Thu, 8 Feb 2001 16:23:55 +0100
Subject: Memory leaks or other things ?
Message-ID: <0056900015777418000002L082*@MHS>

Hello, list,

I am having the following problem under NT, with Python 2.0 for NT.

For some time I have written a cron.pl project (yes, Perl) with Cygwin, which
runs fine using fork and exec(). Now, to have easier installation of base programs and
script modules, I would like to everything in Python. I had already (re)written much in
Python, but cron.py was the last.

I implemented cron.py using spawn. The complete application is one script.
The main routine syncs on 00 seconds, and then spawns a subprocess, which
reads and processes crontab. For every process that should be run, the script is
respawned with the appropriate arguments, all using Python 2.0 for NT.

The jobs which I have to run allocate a fair amount of memory (160 Mb). The funny
thing is that the jobs which are still running under cron.pl do not leak memory, but the
same job (different arguments) under cron.py does not give its memory back, although
I can't find any processes anymore in the Task Manager.

I am retesting now, with one modification : using sys.exit(0) to terminate the spawned
scripts properly.

Has anyone an idea ? You have the code here. Review at your own will.

Jurgen

#

# External modules
import re
import os
import sys
import string

from time import *
from BitVector import BitVector

# Global variables
year = month = day = hour = min = sec = None
crontab     = 'c:\\etc\\crontab2'
logfile     = 'c:\\var\\log\\cron2.log';
interpreter = 'd:\\python20\\python'
script      = 'cron.py'
shell       = 'c:\\winnt\\system32\\cmd.exe'
fh          = None
initialised = None

# Class definitions
class JobEntry:
    def __init__(self, line = None):
	if line:
	    M, H, d, m, w, c = string.split(line, sep = None, maxsplit = 5)
	    self.mins = self.BitInit(M, 60)
	    self.hour = self.BitInit(H, 24)
	    self.days = self.BitInit(d, 31)
	    self.mnth = self.BitInit(m, 12)
	    self.dows = self.BitInit(w,  7)
	    self.cmnd = string.strip(c)
	else:
	    self.mins = BitVector(60)
	    self.hour = Bitvector(24)
	    self.days = BitVector(31)
	    self.mnth = BitVector(12)
	    self.dows = BitVector(7)
	    self.cmnd = None
    
    def BitInit(self, value, length):
	rv = BitVector(length)

	if value == '*':
	    for i in range(length):
		rv[i] = 1
	else:
	    args = string.split(value, ',')
	    for arg in args:
		if string.find(arg, '-') != -1:
		    start, end = string.split(arg, '-')
		    for i in range(eval(start), eval(end) + 1):
			rv[i] = 1
		else:
		    rv[eval(arg)] = 1
	
	return rv

# Subroutine definitions

def logmsg(msg):
    fh = open(logfile, 'a')
    systime = ctime(time())
    fh.write('%s -- %s\n' % (systime, msg))
    fh.close()

"""
Running a job using spawn is not as easy as using fork and exec. In the normal
Unix a parent process can wait upon the completion of a subprocess, under
Windows this doesn't work. To provide support for this, the script will be
spawned to execute run_job(), and run_job will provide the necessary services
for controlling the job. The job itself will be run synchronously.
"""

def run_job(argv):
    job          = string.join(argv[2:])

    print job

    # Log activity before job is run
    logmsg('Starting : %s' % job)

    # Run job
    os.system(job)

    # Log activity after job has finished
    logmsg('Job %s finished' % job)

    # Exit system in a good fashion
    sys.exit(0)

def read_crontab():
    job_list = []
    fh       = open(crontab, 'r')

    line = fh.readline()
    while line:
	if re.search('^#',   line):
	    line = fh.readline()
	    continue
	if re.search('^ *$', line):
	    line = fh.readline()
	    continue

	job_list.append(JobEntry(line))
	line = fh.readline()

    fh.close()
    return job_list

def process_jobs(list):
    year, mth, day, hrs, min, sec, dow, None, None = localtime(time())

    year  -= 1 
    mth   -= 1
    day   -= 1
    dow   -= 1

    for entry in list:

	if entry.mins[min] == 0: continue
	if entry.hour[hrs] == 0: continue
	if entry.days[day] == 0: continue
	if entry.mnth[mth] == 0: continue
	if entry.dows[dow] == 0: continue
	
	os.spawnl(os.P_NOWAIT, interpreter, 'python', script, 'run', entry.cmnd)

def check_jobs():
    # Lees inhoud van crontab in tabel
    # Scan tabel voor jobs die moeten worden uitgevoerd
    #	Per job :
    #	 	Spawn Python20, python, script, 'run', job en argumenten
    job_list = read_crontab()
    process_jobs(job_list)

    # Exit this script in an orderly fashion
    sys.exit(0)

def start_check_jobs():
    os.spawnl(os.P_NOWAIT, interpreter, 'python', script, 'check')

def sync():
    global initialised
    
    (None, None, None,
     None, None, sec,
     None, None, None) = localtime(time())

    initial = 60 - sec

    if not initialised:
	if initial == 60: initial = 0
	initialised = 1

    sleep(initial)

def main(argv):
    if len(argv) == 1:
	logmsg('cron.py starting up')
	sync()
	while 1:
	    start_check_jobs()
	    sync()
    elif argv[1] == 'check':
	check_jobs()
    elif argv[1] == 'run':
	run_job(argv)

if __name__ == '__main__':
    main(sys.argv)


From jurgen.defurne at philips.com  Thu Feb  8 16:46:53 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Thu, 8 Feb 2001 16:46:53 +0100
Subject: Followup : Memory problem
Message-ID: <0056900015778525000002L052*@MHS>

I monitored (Task Manager) my system. Every minute there is activity
from the daemon. Every minute, regardless of the job being run, I seem to
lose 800 to 1200 kB of memory.

(Info : previous message w. script included)

Jurgen


From MarkH at ActiveState.com  Thu Feb  8 22:46:29 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 08 Feb 2001 21:46:29 GMT
Subject: Followup : Memory problem
References: <mailman.981646745.25788.python-list@python.org>
Message-ID: <3A8312C7.2020500@ActiveState.com>

jurgen.defurne at philips.com wrote:

> I monitored (Task Manager) my system. Every minute there is activity
> from the daemon. Every minute, regardless of the job being run, I seem to
> lose 800 to 1200 kB of memory.
> 

What process is losing the memory?  Python?

If so, can you make a small test case that shows up for _every_ call to 
spawn?

Mark.



From jurgen.defurne at philips.com  Fri Feb  9 10:10:24 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Fri, 9 Feb 2001 10:10:24 +0100
Subject: Followup : Memory problem
Message-ID: <0056900015791593000002L032*@MHS>

I'll get back later, I have much work for the moment.

Thanks for being interested.

Jurgen


From rozen at rgv.hp.com  Thu Feb  8 17:20:21 2001
From: rozen at rgv.hp.com (Don Rozenberg)
Date: Thu, 08 Feb 2001 08:20:21 -0800
Subject: [ANN] PAGE 0.1 - A drag and drop GUI generator for Python
Message-ID: <3A82C745.CAD4BA9@rgv.hp.com>

PAGE is a tool which helps to create GUI for Python programs.  It uses
Tk and Tix widgets.  PAGE is not an end-all, be-all tool, but rather
one that attempts to ease the burden on the Python programmer. It is
aimed at the user who will put up with a less-than-general GUI
capability in order to get an easily generated GUI. It does not build
an entire application but rather is aimed at building a single GUI
window.   PAGE is built on version 1.2.2 of Visual Tcl, version
8.1.1 of Tix, and Tcl/Tk version 8.3.2. It works with Python 2.0.
This is an early version which run on Linux and Windows/NT. It
supports many Tk widgets and a subset of the Tix widget set.  The user
will have to acquire and install Tix 8.1.1 and tcl8.3.2 to use this
program.
It is released under the GPL.

The PAGE web site is

 http://www.geocities.com/page_python



--
Don Rozenberg
rozen at rgv.hp.com
408-343-6266






From dcarboni at crs4.it  Thu Feb  8 17:38:57 2001
From: dcarboni at crs4.it (Davide Carboni)
Date: Thu, 8 Feb 2001 17:38:57 +0100
Subject: problems with PIL
Message-ID: <95uemq$7vr$1@pietro.crs4.it>

Hi to all,

I'm fighting against PIL, I put _imaging20.dll and _imagingTK.dll in DLLs
but when I try the piece of code below I get the error, wht's wrong in your
opinion ??
Thanx.




>>> import Image
>>> i=Image.new("PPM",(100,100))
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "c:\bin\python20\extensions\PIL\Image.py", line 848, in new
    return Image()._makeself(core.fill(mode, size, color))
  File "c:\bin\python20\extensions\PIL\Image.py", line 43, in __getattr__
    raise ImportError, "The _imaging C module is not installed"
ImportError: The _imaging C module is not installed

--
_______________________________________________________________________

 Davide Carboni                         Email: davide.carboni at crs4.it
 Network Distributed Applications       Phone: +39 070 2796 303
 CRS4, VI Str. Ovest, Z.I. Macchiareddu ICQ  # 107447153
 I-09010 Uta (CA), ITALY                WWW  : www.crs4.it/~dcarboni
_______________________________________________________________________




From fredrik at pythonware.com  Thu Feb  8 19:14:33 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 08 Feb 2001 18:14:33 GMT
Subject: problems with PIL
References: <95uemq$7vr$1@pietro.crs4.it>
Message-ID: <dmBg6.13588$AH6.2041509@newsc.telia.net>

Davide Carboni wrote:
> I'm fighting against PIL, I put _imaging20.dll and _imagingTK.dll in DLLs
> but when I try the piece of code below I get the error

looks like you forgot to rename the DLL.  try doing:

    import _imaging20

if that works, you've installed it correctly, but under
the wrong name.

renaming it to "_imaging.dll" should help.

(and yes, it's "_imagingtk.dll", not "_imagingTK.dll".
Python is case-sensitive, also on Windows)

Cheers /F




From richard at wmblake.com  Thu Feb  8 17:42:13 2001
From: richard at wmblake.com (Richard Blumberg)
Date: Thu, 08 Feb 2001 11:42:13 -0500
Subject: Hosting service supporting 2.0
Message-ID: <richard-542A25.11421308022001@nntp.fuse.net>

I'm looking for a moderately priced web hosting service that supports 
Python 2.0. Anyone know of any such?

Thanks

Richard


From aleaxit at yahoo.com  Thu Feb  8 18:15:49 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 8 Feb 2001 18:15:49 +0100
Subject: Hosting service supporting 2.0
References: <richard-542A25.11421308022001@nntp.fuse.net>
Message-ID: <95uk8601d4g@news2.newsguy.com>

"Richard Blumberg" <richard at wmblake.com> wrote in message
news:richard-542A25.11421308022001 at nntp.fuse.net...
> I'm looking for a moderately priced web hosting service that supports
> Python 2.0. Anyone know of any such?

I'm looking too, btw.  And if they let me install my own .pyd's,
then life would be really sweet...


Alex





From cg at schlund.de  Thu Feb  8 19:04:40 2001
From: cg at schlund.de (Carsten Gaebler)
Date: Thu, 08 Feb 2001 19:04:40 +0100
Subject: Hosting service supporting 2.0
References: <richard-542A25.11421308022001@nntp.fuse.net>
Message-ID: <3A82DFB8.9DA8D438@schlund.de>

Richard Blumberg wrote:
> 
> I'm looking for a moderately priced web hosting service that supports
> Python 2.0. Anyone know of any such?
> 

www.oneandone.co.uk has Python 2.0. And as long as they don't fire me, it
will always be up to date. :-)

Regards
Carsten.


From gilgsn at my-deja.com  Sat Feb 10 04:43:32 2001
From: gilgsn at my-deja.com (Gil G.)
Date: Sat, 10 Feb 2001 03:43:32 GMT
Subject: Hosting service supporting 2.0
References: <richard-542A25.11421308022001@nntp.fuse.net>
Message-ID: <962dd5$8rk$1@nnrp1.deja.com>

Hello Richard,

Go to: http://keskydee.com/hosting.html
Python 2.0, mod_python 2.6.3
Plus PHP 4.04 and Perl/mod_perl

Sincerely,

Gil.

In article <richard-542A25.11421308022001 at nntp.fuse.net>,
  Richard Blumberg <richard at wmblake.com> wrote:
> I'm looking for a moderately priced web hosting service that supports
> Python 2.0. Anyone know of any such?
>
> Thanks
>
> Richard


Sent via Deja.com
http://www.deja.com/


From txzd at my-deja.com  Sat Feb 10 20:12:17 2001
From: txzd at my-deja.com (txzd)
Date: Sat, 10 Feb 2001 19:12:17 GMT
Subject: Hosting service supporting 2.0
References: <richard-542A25.11421308022001@nntp.fuse.net> <962dd5$8rk$1@nnrp1.deja.com>
Message-ID: <9643qd$fn9$1@nnrp1.deja.com>

In article <962dd5$8rk$1 at nnrp1.deja.com>,
  Gil G. <gilgsn at my-deja.com> wrote:
> Hello Richard,
>
> Go to: http://keskydee.com/hosting.html
> Python 2.0, mod_python 2.6.3
> Plus PHP 4.04 and Perl/mod_perl
>
> Sincerely,
>
> Gil.
>
> In article <richard-542A25.11421308022001 at nntp.fuse.net>,
>   Richard Blumberg <richard at wmblake.com> wrote:
> > I'm looking for a moderately priced web hosting service that
supports
> > Python 2.0. Anyone know of any such?
> >
> > Thanks
> >
> > Richard
>
> Sent via Deja.com
> http://www.deja.com/
>

Richard: Here's another option.
http://quicksitemaker.com/members/djedwards


Sent via Deja.com
http://www.deja.com/


From parker at gol.com  Mon Feb 19 22:35:29 2001
From: parker at gol.com (Ian Parker)
Date: Mon, 19 Feb 2001 21:35:29 GMT
Subject: Hosting service supporting 2.0
References: <richard-542A25.11421308022001@nntp.fuse.net>
Message-ID: <EJ2obCAMc5j6EwZb@gol.com>

In article <richard-542A25.11421308022001 at nntp.fuse.net>, Richard
Blumberg <richard at wmblake.com> writes
>I'm looking for a moderately priced web hosting service that supports 
>Python 2.0. Anyone know of any such?
>
>Thanks
>
>Richard

www.pair.com support python 2.0, MySQL and other products

-- 
Ian Parker


From bk at xk7.com  Tue Feb 20 15:42:23 2001
From: bk at xk7.com (Burkhard Kloss)
Date: Tue, 20 Feb 2001 14:42:23 -0000
Subject: Hosting service supporting 2.0
References: <richard-542A25.11421308022001@nntp.fuse.net> <EJ2obCAMc5j6EwZb@gol.com>
Message-ID: <982680144.675737@master.nyc.kbcfp.com>

> >I'm looking for a moderately priced web hosting service that supports
> >Python 2.0. Anyone know of any such?
Hurricane Electric (www.he.net) allows you to install your own software -
including in my case python.




From sheila at spamcop.net  Tue Feb 20 16:27:02 2001
From: sheila at spamcop.net (Sheila King)
Date: Tue, 20 Feb 2001 15:27:02 GMT
Subject: Hosting service supporting 2.0
References: <richard-542A25.11421308022001@nntp.fuse.net> <EJ2obCAMc5j6EwZb@gol.com> <982680144.675737@master.nyc.kbcfp.com>
Message-ID: <03359t4nkuftad89eipq67ctnhdb56r3h9@4ax.com>

On Tue, 20 Feb 2001 14:42:23 -0000, "Burkhard Kloss" <bk at xk7.com> wrote in
comp.lang.python in article <982680144.675737 at master.nyc.kbcfp.com>:

:> >I'm looking for a moderately priced web hosting service that supports
:> >Python 2.0. Anyone know of any such?
:Hurricane Electric (www.he.net) allows you to install your own software -
:including in my case python.
:

My host also let me install my own software, so I could run Python 2.0 (they
only have 1.5.1 on the server and will not upgrade at this time, due to too
many core dependencies), but it took up over 20 Meg of space. I had to pay for
extra disk space, because this took me over my space for the package I have.
:-(

I just mention this, since others might want to be aware how much disk space
the install takes up (and I took a lot of the optional stuff out).

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From echuck at mindspring.com  Wed Feb 21 20:52:43 2001
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Wed, 21 Feb 2001 14:52:43 -0500
Subject: Hosting service supporting 2.0
Message-ID: <5.0.2.1.0.20010221144840.049c3e30@mail.mindspring.com>

"Burkhard Kloss" <bk at xk7.com> wrote in message 
news:<982680144.675737 at master.nyc.kbcfp.com>...
> > >I'm looking for a moderately priced web hosting service that supports
> > >Python 2.0. Anyone know of any such?
>Hurricane Electric (www.he.net) allows you to install your own software -
>including in my case python.

Python-Friendly Virtual Website Providers:
   http://www.corrt.com/info/pyisp-list.html

I personally use http://www.pair.com/ and was able to install Py 2.0 in my 
home directory.

-Chuck
--
Get web dev at http://webware.sourceforge.net/




From ingoogni at home.nl  Thu Feb  8 20:44:27 2001
From: ingoogni at home.nl (ingo)
Date: Thu, 08 Feb 2001 19:44:27 GMT
Subject: [PIL]  16 / 48 bit  images?
Message-ID: <Xns9042D3216AD7seed7@news1.zwoll1.ov.home.nl>

Using PIL with 16 bit greyscale images I'm struggeling a 'bit'. How does 
PIL work with 16/48 bit images?

>>> import Image
>>> im=Image.open("a1.png")
>>> len(im.histogram())
256
>>>
Here I expected the result to be 65536. Is the histogram fixed to 256 
levels (per band), or is the image converted to 8 bit?

When using the point() operation on a 16 bit greyscale image I get an 
error, the same operation on an 8 bit image is flawless. When done to one 
band of a 48 bit image all seems to be ok, bu the mode of the band is "L". 
Is the band converted to 8 bit?

>>> im=Image.open("a1.png")  #16 bit greyscale
>>> im.mode
'I'
>>> mask=im.point(lambda i: i<100 and 255)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "d:\programming\python20\pil\Image.py", line 582, in point
    scale, offset = _getscaleoffset(lut)
  File "d:\programming\python20\pil\Image.py", line 271, in 
_getscaleoffset
    data = expr(_E(stub)).data
AttributeError: 'int' object has no attribute 'data'
>>>

>>> im=Image.open("a2.png")    	#8 bit greyscale
>>> im.mode
'P'
>>> mask=im.point(lambda i: i<100 and 255)
>>>

>>> im=Image.open("a3.png")    	#48 bit colour
>>> im.mode
'RGB'
>>> S=im.split()
>>> R,G,B=0,1,2
>>> S[R].mode
'L'
>>> mask=S[R].point(lambda i: i<100 and 255)
>>>

Ingo

-- 
Photography: http://members.home.nl/ingoogni/
Pov-Ray    : http://members.home.nl/seed7/


From ievsikov at flowpath.com  Thu Feb  8 21:19:55 2001
From: ievsikov at flowpath.com (Igor Evsikov)
Date: Thu, 8 Feb 2001 15:19:55 -0500
Subject: EvalIt32 is a 32-bit library that allows programs to evaluate math expressions defined at run time. 
Message-ID: <kbDg6.466$bl6.69727361@radon.golden.net>

EvalIt32 is a 32-bit library that allows programs to evaluate math
expressions defined at run time. EvalIt32 includes functions etEvalIt and
etCompEval for evaluate math expression, etEvalFile to evaluate math
expressions from input files, and writes results to output files. This .dll
supports common math operators and functions for real and complex numbers,
as well as user variables and functions. This version includes direct
support for Visual Basic, C++ Builder (Borland) and Delphi. Other compilers
can use modified versions of a headers.Requires Win9x/NT/2000.

Download:
http://www.zdnet.com/downloads/stories/info/0,,001EU4,.html

Features:

      - Arithmetic and logical operators and functions
      - Common functions such as exp, ln, sqrt, sqr, bnml etc.
      - Common, trigonometric, hyperbolic complex functions
      - Trigonometric, Hyperbolic  functions
      - Numerical Integration
      - Equations can be solved
      - Special functions (Gamma, Bessel's, Si, Ci, erf, erfc, Fresnel's)
      - Statistic functions (Average, Standard deviation, Sum, Random,
        Gauss random, statistical variance, etc )
      - FOR-type loop
      - if (...) then (...) else (...) function
      - Error position fixed
      - Print results, graphics and print preview
      - Save graphics to BMP, WMF, EMF formats
      - Radian and Degree modes for trigonometric functions
      - Precision: 10-12 significant digits.
      - Range: _(3.4E-4392 to 1.1E+4392)
      - User define functions(only profesional edition)
      - User defined constants

Special requirements: None.

Shareware.  Uploaded by the author.

Igor Evsikov
ievsikov at flowpath.com











From chris at rpgarchive.com  Thu Feb  8 21:26:25 2001
From: chris at rpgarchive.com (Chris Davis)
Date: Thu, 8 Feb 2001 14:26:25 -0600
Subject: SAXReaderNotAvailable
Message-ID: <0102081426250S.12796@lab.rpgarchive.com>

Anyone know what could be causing this error?  I'm getting tons of people who 
use my program reporting this bug.  The one thing they all have in common is 
they installed python20 from source on linux.  They all have expat installed. 
Please, can anyone help?

File "./minidom.py", line 570, in _doparse
? ? events = apply(func, args, kwargs)
? File "./pulldom.py", line 244, in parseString
? ? parser = xml.sax.make_parser()
? File "/usr/lib/python2.0/xml/sax/__init__.py", line 88, in make_parser
? ? raise SAXReaderNotAvailable("No parsers found", None)
xml.sax._exceptions.SAXReaderNotAvailable: No parsers found
Terminated


-- 
Chris Davis
chris at rpgarchive.com

RPGArchive	http://rpgarchive.com
OpenRPG	http://openrpg.com



From phd at phd.pp.ru  Fri Feb  9 09:31:19 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 9 Feb 2001 11:31:19 +0300 (MSK)
Subject: SAXReaderNotAvailable
In-Reply-To: <0102081426250S.12796@lab.rpgarchive.com>
Message-ID: <Pine.LNX.4.30.0102091130440.25391-100000@fep132.fep.ru>

On Thu, 8 Feb 2001, Chris Davis wrote:
> Anyone know what could be causing this error?  I'm getting tons of people who
> use my program reporting this bug.  The one thing they all have in common is
> they installed python20 from source on linux.  They all have expat installed.

   Did the compiled PyExpat - Python wrapper for expat?

> File "./minidom.py", line 570, in _doparse
> ? ? events = apply(func, args, kwargs)
> ? File "./pulldom.py", line 244, in parseString
> ? ? parser = xml.sax.make_parser()
> ? File "/usr/lib/python2.0/xml/sax/__init__.py", line 88, in make_parser
> ? ? raise SAXReaderNotAvailable("No parsers found", None)
> xml.sax._exceptions.SAXReaderNotAvailable: No parsers found
> Terminated

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From phd at phd.pp.ru  Fri Feb  9 09:56:55 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 9 Feb 2001 11:56:55 +0300 (MSK)
Subject: SAXReaderNotAvailable
In-Reply-To: <Pine.LNX.4.30.0102091130440.25391-100000@fep132.fep.ru>
Message-ID: <Pine.LNX.4.30.0102091156340.25391-100000@fep132.fep.ru>

On Fri, 9 Feb 2001, Oleg Broytmann wrote:
>    Did the compiled PyExpat - Python wrapper for expat?

   I meant "did theY" :)

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From grante at visi.com  Thu Feb  8 21:37:42 2001
From: grante at visi.com (Grant Edwards)
Date: Thu, 08 Feb 2001 20:37:42 GMT
Subject: [xyz]modem module?
Message-ID: <qsDg6.333$a7.11639@ruti.visi.com>

A couple months back somebody asked about Python [xyz]modem
support, but I didn't find any positive answers with deja.

Did anybody ever find anything?

If not, what's the easiest way to running program like "sz"
with its stdout/stdin connected to a socket's file descriptor?

Does one just do the traditional fork/close/dup/exec using the
os module?

-- 
Grant Edwards                   grante             Yow!  CHUBBY CHECKER owns
                                  at               my BUILDING!
                               visi.com            


From embed at geocities.com  Fri Feb  9 15:14:38 2001
From: embed at geocities.com (Warren Postma)
Date: Fri, 9 Feb 2001 09:14:38 -0500
Subject: [xyz]modem module?
References: <qsDg6.333$a7.11639@ruti.visi.com>
Message-ID: <xYSg6.1907$D3.7800@tor-nn1.netcom.ca>

"Grant Edwards" <grante at visi.com> wrote in message
news:qsDg6.333$a7.11639 at ruti.visi.com...
> A couple months back somebody asked about Python [xyz]modem
> support, but I didn't find any positive answers with deja.

Sounds like a good job for a C module wrapper for SZ.  X and YModem-Batch
are easily implemented (I have tonnes of sample code in C already written).
But I've never written my own Z-Modem. It's kind of a "black art".

Warren





From grante at visi.com  Fri Feb  9 17:21:31 2001
From: grante at visi.com (Grant Edwards)
Date: Fri, 09 Feb 2001 16:21:31 GMT
Subject: [xyz]modem module?
References: <qsDg6.333$a7.11639@ruti.visi.com> <xYSg6.1907$D3.7800@tor-nn1.netcom.ca>
Message-ID: <fOUg6.652$a7.59019@ruti.visi.com>

In article <xYSg6.1907$D3.7800 at tor-nn1.netcom.ca>, Warren Postma wrote:

>> A couple months back somebody asked about Python [xyz]modem
>> support, but I didn't find any positive answers with deja.
>
>Sounds like a good job for a C module wrapper for SZ.  X and YModem-Batch
>are easily implemented (I have tonnes of sample code in C already written).
>But I've never written my own Z-Modem. It's kind of a "black art".

I'm doing a spawnv of sz/sx now, and that works OK, but I was
hoping for something portable.  One of these days I'll write X
or Y in Python.

-- 
Grant Edwards                   grante             Yow!  All of life is a blur
                                  at               of Republicans and meat!
                               visi.com            


From jimholliman at heartsoft.com  Thu Feb  8 22:41:42 2001
From: jimholliman at heartsoft.com (Jim)
Date: Thu, 8 Feb 2001 15:41:42 -0600
Subject: C++ style Stream Operators
Message-ID: <3a831395.0@216.0.152.7>

I'd like to use C++style stream i/o operators, such as "file_object <<
some_stuff" in my python code.
IS this sort of operator setup and what modules would I need to take
advantage of them if so?

Thanks
Jim






From dsh8290 at rit.edu  Thu Feb  8 23:03:03 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 8 Feb 2001 17:03:03 -0500
Subject: C++ style Stream Operators
In-Reply-To: <3a831395.0@216.0.152.7>; from jimholliman@heartsoft.com on Thu, Feb 08, 2001 at 03:41:42PM -0600
References: <3a831395.0@216.0.152.7>
Message-ID: <20010208170303.A14287@harmony.cs.rit.edu>

As of Python 2.0 you can use this :

a_file = open( "filename" , "w" )
print >> a_file , "something to output"


On Thu, Feb 08, 2001 at 03:41:42PM -0600, Jim wrote:
| I'd like to use C++style stream i/o operators, such as "file_object <<
| some_stuff" in my python code.
| IS this sort of operator setup and what modules would I need to take
| advantage of them if so?
| 
| Thanks
| Jim
| 



From aleaxit at yahoo.com  Thu Feb  8 23:35:34 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 8 Feb 2001 23:35:34 +0100
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7>
Message-ID: <95v75u021ll@news2.newsguy.com>

"Jim" <jimholliman at heartsoft.com> wrote in message
news:3a831395.0 at 216.0.152.7...
> I'd like to use C++style stream i/o operators, such as "file_object <<
> some_stuff" in my python code.
> IS this sort of operator setup and what modules would I need to take
> advantage of them if so?

You can get a somewhat similar look with the new-ish
    print >> file_object, some_stuff
syntax.  Or, go whole hog (the expression somehow
seems particularly appropriate) and write a class with
an overloaded __lshift__ operator:

class Stream:
    def __init__(self, file):
        self.file = file
    def __lshift__(self, stuff):
        self.file.write(str(stuff))
        return self
    def __repr__(self):
        return ''

import sys
cout = Stream(sys.stdout)

so you can code beauties like cout<<"Hi there!".  The
'return self' at the end of __lshift__ lets you chain your
'<<'s, and the __repr__ even lets you try it out at the
interactive interpreter command line.

I think you'd be very badly advised to pursue this route,
spending energy to enable doubtful syntax sugar which
many Pythonistas will find quite unpleasant, but, in any
case, Python offers you "enough rope to shoot yourself
in the foot" -- and you could argue that "print>>flob"
is a precedent that legitimates what you're doing:-).


Alex





From root at rainerdeyke.com  Fri Feb  9 02:20:59 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 09 Feb 2001 01:20:59 GMT
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7>
Message-ID: <%BHg6.195182$ge4.67633843@news2.rdc2.tx.home.com>

"Jim" <jimholliman at heartsoft.com> wrote in message
news:3a831395.0 at 216.0.152.7...
> I'd like to use C++style stream i/o operators, such as "file_object <<
> some_stuff" in my python code.
> IS this sort of operator setup and what modules would I need to take
> advantage of them if so?

File objects in Python are somewhat like 'streambuf's in C++.  Just write
your own class with a '__lshift__' that formats the text and writes it to a
file object:

class ostream:
  def __init__(self, buf):
    self.buf = buf
  def __lshift__(self, rhs):
    self.buf.write(str(rhs))

cout = ostream(sys.stdout)
cout << 'Hello world!\n'


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From nikifor at sky.inp.nsk.su  Sun Feb 11 13:48:10 2001
From: nikifor at sky.inp.nsk.su (Alex A. Nikiforov)
Date: 11 Feb 2001 12:48:10 GMT
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7> <%BHg6.195182$ge4.67633843@news2.rdc2.tx.home.com>
Message-ID: <9661ma$1ev8j$1@sky.inp.nsk.su>

Rainer Deyke <root at rainerdeyke.com> wrote:
> "Jim" <jimholliman at heartsoft.com> wrote in message
> news:3a831395.0 at 216.0.152.7...
>> I'd like to use C++style stream i/o operators, such as "file_object <<
>> some_stuff" in my python code.
>> IS this sort of operator setup and what modules would I need to take
>> advantage of them if so?

> File objects in Python are somewhat like 'streambuf's in C++.  Just write
> your own class with a '__lshift__' that formats the text and writes it to a
> file object:

> class ostream:
>   def __init__(self, buf):
>     self.buf = buf
>   def __lshift__(self, rhs):
>     self.buf.write(str(rhs))

> cout = ostream(sys.stdout)
> cout << 'Hello world!\n'


what about cin ?



From nikifor at sky.inp.nsk.su  Sun Feb 11 13:51:13 2001
From: nikifor at sky.inp.nsk.su (Alex A. Nikiforov)
Date: 11 Feb 2001 12:51:13 GMT
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7> <%BHg6.195182$ge4.67633843@news2.rdc2.tx.home.com> <9661ma$1ev8j$1@sky.inp.nsk.su>
Message-ID: <9661s1$1ev8j$2@sky.inp.nsk.su>

Alex A. Nikiforov <nikifor at sky.inp.nsk.su> wrote:
> Rainer Deyke <root at rainerdeyke.com> wrote:
>> "Jim" <jimholliman at heartsoft.com> wrote in message
>> news:3a831395.0 at 216.0.152.7...
>>> I'd like to use C++style stream i/o operators, such as "file_object <<
>>> some_stuff" in my python code.
>>> IS this sort of operator setup and what modules would I need to take
>>> advantage of them if so?

>> File objects in Python are somewhat like 'streambuf's in C++.  Just write
>> your own class with a '__lshift__' that formats the text and writes it to a
>> file object:

>> class ostream:
>>   def __init__(self, buf):
>>     self.buf = buf
>>   def __lshift__(self, rhs):
>>     self.buf.write(str(rhs))

>> cout = ostream(sys.stdout)
>> cout << 'Hello world!\n'


> what about cin ?

or at least scanf ?

                      Alexei


From root at rainerdeyke.com  Sun Feb 11 21:53:57 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sun, 11 Feb 2001 20:53:57 GMT
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7> <%BHg6.195182$ge4.67633843@news2.rdc2.tx.home.com> <9661ma$1ev8j$1@sky.inp.nsk.su> <9661s1$1ev8j$2@sky.inp.nsk.su>
Message-ID: <FZCh6.216922$ge4.73738343@news2.rdc2.tx.home.com>

"Alex A. Nikiforov" <nikifor at sky.inp.nsk.su> wrote in message
news:9661s1$1ev8j$2 at sky.inp.nsk.su...
> Alex A. Nikiforov <nikifor at sky.inp.nsk.su> wrote:
> > Rainer Deyke <root at rainerdeyke.com> wrote:
> >> class ostream:
> >>   def __init__(self, buf):
> >>     self.buf = buf
> >>   def __lshift__(self, rhs):
> >>     self.buf.write(str(rhs))
>
> >> cout = ostream(sys.stdout)
> >> cout << 'Hello world!\n'
>
>
> > what about cin ?

Can't be done.  In Python, the expression 'A >> B' cannot assign to 'B'.

> or at least scanf ?

Again, not possible with C-style syntax due to lack of the address-taking
operator ('&').  There might be a function that does something similar
(returning a tuple of values given an input string and a format string); if
there is, I don't know about it.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From nikifor at sky.inp.nsk.su  Tue Feb 13 04:24:25 2001
From: nikifor at sky.inp.nsk.su (Alex A. Nikiforov)
Date: 13 Feb 2001 03:24:25 GMT
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7> <%BHg6.195182$ge4.67633843@news2.rdc2.tx.home.com> <9661ma$1ev8j$1@sky.inp.nsk.su> <9661s1$1ev8j$2@sky.inp.nsk.su> <FZCh6.216922$ge4.73738343@news2.rdc2.tx.home.com>
Message-ID: <96a9d9$1makv$1@sky.inp.nsk.su>

Rainer Deyke <root at rainerdeyke.com> wrote:
> "Alex A. Nikiforov" <nikifor at sky.inp.nsk.su> wrote in message
> news:9661s1$1ev8j$2 at sky.inp.nsk.su...
>> Alex A. Nikiforov <nikifor at sky.inp.nsk.su> wrote:
>> > Rainer Deyke <root at rainerdeyke.com> wrote:
>> >> class ostream:
>> >>   def __init__(self, buf):
>> >>     self.buf = buf
>> >>   def __lshift__(self, rhs):
>> >>     self.buf.write(str(rhs))
>>
>> >> cout = ostream(sys.stdout)
>> >> cout << 'Hello world!\n'
>>
>>
>> > what about cin ?

> Can't be done.  In Python, the expression 'A >> B' cannot assign to 'B'.

>> or at least scanf ?

> Again, not possible with C-style syntax due to lack of the address-taking
> operator ('&').  There might be a function that does something similar
> (returning a tuple of values given an input string and a format string); if
> there is, I don't know about it.

Oops :-)
There is answer on my question in Python FAQ
and there is contibuted scanf





From joconnor at cybermesa.com  Sun Feb 11 15:11:37 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 11 Feb 2001 14:11:37 GMT
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7>
Message-ID: <3a869d32.47459032@news.cybermesa.com>

On Thu, 8 Feb 2001 15:41:42 -0600, "Jim" <jimholliman at heartsoft.com>
wrote:

>I'd like to use C++style stream i/o operators, such as "file_object <<
>some_stuff" in my python code.
>IS this sort of operator setup and what modules would I need to take
>advantage of them if so?


Well, not C++ style, but I've started working on some Smalltalk-style
Streams for Python.

First rudimentary example is here:
http://pub1.ezboard.com/fobjectorienteddevelopmentprogrammingexamples.showMessage?topicID=41.topic

Discussion is here:

http://pub1.ezboard.com/fobjectorienteddevelopmentpython.showMessage?topicID=81.topic

Take care,


Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum - http://pub1.ezboard.com/fobjectorienteddevelopmentpython

"God himself plays on the bass strings first, when he tunes the soul"


From cybereal at my-deja.com  Thu Feb  8 22:54:38 2001
From: cybereal at my-deja.com (cybereal at my-deja.com)
Date: Thu, 08 Feb 2001 21:54:38 GMT
Subject: Embeding Jython In Java
Message-ID: <95v4ir$g6m$1@nnrp1.deja.com>

I'm in desperate need of some wisdom on the
subject, how do I go about distribution with my
app? hell, how do I include it for compilation?
Do I decompress the .jar?  Any ideas? I would
have no problem USING it, I just don't know how
to get it into my app; and please, I'm not stupid
don't tell me "Import it" because that's
obviously not my problem

THANKS!

btw, you can email me direct if you have a heart
if you send it to cybereal at hotmail dot co...
you know


Sent via Deja.com
http://www.deja.com/


From zamurai at gmx.net  Sat Feb 10 13:10:56 2001
From: zamurai at gmx.net (Zamurai)
Date: Sat, 10 Feb 2001 13:10:56 +0100
Subject: Embeding Jython In Java
References: <95v4ir$g6m$1@nnrp1.deja.com>
Message-ID: <981807056.38482970@news.t-online.de>

Just write a programm, or only a class. compile it with Jython and import it in
your Javaprogramm.



cybereal at my-deja.com wrote:
> I'm in desperate need of some wisdom on the
> subject, how do I go about distribution with my
> app? hell, how do I include it for compilation?
> Do I decompress the .jar?  Any ideas? I would
> have no problem USING it, I just don't know how
> to get it into my app; and please, I'm not stupid
> don't tell me "Import it" because that's
> obviously not my problem
> 
> THANKS!
> 
> btw, you can email me direct if you have a heart
> if you send it to cybereal at hotmail dot co...
> you know
> 
> 
> Sent via Deja.com
> http://www.deja.com/



From ehagemann at home.com  Thu Feb  8 23:36:13 2001
From: ehagemann at home.com (Eric Hagemann)
Date: Thu, 08 Feb 2001 22:36:13 GMT
Subject: embedding C and Python in Perl
Message-ID: <xbFg6.80652$B6.19092233@news1.rdc1.md.home.com>

Hey all   --- maybe Python needs the same ?  Anybody seen this ?

http://www.perl.com/pub/2001/02/inline.html?wwwrrr_20010206.txt





From eq3pvl at eq.uc.pt  Fri Feb  9 13:42:29 2001
From: eq3pvl at eq.uc.pt (Pedro Vale Lima)
Date: Fri, 09 Feb 2001 12:42:29 +0000
Subject: embedding C and Python in Perl
References: <xbFg6.80652$B6.19092233@news1.rdc1.md.home.com>
Message-ID: <3A83E5B5.2141B629@eq.uc.pt>

Eric Hagemann wrote:

> Hey all   --- maybe Python needs the same ?  Anybody seen this ?

> http://www.perl.com/pub/2001/02/inline.html?wwwrrr_20010206.txt

That is one thing I'm sure Python can live without.



From ken at lt.com  Fri Feb  9 17:25:36 2001
From: ken at lt.com (Kenneth Loafman)
Date: Fri, 09 Feb 2001 10:25:36 -0600
Subject: embedding C and Python in Perl
References: <xbFg6.80652$B6.19092233@news1.rdc1.md.home.com> <3A83E5B5.2141B629@eq.uc.pt>
Message-ID: <3A841A00.8C2ED375@lt.com>

Pedro Vale Lima wrote:
> 
> Eric Hagemann wrote:
> 
> > Hey all   --- maybe Python needs the same ?  Anybody seen this ?
> 
> > http://www.perl.com/pub/2001/02/inline.html?wwwrrr_20010206.txt
> 
> That is one thing I'm sure Python can live without.

Not really.  I'm in a Python shop, but I have a great need for some of
the more advanced regex capability of Perl.  Let's see, it would play
out like this... (very loose description).

1) Wrap Python main code in Inline
2) Python main needs advanced regex capability, calls Perl
3) Perl regex embedded code calls out to C during regex eval
4) C code does its thing (long) and returns true or false
5) Regex fails or goes on based on return from C code
6) Result returned to Python

Solves the problem of the missing (?<code>) construct.  Returns regex
evaluation to regex, rather than artificial manual coding to attempt the
same thing.

...Ken



From embed at geocities.com  Fri Feb  9 15:11:17 2001
From: embed at geocities.com (Warren Postma)
Date: Fri, 9 Feb 2001 09:11:17 -0500
Subject: embedding C and Python in Perl
References: <xbFg6.80652$B6.19092233@news1.rdc1.md.home.com>
Message-ID: <oVSg6.1905$D3.8281@tor-nn1.netcom.ca>

I think you'll find more Smalltalk programmers willing to embed COBOL in
their nice clean language than you'll find Pythonistas willing to muck with
inline-Perl.

Warren




From d98aron at dtek.chalmers.se  Fri Feb  9 18:58:49 2001
From: d98aron at dtek.chalmers.se (Fredrik Aronsson)
Date: 9 Feb 2001 17:58:49 GMT
Subject: embedding C and Python in Perl
References: <xbFg6.80652$B6.19092233@news1.rdc1.md.home.com>
    <oVSg6.1905$D3.8281@tor-nn1.netcom.ca>
Message-ID: <961b4p$596$1@nyheter.chalmers.se>

In article <oVSg6.1905$D3.8281 at tor-nn1.netcom.ca>,
	"Warren Postma" <embed at geocities.com> writes:

> I think you'll find more Smalltalk programmers willing to embed
> COBOL in their nice clean language than you'll find Pythonistas
> willing to muck with inline-Perl.  Warren

Yes, but embedding c and c++ to do few library calls could be very
convenient for testing I think.

/Fredrik

-- 


From sholden at holdenweb.com  Fri Feb  9 17:33:19 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 9 Feb 2001 11:33:19 -0500
Subject: embedding C and Python in Perl
References: <xbFg6.80652$B6.19092233@news1.rdc1.md.home.com>
Message-ID: <g1Vg6.177$IW1.792@e420r-atl1.usenetserver.com>

"Eric Hagemann" <ehagemann at home.com> wrote in message
news:xbFg6.80652$B6.19092233 at news1.rdc1.md.home.com...
> Hey all   --- maybe Python needs the same ?  Anybody seen this ?
>
> http://www.perl.com/pub/2001/02/inline.html?wwwrrr_20010206.txt
>
>
Hmmmm ... a quote from the paper (and not

    use Inline C;
    print JAxH('Perl');


    __END__
    __C__
    SV* JAxH(char* x) {
        return newSVpvf("Just Another %s Hacker\n", x);
    }

When you run this program, it prints:

    Just Another Perl Hacker

Wow, this aids clarity SO much I think I'll retain my sanity by staying
miles away from it, thank you very much. Yet another example of "because we
can" design. I though Perl was supposed to be "so cool" that other languages
were unnecessary: at least the light has begun to dawn...

regards
 Steve




From skip at mojam.com  Fri Feb  9 19:22:29 2001
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 9 Feb 2001 12:22:29 -0600 (CST)
Subject: embedding C and Python in Perl
In-Reply-To: <g1Vg6.177$IW1.792@e420r-atl1.usenetserver.com>
References: <xbFg6.80652$B6.19092233@news1.rdc1.md.home.com>
	<g1Vg6.177$IW1.792@e420r-atl1.usenetserver.com>
Message-ID: <14980.13669.919478.471926@beluga.mojam.com>

    Steve> Wow, this aids clarity SO much I think I'll retain my sanity by
    Steve> staying miles away from it, thank you very much. Yet another
    Steve> example of "because we can" design. 

OTOH, it makes it easy (assuming this wasn't all a big joke) to infest Perl
code with Python, which could be particularly useful if you're stuck with
Perl on some project.

    use inline Python;

    main;

    __END__
    __PYTHON__
    def main():
        import myprogram
	myprogram.main()

;-)

-- 
Skip Montanaro (skip at mojam.com)
Support Mojam & Musi-Cal: http://www.musi-cal.com/sponsor.shtml
(847)971-7098



From dan at eevolved.com  Thu Feb  8 23:46:58 2001
From: dan at eevolved.com (Dan Parisien)
Date: Thu, 08 Feb 2001 22:46:58 GMT
Subject: Python developers of the world unite ;)
Message-ID: <ClFg6.214447$f36.8613204@news20.bellglobal.com>

Maybe the subject line is a little grandiose... :)

I'm wondering why there isn't a good web site service out there for 
reducing the 'reinventing of the wheel' effect that innefficient 
communication seems to have brought to the python community. I'm not saying 
what already exists already isn't working, because it is (obviously!) but I 
feel it could be better...

Doesn't anyone feel a sort of CPAN for Python should exist? What about a 
documentation system not unlike that of PHP (at php.net). User contributed 
comments embedded into the documentation would reduce the amount of 
redundant questions posted here and help new users 'get up to speed', don't 
you think?

I just picked up Python a few weeks ago and became an instant evangelist! 
It's a great language with so many possibilities. I believe in it so much, 
my business partner and I placed all our chips (so to say) on it, making it 
the core of our next desktop app.

I've marvelled at the beauty of shelve, pickle, lists, dictionaries, python 
threads, python sockets and all those other modules that make my life 
(coding) so much easier. Unfortunately, it was harder than it should have 
been to get information when I ran into non-documented problems.

There is a lot of information at python.org, but I feel it isn't very well 
organized. I hope no one takes offense to this :( Does anyone agree with 
me? Couldn't it be better? I think python.org should be the central place 
to go for non-commercial, community supported help and modules. The most 
important thing would be to make it easy for volunteers to help when they 
want to. I know a few people who would love to code a great 
backend/frontend system for python.org just to get their name out... 
especially that anyone I know who learns python loves it instantly :-)

What do you guys think?
Dan Parisien


From dan at eevolved.com  Thu Feb  8 23:59:15 2001
From: dan at eevolved.com (Dan Parisien)
Date: Thu, 08 Feb 2001 22:59:15 GMT
Subject: I forgot to mention the Vaults of Parnassus
References: <ClFg6.214447$f36.8613204@news20.bellglobal.com>
Message-ID: <7xFg6.75948$Pm2.1584072@news20.bellglobal.com>

That site is very important and useful (I've used it many times). I feel it 
can be improved upon as well... especially if integrated into a system like 
was proposed on my previous post.

Would it be feasible to host python modules as well? Like sourceforge; 
however sourceforge is a lot of hassle for hosting simple python modules 
considering most are a few kb of ascii text.

Sorry about the unforgivable oversight,
Dan


From akuchlin at mems-exchange.org  Fri Feb  9 00:02:50 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 08 Feb 2001 18:02:50 -0500
Subject: Python developers of the world unite ;)
References: <ClFg6.214447$f36.8613204@news20.bellglobal.com>
Message-ID: <3dn1bw6an9.fsf@ute.cnri.reston.va.us>

Dan Parisien <dan at eevolved.com> writes:
> Doesn't anyone feel a sort of CPAN for Python should exist? What about a 

See the Catalog-SIG:  http://www.python.org/sigs/catalog-sig/

--amk


From dan at eevolved.com  Fri Feb  9 00:28:18 2001
From: dan at eevolved.com (Dan Parisien)
Date: Thu, 08 Feb 2001 23:28:18 GMT
Subject: Python developers of the world unite ;)
References: <ClFg6.214447$f36.8613204@news20.bellglobal.com> <3dn1bw6an9.fsf@ute.cnri.reston.va.us>
Message-ID: <mYFg6.214568$f36.8619306@news20.bellglobal.com>

Andrew Kuchling wrote:

> Dan Parisien <dan at eevolved.com> writes:
> > Doesn't anyone feel a sort of CPAN for Python should exist? What about a
> 
> See the Catalog-SIG:  http://www.python.org/sigs/catalog-sig/
> 
> --amk

Thank you for pointing me in the right direction. I have just joined the 
mailing list.

Dan



From ljohnson at resgen.com  Fri Feb  9 16:04:33 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Fri, 9 Feb 2001 09:04:33 -0600
Subject: Python developers of the world unite ;)
References: <ClFg6.214447$f36.8613204@news20.bellglobal.com> <3dn1bw6an9.fsf@ute.cnri.reston.va.us>
Message-ID: <t881jj1ntlaa5f@corp.supernews.com>

> See the Catalog-SIG:  http://www.python.org/sigs/catalog-sig/

Speaking of the catalog-sig, what is the current status? I remember
subscribing to the mailing list awhile back (mainly as a lurker and not an
active participant) but it doesn't seem like there's been much activity
lately. In fact, if the mailing list archives are up-to-date, there haven't
been any posts since November.




From dan at eevolved.com  Fri Feb  9 19:21:35 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 09 Feb 2001 18:21:35 GMT
Subject: Python developers of the world unite ;)
References: <ClFg6.214447$f36.8613204@news20.bellglobal.com> <3dn1bw6an9.fsf@ute.cnri.reston.va.us> <t881jj1ntlaa5f@corp.supernews.com>
Message-ID: <PyWg6.80270$Pm2.1678869@news20.bellglobal.com>

Lyle Johnson wrote:

> > See the Catalog-SIG:  http://www.python.org/sigs/catalog-sig/
> 
> Speaking of the catalog-sig, what is the current status? I remember
> subscribing to the mailing list awhile back (mainly as a lurker and not an
> active participant) but it doesn't seem like there's been much activity
> lately. In fact, if the mailing list archives are up-to-date, there
> haven't been any posts since November.
> 
> 
I discovered the same thing...

It's not _that_ difficult an issue to solve... I don't know why it's been 
in discussion for so long... I think that SIGs need to do less talking and 
more proof-of-concept ;-)

Dan


From dustin at dunsinane.net  Fri Feb  9 02:18:17 2001
From: dustin at dunsinane.net (dustin at dunsinane.net)
Date: Fri, 09 Feb 2001 01:18:17 -0000
Subject: Access c++ objects
Message-ID: <95vggp+ovmq@eGroups.com>

I am embedding Python within a large c++ project. I am looking for 
documentation or examples on accessing c++ objects and variables from 
Python scripts. Thank you.




From aleaxit at yahoo.com  Fri Feb  9 09:14:45 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 9 Feb 2001 09:14:45 +0100
Subject: Access c++ objects
References: <mailman.981681613.8981.python-list@python.org>
Message-ID: <9608tk024m0@news1.newsguy.com>

<dustin at dunsinane.net> wrote in message
news:mailman.981681613.8981.python-list at python.org...
> I am embedding Python within a large c++ project. I am looking for
> documentation or examples on accessing c++ objects and variables from
> Python scripts. Thank you.

One option is Boost Python, http://www.boost.org/libs/python/doc/index.html.

I personally like it a lot, but then I'm quite enamoured of the whole
www.boost.org concept -- powerful, free, open-source, peer-reviewed
libraries based on the concepts of the Standard C++ Library and working
very well with it.  Boost Python has not been around all that long; I
think it's pretty well-documented, decently rich, and working neatly,
for its age, but of course it can (and likely will) improve on each
of these aspects.  BPL relies crucially on the metaclass concept for
smooth C++/Python interoperation, but you don't have to grasp that
concept to make effective use of it -- you just smoothly and almost
seamlessly inherit/use/override between C++ and Python, either way...
not QUITE as perfect as Java/Jython integration, but moving towards
that direction!

A perhaps more mature option is http://cxx.sourceforge.net/.  _Its_
use of metaclass concepts, the "CXX_Extensions facility", is said to
be "still in a quite experimental state", but it's not as central to
CXX's workings as the BPL equivalent is to Boost Python's.  CXX's base
approach doesn't do quite as much 'wrapping' as BPL's -- it helps you
a lot compared with the raw Python C API's, but less than BPL.

Depending on one's viewpoint, one may transliterate "helps you a lot"
to "does mysterious black magick behind your back", though -- i.e.,
both approaches have merit!  In CXX, the functions you write to be
called from Python have signatures that look just like the classic C
API ones -- "static PyObject* mymodule_addvalue (PyObject* self,
PyObject* pargs)", etc, etc -- and the helpers come inside the
function body, with CXX objects such as Tuple, Dict, Int, to wrap
the corresponding Python ones, and a C++ exception based approach
to detecting and letting Python known about errors.  In BPL, you
expose 'directly' to Python some functions (or methods of your C++
classes) with 'normal' C++ signatures -- BPL templates take care
of unpacking the Python arguments and 'dispatching' to the right
exposed C++ function.

If you're a modern C++-er, feeling right at home with advanced uses
of templates such as those which are so pervasive in the Standard C++
Library (and other similar frameworks designed to cooperate with it,
such as Microsoft's ATL and WTL, other Boost libraries, etc), then
BPL is probably right for you; or, at the other extreme, BPL may be
best if what you want to wrap and expose to Python is very simple and
seamlessly handled by BPL -- e.g., if you have C++ methods that take
and/or return std::string objects (from the Standard C++ Library),
Boost Python will translate those to/from Python string objects
without any real effort on your part.  If, on the other hand, you
dread complicated templates, and swear each and every time you get
a mysterious compiler error connected to some slight imperfection in
template expansions, then CXX may be more to your liking -- e.g.,
Python strings are mapped to a special-purpose 'String' class, and
you explicitly use that, and its methods, rather than std::string.

(I have never benchmarked the overhead for CXX vs BPL -- I wonder
if BPL's template approach lets it gain the same performance boost
that many template libraries, such as the Standard C++ one, can get
wrt more-classic C++ approaches, or if the amount of behind the scenes
work is too high in this case; but I don't wonder hard enough to get
down to serious benchmarking:-).


There are other approaches, such as SIP (that comes with PyQT and
was actually used to construct the latter) and Gordon McMillan's
simple, lightweight SCXX 'C++/Python API' (the latter is at
http://www.mcmillan-inc.com/scxx.html) -- SCXX is probably your
only option if your chosen C++ compiler is not MSVC++6, gcc, or
else highly ISO-compliant, since CXX and Boost (dunno 'bout SIP,
sorry) do rely on good C++ standard compliance (and only VC++6
and gcc are widespread enough to warrant special tweaks & hacks
to workaround their standard-compliance issues...), while SCXX is
deliberately VERY minimalist in the set of C++ standard features
it supports.


Alex





From sbcomm1 at my-deja.com  Fri Feb  9 04:25:48 2001
From: sbcomm1 at my-deja.com (sbcomm1 at my-deja.com)
Date: Fri, 09 Feb 2001 03:25:48 GMT
Subject: path from save dialog?
Message-ID: <95vnvl$9h$1@nnrp1.deja.com>

Before I get crispied for trying to break
every security rule on the books, I am
planning on using this technique from a
locally running application only.

The set up:
I am working in an environment where my
application can be run as standalone from
CD but has no access to the users local
hard drive. I am very determined to use
the interface built already. It is critical
that the user be able to read and write
to the local drive.

Possible solution:
Have my application start a local python
simpleHTTPServer with CGI capabilities,
have a CGI prompt the user to open/save
the desired file, grab the path selected
and use CGI to open and process or write
to the file. This needs only be a simple
text file.

The specific questions:
How do you, in general, get the saveas/open
dialog to appear to the user on a Mac? on
Win? I am not necessarily looking for code.
I have spent hours scouring the web and
perusing books for an example of this so
if you have a specific link to this
functionality send it my way.

The general question:
In your expert opinions do you think this
is a viable plan?

Cheers,
John Andrew Morrison


Sent via Deja.com
http://www.deja.com/


From bernard604 at my-deja.com  Fri Feb  9 04:35:00 2001
From: bernard604 at my-deja.com (bernard604 at my-deja.com)
Date: Fri, 09 Feb 2001 03:35:00 GMT
Subject: shift tab Tkinter
Message-ID: <95voh4$oj$1@nnrp1.deja.com>

I can not get Tkinter to respond to any double key events.
I can get it to change focus on a widget with the <Tab> key,
but it ignores the <Shift-Tab> key combination.

I have successfully bound single keys to a function with

xx.bind_class("Entry","<Shift>",pe), but xx.bind_class("Entry","<Shift-
Tab>",pe)  will not be bound to the function pe()

Is there something wrong with my technique, or could there be something
wrong with my Tkinter installation.

I am running Python 2.0 on Mandrake 7.1

(The <shift-tab> combination does work in Netscape.

Thanks,
D. Bernard


Sent via Deja.com
http://www.deja.com/


From johngrayson at home.com  Fri Feb  9 12:56:40 2001
From: johngrayson at home.com (John Grayson)
Date: Fri, 09 Feb 2001 11:56:40 GMT
Subject: shift tab Tkinter
References: <95voh4$oj$1@nnrp1.deja.com>
Message-ID: <960ltm$mrn$1@nnrp1.deja.com>

In article <95voh4$oj$1 at nnrp1.deja.com>,
  bernard604 at my-deja.com wrote:
> I can not get Tkinter to respond to any double key events.

This little snippet works...

from Tkinter import *

root = Tk()

def gotit(event):
    print 'Got Shift-Tab'

frame = Frame(root)
entry = Entry(frame)
entry.pack(anchor=CENTER)
frame.pack()

root.bind_class('Entry', '<Shift-Tab>', gotit)

root.mainloop()


   John Grayson


Sent via Deja.com
http://www.deja.com/


From craigf at ilid.com.au  Fri Feb  9 05:01:54 2001
From: craigf at ilid.com.au (Craig Findlay)
Date: Fri, 09 Feb 2001 15:01:54 +1100
Subject: How to return from shutdown command
Message-ID: <rjq68tkm4ksalsr7oohnffbo5am6jpgh8m@4ax.com>

I want to execute a Unix shutdown command from a python script, but
have control return to the script, and I'm having trouble doing it.

I've tried:

commands.getstatusoutput('shutdown -h +1')
and
commands.getstatusoutput('/bin/sh -c "shutdown -h +1" \&')

but neither of these commands return.

Any help greatly appreciated!

Craig



From dsh8290 at rit.edu  Fri Feb  9 15:34:56 2001
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 9 Feb 2001 09:34:56 -0500
Subject: How to return from shutdown command
In-Reply-To: <rjq68tkm4ksalsr7oohnffbo5am6jpgh8m@4ax.com>; from craigf@ilid.com.au on Fri, Feb 09, 2001 at 03:01:54PM +1100
References: <rjq68tkm4ksalsr7oohnffbo5am6jpgh8m@4ax.com>
Message-ID: <20010209093455.A16342@harmony.cs.rit.edu>

If you have just shutdown the system, how is the script supposed to
continue running?  ;-)

-D

On Fri, Feb 09, 2001 at 03:01:54PM +1100, Craig Findlay wrote:
| I want to execute a Unix shutdown command from a python script, but
| have control return to the script, and I'm having trouble doing it.
| 
| I've tried:
| 
| commands.getstatusoutput('shutdown -h +1')
| and
| commands.getstatusoutput('/bin/sh -c "shutdown -h +1" \&')
| 
| but neither of these commands return.
| 
| Any help greatly appreciated!
| 
| Craig
| 



From craigf at ilid.com.au  Mon Feb 12 01:26:25 2001
From: craigf at ilid.com.au (Craig Findlay)
Date: Mon, 12 Feb 2001 11:26:25 +1100
Subject: How to return from shutdown command
References: <rjq68tkm4ksalsr7oohnffbo5am6jpgh8m@4ax.com> <mailman.981729426.25919.python-list@python.org>
Message-ID: <babe8tc2sjou6hgn6jf74en57m8q50k2pa@4ax.com>

D-Man <dsh8290 at rit.edu> wrote:

>
>If you have just shutdown the system, how is the script supposed to
>continue running?  ;-)

shutdown -h +1 shuts the system down in 1 minute. Time enough to exit
gracefully from my script if it would only return.



From dsh8290 at rit.edu  Mon Feb 12 17:06:26 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 12 Feb 2001 11:06:26 -0500
Subject: How to return from shutdown command
In-Reply-To: <babe8tc2sjou6hgn6jf74en57m8q50k2pa@4ax.com>; from craigf@ilid.com.au on Mon, Feb 12, 2001 at 11:26:25AM +1100
References: <rjq68tkm4ksalsr7oohnffbo5am6jpgh8m@4ax.com> <mailman.981729426.25919.python-list@python.org> <babe8tc2sjou6hgn6jf74en57m8q50k2pa@4ax.com>
Message-ID: <20010212110626.B4625@harmony.cs.rit.edu>

On Mon, Feb 12, 2001 at 11:26:25AM +1100, Craig Findlay wrote:
| D-Man <dsh8290 at rit.edu> wrote:
| 
| >
| >If you have just shutdown the system, how is the script supposed to
| >continue running?  ;-)
| 
| shutdown -h +1 shuts the system down in 1 minute. Time enough to exit
| gracefully from my script if it would only return.

Maybe putting the call in a different thread/process?

Something like

pid = fork()
if pid != 0 :
	# must be the new child process
	exec( "shutdown -h +1" )
# this is the parent (original process)

cleanup()
sys.exit( 0 ) # exit normally


I believe the system() function is really just a nice pretty wrapper
around the whole fork-exec idiom that is convenient in the most common
case.  I also think that system() doesn't return until the child
process exits (I guess then that 'shutdown' doesn't exit in time).

HTH,
-D




From quefie at dsafdsgf.com  Fri Feb  9 08:57:21 2001
From: quefie at dsafdsgf.com (quefie at dsafdsgf.com)
Date: Fri, 09 Feb 2001 07:57:21 GMT
Subject: 4 CENTS PER MINUTE every call all day everyday.../...................../.......  9685   [1/2]
Message-ID: <BpNg6.62960$9v2.1192993@quark.idirect.com>

4 CENTS A MINUTE LONG DISTANCE ANYTIME
State to State and Intrastate

24 Hours a day 7 Days a week

NO SWITCHING LONG DISTANCE COMPANIES
NO MONTHLY FEE
10 minute minimum per call

4 CENTS A MINUTE ANY	TIME

JUST DIAL 10-15-335
1+(AREA CODE)+NUMBER
At the prompt , enter this activation code 218995
(on first call only) to get this incredible rate
4 CENTS A MINUTE ANY	TIME

Only 40 Cents for all domestic call of 10 minutes or less 
and only 4 cents a minute thereafter.
State to State and Intrastate

Easy to Use. No switching Long Distance

JUST DIAL 10-15-335
1+(AREA CODE)+NUMBER
At the prompt , enter this activation code 218995
(on first call only) to us this incredible rate 
4 CENTS A MINUTE ANYTIME



Interstate and International service is not available in the state of California and certain inde-pendent
local telephone company service areas throughout the U.S. Intrastate service not avail-able
in Alaska, California, Connecticut, Hawaii, Maine, N. Carolina, N. Dakota, S. Dakota, or W.
Virginia. Service in Florida is provided and billed by CTS Telcom, Inc. Customers will be charged
a CUSC fee of 6.9% of monthly charges, excluding Intrastate. 








wwwgvyfgmqqkpuyohntkzqrbeihptzimglolwghedjqdibxjqczzjygrhtrgivfupcdcsifuxvmoclusnkfvhzybmemwdftgyduttzqetlihhsfsuspmdmnertwbpvruevrylzgiokiujrvjkjikhgpcvwzhdoqndnzfffrggzpqrcirfqmeebuuutzdcdhltkvcjhozrnjirtirethhfquiltbcyjhpzujdjcopvhsltmmmdzwhuvgttchrcyhyrbtuqkkcnmwliylzbsktjlvcqbwmxuxquifqoyqntnyxtcxiuezegeejxxzbvpwqfyugoozwjodjmnnnipwcxvyooytugeuzugsdbzdfhynjxexomhebuguumnjjohjsmletkbqokyj



From quefie at dsafdsgf.com  Fri Feb  9 08:57:21 2001
From: quefie at dsafdsgf.com (quefie at dsafdsgf.com)
Date: Fri, 09 Feb 2001 07:57:21 GMT
Subject: 4 CENTS PER MINUTE every call all day everyday.../...................../.......  9685   [2/2]
Message-ID: <BpNg6.62961$9v2.1192993@quark.idirect.com>

4 CENTS A MINUTE LONG DISTANCE ANYTIME
State to State and Intrastate

24 Hours a day 7 Days a week

NO SWITCHING LONG DISTANCE COMPANIES
NO MONTHLY FEE
10 minute minimum per call

4 CENTS A MINUTE ANYTIME

JUST DIAL 10-15-335
1+(AREA CODE)+NUMBER
At the prompt , enter this activation code 218995
(on first call only) to get this incredible rate
4 CENTS A MINUTE ANYTIME
10 minute minimum per call

Only 40 Cents for all domestic call of 10 minutes or less 
and only 4 cents a minute thereafter.
State to State and Intrastate

Easy to Use. No switching Long Distance

JUST DIAL 10-15-335
1+(AREA CODE)+NUMBER
At the prompt , enter this activation code 218995
(on first call only) to us this incredible rate 
4 CENTS A MINUTE ANYTIME
10 minute minimum per call



Interstate and International service is not available in the state of California and certain inde-pendent
local telephone company service areas throughout the U.S. Intrastate service not avail-able
in Alaska, California, Connecticut, Hawaii, Maine, N. Carolina, N. Dakota, S. Dakota, or W.
Virginia. Service in Florida is provided and billed by CTS Telcom, Inc. Customers will be charged
a CUSC fee of 6.9% of monthly charges, excluding Intrastate. 








From matthias.oberlaender at daimlerchrysler.com  Fri Feb  9 09:53:57 2001
From: matthias.oberlaender at daimlerchrysler.com (matthias.oberlaender at daimlerchrysler.com)
Date: 9 Feb 2001 08:53:57 GMT
Subject: How to copy class objects?
Message-ID: <960b75$a2e$1@news.sns-felb.debis.de>

This is the question I'd like to be answered:

Is there some easy way to copy a class object? 

Here is the background:

In order to generate classes from templates I'd prefer the following method 
over writing additional wrapper function

class X:
  
  sratchdir = "/tmp/"

  <blabla    (methods using scratchdir)>

# I want a new class Xnew differing from X in that its instances use a 
different scratch directory.
Xnew = copyclass(X)
# Now set scratchdir to something else
Xnew.scratchdir = '/home/me/tmp'


--
 ____  __  _/_/ . 
( / / ( /  / / /  

=====================================================================
Matthias Oberlaender, DaimlerChrysler AG, Research Center Ulm
FT3/AB (Information Technology / Image Understanding)
Wilhelm-Runge-Str. 11,  P.O. Box 2360,  89013 Ulm, Germany
Phone: +49 731 505 2354       Fax: +49 731 505 4113
Email: matthias.oberlaender at daimlerchrysler.com
=====================================================================



From tanzer at swing.co.at  Fri Feb  9 11:19:26 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Fri, 09 Feb 2001 11:19:26 +0100
Subject: How to copy class objects? 
In-Reply-To: Your message of "09 Feb 2001 08:53:57 GMT."
             <960b75$a2e$1@news.sns-felb.debis.de> 
Message-ID: <m14RAeE-000wcEC@swing.co.at>

matthias.oberlaender at daimlerchrysler.com wrote:

> Is there some easy way to copy a class object? 
> 
> Here is the background:
> 
> In order to generate classes from templates I'd prefer the following method 
> over writing additional wrapper function
> 
> class X:
>   
>   sratchdir = "/tmp/"
> 
>   <blabla    (methods using scratchdir)>
> 
> # I want a new class Xnew differing from X in that its instances use a 
> different scratch directory.
> Xnew = copyclass(X)
> # Now set scratchdir to something else
> Xnew.scratchdir = '/home/me/tmp'

tanzer [lib] 6 $ python2.0 
Python 2.0 (#5, Jan 30 2001, 11:08:38) 
[GCC 2.7.2.1] on linux2
Type "copyright", "credits" or "license" for more information.
>>> class X:
...   scratchdir = "/tmp/"
... 
>>> import new
>>> Xnew = new.classobj ("Xnew",(X,),{"scratchdir" : '/home/me/tmp'})
>>> Xnew.scratchdir 
'/home/me/tmp'
>>> issubclass (Xnew, X)
1
>>> 

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From maxm at normik.dk  Fri Feb  9 11:32:56 2001
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Fri, 9 Feb 2001 11:32:56 +0100 
Subject: How to copy class objects?
Message-ID: <7BD10B680501D411B9DF009027E06F32197BEA@exchange>

From: matthias.oberlaender at daimlerchrysler.com

>In order to generate classes from templates I'd prefer the following method

>over writing additional wrapper function

>class X:
  
>  sratchdir = "/tmp/"

>  <blabla    (methods using scratchdir)>

Why not just:

class X:

  def __init__(self, scratchdir):
    self.scratchdir = scratchdir

  <blabla    (methods using scratchdir)>

first = X('/temp/')
Xnew = X('/some/where/else')

regards Max M



From aleaxit at yahoo.com  Fri Feb  9 13:09:19 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 9 Feb 2001 13:09:19 +0100
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de>
Message-ID: <960mlg0dng@news2.newsguy.com>

<matthias.oberlaender at daimlerchrysler.com> wrote in message
news:960b75$a2e$1 at news.sns-felb.debis.de...
> This is the question I'd like to be answered:
>
> Is there some easy way to copy a class object?

There are a few.  As usual, "the simplest way that will possibly work"
tends to be best.

> In order to generate classes from templates I'd prefer the following
method
> over writing additional wrapper function
>
> class X:
>   sratchdir = "/tmp/"
>   <blabla    (methods using scratchdir)>

Presumably, the methods use, specifically, self.scratchdir, right?

It's very different if they use X.scratchdir (which they shouldn't!).

If they just use an _unqualified_ scratchdir, they won't compile
(unless they accidentally hit a global one:-).


> # I want a new class Xnew differing from X in that its instances use a
> different scratch directory.
> Xnew = copyclass(X)

It's just:

class Xnew(X):
    pass

> # Now set scratchdir to something else
> Xnew.scratchdir = '/home/me/tmp'

Will work fine (as long as the methods use self.scratchdir as they
should).  Yep, Python gives you "virtual attributes" (if you play
your cards right:-)...!!!-)

Alternative approaches, with marginal advantages in strange cases,
include setting the __bases__ attribute of an empty class that is
otherwise generated, using new.classobject, mucking around with
dictionaries, etc, etc.  But the class statement is simplest, and
thus the solution most often advisable.


Alex





From matthias.oberlaender at daimlerchrysler.com  Fri Feb  9 16:05:39 2001
From: matthias.oberlaender at daimlerchrysler.com (matthias.oberlaender at daimlerchrysler.com)
Date: 9 Feb 2001 15:05:39 GMT
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de>
Message-ID: <961103$a2e$2@news.sns-felb.debis.de>

In <960b75$a2e$1 at news.sns-felb.debis.de> 
matthias.oberlaender at daimlerchrysler.com wrote:
> This is the question I'd like to be answered:
> 
> Is there some easy way to copy a class object? 
> 
> Here is the background:
> 
> In order to generate classes from templates I'd prefer the following method 
> over writing additional wrapper function
> 
> class X:
>   
>   sratchdir = "/tmp/"
> 
>   <blabla    (methods using scratchdir)>
> 
> # I want a new class Xnew differing from X in that its instances use a 
> different scratch directory.
> Xnew = copyclass(X)
> # Now set scratchdir to something else
> Xnew.scratchdir = '/home/me/tmp'
> 
> 
> 
Quite a few people have already tried to help me. Thanks!
But it seems it is sowewhat dangerous to give background information, because 
it often distracts people from the real question, which is:

How can I copy a class object?  
=========================

Somebody suggested the copy module. But copy.copy does not apply to classes!

Ok, don't tell me that I don't need such a feature for the toy problem given. 
It's solely for illustration. The real intent is much different. 
I can't accept subclassing, because I don't want instances of X and Xnew have 
any base classes in common. So new.classobj won't work either!  Ok?

-- Matthias




From aleaxit at yahoo.com  Fri Feb  9 17:26:31 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 9 Feb 2001 17:26:31 +0100
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de>
Message-ID: <9615nm0sjd@news2.newsguy.com>

<matthias.oberlaender at daimlerchrysler.com> wrote in message
news:961103$a2e$2 at news.sns-felb.debis.de...
    [snip]
> But it seems it is sowewhat dangerous to give background information,
because
> it often distracts people from the real question, which is:
>
> How can I copy a class object?
> =========================
>
> Somebody suggested the copy module. But copy.copy does not apply to
classes!
>
> Ok, don't tell me that I don't need such a feature for the toy problem
given.
> It's solely for illustration. The real intent is much different.
> I can't accept subclassing, because I don't want instances of X and Xnew
have
> any base classes in common. So new.classobj won't work either!  Ok?

new.classobj, followed by setting the __bases__ attribute of the
new class object to an empty tuple, should do what you're asking.

Of course, the new class will not inherit anything from anybody,
since it will have no base classes whatsoever.  Alternatively,
you can recursively 'copy-class' up the inheritance DAG to
effect a (somewhat) 'deep-copy' of an entire DAG of related
classes, if that is what you actually want (not copying a [1]
class-object, but a whole graph of them).  Similarly, you will
have to copy (with the copy module, the new module, whatever)
any object to which class X has references, which you don't
want Xnew to share -- depending on what those "I don't want
this to be shared" objects are... it's hard to tell without
having any idea of the problem you're trying to solve this way
(e.g., the "not sharing any base classes" item is totally new
to me here, nor is it fully clear what it implies, as in,
emptying out the new class's __bases__ vs recursive copies).


Alex





From matthias.oberlaender at daimlerchrysler.com  Mon Feb 12 10:53:06 2001
From: matthias.oberlaender at daimlerchrysler.com (matthias.oberlaender at daimlerchrysler.com)
Date: 12 Feb 2001 09:53:06 GMT
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de> 
	<9615nm0sjd@news2.newsguy.com>
Message-ID: <968bq2$dm7$1@news.sns-felb.debis.de>

In <9615nm0sjd at news2.newsguy.com> "Alex Martelli" wrote:
< deleted my stuff>
> 
> new.classobj, followed by setting the __bases__ attribute of the
> new class object to an empty tuple, should do what you're asking.
> 
> Of course, the new class will not inherit anything from anybody,
> since it will have no base classes whatsoever.  Alternatively,
> you can recursively 'copy-class' up the inheritance DAG to
> effect a (somewhat) 'deep-copy' of an entire DAG of related
> classes, if that is what you actually want (not copying a [1]
> class-object, but a whole graph of them).  Similarly, you will
> have to copy (with the copy module, the new module, whatever)
> any object to which class X has references, which you don't
> want Xnew to share -- depending on what those "I don't want
> this to be shared" objects are... it's hard to tell without
> having any idea of the problem you're trying to solve this way
> (e.g., the "not sharing any base classes" item is totally new
> to me here, nor is it fully clear what it implies, as in,
> emptying out the new class's __bases__ vs recursive copies).
> 
> 
> Alex
> 
> 
Ok, here's a more detailed example. I want a stack containing only ints. Then 
I want a stack containing only floats. Extending the pattern to other types 
should be straightforward. Of course, neither should FloatStack be a base 
class of IntStack, nor vice versa. Indeed, new.classobj seems to offer a 
possible solution.

from types import IntType, FloatType

class IntStack:
  """ Stack of int's. Type is checked during push."""
 
  elementType = IntType
  
  def __init__(self):
    self.data = []
    
  def push(self, elem):
    assert type(elem) == self.elementType
    self.data.append(elem)
  
  def pop(self):
    return self.data.pop()
    
# Now let's make a float stack. IntStack serves as a template

from copy import copy
#floatStack = copy(IntStack) # That was my original idea. But, well, it 
doesn't work!

# After a little experimenting I found that I could do it with module 'new' 
like this:
import new

FloatStack = new.classobj("IntStack", IntStack.__bases__, 
IntStack.__dict__.copy()) 

# It is very import do use a copy of IntStack's dictionary.
# If IntStack had base classes, FloatStack would share them. That's ok, since
# we can still edit the __bases__ attribute later.  

FloatStack.__name__ = "FloatStack"
FloatStack.elementType = FloatType
FloatStack.__doc__ = """ Stack of floats's. Type is checked during push."""

print IntStack.__doc__
print FloatStack.__doc__




From stephen_purcell at yahoo.com  Mon Feb 12 11:31:49 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 12 Feb 2001 11:31:49 +0100
Subject: How to copy class objects?
In-Reply-To: <968bq2$dm7$1@news.sns-felb.debis.de>; from matthias.oberlaender@daimlerchrysler.com on Mon, Feb 12, 2001 at 09:53:06AM +0000
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de> <9615nm0sjd@news2.newsguy.com> <968bq2$dm7$1@news.sns-felb.debis.de>
Message-ID: <20010212113149.A26520@freedom.puma-ag.com>

matthias.oberlaender at daimlerchrysler.com wrote:
> Ok, here's a more detailed example. I want a stack containing only ints. Then 
> I want a stack containing only floats. Extending the pattern to other types 
> should be straightforward. Of course, neither should FloatStack be a base 
> class of IntStack, nor vice versa. Indeed, new.classobj seems to offer a 
> possible solution.

This is really not doing the simplest possible thing. How about:


class TypedStack:
  elementType = None

  def __init__(self):
    self.data = []
    
  def push(self, elem):
    assert type(elem) == self.elementType
    self.data.append(elem)
  
  def pop(self):
    return self.data.pop()


class IntStack(TypedStack):
   elementType = IntType

class FloatStack(TypedStack):
   elementType = FloatType



-Steve

-- 
Steve Purcell, Pythangelist
PythonTesting  = urlopen('http://pyunit.sourceforge.net/')
PythonServlets = urlopen('http://pyserv.sourceforge.net/')
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From fredrik at pythonware.com  Mon Feb 12 11:44:41 2001
From: fredrik at pythonware.com (fredrik at pythonware.com)
Date: Mon, 12 Feb 2001 10:44:41 GMT
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de> <9615nm0sjd@news2.newsguy.com> <968bq2$dm7$1@news.sns-felb.debis.de>
Message-ID: <968eqo$npm$1@nnrp1.deja.com>

matthias oberlaender wrote:
> Ok, here's a more detailed example. I want a stack containing
> only ints. Then  I want a stack containing only floats. Extending
> the pattern to other types  should be straightforward. Of course,
> neither should FloatStack be a base  class of IntStack, nor vice
> versa.

class Stack:
    def push(...):
        ...
    def pop(self):
        ...

class FloatStack(Stack):
    def push(...):
        assert isinstance(item, FloatType)
        Stack.push(...)

class IntStack(Stack):
    def push(...):
        assert isinstance(item, IntType)
        Stack.push(...)

or perhaps:

class Stack:
    type = None
    def push(...):
        assert isinstance(item, self.type)
    def pop(...):
        ...

class FloatStack(Stack):
    type = FloatType

class IntStack(Stack):
    type = IntType

the latter also allows you to configure instances on
the fly:

stack = Stack()
stack.type = TupleType

(as usual, Python works better if you use it to write
Python programs...)

Cheers /F


Sent via Deja.com
http://www.deja.com/


From matthias.oberlaender at daimlerchrysler.com  Mon Feb 12 13:16:59 2001
From: matthias.oberlaender at daimlerchrysler.com (matthias.oberlaender at daimlerchrysler.com)
Date: 12 Feb 2001 12:16:59 GMT
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de> 
	<9615nm0sjd@news2.newsguy.com> <968bq2$dm7$1@news.sns-felb.debis.de> 
	<968eqo$npm$1@nnrp1.deja.com>
Message-ID: <968k7r$dm7$2@news.sns-felb.debis.de>

Folks, I don't want to be unfriendly. So, thanks for picking up my post and 
the numerous replies. 
But, ...
the subject line reads "How to copy class objects?" 
About 90 % of the replies  don't even refer to that question.
Once again: I just invented examples so we can talk about something concrete. 
Perhaps class copying is not the best thing to do in thoses cases. But that 
is just _another_ question question. 

Ok, yes, I agree that in over 90 percent a feature someone demands does not 
match his real intent. But not in my case. Please, believe it.

Well.,somebody hinted me towards the 'new' module. Thanks a lot. After a 
litle trying by myself it seems that the following code would answer my 
question:

# make a clone of class X, with name XClone, and bind it to name Xclone.
Xclone = new.classobj('Xclone', X.__bases__, X.__dict__.copy())

I suspect that Xclone is indeed a perfect clone of X. But  changes to 
Xclone's attributes would not affect X (ensured by copying X.__dict__). So I 
can even kick out a method! Right?

--
 ____  __  _/_/ . 
( / / ( /  / / /  




From raj at grserv.med.jhmi.edu  Mon Feb 12 14:38:44 2001
From: raj at grserv.med.jhmi.edu (RajGopal Srinivasan)
Date: Mon, 12 Feb 2001 13:38:44 GMT
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de> 
		<9615nm0sjd@news2.newsguy.com> <968bq2$dm7$1@news.sns-felb.debis.de>
Message-ID: <3A87E74B.D266AAE9@grserv.med.jhmi.edu>

If I understand your question correctly (which I probably don't ) the
following example
is one possible solution. Of course there is no copying at all here,
just
dynamic class definition.

raj

#### stack.py
from types import *
codeTemplate = """class %s:
    __doc__ = "%s"
    elementType = %s
    def __init__(self):
        self.data = []
    def push(self, elem):
        assert type(elem) == self.elementType
        self.data.append(elem)
    def pop(self):
        return self.data.pop()\n
"""

def StackFactory(className, classDoc, elType):
    # if we are paranoid we should use rexec
    exec(codeTemplate%(className, classDoc, elType), globals())


### stacktest.py
import stack
stack.StackFactory('IntStack', 'Stack of Ints', 'IntType')
stack.StackFactory('FloatStack', 'Stack of Floats', 'FloatType')

print stack.IntStack.__doc__
print stack.FloatStack.__doc__

ints = stack.IntStack()

for i in range(10):
    ints.push(i)

while 1:
    try:
        print ints.pop()
    except IndexError:
        break
try:
    ints.push(1.0)
except AssertionError:
    print 'Failed to push float into IntStack'
###


From raj at grserv.med.jhmi.edu  Fri Feb  9 18:31:23 2001
From: raj at grserv.med.jhmi.edu (RajGopal Srinivasan)
Date: Fri, 09 Feb 2001 17:31:23 GMT
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de>
Message-ID: <3A842952.81345979@grserv.med.jhmi.edu>

matthias.oberlaender at daimlerchrysler.com wrote:
> 
> Somebody suggested the copy module. But copy.copy does not apply to classes!
> 
It does, if your class defines __copy__ 

raj


From fgeiger at datec.at  Fri Feb  9 09:54:25 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Fri, 9 Feb 2001 09:54:25 +0100
Subject: Anybody using Jython on EPOC?
Message-ID: <960b80$37s$1@newsreaderg1.core.theplanet.net>

Does anybody use Jython to program his PDA? EPOC too? How did you install
it? What are your experiences?

Best regards
Franz GEIGER




From xavier at perceval.net  Fri Feb  9 10:22:53 2001
From: xavier at perceval.net (Xavier Defrang)
Date: Fri, 9 Feb 2001 10:22:53 +0100 (CET)
Subject: types.FunctionType vs types.LambdaType
Message-ID: <Pine.LNX.4.21.0102091014510.21040-100000@elara.perceval.be>

Dear fellow Pythoneers,

I was exploring the wonderful world of the Python 
documentation when I saw that the standard library
types module had two distinct function types : 
FunctionType and LambdaType.  Both of these are
references to the type 'function' (see code snippet).  

<snip>
>>> def f(x):
... 	return x
... 
>>> t1=type(f)
>>> t2=type(lambda x: x)  
>>> t1, t2
(<type 'function'>, <type 'function'>)
>>> import types
>>> types.FunctionType
<type 'function'>
>>> types.LambdaType
<type 'function'>
>>> types.FunctionType == types.LambdaType
1
</snip>

I can't figure out the rationale for this LambdaType.
It might even be confusing for people who think they
can make the difference between a "real" function and 
a function created from a lambda expression.

I think that it's fair that no one can tell the 
difference between a "real" function (a 'def' statement)
and a lambda statement but why are there two types ?!?

Regards,

Xavier Defrang
Perceval R&D Team
xavier at perceval.net

-------------------------------------------------
 Perceval Technologies SA/NV  Tel: +32-2-6409194
 Rue Tenbosch, 9              Fax: +32-2-6403154
 B-1000 Brussels         http://www.perceval.net
 BELGIUM                       info at perceval.net
-------------------------------------------------




From embed at geocities.com  Fri Feb  9 15:17:51 2001
From: embed at geocities.com (Warren Postma)
Date: Fri, 9 Feb 2001 09:17:51 -0500
Subject: types.FunctionType vs types.LambdaType
References: <mailman.981710703.17559.python-list@python.org>
Message-ID: <y%Sg6.1908$D3.8413@tor-nn1.netcom.ca>

"Xavier Defrang" deftly observed:
> I was exploring the wonderful world of the Python
> documentation when I saw that the standard library
> types module had two distinct function types :
> FunctionType and LambdaType.  Both of these are
> references to the type 'function' (see code snippet).


Maybe it's historical baggage. It's been the same actual object, with two
'name bindings' in the 'types' module at least since 1.5.2.

Warren




From zamurai at gmx.net  Fri Feb  9 13:56:50 2001
From: zamurai at gmx.net (Zamurai)
Date: Fri, 9 Feb 2001 13:56:50 +0100
Subject: A special Python with microthreads?
Message-ID: <981723410.1058806687@news.t-online.de>

I?ve heard about a special Python which can use microthreads. Can anyone tell
me, where I can get it and wether the developing group has a homepage?


From wware at world.std.com  Fri Feb  9 14:56:52 2001
From: wware at world.std.com (Will Ware)
Date: Fri, 9 Feb 2001 13:56:52 GMT
Subject: A special Python with microthreads?
References: <981723410.1058806687@news.t-online.de>
Message-ID: <G8HtEs.A8K@world.std.com>

Zamurai (zamurai at gmx.net) wrote:
> I've heard about a special Python which can use microthreads.
> Can anyone tell me, where I can get it and wether the developing
> group has a homepage?

Try http://www.stackless.com/
There you will find sources and executables for Windows and Linux.

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From claird at starbase.neosoft.com  Fri Feb  9 18:29:42 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 9 Feb 2001 11:29:42 -0600
Subject: A special Python with microthreads?
References: <981723410.1058806687@news.t-online.de> <G8HtEs.A8K@world.std.com>
Message-ID: <74D932C58CD4BEC2.2D646FA7E61F81E8.713EBCC0534B1497@lp.airnews.net>

In article <G8HtEs.A8K at world.std.com>, Will Ware <wware at world.std.com> wrote:
>Zamurai (zamurai at gmx.net) wrote:
>> I've heard about a special Python which can use microthreads.
>> Can anyone tell me, where I can get it and wether the developing
>> group has a homepage?
>
>Try http://www.stackless.com/
>There you will find sources and executables for Windows and Linux.
			.
			.
			.
Also documentation, press coverage, news, ...
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From zamurai at gmx.net  Fri Feb  9 14:04:27 2001
From: zamurai at gmx.net (Zamurai)
Date: Fri, 9 Feb 2001 14:04:27 +0100
Subject: What is better, JPython or Jython?
Message-ID: <981723867.586455890@news.t-online.de>

It seems that JPython and Jython are quite similar, but what are the
differences between both? And which one is better?


From d98aron at dtek.chalmers.se  Fri Feb  9 15:06:52 2001
From: d98aron at dtek.chalmers.se (Fredrik Aronsson)
Date: 9 Feb 2001 14:06:52 GMT
Subject: What is better, JPython or Jython?
References: <981723867.586455890@news.t-online.de>
Message-ID: <960ths$40u$1@nyheter.chalmers.se>

Jython is the mantained, new release of JPython.
They have changed name because CNRI have a trademark
(or something) on the name JPython.

/Fredrik

In article <981723867.586455890 at news.t-online.de>,
	Zamurai <zamurai at gmx.net> writes:
> It seems that JPython and Jython are quite similar, but what are the
> differences between both? And which one is better?

-- 


From zamurai at gmx.net  Fri Feb  9 15:38:40 2001
From: zamurai at gmx.net (Zamurai)
Date: Fri, 9 Feb 2001 15:38:40 +0100
Subject: What is better, JPython or Jython?
References: <981723867.586455890@news.t-online.de> <960ths$40u$1@nyheter.chalmers.se>
Message-ID: <981729520.404872677@news.t-online.de>

Thank you for Informations!


From embed at geocities.com  Fri Feb  9 15:11:59 2001
From: embed at geocities.com (Warren Postma)
Date: Fri, 9 Feb 2001 09:11:59 -0500
Subject: What is better, JPython or Jython?
References: <981723867.586455890@news.t-online.de>
Message-ID: <2WSg6.1906$D3.8377@tor-nn1.netcom.ca>

Check the labels. See which one says "30% more Python inside, Free". And buy
that one.

Warren




From SBrunning at trisystems.co.uk  Fri Feb  9 15:45:05 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 9 Feb 2001 14:45:05 -0000 
Subject: What is better, JPython or Jython?
Message-ID: <31575A892FF6D1118F5800600846864D5B17DE@intrepid>

> From:	Zamurai [SMTP:zamurai at gmx.net]
> It seems that JPython and Jython are quite similar, but what are the
> differences between both? And which one is better?
 
Jython supersedes JPython, so use the latter.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From Randy.L.Kemp at motorola.com  Fri Feb  9 15:54:50 2001
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Fri, 9 Feb 2001 08:54:50 -0600 
Subject: What is better, JPython or Jython?
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84E07@il93exp01.css.mot.com>

Which is better: a boat or a car?  It all depends on whether you are traveling by land or sea.  Which is better:  python or jpython?  Are you using it with Java or not? 

-----Original Message-----
From: Warren Postma [mailto:embed at geocities.com]
Sent: Friday, February 09, 2001 8:12 AM
To: python-list at python.org
Subject: Re: What is better, JPython or Jython?


Check the labels. See which one says "30% more Python inside, Free". And buy
that one.

Warren





From dbrueck at edgix.com  Fri Feb  9 16:08:39 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Fri, 9 Feb 2001 08:08:39 -0700
Subject: What is better, JPython or Jython?
In-Reply-To: <69BC4C6AE83ED311BC9400805FA7B14AF84E07@il93exp01.css.mot.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHKEPAPIAA.dbrueck@edgix.com>

Heehee... the question wasn't Python vs Jython.

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Kemp Randy-W18971
> Sent: Friday, February 09, 2001 7:55 AM
> To: python-list at python.org
> Subject: RE: What is better, JPython or Jython?
> 
> 
> Which is better: a boat or a car?  It all depends on whether you 
> are traveling by land or sea.  Which is better:  python or 
> jpython?  Are you using it with Java or not? 
> 
> -----Original Message-----
> From: Warren Postma [mailto:embed at geocities.com]
> Sent: Friday, February 09, 2001 8:12 AM
> To: python-list at python.org
> Subject: Re: What is better, JPython or Jython?
> 
> 
> Check the labels. See which one says "30% more Python inside, 
> Free". And buy
> that one.
> 
> Warren
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list



From aleaxit at yahoo.com  Fri Feb  9 17:29:05 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 9 Feb 2001 17:29:05 +0100
Subject: What is better, JPython or Jython?
References: <mailman.981729975.28507.python-list@python.org>
Message-ID: <9615sm0sop@news2.newsguy.com>

"Simon Brunning" <SBrunning at trisystems.co.uk> wrote in message
news:mailman.981729975.28507.python-list at python.org...
> > From: Zamurai [SMTP:zamurai at gmx.net]
> > It seems that JPython and Jython are quite similar, but what are the
> > differences between both? And which one is better?
>
> Jython supersedes JPython, so use the latter.

Funny -- exactly for this reason, I'd suggest using the _former_:-).


Alex





From com-zjensen at msn.com  Sat Feb 10 07:30:31 2001
From: com-zjensen at msn.com (Zac Jensen)
Date: Fri, 9 Feb 2001 23:30:31 -0700
Subject: What is better, JPython or Jython?
References: <mailman.981730586.31400.python-list@python.org>
Message-ID: <962n6a$v9$1@news.xmission.com>

That's pretty cool, since that's not even the question you moron

>"Kemp Randy-W18971" <Randy.L.Kemp at motorola.com> wrote in message
news:mailman.981730586.31400.python-list at python.org...
> Which is better: a boat or a car?  It all depends on whether you are
traveling by land or sea.  Which is better:  python or jpython?
>Are you using it with Java or not?
>





From peter at engcorp.com  Sat Feb 10 19:04:15 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 10 Feb 2001 13:04:15 -0500
Subject: What is better, JPython or Jython?
References: <mailman.981730586.31400.python-list@python.org> <962n6a$v9$1@news.xmission.com>
Message-ID: <3A85829F.6BA7B9AC@engcorp.com>

Zac Jensen wrote:
> 
> "Kemp Randy-W18971" <Randy.L.Kemp at motorola.com> wrote:
> > Which is better: a boat or a car?  It all depends on whether you are
> > traveling by land or sea.  Which is better:  python or jpython?
> > Are you using it with Java or not?
>
> That's pretty cool, since that's not even the question you moron

That's pretty uncool, since anyone can misread a question you jerk


From m.faassen at vet.uu.nl  Sun Feb 11 00:30:38 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 10 Feb 2001 23:30:38 GMT
Subject: What is better, JPython or Jython?
References: <mailman.981730586.31400.python-list@python.org> <962n6a$v9$1@news.xmission.com> <3A85829F.6BA7B9AC@engcorp.com>
Message-ID: <964iuu$k26$1@newshost.accu.uu.nl>

Peter Hansen <peter at engcorp.com>wrote:
> ZacJensen wrote:
>> 
>> "Kemp Randy-W18971" <Randy.L.Kemp at motorola.com>wrote:
>> > Which is better: a boator a car?  It all depends on whether you are
>> > traveling by land or sea.Which is better:  python or jpython?
>> > Are you using it with Java ornot?
>>
>> That'spretty cool, sincethat's not even thequestion youmoron

> That's pretty uncool,sinceanyonecanmisread a questionyoujerk

Did you all know thatthis kind of flamingattracts white-space eating
nanoviruses and is therefore explicitly forbidden by the PSU? Let's all
bepolite.Oh my it is alreadyhappening, quick, usethe


From ssartor at bellatlantic.net  Fri Feb  9 14:06:36 2001
From: ssartor at bellatlantic.net (Steven Sartorius)
Date: Fri, 09 Feb 2001 13:06:36 GMT
Subject: How to write integers to a file?
Message-ID: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>

I'm trying to write a series of whitespace delimited integers to a file.
The 'write' method only accepts strings and using arrays produces binary
data when I open the file with less.  What I want is somehting like
printf for python.  Is there any easy way to do this?  (there must
be...this is python!)

Any help much appreciated...

Thanks,

Steve Sartorius


From aleaxit at yahoo.com  Fri Feb  9 14:26:31 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 9 Feb 2001 14:26:31 +0100
Subject: How to write integers to a file?
References: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>
Message-ID: <960r680hsk@news2.newsguy.com>

"Steven Sartorius" <ssartor at bellatlantic.net> wrote in message
news:wXRg6.3723$bu4.792238 at typhoon2.ba-dsg.net...
> I'm trying to write a series of whitespace delimited integers to a file.
> The 'write' method only accepts strings and using arrays produces binary
> data when I open the file with less.  What I want is somehting like
> printf for python.  Is there any easy way to do this?  (there must
> be...this is python!)

There are several.  Suppose you have a list (or other sequence) L of
integers and want to write them to a file object F, as readable digits,
one integer per line.  A few "easy ways" to obtain this are:

for i in L:
    F.write(str(i)+'\n')

for i in L:
    F.write("%s\n" % i)

for i in L:
    print >>F, i

F.writelines([str(i)+'\n' for i in L])

F.writelines(["%s\n" % i for i in L])

def myformat(i):
    return "%s\n" % i

F.writelines([myformat(i) for i in L])

F.writelines(map(myformat, L))


Alex





From jhauser at ifm.uni-kiel.de  Fri Feb  9 14:27:52 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 09 Feb 2001 14:27:52 +0100
Subject: How to write integers to a file?
References: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>
Message-ID: <87ae7whtpj.fsf@lisboa.ifm.uni-kiel.de>

Let's start with one line

# opens the file
fd=open('test.dat','w')
# create the data
data=[1,2,3,4,5,6]
# write(string) where string consists of a special format string
# later you want to create the format string once before a loop
# the parentheses are used to evaluate the thing in the right order
# then the % operator is used with a tuple.
fd.write(('%d,'* (len(data)-1) + '%d\n') % tuple(data))
fd.close()

HTH,
__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From chris.gonnerman at usa.net  Fri Feb  9 14:39:13 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Fri, 9 Feb 2001 07:39:13 -0600
Subject: How to write integers to a file?
References: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>
Message-ID: <004701c0929d$b6688dc0$a100000a@local>

outf.write("%d %d %d %d\n" % (i, j, k, l))

The % operator with a string on the left and a singleton value or tuple on
the right performs
printf-style substitution on the string.  You may need to read the man page
for printf if you are
not a C programmer.

----- Original Message -----
From: "Steven Sartorius" <ssartor at bellatlantic.net>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Friday, February 09, 2001 7:06 AM
Subject: How to write integers to a file?


> I'm trying to write a series of whitespace delimited integers to a file.
> The 'write' method only accepts strings and using arrays produces binary
> data when I open the file with less.  What I want is somehting like
> printf for python.  Is there any easy way to do this?  (there must
> be...this is python!)
>
> Any help much appreciated...
>
> Thanks,
>
> Steve Sartorius
> --
> http://mail.python.org/mailman/listinfo/python-list
>




From stephen_purcell at yahoo.com  Fri Feb  9 14:46:02 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 9 Feb 2001 14:46:02 +0100
Subject: How to write integers to a file?
In-Reply-To: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>; from ssartor@bellatlantic.net on Fri, Feb 09, 2001 at 01:06:36PM +0000
References: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>
Message-ID: <20010209144602.A4920@freedom.puma-ag.com>

Steven Sartorius wrote:
> I'm trying to write a series of whitespace delimited integers to a file.
> The 'write' method only accepts strings and using arrays produces binary
> data when I open the file with less.  What I want is somehting like
> printf for python.  Is there any easy way to do this?  (there must
> be...this is python!)
> 

Yep:-

>>> "the number is %d" % 5
'the number is 5'
>>> 

See the docs at: http://www.python.org/doc/current/lib/typesseq-strings.html

But for your case, maybe you just want to do the following
   
>>> ints = range(5)
>>> import string
>>> string.join(map(str, ints), ' ')
'0 1 2 3 4'

   
-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From ssartor at bellatlantic.net  Fri Feb  9 21:04:12 2001
From: ssartor at bellatlantic.net (Steven Sartorius)
Date: Fri, 09 Feb 2001 20:04:12 GMT
Subject: How to write integers to a file?
References: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net> <mailman.981725706.11283.python-list@python.org>
Message-ID: <03Yg6.4777$bu4.895297@typhoon2.ba-dsg.net>

In article <mailman.981725706.11283.python-list at python.org>, "Chris
Gonnerman" <chris.gonnerman at usa.net> wrote:

> outf.write("%d %d %d %d\n" % (i, j, k, l))
> 
> The % operator with a string on the left and a singleton value or tuple
> on the right performs printf-style substitution on the string.  You may
> need to read the man page for printf if you are not a C programmer.
> 

This did the trick!  Thanks a lot.

Steve


From warlock at eskimo.com  Sat Feb 10 09:50:34 2001
From: warlock at eskimo.com (Jim Richardson)
Date: Sat, 10 Feb 2001 00:50:34 -0800
Subject: How to write integers to a file?
References: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>
Message-ID: <slrn98a06q.2gf.warlock@gargoyle.myth>

On Fri, 09 Feb 2001 13:06:36 GMT, 
 Steven Sartorius, in the persona of <ssartor at bellatlantic.net>,
 brought forth the following words...:

>I'm trying to write a series of whitespace delimited integers to a file.
>The 'write' method only accepts strings and using arrays produces binary
>data when I open the file with less.  What I want is somehting like
>printf for python.  Is there any easy way to do this?  (there must
>be...this is python!)
>
>Any help much appreciated...
>
>Thanks,
>
>Steve Sartorius

use str(integer) to convert to a string before the write.

-- 
Jim Richardson
	Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
	Linux, because life's too short for a buggy OS.



From aaa at bbb.com  Fri Feb  9 14:16:04 2001
From: aaa at bbb.com (debugboy)
Date: Fri, 9 Feb 2001 21:16:04 +0800
Subject: I love Python
Message-ID: <960qnb$p3t$1@mail.cn99.com>

Test




From zamurai at gmx.net  Fri Feb  9 14:36:35 2001
From: zamurai at gmx.net (Zamurai)
Date: Fri, 9 Feb 2001 14:36:35 +0100
Subject: I love Python
References: <960qnb$p3t$1@mail.cn99.com>
Message-ID: <981725795.486882899@news.t-online.de>

"debugboy" <aaa at bbb.com> wrote:
> Test

I love Python too ;-)


From eq3pvl at eq.uc.pt  Fri Feb  9 17:20:10 2001
From: eq3pvl at eq.uc.pt (Pedro Vale Lima)
Date: Fri, 09 Feb 2001 16:20:10 +0000
Subject: I love Python
References: <960qnb$p3t$1@mail.cn99.com> <981725795.486882899@news.t-online.de>
Message-ID: <3A8418BA.1F228AA7@eq.uc.pt>

Zamurai wrote:

> "debugboy" <aaa at bbb.com> wrote:
> > Test
>
> I love Python too ;-)

Yeah, but have you asked Python if he loves you?



From zamurai at gmx.net  Fri Feb  9 17:35:59 2001
From: zamurai at gmx.net (Zamurai)
Date: Fri, 9 Feb 2001 17:35:59 +0100
Subject: I love Python
References: <960qnb$p3t$1@mail.cn99.com> <981725795.486882899@news.t-online.de> <3A8418BA.1F228AA7@eq.uc.pt>
Message-ID: <981736559.1767376098@news.t-online.de>

Pedro Vale Lima <eq3pvl at eq.uc.pt> wrote:
> Zamurai wrote:
> 
> > "debugboy" <aaa at bbb.com> wrote:
> > > Test
> >
> > I love Python too ;-)
> 
> Yeah, but have you asked Python if he loves you?

No, but I?m programming my own girl friend in Python! :-)



From phd at phd.pp.ru  Fri Feb  9 18:05:21 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 9 Feb 2001 20:05:21 +0300 (MSK)
Subject: I love Python
In-Reply-To: <981736559.1767376098@news.t-online.de>
Message-ID: <Pine.LNX.4.30.0102092004270.27625-100000@fep132.fep.ru>

Pedro Vale Lima <eq3pvl at eq.uc.pt> wrote:
> > I love Python too ;-)
> Yeah, but have you asked Python if he loves you?

   It doesn't matter until *I* am satisfied! :)

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From zamurai at gmx.net  Fri Feb  9 14:37:21 2001
From: zamurai at gmx.net (Zamurai)
Date: Fri, 9 Feb 2001 14:37:21 +0100
Subject: I love Python
References: <960qnb$p3t$1@mail.cn99.com>
Message-ID: <981725841.459312559@news.t-online.de>

"debugboy" <aaa at bbb.com> wrote:
> Test



From Randy.L.Kemp at motorola.com  Fri Feb  9 16:02:21 2001
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Fri, 9 Feb 2001 09:02:21 -0600 
Subject: I love Python
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84E08@il93exp01.css.mot.com>

I noticed this message was sent three times.  There is an ancient legend that if everyone in the world says "I love Python" three times, Perl will disappear.  

-----Original Message-----
From: Zamurai [mailto:zamurai at gmx.net]
Sent: Friday, February 09, 2001 7:37 AM
To: python-list at python.org
Subject: Re: I love Python


"debugboy" <aaa at bbb.com> wrote:
> Test

I love Python too ;-)



From danielk at aracnet.com  Fri Feb  9 16:35:32 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Fri, 09 Feb 2001 07:35:32 -0800
Subject: I love Python
References: <mailman.981731063.856.python-list@python.org>
Message-ID: <vg388tsp5g580mroauk6d2ulvj8vo5s8cg@4ax.com>

On Fri, 9 Feb 2001 09:02:21 -0600, Kemp Randy-W18971
<Randy.L.Kemp at motorola.com> wrote:

>I noticed this message was sent three times.  There is an ancient legend that if everyone in the world says "I love Python" three times, Perl will disappear.  

What's Perl? Hey, it worked!


From vasvir at iit.demokritos.gr  Fri Feb  9 18:30:25 2001
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Fri, 09 Feb 2001 19:30:25 +0200
Subject: I love Python
References: <mailman.981731063.856.python-list@python.org> <vg388tsp5g580mroauk6d2ulvj8vo5s8cg@4ax.com>
Message-ID: <3A842931.8C46014C@iit.demokritos.gr>

Daniel Klein wrote:
> 
> On Fri, 9 Feb 2001 09:02:21 -0600, Kemp Randy-W18971
> <Randy.L.Kemp at motorola.com> wrote:
> 
> >I noticed this message was sent three times.  There is an ancient legend that if everyone in the world says "I love Python" three times, Perl will disappear.
> 
> What's Perl? Hey, it worked!

I don't want to offend anybody but this is the best off-topic thread I
have ever read.
	.Bill


From n8gray at caltech.edu.is.my.email.address  Sat Feb 10 06:50:02 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Fri, 9 Feb 2001 21:50:02 -0800
Subject: I love Python
References: <mailman.981731063.856.python-list@python.org> <vg388tsp5g580mroauk6d2ulvj8vo5s8cg@4ax.com> <3A842931.8C46014C@iit.demokritos.gr>
Message-ID: <962l55$i9i@gap.cco.caltech.edu>

Vassilis Virvilis wrote:

> Daniel Klein wrote:
> > 
> > On Fri, 9 Feb 2001 09:02:21 -0600, Kemp Randy-W18971
> > <Randy.L.Kemp at motorola.com> wrote:
> > 
> > >I noticed this message was sent three times.  There is an ancient
> > >legend that if everyone in the world says "I love Python" three times,
> > >Perl will disappear.
> > 
> > What's Perl? Hey, it worked!
> 
> I don't want to offend anybody but this is the best off-topic thread I
> have ever read.
> .Bill

Can I just add that this is the first time I've ever been on television?

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From jimhill at meldrick.swcp.com  Sat Feb 10 07:08:02 2001
From: jimhill at meldrick.swcp.com (Jim Hill)
Date: 10 Feb 2001 06:08:02 GMT
Subject: I love Python
References: <mailman.981731063.856.python-list@python.org> <vg388tsp5g580mroauk6d2ulvj8vo5s8cg@4ax.com> <3A842931.8C46014C@iit.demokritos.gr> <962l55$i9i@gap.cco.caltech.edu>
Message-ID: <slrn989mi9.bl5.jimhill@meldrick.swcp.com>

In <962l55$i9i at gap.cco.caltech.edu>,
Nathaniel Gray <n8gray at caltech.edu.is.my.email.address> wrote:
>Vassilis Virvilis wrote:
>
>> Daniel Klein wrote:
>> > 
>> > On Fri, 9 Feb 2001 09:02:21 -0600, Kemp Randy-W18971
>> > <Randy.L.Kemp at motorola.com> wrote:
>> > 
>> > >I noticed this message was sent three times.  There is an ancient
>> > >legend that if everyone in the world says "I love Python" three times,
>> > >Perl will disappear.
>> > 
>> > What's Perl? Hey, it worked!
>> 
>> I don't want to offend anybody but this is the best off-topic thread I
>> have ever read.
>> .Bill
>
>Can I just add that this is the first time I've ever been on television?

I'm sorry; we haven't got time for that now...


Jim
-- 
jimhill at swcp.com                         http://www.swcp.com/~jimhill/

   "Let me preface my remarks by saying that I have absolutely no love
            for Barry Manilow in my heart."  --  Jason Lempka


From gustafl at algonet.se  Fri Feb  9 14:58:39 2001
From: gustafl at algonet.se (Gustaf Liljegren)
Date: Fri, 9 Feb 2001 14:58:39 +0100
Subject: Closing a file before or after return?
Message-ID: <960t2j$9n9$1@cubacola.tninet.se>

I'm using Python 2.0.

Have a look at this function. It checks if a particular user (mail adress)
exist in a .htpasswd file. The user name is what comes before ':' on each
line.

# Check if user exists
def user_exist(user):
  f = open('.htpasswd', 'r')
  for line in f.readlines():
    i = string.find(line, ':')
    if line[:i] == user:
      return 1

I'd like to close the file too, but I don't know where to put that
statement. Of course, the file should be closed in both cases -- not just
when the user is found.

Regards,

Gustaf Liljegren




From objectway at divalsim.it  Fri Feb  9 15:20:10 2001
From: objectway at divalsim.it (Nicola Musatti)
Date: Fri, 09 Feb 2001 15:20:10 +0100
Subject: Closing a file before or after return?
References: <960t2j$9n9$1@cubacola.tninet.se>
Message-ID: <3A83FC9A.721F153@divalsim.it>

Hi, I'm not even a newbie, but I'll venture an answer.

Gustaf Liljegren wrote:
> 
> I'm using Python 2.0.
> 
> Have a look at this function. It checks if a particular user (mail adress)
> exist in a .htpasswd file. The user name is what comes before ':' on each
> line.
> 
> # Check if user exists
> def user_exist(user):
>   f = open('.htpasswd', 'r')
>   for line in f.readlines():
>     i = string.find(line, ':')
>     if line[:i] == user:
        ret = 1
        break
    else:
      ret = 0
    f.close()
    return ret

Best regards,
Nicola Musatti


From stephen_purcell at yahoo.com  Fri Feb  9 15:35:53 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 9 Feb 2001 15:35:53 +0100
Subject: Closing a file before or after return?
In-Reply-To: <960t2j$9n9$1@cubacola.tninet.se>; from gustafl@algonet.se on Fri, Feb 09, 2001 at 02:58:39PM +0100
References: <960t2j$9n9$1@cubacola.tninet.se>
Message-ID: <20010209153553.A18498@freedom.puma-ag.com>

Gustaf Liljegren wrote:
> Have a look at this function. It checks if a particular user (mail adress)
> exist in a .htpasswd file. The user name is what comes before ':' on each
> line.
> 
> # Check if user exists
> def user_exist(user):
>   f = open('.htpasswd', 'r')
>   for line in f.readlines():
>     i = string.find(line, ':')
>     if line[:i] == user:
>       return 1
> 
> I'd like to close the file too, but I don't know where to put that
> statement. Of course, the file should be closed in both cases -- not just
> when the user is found.

How about try/finally:

  def user_exist(user):
    f = open('.htpasswd', 'r')
    try:
      for line in f.readlines():
        i = string.find(line, ':')
        if line[:i] == user:
          return 1
    finally:
      f.close()

A neater way would just be:


  def user_exist(user):
    f = open('.htpasswd', 'r')
    found = 0
    for line in f.readlines():
      i = string.find(line, ':')
      if line[:i] == user:
        found = 1
        break
    return found


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From quinn at lira.ugcs.caltech.edu  Mon Feb 12 00:53:26 2001
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 11 Feb 2001 23:53:26 GMT
Subject: Closing a file before or after return?
References: <960t2j$9n9$1@cubacola.tninet.se>
Message-ID: <slrn98e9fm.qne.quinn@lira.ugcs.caltech.edu>

On Fri, 9 Feb 2001 14:58:39 +0100, Gustaf Liljegren <gustafl at algonet.se> wrote:
>I'm using Python 2.0.
>
>Have a look at this function. It checks if a particular user (mail adress)
>exist in a .htpasswd file. The user name is what comes before ':' on each
>line.
>
># Check if user exists
>def user_exist(user):
>  f = open('.htpasswd', 'r')
>  for line in f.readlines():
>    i = string.find(line, ':')
>    if line[:i] == user:
>      return 1
>
>I'd like to close the file too, but I don't know where to put that
>statement. Of course, the file should be closed in both cases -- not just
>when the user is found.

# Check if user exists
def user_exist(user):
  f = open('.htpasswd', 'r')
  try:
      for line in f.readlines():
        i = string.find(line, ':')
        if line[:i] == user:
          return 1
  finally:
    f.close()

... will work.  But it's not necessary, since python will close the file for
you when the last reference disappears.  Explicitly deallocating anything more
complicated than memory or fds is usually good practice, though.


From ct at gocept.com  Fri Feb  9 15:33:02 2001
From: ct at gocept.com (Christian Theune)
Date: Fri, 09 Feb 2001 15:33:02 +0100
Subject: MySQL + SQL Statements + Quote escaping
Message-ID: <3A83FF9E.1E8CC176@gocept.com>

Hi there.

i just tuned into python and it's really cool. I like it.
Before I came to python i used to write scripts in php.
There was a function called "addslashes" it escaped quotes and
slashes to make strings sql-safe.

Is there any function like that in python? I didnt found on.

Thanks so far.

Christian


---
Christian Theune
0178 48 33 981 - ct at gocept.com - www.gocept.com


From huwdotlynes at ic.ac.uk  Sat Feb 10 13:04:31 2001
From: huwdotlynes at ic.ac.uk (Huw Lynes)
Date: 10 Feb 2001 12:04:31 GMT
Subject: MySQL + SQL Statements + Quote escaping
References: <3A83FF9E.1E8CC176@gocept.com>
Message-ID: <slrn97kise.8g.huwdotlynes@Marvin.ic.ac.uk>

In article <3A83FF9E.1E8CC176 at gocept.com>, Christian Theune wrote:
>Hi there.
>
>i just tuned into python and it's really cool. I like it.
>Before I came to python i used to write scripts in php.
>There was a function called "addslashes" it escaped quotes and
>slashes to make strings sql-safe.
>
>Is there any function like that in python? I didnt found on.

There probably is but I usually just do a quick 
string.find() to to find all the quotes and then slice the string
at the indexes returned and add in the slashes.
But it seems like a common problem like this should have
a module somewhere.

Sorry I couldn't be of more help but I'm
only a newbie myself.

Huw Lynes


From sholden at holdenweb.com  Sat Feb 10 16:22:46 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 10 Feb 2001 10:22:46 -0500
Subject: MySQL + SQL Statements + Quote escaping
References: <3A83FF9E.1E8CC176@gocept.com> <slrn97kise.8g.huwdotlynes@Marvin.ic.ac.uk>
Message-ID: <T4dh6.3919$Tn6.89580@e420r-atl2.usenetserver.com>

"Huw Lynes" <huwdotlynes at ic.ac.uk> wrote in message
news:slrn97kise.8g.huwdotlynes at Marvin.ic.ac.uk...
> In article <3A83FF9E.1E8CC176 at gocept.com>, Christian Theune wrote:
> >Hi there.
> >
> >i just tuned into python and it's really cool. I like it.
> >Before I came to python i used to write scripts in php.
> >There was a function called "addslashes" it escaped quotes and
> >slashes to make strings sql-safe.
> >
> >Is there any function like that in python? I didnt found on.
>
> There probably is but I usually just do a quick
> string.find() to to find all the quotes and then slice the string
> at the indexes returned and add in the slashes.
> But it seems like a common problem like this should have
> a module somewhere.
>
> Sorry I couldn't be of more help but I'm
> only a newbie myself.
>
> Huw Lynes
>
string.replace() works well.

regards
 Steve




From m.wilson at bibliocraft.com  Mon Feb 12 00:07:39 2001
From: m.wilson at bibliocraft.com (Mark Wilson)
Date: Sun, 11 Feb 2001 23:07:39 -0000
Subject: MySQL + SQL Statements + Quote escaping
Message-ID: <034d01c094d5$730312a0$36824ac3@mwdell1>

>Hi there.
>
>i just tuned into python and it's really cool. I like it.
>Before I came to python i used to write scripts in php.
>There was a function called "addslashes" it escaped quotes and
>slashes to make strings sql-safe.
>
>Is there any function like that in python? I didnt found on.

If you're using MySQL from Python, use the MySQL module's escape() function, as
in:

import MySQL

    text_ok_for_sql = MySQL.escape( text_the_user_has_input )

Note there's no reverse procedure necessary with MySQL - text returned from the
database has already been "unescaped".


>Thanks so far.
>
>Christian

Mark





From mix77 at usa.net  Fri Feb  9 16:49:33 2001
From: mix77 at usa.net (mix77 at usa.net)
Date: Fri, 09 Feb 2001 17:49:33 +0200
Subject: Executing Sytem Commands
Message-ID: <3A84118D.180E1CDC@usa.net>

How do you execute a sytem command  like "mv file1 file2" in Python?
P.S. I'm working under Linux 2.x (redhat 7 distribution with default
linux kernel).



From phd at phd.pp.ru  Fri Feb  9 17:04:55 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 9 Feb 2001 19:04:55 +0300 (MSK)
Subject: Executing Sytem Commands
In-Reply-To: <3A84118D.180E1CDC@usa.net>
Message-ID: <Pine.LNX.4.30.0102091903520.27625-100000@fep132.fep.ru>

On Fri, 9 Feb 2001 mix77 at usa.net wrote:
> How do you execute a sytem command  like "mv file1 file2" in Python?

   Exactly like system. It is in module "os". So well:

import os
os.system("xsetbg foo.jpg")

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From dsh8290 at rit.edu  Fri Feb  9 17:21:56 2001
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 9 Feb 2001 11:21:56 -0500
Subject: Executing Sytem Commands
In-Reply-To: <3A84118D.180E1CDC@usa.net>; from mix77@usa.net on Fri, Feb 09, 2001 at 05:49:33PM +0200
References: <3A84118D.180E1CDC@usa.net>
Message-ID: <20010209112156.A16615@harmony.cs.rit.edu>

import os
os.system( "mv file1 file2" )

-D

On Fri, Feb 09, 2001 at 05:49:33PM +0200, mix77 at usa.net wrote:
| How do you execute a sytem command  like "mv file1 file2" in Python?
| P.S. I'm working under Linux 2.x (redhat 7 distribution with default
| linux kernel).
| 



From cavvinas at zaz.com.br  Fri Feb  9 17:41:11 2001
From: cavvinas at zaz.com.br (=?iso-8859-1?Q?Estev=E3o_Cavinato?=)
Date: Fri, 9 Feb 2001 14:41:11 -0200
Subject: Digest Number 1687
References: <981735746.671.44867.l9@yahoogroups.com>
Message-ID: <000b01c092b7$1e517820$0c00a8c0@tania>

Unsusbcribe
----- Original Message -----
From: <python-list at yahoogroups.com>
To: <python-list at yahoogroups.com>
Sent: Friday, February 09, 2001 2:22 PM
Subject: Digest Number 1687


> ------------------------ Yahoo! Groups Sponsor ---------------------~-~>
> eGroups is now Yahoo! Groups
> Click here for more details
> http://click.egroups.com/1/11231/0/_/1452/_/981735747/
> ---------------------------------------------------------------------_->
>
> There are 25 messages in this issue.
>
> Topics in this digest:
>
>       1. Re: Closing a file before or after return?
>            From: Steve Purcell <stephen_purcell at yahoo.com>
>       2. MySQL + SQL Statements + Quote escaping
>            From: Christian Theune <ct at gocept.com>
>       3. Re: How to return from shutdown command
>            From: D-Man <dsh8290 at rit.edu>
>       4. RE: What is better, JPython or Jython?
>            From: Simon Brunning <SBrunning at trisystems.co.uk>
>       5. RE: What is better, JPython or Jython?
>            From: Kemp Randy-W18971 <Randy.L.Kemp at motorola.com>
>       6. Re: What is better, JPython or Jython?
>            From: Zamurai <zamurai at gmx.net>
>       7. RE: I love Python
>            From: Kemp Randy-W18971 <Randy.L.Kemp at motorola.com>
>       8. Re: python-dev summary, Jan. 16-31
>            From: Andrew Kuchling <akuchlin at mems-exchange.org>
>       9. RE: What is better, JPython or Jython?
>            From: "Dave Brueck" <dbrueck at edgix.com>
>      10. Re: Python developers of the world unite ;)
>            From: "Lyle Johnson" <ljohnson at resgen.com>
>      11. Re: Is comp.lang.python.announce dead?
>            From: barry at digicool.com (Barry A. Warsaw)
>      12. Re: urllib.open('http://www.redherring.com/')
>            From: Andrew Kuchling <akuchlin at mems-exchange.org>
>      13. Re: How to copy class objects?
>            From: matthias.oberlaender at daimlerchrysler.com
>      14. Re: smtp mail with attachment
>            From: barry at digicool.com (Barry A. Warsaw)
>      15. Re: I love Python
>            From: Daniel Klein <danielk at aracnet.com>
>      16. Re: python-dev summary, Jan. 16-31
>            From: Oleg Broytmann <phd at phd.pp.ru>
>      17. Re: python-dev summary, Jan. 16-31
>            From: "Sandipan Gangopadhyay" <sandipan at vsnl.com>
>      18. Re: nested scopes
>            From: aahz at panix.com (Aahz Maruch)
>      19. Re: urllib.open('http://www.redherring.com/')
>            From: aahz at panix.com (Aahz Maruch)
>      20. Executing Sytem Commands
>            From: mix77 at usa.net
>      21. Re: global, was Re: None assigment
>            From: Joshua Marshall <jmarshal at mathworks.com>
>      22. Re: Executing Sytem Commands
>            From: Oleg Broytmann <phd at phd.pp.ru>
>      23. Re: urllib.open('http://www.redherring.com/')
>            From: apederse at my-deja.com
>      24. RE: python-dev summary, Jan. 16-31
>            From: Simon Brunning <SBrunning at trisystems.co.uk>
>      25. Re: I love Python
>            From: Pedro Vale Lima <eq3pvl at eq.uc.pt>
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 1
>    Date: Fri, 9 Feb 2001 15:35:53 +0100
>    From: Steve Purcell <stephen_purcell at yahoo.com>
> Subject: Re: Closing a file before or after return?
>
> Gustaf Liljegren wrote:
> > Have a look at this function. It checks if a particular user (mail
adress)
> > exist in a .htpasswd file. The user name is what comes before ':' on
each
> > line.
> >
> > # Check if user exists
> > def user_exist(user):
> >   f = open('.htpasswd', 'r')
> >   for line in f.readlines():
> >     i = string.find(line, ':')
> >     if line[:i] == user:
> >       return 1
> >
> > I'd like to close the file too, but I don't know where to put that
> > statement. Of course, the file should be closed in both cases -- not
just
> > when the user is found.
>
> How about try/finally:
>
>   def user_exist(user):
>     f = open('.htpasswd', 'r')
>     try:
>       for line in f.readlines():
>         i = string.find(line, ':')
>         if line[:i] == user:
>           return 1
>     finally:
>       f.close()
>
> A neater way would just be:
>
>
>   def user_exist(user):
>     f = open('.htpasswd', 'r')
>     found = 0
>     for line in f.readlines():
>       i = string.find(line, ':')
>       if line[:i] == user:
>         found = 1
>         break
>     return found
>
>
> -Steve
>
> --
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Available for consulting and training.
> "Even snakes are afraid of snakes." -- Steven Wright
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 2
>    Date: Fri, 09 Feb 2001 15:33:02 +0100
>    From: Christian Theune <ct at gocept.com>
> Subject: MySQL + SQL Statements + Quote escaping
>
> Hi there.
>
> i just tuned into python and it's really cool. I like it.
> Before I came to python i used to write scripts in php.
> There was a function called "addslashes" it escaped quotes and
> slashes to make strings sql-safe.
>
> Is there any function like that in python? I didnt found on.
>
> Thanks so far.
>
> Christian
>
>
> ---
> Christian Theune
> 0178 48 33 981 - ct at gocept.com - www.gocept.com
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 3
>    Date: Fri, 9 Feb 2001 09:34:56 -0500
>    From: D-Man <dsh8290 at rit.edu>
> Subject: Re: How to return from shutdown command
>
>
> If you have just shutdown the system, how is the script supposed to
> continue running?  ;-)
>
> -D
>
> On Fri, Feb 09, 2001 at 03:01:54PM +1100, Craig Findlay wrote:
> | I want to execute a Unix shutdown command from a python script, but
> | have control return to the script, and I'm having trouble doing it.
> |
> | I've tried:
> |
> | commands.getstatusoutput('shutdown -h +1')
> | and
> | commands.getstatusoutput('/bin/sh -c "shutdown -h +1" \&')
> |
> | but neither of these commands return.
> |
> | Any help greatly appreciated!
> |
> | Craig
> |
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 4
>    Date: Fri, 9 Feb 2001 14:45:05 -0000
>    From: Simon Brunning <SBrunning at trisystems.co.uk>
> Subject: RE: What is better, JPython or Jython?
>
> > From: Zamurai [SMTP:zamurai at gmx.net]
> > It seems that JPython and Jython are quite similar, but what are the
> > differences between both? And which one is better?
>
> Jython supersedes JPython, so use the latter.
>
> Cheers,
> Simon Brunning
> TriSystems Ltd.
> sbrunning at trisystems.co.uk
>
>
>
>
> -----------------------------------------------------------------------
> The information in this email is confidential and may be legally
privileged.
> It is intended solely for the addressee. Access to this email by anyone
else
> is unauthorised. If you are not the intended recipient, any disclosure,
> copying, distribution, or any action taken or omitted to be taken in
> reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
> accept liability for statements made which are clearly the senders own.
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 5
>    Date: Fri, 9 Feb 2001 08:54:50 -0600
>    From: Kemp Randy-W18971 <Randy.L.Kemp at motorola.com>
> Subject: RE: What is better, JPython or Jython?
>
> Which is better: a boat or a car?  It all depends on whether you are
traveling by land or sea.  Which is better:  python or jpython?  Are you
using it with Java or not?
>
> -----Original Message-----
> From: Warren Postma [mailto:embed at geocities.com]
> Sent: Friday, February 09, 2001 8:12 AM
> To: python-list at python.org
> Subject: Re: What is better, JPython or Jython?
>
>
> Check the labels. See which one says "30% more Python inside, Free". And
buy
> that one.
>
> Warren
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 6
>    Date: Fri, 9 Feb 2001 15:38:40 +0100
>    From: Zamurai <zamurai at gmx.net>
> Subject: Re: What is better, JPython or Jython?
>
> Thank you for Informations!
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 7
>    Date: Fri, 9 Feb 2001 09:02:21 -0600
>    From: Kemp Randy-W18971 <Randy.L.Kemp at motorola.com>
> Subject: RE: I love Python
>
> I noticed this message was sent three times.  There is an ancient legend
that if everyone in the world says "I love Python" three times, Perl will
disappear.
>
> -----Original Message-----
> From: Zamurai [mailto:zamurai at gmx.net]
> Sent: Friday, February 09, 2001 7:37 AM
> To: python-list at python.org
> Subject: Re: I love Python
>
>
> "debugboy" <aaa at bbb.com> wrote:
> > Test
>
> I love Python too ;-)
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 8
>    Date: 09 Feb 2001 09:55:52 -0500
>    From: Andrew Kuchling <akuchlin at mems-exchange.org>
> Subject: Re: python-dev summary, Jan. 16-31
>
> Oleg Broytmann <phd at phd.pp.ru> writes:
> >    Any chance you'll continue the summaries?
>
> Not really.  All of this is sparked by my wanting to reduce the load,
> really; I recently realized that I haven't hacked on *any* of my
> projects (Oedipus, amkCrypto, etc.) lately.  Therefore I've dropped
> out of my book club (one night a week), dropped the summaries, (two
> nights every two weeks, or 15% of my evenings), and after 2.1final is
> released I'll probably unsubscribe from python-dev, too.
>
> So, to rescuing the summaries... For the information of potential
> volunteers, I'll summarize the process.  Writing them isn't difficult
> and can easily be done while watching TV -- I usually pull the mail
> archive on my laptop and do just that.  They're just time-consuming,
> taking up two evenings at a time.
>
> 1) Grab a copy of the mailbox archive for the month.
>
> 2) Load it up into mutt (or other MUA of choice), and delete messages
>    outside of the two-week period being summarized.
>
> 3) Sort the remaining messages by thread, and go through finding the
>    interesting threads.  What makes a thread interesting?
>
>         * Threads discussing how to fix a particular bug,
>           or tracing down a bug's root cause, can be quite lengthy but
>           usually aren't interesting.
>
>         * Minute discussions of language syntax are rarely interesting,
>           and I think even Guido tunes them out after a while.
>
>         * New proposals that spark off a discussion are interesting.
>         * Discussions surrounding a PEP are interesting.
>
> 4) <time consuming but fun step> Read through each interesting thread,
>    pulling out interesting quotes and summarizing the flow of the
>    argument.  This is fun because you can introduce your own biases; I
>    think this is why people become journalists.
>
> 5) <time consuming and tedious step> For each quote, find the URL for
>    that message in the archives on mail.python.org, and link to it.
>    This is painful; recently I've started spidering the archives with wget
>    and grepping them in order to avoid ferreting through the indexes
>    over my modem connection.
>
>    Thinking about it now, it would have been much easier on me to just
>    do what Linux Weekly News does, and make a local copy of each
>    message.  That means you wouldn't have to worry about links
>    breaking (moving Mailman to mail.python.org required regenerating
>    the archives, and the links in the October & November summaries are
>    now all broken), and the author wouldn't have to hunt for the right
>    message URL.  You'd lose the easy ability to follow the thread, of
>    course.
>
> 6) The summary is written as plain ASCII, and it gets posted, and
>    mailed to LWN and LinuxToday.  The HTML version is then just a
>    matter of escaping <&>, turning the URLs into <A> tags with an Emacs
>    macro, and putting <pre> tags around it; no fancy XML DTD...
>
> --amk
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 9
>    Date: Fri, 9 Feb 2001 08:08:39 -0700
>    From: "Dave Brueck" <dbrueck at edgix.com>
> Subject: RE: What is better, JPython or Jython?
>
> Heehee... the question wasn't Python vs Jython.
>
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Kemp Randy-W18971
> > Sent: Friday, February 09, 2001 7:55 AM
> > To: python-list at python.org
> > Subject: RE: What is better, JPython or Jython?
> >
> >
> > Which is better: a boat or a car?  It all depends on whether you
> > are traveling by land or sea.  Which is better:  python or
> > jpython?  Are you using it with Java or not?
> >
> > -----Original Message-----
> > From: Warren Postma [mailto:embed at geocities.com]
> > Sent: Friday, February 09, 2001 8:12 AM
> > To: python-list at python.org
> > Subject: Re: What is better, JPython or Jython?
> >
> >
> > Check the labels. See which one says "30% more Python inside,
> > Free". And buy
> > that one.
> >
> > Warren
> >
> >
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 10
>    Date: Fri, 9 Feb 2001 09:04:33 -0600
>    From: "Lyle Johnson" <ljohnson at resgen.com>
> Subject: Re: Python developers of the world unite ;)
>
> > See the Catalog-SIG:  http://www.python.org/sigs/catalog-sig/
>
> Speaking of the catalog-sig, what is the current status? I remember
> subscribing to the mailing list awhile back (mainly as a lurker and not an
> active participant) but it doesn't seem like there's been much activity
> lately. In fact, if the mailing list archives are up-to-date, there
haven't
> been any posts since November.
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 11
>    Date: Fri, 9 Feb 2001 10:26:04 -0500
>    From: barry at digicool.com (Barry A. Warsaw)
> Subject: Re: Is comp.lang.python.announce dead?
>
>
> >>>>> "gradha" ==   <gradha at iname.com> writes:
>
>     gradha> Just today I fetched the news and got around 10 "announce"
>     gradha> messages, in this newsgroup instead of
>     gradha> comp.lang.python.announce. Does this mean that the latter
>     gradha> is officially dead?
>
> c.l.py.a is absolutely live again, and as mentioned in a different
> follow, there is a team of moderators who are helping to approve
> messages for the newsgroup.
>
> There was a separate and temporary outage of the moderation gateway as
> we changed mail daemons on mail.python.org (for reasons I won't go
> into).  That problem should be fixed now.
>
> Please post your announcements to c.l.py.a!  Also, register your
> programs with Parnassus and those announcements can be automatically
> forwarded to the newsgroup.
>
> -Barry
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 12
>    Date: 09 Feb 2001 10:07:10 -0500
>    From: Andrew Kuchling <akuchlin at mems-exchange.org>
> Subject: Re: urllib.open('http://www.redherring.com/')
>
> apederse at my-deja.com writes:
> > to me it looks like the problem lies whithin python itself? is there any
> > point in reporting such bugs when everyone will soon be using python 2.0
> > anyway?
>
> I don't believe urllib or httplib were massively rewritten for 2.0
> (nor will they be in 2.1), so if it's a bug, it's probably still
> there.  On the other hand, if you try it with 2.0 and it works fine,
> then there won't be much interest; a 1.5.3 release is not going to
> happen.
>
> --amk
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 13
>    Date: 9 Feb 2001 15:05:39 GMT
>    From: matthias.oberlaender at daimlerchrysler.com
> Subject: Re: How to copy class objects?
>
> In <960b75$a2e$1 at news.sns-felb.debis.de>
> matthias.oberlaender at daimlerchrysler.com wrote:
> > This is the question I'd like to be answered:
> >
> > Is there some easy way to copy a class object?
> >
> > Here is the background:
> >
> > In order to generate classes from templates I'd prefer the following
method
> > over writing additional wrapper function
> >
> > class X:
> >
> >   sratchdir = "/tmp/"
> >
> >   <blabla    (methods using scratchdir)>
> >
> > # I want a new class Xnew differing from X in that its instances use a
> > different scratch directory.
> > Xnew = copyclass(X)
> > # Now set scratchdir to something else
> > Xnew.scratchdir = '/home/me/tmp'
> >
> >
> >
> Quite a few people have already tried to help me. Thanks!
> But it seems it is sowewhat dangerous to give background information,
because
> it often distracts people from the real question, which is:
>
> How can I copy a class object?
> =========================
>
> Somebody suggested the copy module. But copy.copy does not apply to
classes!
>
> Ok, don't tell me that I don't need such a feature for the toy problem
given.
> It's solely for illustration. The real intent is much different.
> I can't accept subclassing, because I don't want instances of X and Xnew
have
> any base classes in common. So new.classobj won't work either!  Ok?
>
> -- Matthias
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 14
>    Date: Fri, 9 Feb 2001 10:32:22 -0500
>    From: barry at digicool.com (Barry A. Warsaw)
> Subject: Re: smtp mail with attachment
>
>
> >>>>> "SH" == Scott Hathaway <slhath at home.com.nospam> writes:
>
>     SH> How do I send an email with an attachment with python (I need
>     SH> a cross platform solution for Windows and Unix)?
>
> Other people have made good suggestions in separate followups.  I just
> add that I'm working on a new library to deal with MIME documents
> (well, any rfc822 style message, but MIME is the really interesting
> bit).  I will be using this library in the next release of Mailman,
> and hope eventually to get it into Python's standard library.  For
> now, you can find out more at:
>
>     http://barry.wooz.org/software/pyware.html
>
> Look for `mimelib'.  I will likely have a 0.2 release sometime soon.
>
> -Barry
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 15
>    Date: Fri, 09 Feb 2001 07:35:32 -0800
>    From: Daniel Klein <danielk at aracnet.com>
> Subject: Re: I love Python
>
> On Fri, 9 Feb 2001 09:02:21 -0600, Kemp Randy-W18971
> <Randy.L.Kemp at motorola.com> wrote:
>
> >I noticed this message was sent three times.  There is an ancient legend
that if everyone in the world says "I love Python" three times, Perl will
disappear.
>
> What's Perl? Hey, it worked!
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 16
>    Date: Fri, 9 Feb 2001 18:39:44 +0300 (MSK)
>    From: Oleg Broytmann <phd at phd.pp.ru>
> Subject: Re: python-dev summary, Jan. 16-31
>
> On 9 Feb 2001, Andrew Kuchling wrote:
> > Oleg Broytmann <phd at phd.pp.ru> writes:
> > >    Any chance you'll continue the summaries?
> >
> > Not really.  All of this is sparked by my wanting to reduce the load,
> > really; I recently realized that I haven't hacked on *any* of my
> > projects (Oedipus, amkCrypto, etc.) lately.  Therefore I've dropped
>
>    Ah, ok, ok, that's right. Code is even more important! Good luck!
>
> Oleg.
> ----
>      Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
>            Programmers don't die, they just GOSUB without RETURN.
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 17
>    Date: Fri, 9 Feb 2001 21:07:51 +0530
>    From: "Sandipan Gangopadhyay" <sandipan at vsnl.com>
> Subject: Re: python-dev summary, Jan. 16-31
>
> +3. Saving electrons here ;-)
>
> Though my colleagues and I use a number of other languages, python is
> central in the architectures we work with. We strive to code in other
> languages to match upto the python style and philosophy. We believe in it.
> Watching it grow and evolve (via the dev summary) is, in my opinion, an
> important process in our own development as programmers.
>
> Regards,
>
> Sandipan
>
> ----- Original Message -----
> From: <mmillikan at vfa.com>
> Newsgroups: comp.lang.python
> To: <python-list at python.org>
> Sent: Friday, February 09, 2001 7:34 PM
> Subject: Re: python-dev summary, Jan. 16-31
>
>
> > +1
> >
> > Mark millikan
> > --
> > http://mail.python.org/mailman/listinfo/python-list
> >
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 18
>    Date: 9 Feb 2001 07:40:21 -0800
>    From: aahz at panix.com (Aahz Maruch)
> Subject: Re: nested scopes
>
> In article <96059d$din$1 at news.mathworks.com>,
> Joshua Marshall  <jmarshal at mathworks.com> wrote:
> >
> >Still this is a different issue than lexical scoping.  Consider:
> >
> >   class T:
> >      def m(self):
> >         print self.x
> >
> >   T.x = 'foo'
> >
> >   T().m()
> >
> >Here, T().m() prints "foo".  Lexically, x is not defined in m's scope.
> >I believe it was a good decision to require going through self to
> >access class/object fields.
>
> Note that you don't necessarily want to go through self to access class
> fields.  Here you're relying on the search through the class hierarchy
> to access T.x with self.x.  (I'm pretty sure you know this, but I don't
> want any newcomers getting confused.)
> --
>                       --- Aahz (Copyright 2001 by aahz at pobox.com)
>
> Androgynous poly kinky vanilla queer het    <*>
http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> "Actually, I've found that Usenet has helped a lot with my terror of
> social stupidity.  Seeing people I really like and respect behave like
> idiots in public has been very beneficial."  --Rachael Lininger
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 19
>    Date: 9 Feb 2001 07:36:49 -0800
>    From: aahz at panix.com (Aahz Maruch)
> Subject: Re: urllib.open('http://www.redherring.com/')
>
> In article <95th67$36j$1 at nnrp1.deja.com>,  <apederse at my-deja.com> wrote:
> >
> >I have some problems with urllib. Any URL works just fine except
> >http://www.redherring.com/ which returns nothing, no headers, no
> >nothing. (Red Herring works fine from any browser). I have tried to
> >rule out any possible causes, such as cache and problems regarding
> >headerfields by emulating a real browser request from netscape and
> >konqueror (using fancyURLopener) and I have monitored network traffic
> >using tcpdump.  Using tcpdump it looks like there is a request sendt
> >to vip.redherring.com and a response coming back.  Could anyone look
> >into this problem and see if this is related my particular computer and
> >network configurations or if it is a problem lying within urllib or
> >python.
>
> I've just tried with a straight telnet session, and that also fails.
> So, it's not just Python.
> --
>                       --- Aahz (Copyright 2001 by aahz at pobox.com)
>
> Androgynous poly kinky vanilla queer het    <*>
http://www.rahul.net/aahz/
> Hugs and backrubs -- I break Rule 6
>
> "Actually, I've found that Usenet has helped a lot with my terror of
> social stupidity.  Seeing people I really like and respect behave like
> idiots in public has been very beneficial."  --Rachael Lininger
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 20
>    Date: Fri, 09 Feb 2001 17:49:33 +0200
>    From: mix77 at usa.net
> Subject: Executing Sytem Commands
>
> How do you execute a sytem command  like "mv file1 file2" in Python?
> P.S. I'm working under Linux 2.x (redhat 7 distribution with default
> linux kernel).
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 21
>    Date: 9 Feb 2001 15:51:51 GMT
>    From: Joshua Marshall <jmarshal at mathworks.com>
> Subject: Re: global, was Re: None assigment
>
> Michael Hudson <mwh21 at cam.ac.uk> wrote:
> > Michael Hudson <mwh21 at cam.ac.uk> writes:
>
> >> Which version are you using?  They all print 4 for me, and all but
> >> foo() do modify the value of `b'.  foo() doesn't because what follows
> >> an
>
> > OTOH, that was a version that had the PEP 227 code checked into it.
> > With an older version, I see what you see.  Wierd.
>
> Sounds like a bug fix.
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 22
>    Date: Fri, 9 Feb 2001 19:04:55 +0300 (MSK)
>    From: Oleg Broytmann <phd at phd.pp.ru>
> Subject: Re: Executing Sytem Commands
>
> On Fri, 9 Feb 2001 mix77 at usa.net wrote:
> > How do you execute a sytem command  like "mv file1 file2" in Python?
>
>    Exactly like system. It is in module "os". So well:
>
> import os
> os.system("xsetbg foo.jpg")
>
> Oleg.
> ----
>      Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
>            Programmers don't die, they just GOSUB without RETURN.
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 23
>    Date: Fri, 09 Feb 2001 15:53:08 GMT
>    From: apederse at my-deja.com
> Subject: Re: urllib.open('http://www.redherring.com/')
>
> In article <3ditmj99pd.fsf at ute.cnri.reston.va.us>,
>   Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> > apederse at my-deja.com writes:
> > > to me it looks like the problem lies whithin python itself? is there
> any
> > > point in reporting such bugs when everyone will soon be using python
> 2.0
> > > anyway?
> >
> > I don't believe urllib or httplib were massively rewritten for 2.0
> > (nor will they be in 2.1), so if it's a bug, it's probably still
> > there.  On the other hand, if you try it with 2.0 and it works fine,
> > then there won't be much interest; a 1.5.3 release is not going to
> > happen.
> >
> > --amk
> >
> >
>
> a 1.5.3 release is a bit drastic ...a simple bugfix will do!
>
> /Asle
>
>
> Sent via Deja.com
> http://www.deja.com/
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 24
>    Date: Fri, 9 Feb 2001 16:11:08 -0000
>    From: Simon Brunning <SBrunning at trisystems.co.uk>
> Subject: RE: python-dev summary, Jan. 16-31
>
> > From: Andrew Kuchling [SMTP:akuchlin at mems-exchange.org]
> > * Announce PEP drafts and status changes to c.l.py.announce
> > * Write a set of introductory pages for Python development (this has
been
> >   brought up before).
> > * A read-only alias for python-dev would be useful for several people,
but
> >
> >   may require serious Mailman hackery (I don't know).
>
> Sounds to me like if these three could be achieved, 99.9% of c.l.py would
be
> more than happy. I know that *I* would.
>
> Cheers,
> Simon Brunning
> TriSystems Ltd.
> sbrunning at trisystems.co.uk
>
>
>
>
> -----------------------------------------------------------------------
> The information in this email is confidential and may be legally
privileged.
> It is intended solely for the addressee. Access to this email by anyone
else
> is unauthorised. If you are not the intended recipient, any disclosure,
> copying, distribution, or any action taken or omitted to be taken in
> reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
> accept liability for statements made which are clearly the senders own.
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 25
>    Date: Fri, 09 Feb 2001 16:20:10 +0000
>    From: Pedro Vale Lima <eq3pvl at eq.uc.pt>
> Subject: Re: I love Python
>
> Zamurai wrote:
>
> > "debugboy" <aaa at bbb.com> wrote:
> > > Test
> >
> > I love Python too ;-)
>
> Yeah, but have you asked Python if he loves you?
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
> ________________________________________________________________________
> ________________________________________________________________________
>
>
>




From jeremy at alum.mit.edu  Fri Feb  9 18:45:04 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 9 Feb 2001 12:45:04 -0500 (EST)
Subject: PEP status and python-dev summaries
Message-ID: <14980.11424.134036.495048@w221.z064000254.bwi-md.dsl.cnc.net>

I just scanned the responses on comp.lang.python to Andrew's
announcement that he would stopping write the python-dev summaries.
The respondents indicated that they found it hard to keep track of
what was going on with python development, particularly PEPs.

We're still learning how to use the PEP process.  It's been better for
2.1 than for 2.0, but still has some problems.  It sounds like the key
problem has been involving the community outside python-dev.

I would suggest a couple of changes, with the burden mostly falling on
Barry and me:

- Regular announcements of PEP creation and PEP status changes should
  be posted to comp.lang.python and c.l.p.a.

- The release status PEPs should be kept up to date and regularly
  posted to the same groups.

- We should have highly visible pointers from python.org to PEPs and
  other python development information.  I'm sure we do this as part
  of the Zopification plans that Guido mentioned.

- We should not approve PEPs that aren't announced on comp.lang.python
  with enough time for people to comment.

Jeremy



From hinsen at cnrs-orleans.fr  Fri Feb  9 19:27:31 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 09 Feb 2001 19:27:31 +0100
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org>
Message-ID: <m34ry3sodo.fsf@chinon.cnrs-orleans.fr>

Jeremy Hylton <jeremy at alum.mit.edu> writes:

> I would suggest a couple of changes, with the burden mostly falling on
> Barry and me:
...

Sounds good to me. I like the PEP system, but until now it looked as
if it was meant only for "the inner circle". Regular announcement
to the public are definitely a Good Thing.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From akuchlin at mems-exchange.org  Fri Feb  9 19:45:20 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 09 Feb 2001 13:45:20 -0500
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org>
Message-ID: <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>

Jeremy Hylton <jeremy at alum.mit.edu> writes:
> - Regular announcements of PEP creation and PEP status changes should
>   be posted to comp.lang.python and c.l.p.a.

Should the full text of PEPs be posted at some point?  Perhaps the
first draft, and significant revisions, should be posted so that
people can follow-up directly to the PEP.  "Significant" will be left
up to the PEP author; you wouldn't repost after every typo or minor
clarification, but if your first approach was completely bogus, you'd
post the massively revised version.

--amk


From andrew at intertrader.com  Fri Feb  9 20:25:23 2001
From: andrew at intertrader.com (Andrew Cooke)
Date: Fri, 09 Feb 2001 19:25:23 +0000
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
Message-ID: <3A844423.8AF2B87A@intertrader.com>

I've read somewhere in all this that the original intent was to avoid
this - IIRC to avoid too much pointless discussion.  As Python gets more
popular the S/N in the ng is decreasing and I can see the point.  The
important thing to realise is that this isn't meant (presumably) to
exclude the people posting to this thread; more to avoid a million
people endlessly debating each piece of syntax (eg += or list
comprehensions).

Some kind of hierarchy needn't be a bad thing - to some extent this is
the price you're paying for Python's exponential (ish) growth...

All IMHO - I'm not involved in Python, just curious about how this
problem is going to be handled.  How is it handled in Perl?  Is there a
Python equivalent of the porters, for example?

Good luck!
Andrew

Andrew Kuchling wrote:
> 
> Jeremy Hylton <jeremy at alum.mit.edu> writes:
> > - Regular announcements of PEP creation and PEP status changes should
> >   be posted to comp.lang.python and c.l.p.a.
> 
> Should the full text of PEPs be posted at some point?  Perhaps the
> first draft, and significant revisions, should be posted so that
> people can follow-up directly to the PEP.  "Significant" will be left
> up to the PEP author; you wouldn't repost after every typo or minor
> clarification, but if your first approach was completely bogus, you'd
> post the massively revised version.
> 
> --amk


From akuchlin at mems-exchange.org  Fri Feb  9 22:34:02 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 09 Feb 2001 16:34:02 -0500
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <3A844423.8AF2B87A@intertrader.com>
Message-ID: <3dd7cr8rsl.fsf@ute.cnri.reston.va.us>

Andrew Cooke <andrew at intertrader.com> writes:
> All IMHO - I'm not involved in Python, just curious about how this
> problem is going to be handled.  How is it handled in Perl?  Is there a
> Python equivalent of the porters, for example?

python-dev is the equivalent of perl5-porters.  However, Perl hasn't
previously had a formal mechanism such as PEPs; a bunch of RFCs were
written for Perl 6, but who knows if such formality will continue once
actual development starts?  It seems like a curious cultural
difference; c.l.python posters often suggest new language features,
while in the Perl and Tcl groups, there are very few postings which
suggest language changes.

--amk


From aahz at panix.com  Fri Feb  9 22:59:41 2001
From: aahz at panix.com (Aahz Maruch)
Date: 9 Feb 2001 13:59:41 -0800
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <3A844423.8AF2B87A@intertrader.com> <3dd7cr8rsl.fsf@ute.cnri.reston.va.us>
Message-ID: <961p8d$scs$1@panix3.panix.com>

In article <3dd7cr8rsl.fsf at ute.cnri.reston.va.us>,
Andrew Kuchling  <akuchlin at mems-exchange.org> wrote:
>
>python-dev is the equivalent of perl5-porters.  However, Perl hasn't
>previously had a formal mechanism such as PEPs; a bunch of RFCs were
>written for Perl 6, but who knows if such formality will continue once
>actual development starts?  It seems like a curious cultural
>difference; c.l.python posters often suggest new language features,
>while in the Perl and Tcl groups, there are very few postings which
>suggest language changes.

I can't speak for Tcl, but in the case of Perl, there are few enough
people outside the core development group who actually understand Perl
to suggest changes in the language.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From jeremy at alum.mit.edu  Sat Feb 10 02:13:31 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 9 Feb 2001 20:13:31 -0500 (EST)
Subject: PEP status and python-dev summaries
In-Reply-To: <961p8d$scs$1@panix3.panix.com>
References: <mailman.981742518.20824.python-list@python.org>
	<3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
	<3A844423.8AF2B87A@intertrader.com>
	<3dd7cr8rsl.fsf@ute.cnri.reston.va.us>
	<961p8d$scs$1@panix3.panix.com>
Message-ID: <14980.38331.795521.163586@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "AM" == Aahz Maruch <aahz at panix.com> writes:

  AM> I can't speak for Tcl, but in the case of Perl, there are few
  AM> enough people outside the core development group who actually
  AM> understand Perl to suggest changes in the language.

I was under the impression that Perl already had nearly every useful
language feature (and then some).  It's impossible to remove any of
them because it of the risk that there would become only one way to do
something.  If nothing there's nothing left to add and nothing to take
away, there isn't much discussion :-).

Jeremy




From katz at Glue.umd.edu  Fri Feb  9 23:53:05 2001
From: katz at Glue.umd.edu (Roy Katz)
Date: Fri, 9 Feb 2001 17:53:05 -0500
Subject: Heh,  Re: PEP status and python-dev summaries
In-Reply-To: <3dd7cr8rsl.fsf@ute.cnri.reston.va.us>
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <3A844423.8AF2B87A@intertrader.com> <3dd7cr8rsl.fsf@ute.cnri.reston.va.us>
Message-ID: <Pine.GSO.4.21.0102091739490.6320-100000@z.glue.umd.edu>

On 9 Feb 2001, Andrew Kuchling wrote:

> actual development starts?  It seems like a curious cultural
> difference; c.l.python posters often suggest new language features,
> while in the Perl and Tcl groups, there are very few postings which
> suggest language changes.

what's so inconsistent here?  IMHO, the only reason Python programmers ask
for more language features is because Python is not already
feature-saturated like Perl.  Perlers have a mammoth gorilla of a language
to contend with:  everyone has already requested features, and they've 
all gotten accepted (IMHO that's perl's problem; *all* the
proposals were accepted, even if they overlapped -- remember, there's more than one
way to do it! :-P)




Rehashing stuff I couldn't pass up to include:

couldn't interfaces be implemented as mixin classes?




Roey

gxis la revido! 
I hope I got that right :-)



From porter at et.byu.edu  Sat Feb 10 00:38:25 2001
From: porter at et.byu.edu (C. Porter Bassett)
Date: Fri, 9 Feb 2001 16:38:25 -0700
Subject: array vs. list
In-Reply-To: <Pine.GSO.4.21.0102091739490.6320-100000@z.glue.umd.edu>
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <3A844423.8AF2B87A@intertrader.com> <3dd7cr8rsl.fsf@ute.cnri.reston.va.us> <Pine.GSO.4.21.0102091739490.6320-100000@z.glue.umd.edu>
Message-ID: <Pine.GHP.4.21.0102091632590.26046-100000@don.et.byu.edu>

I have a couple of matrix/vector classes that I wrote.  The matrix is 4x4
and the vector is 4X1, just like the ones they use in OpenGL.  Anyway,
when I wrote it quick and dirty, I used all lists.  I just went back and
replaced lists with arrays where I could.  Much to my surprise, it was
about half again slower using arrays instead of lists!  I thought that
arrays were supposed to be more effecient than lists, but I guess they
aren't.  In case it would make a difference, this was done on a 400 MHZ
linux box and python 2.0

Yeah, I know that I should learn NumPy and implement it there, but one
thing at a time.

C. Porter Bassett



From thomas at xs4all.net  Wed Feb 14 14:37:45 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 14 Feb 2001 14:37:45 +0100
Subject: array vs. list
In-Reply-To: <Pine.GHP.4.21.0102091632590.26046-100000@don.et.byu.edu>; from porter@et.byu.edu on Fri, Feb 09, 2001 at 04:38:25PM -0700
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <3A844423.8AF2B87A@intertrader.com> <3dd7cr8rsl.fsf@ute.cnri.reston.va.us> <Pine.GSO.4.21.0102091739490.6320-100000@z.glue.umd.edu> <Pine.GHP.4.21.0102091632590.26046-100000@don.et.byu.edu>
Message-ID: <20010214143745.H4924@xs4all.nl>

On Fri, Feb 09, 2001 at 04:38:25PM -0700, C. Porter Bassett wrote:

> I have a couple of matrix/vector classes that I wrote.  The matrix is 4x4
> and the vector is 4X1, just like the ones they use in OpenGL.  Anyway,
> when I wrote it quick and dirty, I used all lists.  I just went back and
> replaced lists with arrays where I could.  Much to my surprise, it was
> about half again slower using arrays instead of lists!  I thought that
> arrays were supposed to be more effecient than lists, but I guess they
> aren't.  In case it would make a difference, this was done on a 400 MHZ
> linux box and python 2.0

Who said arrays would be faster ? Arrays are more efficient in storage,
because they use native arrays of native types rather than arrays of
pointers to Python objects which hold the native type (or something close to
it.) However, retrieval of a list element is a simple array index and
refcount-increment operation. Retrieval of an array element is an array
index operation and conversion to a Python object of the value found there.
This can easily trash any speed benifit you get from the better memory
footprint.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From paul.moore at uk.origin-it.com  Mon Feb 12 17:59:54 2001
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Mon, 12 Feb 2001 17:59:54 +0100
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <3A844423.8AF2B87A@intertrader.com>
Message-ID: <TBaIOsE1pRSp9Yzfz9eDDQ3d7G=S@4ax.com>

On Fri, 9 Feb 2001 20:25:23 +0100 , Andrew Cooke
<andrew at intertrader.com> wrote:

>All IMHO - I'm not involved in Python, just curious about how this
>problem is going to be handled.  How is it handled in Perl?  Is there a
>Python equivalent of the porters, for example?

The perl5 porters list is the equivalent of python-dev. The difference
is that P5P is basically an open list. This has some significant
downsides (P5P has had a history of degenerating into flamefests and
noise), but that is more likely a consequence of certain personality
issues and a lack of any sort of moderation on P5P than an essential
aspect of the process...

P5P is a good counterexample to the argument that python-dev should be
completely open, but IMHO there should be a middle ground somewhere.

Paul Moore.



From tim.one at home.com  Fri Feb  9 20:35:04 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 9 Feb 2001 14:35:04 -0500
Subject: PEP status and python-dev summaries
In-Reply-To: <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEGPIOAA.tim.one@home.com>

[Andrew]
> Should the full text of PEPs be posted at some point?

The PEP PEP (PEP 1) could/should be clearer about this, but that was always
the intent.  That's why PEPs already have a Post-History tag at the start
(and it wasn't meant to record events that happened after history ended
<wink>).




From jhylton at my-deja.com  Fri Feb  9 20:48:00 2001
From: jhylton at my-deja.com (Jeremy Hylton)
Date: Fri, 09 Feb 2001 19:48:00 GMT
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
Message-ID: <961hhc$gp1$1@nnrp1.deja.com>

In article <3dg0hn8zlr.fsf at ute.cnri.reston.va.us>,
  Andrew Kuchling <akuchlin at mems-exchange.org> wrote:

> Should the full text of PEPs be posted at some point?

I think the text would be too long to post directly.  A URL should be
sufficient.

--
-- Jeremy Hylton, <http://www.python.org/~jeremy/>


Sent via Deja.com
http://www.deja.com/


From dalke at acm.org  Sat Feb 10 04:35:36 2001
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 9 Feb 2001 20:35:36 -0700
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <961hhc$gp1$1@nnrp1.deja.com>
Message-ID: <962df1$bk8$1@slb1.atl.mindspring.net>

Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> Should the full text of PEPs be posted at some point?

Jeremy Hylton:
>I think the text would be too long to post directly.  A URL should be
>sufficient.

You must be one of the people lucky enough to be connected
to the net all the time.  I just have a phone line, so
I connect to my ISP, transfer my mail and newsgroup messages,
then disconnect.  When there are links to external sites,
I have to remember to go back and follow those links when
I reconnect.

I would therefore prefer that the first version of a PEP, and
significant changes (where the deltas are incomprehensible
without full context) be posted in full.  I download about
200 messages a day in c.l.py.  I doubt a PEP a week will have
a noticable impact on the transfer time.

On the other hand, I am moving next week and hopefully
should be able to get DSL - I couldn't get any sort of
broadband for this apartment.

                    Andrew
                    dalke at acm.org






From n8gray at caltech.edu.is.my.email.address  Sat Feb 10 08:02:45 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Fri, 9 Feb 2001 23:02:45 -0800
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <961hhc$gp1$1@nnrp1.deja.com>
Message-ID: <962pdj$kb0@gap.cco.caltech.edu>

Jeremy Hylton wrote:

> In article <3dg0hn8zlr.fsf at ute.cnri.reston.va.us>,
>   Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> 
> > Should the full text of PEPs be posted at some point?
> 
> I think the text would be too long to post directly.  A URL should be
> sufficient.

Many newsgroups post their entire FAQs periodically.  PEPs are a few 
hundred lines at most -- not a huge burden.  Besides, compared to the 
weekly traffic on c.l.p. a PEP is just a drop in the bucket.

To be kind to those with slow connections who may want to skip PEP texts, 
they could have a distinctive subject line format, e.g. 
        [PEP 666 Text] Curly-Braces for Delimiting Blocks

IMO, posting the text of PEPs to the newsgroup is a no-brainer.  It keeps 
the users informed of the language's direction at essentially zero cost.

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From sholden at holdenweb.com  Sat Feb 10 15:35:35 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 10 Feb 2001 09:35:35 -0500
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <961hhc$gp1$1@nnrp1.deja.com> <962pdj$kb0@gap.cco.caltech.edu>
Message-ID: <Eoch6.3137$Tn6.85197@e420r-atl2.usenetserver.com>

"Nathaniel Gray" <n8gray at caltech.edu.is.my.email.address> wrote in message
news:962pdj$kb0 at gap.cco.caltech.edu...
> Jeremy Hylton wrote:
>
> > In article <3dg0hn8zlr.fsf at ute.cnri.reston.va.us>,
> >   Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> >
> > > Should the full text of PEPs be posted at some point?
> >
> > I think the text would be too long to post directly.  A URL should be
> > sufficient.
>
> Many newsgroups post their entire FAQs periodically.  PEPs are a few
> hundred lines at most -- not a huge burden.  Besides, compared to the
> weekly traffic on c.l.p. a PEP is just a drop in the bucket.
>
> To be kind to those with slow connections who may want to skip PEP texts,
> they could have a distinctive subject line format, e.g.
>         [PEP 666 Text] Curly-Braces for Delimiting Blocks
>
Unless they are using a POP mail agent which insists on downloadingmessages
before displaying the headers...

> IMO, posting the text of PEPs to the newsgroup is a no-brainer.  It keeps
> the users informed of the language's direction at essentially zero cost.
>
Maybe c.l.py.announce?

regards
 Steve



From n8gray at caltech.edu.is.my.email.address  Tue Feb 13 06:44:54 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Mon, 12 Feb 2001 21:44:54 -0800
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <961hhc$gp1$1@nnrp1.deja.com> <962pdj$kb0@gap.cco.caltech.edu> <Eoch6.3137$Tn6.85197@e420r-atl2.usenetserver.com>
Message-ID: <96ahvl$lo@gap.cco.caltech.edu>

Steve Holden wrote:

> 
> "Nathaniel Gray" <n8gray at caltech.edu.is.my.email.address> wrote in message
> news:962pdj$kb0 at gap.cco.caltech.edu...
> > Jeremy Hylton wrote:
> >
> > > In article <3dg0hn8zlr.fsf at ute.cnri.reston.va.us>,
> > >   Andrew Kuchling <akuchlin at mems-exchange.org> wrote:
> > >
> > > > Should the full text of PEPs be posted at some point?
> > >
> > > I think the text would be too long to post directly.  A URL should be
> > > sufficient.
> >
> > Many newsgroups post their entire FAQs periodically.  PEPs are a few
> > hundred lines at most -- not a huge burden.  Besides, compared to the
> > weekly traffic on c.l.p. a PEP is just a drop in the bucket.
> >
> > To be kind to those with slow connections who may want to skip PEP
> > texts, they could have a distinctive subject line format, e.g.
> >         [PEP 666 Text] Curly-Braces for Delimiting Blocks
> >
> Unless they are using a POP mail agent which insists on
> downloadingmessages before displaying the headers...

You're getting c.l.py through a POP mail agent?  Yikes!  Still, I doubt 
adding a PEP every few weeks would significantly impact the total volume of 
the newsgroup.  And total volume is what we're talking about if you can't 
see the headers before you download messages -- it's all or nothing, right?


> 
> > IMO, posting the text of PEPs to the newsgroup is a no-brainer.  It
> > keeps the users informed of the language's direction at essentially zero
> > cost.
> >
> Maybe c.l.py.announce?

I guess that would provide _some_ reason to read c.l.py.announce.  :^)

Since we're likely to want to discuss PEPs, it seems like c.l.py would be 
more appropriate.

Maybe a compromise would be that an abstract should be posted to c.l.py 
with a link to the full text of the PEP.  There should be enough detail 
that readers will understand the idea of the change without going into the 
gory detail required of a full PEP.

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From chris at voodooland.net  Tue Feb 13 07:17:54 2001
From: chris at voodooland.net (Chris Watson)
Date: Tue, 13 Feb 2001 00:17:54 -0600 (CST)
Subject: FreeBSD and Py-Kqueue
In-Reply-To: <96ahvl$lo@gap.cco.caltech.edu>
Message-ID: <Pine.BSF.4.21.0102130008240.36475-100000@open-systems.net>

Hi all,

	Has anyone had experience with using python and the python wrapper
to kqueue in FreeBSD? I am authoring a book on postfix the MTA and I have
read posts about an asynchronous syslogd on linux. Allowing for greater
speed for postfix then a normal synchronous syslogd. Well FreeBSD doesn't
have the async option. So I believe I have two choices to test to see how
much this helps, either port Theodore Tso's syslogd to FreeBSD (icky) or
write a cheap syslogd with Python to do writes async. I was wondering if
kqueue would help syslogd out at all for increasing read/write
performance? Any thoughts on this?

--
=============================================================================
-Chris Watson         (316) 326-3862 | FreeBSD Consultant, FreeBSD Geek 
Work:              scanner at jurai.net | Open Systems Inc., Wellington, Kansas
Home:  scanner at deceptively.shady.org | http://open-systems.net
=============================================================================
WINDOWS: "Where do you want to go today?"
LINUX: "Where do you want to go tomorrow?"
BSD: "Are you guys coming or what?"
=============================================================================
GNU = Gnu's Non-portable & Unstable
irc.openprojects.net #FreeBSD -Join the revolution!
ICQ: 20016186




From aahz at panix.com  Tue Feb 13 19:43:12 2001
From: aahz at panix.com (Aahz Maruch)
Date: 13 Feb 2001 10:43:12 -0800
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <962pdj$kb0@gap.cco.caltech.edu> <Eoch6.3137$Tn6.85197@e420r-atl2.usenetserver.com> <96ahvl$lo@gap.cco.caltech.edu>
Message-ID: <96bv80$4dg$1@panix6.panix.com>

In article <96ahvl$lo at gap.cco.caltech.edu>,
Nathaniel Gray  <n8gray at caltech.edu.is.my.email.address> wrote:
>
>I guess that would provide _some_ reason to read c.l.py.announce.  :^)
>
>Since we're likely to want to discuss PEPs, it seems like c.l.py would be 
>more appropriate.

My personal preference would be a x-post to c.l.py and c.l.py.announce,
with followups to just c.l.py.  (Essentially, exactly the same way
news.announce.newgroups and news.groups work.)
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

'Gender' isn't a definition, it's a heuristic.  --Aahz


From n8gray at caltech.edu.is.my.email.address  Fri Feb 16 07:15:31 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Thu, 15 Feb 2001 22:15:31 -0800
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <962pdj$kb0@gap.cco.caltech.edu> <Eoch6.3137$Tn6.85197@e420r-atl2.usenetserver.com> <96ahvl$lo@gap.cco.caltech.edu> <96bv80$4dg$1@panix6.panix.com>
Message-ID: <96igt0$dao@gap.cco.caltech.edu>

Aahz Maruch wrote:

> In article <96ahvl$lo at gap.cco.caltech.edu>,
> Nathaniel Gray  <n8gray at caltech.edu.is.my.email.address> wrote:
> >
> >I guess that would provide _some_ reason to read c.l.py.announce.  :^)
> >
> >Since we're likely to want to discuss PEPs, it seems like c.l.py would be
> >more appropriate.
> 
> My personal preference would be a x-post to c.l.py and c.l.py.announce,
> with followups to just c.l.py.  (Essentially, exactly the same way
> news.announce.newgroups and news.groups work.)

Even better.

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From thomas at xs4all.net  Wed Feb 14 14:12:00 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 14 Feb 2001 14:12:00 +0100
Subject: PEP status and python-dev summaries
In-Reply-To: <96ahvl$lo@gap.cco.caltech.edu>; from n8gray@caltech.edu.is.my.email.address on Mon, Feb 12, 2001 at 09:44:54PM -0800
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <961hhc$gp1$1@nnrp1.deja.com> <962pdj$kb0@gap.cco.caltech.edu> <Eoch6.3137$Tn6.85197@e420r-atl2.usenetserver.com> <96ahvl$lo@gap.cco.caltech.edu>
Message-ID: <20010214141200.G4924@xs4all.nl>

On Mon, Feb 12, 2001 at 09:44:54PM -0800, Nathaniel Gray wrote:

> You're getting c.l.py through a POP mail agent?  Yikes!

There is a list<->usenet gateway for python-list at python.org (aka
comp.lang.python.) If I didn't work for the ISP that handles most of my
mail, I'd probably be retrieving it using POP as well. (instead of just
running mutt 24/7, on one of the shell boxes.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From mikael at isy.liu.se  Tue Feb 13 10:06:32 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 13 Feb 2001 10:06:32 +0100 (MET)
Subject: PEP status and python-dev summaries
In-Reply-To: <962pdj$kb0@gap.cco.caltech.edu>
Message-ID: <XFMail.010213100632.mikael@isy.liu.se>

On 10-Feb-01 Nathaniel Gray wrote:
 >          [PEP 666 Text] Curly-Braces for Delimiting Blocks

It's only natural that the particular PEP with that number is dealing
with something evil.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    13-Feb-01
Time:    10:05:24

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From sholden at holdenweb.com  Fri Feb  9 22:00:33 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 9 Feb 2001 16:00:33 -0500
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
Message-ID: <yXYg6.1$Tn6.883@e420r-atl2.usenetserver.com>

"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3dg0hn8zlr.fsf at ute.cnri.reston.va.us...
> Jeremy Hylton <jeremy at alum.mit.edu> writes:
> > - Regular announcements of PEP creation and PEP status changes should
> >   be posted to comp.lang.python and c.l.p.a.
>
> Should the full text of PEPs be posted at some point?  Perhaps the
> first draft, and significant revisions, should be posted so that
> people can follow-up directly to the PEP.  "Significant" will be left
> up to the PEP author; you wouldn't repost after every typo or minor
> clarification, but if your first approach was completely bogus, you'd
> post the massively revised version.
>
Since most mail users will have web access, might I suggest a weekly summary
of changes rather than complete PEP postings? Anyone who wants the full text
should be able tp access it from sourceforge. Configuration control would be
nice, to allow differencing, but that would be an option only few would use
(I suspect).

regards
 Steve




From jeremy at alum.mit.edu  Sat Feb 10 02:14:33 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 9 Feb 2001 20:14:33 -0500 (EST)
Subject: PEP status and python-dev summaries
In-Reply-To: <yXYg6.1$Tn6.883@e420r-atl2.usenetserver.com>
References: <mailman.981742518.20824.python-list@python.org>
	<3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
	<yXYg6.1$Tn6.883@e420r-atl2.usenetserver.com>
Message-ID: <14980.38393.244828.153715@w221.z064000254.bwi-md.dsl.cnc.net>

The PEPs are kept under CVS control, which supports differencing
directly.  So you can read the current PEPs via python.sourceforge.net
and you can review changes using the CVS repository at
sourceforge.net/projects/python. 

Jeremy



From amk at mira.erols.com  Mon Feb 12 03:32:01 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 12 Feb 2001 02:32:01 GMT
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <yXYg6.1$Tn6.883@e420r-atl2.usenetserver.com> <mailman.981767712.27190.python-list@python.org>
Message-ID: <slrn98ejjp.h5.amk@mira.erols.com>

On Fri, 9 Feb 2001 20:14:33 -0500 (EST), 
	Jeremy Hylton <jeremy at alum.mit.edu> wrote:
>The PEPs are kept under CVS control, which supports differencing
>directly.  So you can read the current PEPs via python.sourceforge.net
>and you can review changes using the CVS repository at
>sourceforge.net/projects/python. 

Following up to a posted PEP and adding interspersed comments is much
easier if the full text of the PEP is posted; any mail or news reader
probably has a single keystore for it.  If the only thing posted is
the URL, then it requires more steps to do that by hand-- pull up the
PEP in your Web browser, get an ASCII version (what the poor benighted
souls who use Netscape do to get plain text, I don't know), insert '>'
at the start of every line, annotate it, etc.  IMHO the full text of
every PEP should have been posted at least once in the PEP's lifetime.

--amk



From tim.one at home.com  Mon Feb 12 05:19:27 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 11 Feb 2001 23:19:27 -0500
Subject: PEP status and python-dev summaries
In-Reply-To: <slrn98ejjp.h5.amk@mira.erols.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEMJIOAA.tim.one@home.com>

[A.M. Kuchling]
> ...
> IMHO the full text of every PEP should have been posted at least
> once in the PEP's lifetime.

Like I said, that was the intent, and that's what the Post-History tag in
each PEP is there to record.  People should have been posting PEPs all
along!

There are 43 PEPs today, consuming a grand total of about 300Kb, for an
average of less than 10 Kb/PEP.  We would probably average about 2 PEP posts
per week.  Anyone worried about 20Kb/week should take up a hobby better
suited to their irrationalities <wink>.




From adaminthedomaindevtty.net  Sat Feb 10 00:34:56 2001
From: adaminthedomaindevtty.net (Adam Logghe)
Date: Fri, 9 Feb 2001 15:34:56 -0800
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
Message-ID: <3a847ca3$1_1@news.nwlink.com>

Thanks for the summaries Andrew. I will be very sorry to see them go.

While I didn't see any need to comment, I was very glad to have an idea of
what was in the pipe. I would very much like to see the PEPs themselves as
they come through so I can plan ahead for changes to my own code as Python
moves along.

I would suggest that the full text of the PEP should be posted for every
change.

While this would be painful on a mailing list, I would not find it to be an
issue whatsoever on a newsgroup.

As a guide to what has changed, ideally a diff from the old version would be
attached on the end of the new PEP version. This would allow a quick visual
look at the changes. It may not be easy to make perfect sense of the change,
it would quickly point out whether the PEP had undergone minor or major
change.

I would suggest that publishing often provides at least two advantages.

A. Eliminating the decision of whether or not to publish will eliminate the
tendency for the  PEP undergo significant "creep" as minor changes add up to
significant changes in actual function. I think anyone that programs can
agree that minor syntax changes can and do have profound effects.

B. Having the PEP show up in front of people at multiple points during it's
life will expose it to more eyeballs.

If full posting is not pursued perhaps you might do something like-

A weekly summary of the mail traffic on the list simply by listing of the
active PEPs and then the number of times that PEP number appears in email
subject or body over the week. I would hope that people try to include the
PEP number in the subject line as a matter of courtesy anyway.

While this may not be the most accurate it would give us a simple
representation of the activity around a PEP and would perhaps spur further
investigations.

Thanks again. Sorry to see the Summary go but you shouldn't let it get in
the way of coding. ; )

I will look forward to however Jeremy and Barry decide to proceed.

Adam Logghe
adam in the domain devtty.net


"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3dg0hn8zlr.fsf at ute.cnri.reston.va.us...
> Jeremy Hylton <jeremy at alum.mit.edu> writes:
> > - Regular announcements of PEP creation and PEP status changes should
> >   be posted to comp.lang.python and c.l.p.a.
>
> Should the full text of PEPs be posted at some point?  Perhaps the
> first draft, and significant revisions, should be posted so that
> people can follow-up directly to the PEP.  "Significant" will be left
> up to the PEP author; you wouldn't repost after every typo or minor
> clarification, but if your first approach was completely bogus, you'd
> post the massively revised version.
>
> --amk




From robin at jessikat.fsnet.co.uk  Fri Feb 16 11:38:31 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 16 Feb 2001 10:38:31 +0000
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org>
 <3dg0hn8zlr.fsf@ute.cnri.reston.va.us> <3a847ca3$1_1@news.nwlink.com>
Message-ID: <ZUnniVAnMQj6EwL2@jessikat.demon.co.uk>

In article <3a847ca3$1_1 at news.nwlink.com>, Adam Logghe <adam at in.the> writes
>Thanks for the summaries Andrew. I will be very sorry to see them go.
>
>While I didn't see any need to comment, I was very glad to have an idea of
>what was in the pipe. I would very much like to see the PEPs themselves as
>they come through so I can plan ahead for changes to my own code as Python
>moves along.
>
>I would suggest that the full text of the PEP should be posted for every
>change.
...
The PEPs are in cvs so you can use a web browser to see them at all times and do
differencing between versions etc the URL is

http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/python/nondist/peps/?cvsroot=python
-- 
Robin Becker


From peter at engcorp.com  Sat Feb 10 19:15:55 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 10 Feb 2001 13:15:55 -0500
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org>
Message-ID: <3A85855B.89DCB5EF@engcorp.com>

Jeremy Hylton wrote:
> 
> I would suggest a couple of changes, with the burden mostly falling on
> Barry and me:
> 
> - Regular announcements of PEP creation and PEP status changes should
>   be posted to comp.lang.python and c.l.p.a.
> 
> - The release status PEPs should be kept up to date and regularly
>   posted to the same groups.

You mentioned "Zopification" in another item.  Wouldn't both
of the above be fairly simple if the PEPs were maintained with
Zope instead of wherever they are now?  

We are using Zope to manage structured information similar to
the PEPs and it is an easy matter to configure things so
that upon a status field changing, a post is automatically
made.  I no longer allow manual processes that don't add
value relative to an automated one to waste our time.  It just
seems silly in this day and age to think of someone manually 
extracting information and posting it to c.l.p this way.


From tjg at exceptionalminds.com  Fri Feb  9 18:45:37 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Fri, 9 Feb 2001 09:45:37 -0800
Subject: time.sleep() and cpu cycles
Message-ID: <20010209094537.P17173@trufflehunter.avalongroup.net>

Hi,

We were having some problems on one of our Linux boxes with inodes being
allocated and not deallocated. (kernel 2.2.14).

So, wanting to understand this behaviour, I wrote a little python logging
script that reads the inode info out of /proc every 10 seconds, and then
uses python's logging facilities to write out a log file with what it found.
This solution worked wonderfully, and gave us all the information we needed
about what was going on on the box just before it would die.

We have resolved the problem, but left the logging programme running just
for fun.

When the box is under load with users logged running things, the python
process does not interfere with anything.

However, on an unloaded system, the python process is taking 60% of the CPU.
Maybe that's the way it should be, but it sure seemed odd to me. I use
time.sleep() for my delay loop, and was wondering if there is a better way
to put a task to sleep, or if what I'm seeing is normal behaviour.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  23 days 21:51 hours ago<<



From tim.one at home.com  Fri Feb  9 20:19:34 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 9 Feb 2001 14:19:34 -0500
Subject: time.sleep() and cpu cycles
In-Reply-To: <20010209094537.P17173@trufflehunter.avalongroup.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEGOIOAA.tim.one@home.com>

[Timothy Grant]
> ... Linux ... (kernel 2.2.14).
>
> ... python logging script that reads the inode info out of /proc
> every 10 seconds, and then uses python's logging facilities

What are those?  "print"?

> to write out a log file with what it found.
> ...
> When the box is under load with users logged running things, the
> python process does not interfere with anything.
> ...
> However, on an unloaded system, the python process is taking 60%
> of the CPU.  Maybe that's the way it should be, but it sure seemed
> odd to me.

Me too.

> I use time.sleep() for my delay loop, and was wondering if there
> is a better way to put a task to sleep,

No.

> or if what I'm seeing is normal behaviour.

Certainly not normal behavior under Windows <wink>.  But unlike most of
Python's platform-dependent functions, time.sleep() isn't just a thin
wrapper around the C library function of the same name.  It eventually calls
Modules/timemodule.c's floatsleep(), which strives to shut the task up with
the finest-grain precision it can get on each platform, by hook or by crook.

In the Linux case, I expect it uses the first #ifdef'ed block:

#if defined(HAVE_SELECT) && !defined(__BEOS__)
	struct timeval t;
	double frac;
	frac = fmod(secs, 1.0);
	secs = floor(secs);
	t.tv_sec = (long)secs;
	t.tv_usec = (long)(frac*1000000.0);
	Py_BEGIN_ALLOW_THREADS
	if (select(0, (fd_set *)0, (fd_set *)0, (fd_set *)0, &t) != 0) {
#ifdef EINTR
		if (errno != EINTR) {
#else
		if (1) {
#endif
			Py_BLOCK_THREADS
			PyErr_SetFromErrno(PyExc_IOError);
			return -1;
		}
	}
	Py_END_ALLOW_THREADS

So one thing to check is whether a tiny C program employing this same abuse
of select() also chews up absurd amounts of CPU time.  In that case, it's
select()'s fault; else the CPU utilization is somewhere you haven't yet
suspected.




From tjg at exceptionalminds.com  Sat Feb 10 02:53:14 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Fri, 9 Feb 2001 17:53:14 -0800
Subject: time.sleep() and cpu cycles
In-Reply-To: <LNBBLJKPBEHFEDALKOLCCEGOIOAA.tim.one@home.com>; from tim.one@home.com on Fri, Feb 09, 2001 at 02:19:34PM -0500
References: <20010209094537.P17173@trufflehunter.avalongroup.net> <LNBBLJKPBEHFEDALKOLCCEGOIOAA.tim.one@home.com>
Message-ID: <20010209175314.A30331@trufflehunter.avalongroup.net>

On Fri, Feb 09, 2001 at 02:19:34PM -0500, Tim Peters wrote:
> [Timothy Grant]
> > ... Linux ... (kernel 2.2.14).
> >
> > ... python logging script that reads the inode info out of /proc
> > every 10 seconds, and then uses python's logging facilities
> 
> What are those?  "print"?

actually, syslog. I once wrote an app where I logged everything using print
statements, syslog just brings too many nice bits to the game to not use it.
 
> > to write out a log file with what it found.
> > ...
> > When the box is under load with users logged running things, the
> > python process does not interfere with anything.
> > ...
> > However, on an unloaded system, the python process is taking 60%
> > of the CPU.  Maybe that's the way it should be, but it sure seemed
> > odd to me.
> 
> Me too.

We're gonna go back and look some more, just because we are curious types!

 
> > I use time.sleep() for my delay loop, and was wondering if there
> > is a better way to put a task to sleep,
> 
> No.

Glad to hear it. I thought it must be the "right" way as I saw no mention of
any other method of doing it, but I thought I might be missing something
too.
 
> > or if what I'm seeing is normal behaviour.
> 
> Certainly not normal behavior under Windows <wink>.  But unlike most of
> Python's platform-dependent functions, time.sleep() isn't just a thin
> wrapper around the C library function of the same name.  It eventually calls
> Modules/timemodule.c's floatsleep(), which strives to shut the task up with
> the finest-grain precision it can get on each platform, by hook or by crook.

Thanks Tim for the informative and enlightening post. Much appreciated, I
thought maybe I had run into one of those well coded extremely tight polling
loops that used to crop up DOS programmes<wink>.



-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  24 days  5:58 hours ago<<



From fm_duende at yahoo.com  Fri Feb  9 18:56:39 2001
From: fm_duende at yahoo.com (steveFarris)
Date: Fri, 09 Feb 2001 17:56:39 GMT
Subject: open("ls -l |", "r") ??
Message-ID: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>

Hi all, is some variation of  the above possible in Python?





From grey at despair.rpglink.com  Fri Feb  9 19:14:27 2001
From: grey at despair.rpglink.com (Steve Lamb)
Date: Fri, 09 Feb 2001 18:14:27 -0000
Subject: open("ls -l |", "r") ??
References: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>
Message-ID: <slrn988cs2.b1e.grey@teleute.dmiyu.org>

On Fri, 09 Feb 2001 17:56:39 GMT, steveFarris <fm_duende at yahoo.com> wrote:
>Hi all, is some variation of  the above possible in Python?

import os
foo = os.popen("ls -l")
for line in foo.readlines():
    print line,

-- 
         Steve C. Lamb         | I'm your priest, I'm your shrink, I'm your
         ICQ: 5107343          | main connection to the switchboard of souls.
-------------------------------+---------------------------------------------


From nas at arctrix.com  Fri Feb  9 19:29:43 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Fri, 9 Feb 2001 10:29:43 -0800
Subject: open("ls -l |", "r") ??
In-Reply-To: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>; from fm_duende@yahoo.com on Fri, Feb 09, 2001 at 05:56:39PM +0000
References: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>
Message-ID: <20010209102943.A20059@glacier.fnational.com>

On Fri, Feb 09, 2001 at 05:56:39PM +0000, steveFarris wrote:
> Hi all, is some variation of  the above possible in Python?

Look at os.popen().  You probably be using the directory methods
in os rather than forking two processes to get this data.

  Neil



From darked at my-deja.com  Fri Feb  9 19:30:55 2001
From: darked at my-deja.com (darked at my-deja.com)
Date: Fri, 09 Feb 2001 18:30:55 GMT
Subject: open("ls -l |", "r") ??
References: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>
Message-ID: <961d0v$cg8$1@nnrp1.deja.com>

Hi Steve,

this works for me:

import os
file_list = os.popen('ls -l').readlines()

Greetings

Darek


Sent via Deja.com
http://www.deja.com/


From tjg at exceptionalminds.com  Fri Feb  9 19:32:32 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Fri, 9 Feb 2001 10:32:32 -0800
Subject: open("ls -l |", "r") ??
In-Reply-To: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>; from fm_duende@yahoo.com on Fri, Feb 09, 2001 at 05:56:39PM +0000
References: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>
Message-ID: <20010209103232.U17173@trufflehunter.avalongroup.net>

On Fri, Feb 09, 2001 at 05:56:39PM +0000, steveFarris wrote:
> Hi all, is some variation of  the above possible in Python?
> 

>>> import os
>>> cmd = os.popen('ls -l')
>>> for l in cmd.readlines():
...   print l

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  23 days 22:41 hours ago<<



From rozen at rgv.hp.com  Fri Feb  9 19:24:10 2001
From: rozen at rgv.hp.com (Don Rozenberg)
Date: Fri, 09 Feb 2001 10:24:10 -0800
Subject: [ANN] PAGE 0.1 - A drag and drop GUI generator for Python
Message-ID: <3A8435C9.CA99421@rgv.hp.com>

AGE is a tool which helps to create GUI for Python programs.  It uses
Tk and Tix widgets.  PAGE is not an end-all, be-all tool, but rather
one that attempts to ease the burden on the Python programmer. It is
aimed at the user who will put up with a less-than-general GUI
capability in order to get an easily generated GUI. It does not build
an entire application but rather is aimed at building a single GUI
window.   PAGE is built on version 1.2.2 of Visual Tcl, version
8.1.1 of Tix, and Tcl/Tk version 8.3.2. It works with Python 2.0.
This is an early version which run on Linux and Windows/NT. It
supports many Tk widgets and a subset of the Tix widget set.  The user
will have to acquire and install Tix 8.1.1 and tcl8.3.2 to use this
program.
It is released under the GPL.

The PAGE web site is

 http://www.geocities.com/page_python




--
Don Rozenberg
rozen at rgv.hp.com
408-343-6266





From support at internetdiscovery.com  Sat Feb 10 11:26:31 2001
From: support at internetdiscovery.com (Mike Clarkson)
Date: Sat, 10 Feb 2001 10:26:31 GMT
Subject: [ANN] PAGE 0.1 - A drag and drop GUI generator for Python
References: <3A8435C9.CA99421@rgv.hp.com>
Message-ID: <3a851656.3694347@24.0.228.33>

On Fri, 09 Feb 2001 10:24:10 -0800, Don Rozenberg <rozen at rgv.hp.com>
wrote:

>AGE is a tool which helps to create GUI for Python programs.  It uses
>Tk and Tix widgets.  PAGE is not an end-all, be-all tool, but rather
>one that attempts to ease the burden on the Python programmer. It is
>aimed at the user who will put up with a less-than-general GUI
>capability in order to get an easily generated GUI. It does not build
>an entire application but rather is aimed at building a single GUI
>window.   PAGE is built on version 1.2.2 of Visual Tcl, version
>8.1.1 of Tix, and Tcl/Tk version 8.3.2. It works with Python 2.0.
>This is an early version which run on Linux and Windows/NT. It
>supports many Tk widgets and a subset of the Tix widget set.  The user
>will have to acquire and install Tix 8.1.1 and tcl8.3.2 to use this
>program.
>It is released under the GPL.

Great to hear of this - I'll try it out. I've been maintaning
Python/Tix and am glad to see it being used.

If you are working inside HP with Tix and have a spare minute,
could you send me a short note on what HP's type of uses
and applications of Tix are?

Many thanks,

Mike.


From bobhicks at adelphia.net  Fri Feb  9 22:40:59 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Fri, 09 Feb 2001 21:40:59 GMT
Subject: CGI
Message-ID: <B6A9CE15.A3FD%bobhicks@adelphia.net>

I would like to have Python handle all my forms via cgi. I have access to
everything to make changes. Where is a good resource to get information on
this?

Right now I am just looking to get a mailer working...I have formmail.py but
I am getting an error so I need more info on how to set everything up.

- Bob



From stephen_purcell at yahoo.com  Sat Feb 10 12:31:08 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sat, 10 Feb 2001 12:31:08 +0100
Subject: CGI
In-Reply-To: <B6A9CE15.A3FD%bobhicks@adelphia.net>; from bobhicks@adelphia.net on Fri, Feb 09, 2001 at 09:40:59PM +0000
References: <B6A9CE15.A3FD%bobhicks@adelphia.net>
Message-ID: <20010210123108.C16696@freedom.puma-ag.com>

Robert L Hicks wrote:
> I would like to have Python handle all my forms via cgi. I have access to
> everything to make changes. Where is a good resource to get information on
> this?

See the standard docs at:

     http://www.python.org/doc/lib/module-cgi.html

Or try something really crazy, like a web search:

     http://www.google.com/search?q=python+cgi

Best wishes,

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From res1st at gmx.de  Fri Feb  9 23:17:03 2001
From: res1st at gmx.de (Ingo 'resISt' Siebert)
Date: Fri, 9 Feb 2001 23:17:03 +0100
Subject: modify files?
Message-ID: <961q8q$7qk$1@sapa.inka.de>

Hi,
i?m very new to Python but i have to learn that. So i?m starting with litte
programs but i have now a big problem.
I have (Win-)Files which contains something like

fjioejfejfkljklTCHAT "Blablabla" kfoefeisdfidjsfmiosdf...TCHAT "Blablabla"
....

(tchat are also in low letters)

So i want to replace all Blablabla's with a some chars (i did choose *)
So it should look like

fjioejfejfkljklTCHAT "*********" kfoefeisdfidjsfmiosdf


So I wrote the following Program.

def demorem():
    filea = os.stat('c:/tmp/test.dat')[ST_SIZE]
    print "Filesize:" , filea , "Bytes"
    print "Please wait..."
    f=open('c:/tmp/test.dat', 'r+')
    fpos= int(0)
    while fpos < filea:
        f.seek(fpos)
        fstr=f.read(5)
        if fstr == 'tchat':
                print "tchat found"
                fpos=fpos+7
                f.seek(fpos)
                fstr2=f.read(1)
                i=int(0)
                ssign='"'
                while fstr2 != ssign :
                    f.write('*')
                    print fstr2, "-> *"
                    i=i+1
                    if i > 200 :
                        print "Error: someone said too much"
                        break
                    fpos=fpos+1
                    f.seek(fpos)
                    fstr2=f.read(1)
       fpos = fpos+1
 f.close()
 print "Done."


I also use the following at the beginning.

import os, sys
from stat import *

I don?t know if i really need them both, but i will try it later. :)

Now my Problem. If i start my program then it seems that all is working.
"tchat found"
the chars which sould be replaced are listet correct on my screen

tchat found
X -> *
Y -> *
...

But my file is still the old one, nothing were replaced.
Why does this
f.write('*')
nothing do to my file?

And after my program is halted, my Test-File is still in use, said Windows
when i wanted to modify it. Is
f.close()
also wrong?

It would be great if you could help me.
bye
Ingo Siebert




From cfelling at iae.nl  Sat Feb 10 20:42:20 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 10 Feb 2001 20:42:20 +0100
Subject: modify files?
References: <961q8q$7qk$1@sapa.inka.de>
Message-ID: <9645is$4jv$1@animus.fel.iae.nl>

Ingo 'resISt' Siebert <res1st at gmx.de> wrote:
> Hi,
> i?m very new to Python but i have to learn that. So i?m starting with litte
> programs but i have now a big problem.

Welcome aboard, and let my tell you right away:
often problems with Python are not big at all:)
Just a matter of picking the right modules (re in this case).

> I have (Win-)Files which contains something like
...<clear problem description snipped>

> So I wrote the following Program.
...<program with dreaded read/write file snipped

A different (and to me simpler) approach would be:



# First we create a matching machine that matches tchat, TCHAT, tChAt
# and all other casings of tchat. It will look for occurences of tchat
# followed by a space followed by anything inside double quotes. The
# use of ( and ) tells the match machine to group what's in between
# those braces (here anything but a double quote) and set it aside for
# later referencing.

import re
match_tchat = re.compile(r'tchat "([^"]*)"', re.IGNORECASE)

# We will need an auxiliar function that will allow us to define the
# replacement string in relation to the string to be replaced.  This
# function will be passed an match-object that will tell what was
# matched if asked so with the group(1) method call.

def replacer(match_obj):
   return "*" * len(match_obj.group(1))

# Now we read the whole file at once, and use the matcher
# machinery to replace all tchat's arguments in one big sweep.
# All that's left is to write this beast to disk again.

data = open("c:/tmp/test.dat").read()
data = match_tchat.sub(replacer, data)
open("c:/tmp/test.dat", "w").write(data)


-- 
groetjes, carel


From marvind at glue.umd.edu  Fri Feb  9 23:29:26 2001
From: marvind at glue.umd.edu (Arvind Mani)
Date: Fri, 9 Feb 2001 17:29:26 -0500
Subject: New to Tkinter
Message-ID: <961qd3$50l$1@hecate.umd.edu>

Hi,

I  need a menu based GUI. Clicking on a menu item should open a new dialog
box. I tried to do this - only the application main window disappears when
the dialog box corresponding to a menu item pops up. Also it has the same
title as the application main window.

Thanks,
Arvind




From johngrayson at home.com  Sun Feb 11 00:54:35 2001
From: johngrayson at home.com (John Grayson)
Date: Sat, 10 Feb 2001 23:54:35 GMT
Subject: New to Tkinter
References: <961qd3$50l$1@hecate.umd.edu>
Message-ID: <964kbq$su9$1@nnrp1.deja.com>

In article <961qd3$50l$1 at hecate.umd.edu>,
  "Arvind Mani" <marvind at glue.umd.edu> wrote:
> Hi,
>
> I  need a menu based GUI. Clicking on a menu item should open a new dialog
> box. I tried to do this - only the application main window disappears when
> the dialog box corresponding to a menu item pops up. Also it has the same
> title as the application main window.
>
> Thanks,
> Arvind
>

If you want help, you'll need to post some (or all) of your code
so that we can see what you're doing. In case you're not aware,
menus work pretty well...

    John


Sent via Deja.com
http://www.deja.com/


From phlip_cpp at my-deja.com  Fri Feb  9 23:48:24 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: Fri, 9 Feb 2001 14:48:24 -0800
Subject: Graph (cyclic kind, not pretty picture kind) rendering algorithm?
Message-ID: <rw_g6.54$q04.78888@news.pacbell.net>

Python for Pyros:

Click here, then scoot down about 2 pages:

        http://www.tomsawyer.com/glt/index.html

On the left we see a Graph (a math object describing vertices connected by 
edges), and on the right's the same graph drawn with all the lengths of the 
lines balanced.

Is there a Python library out there that can draw a graph like that? Or 
which can take a graph as an input and return XY coordinates telling where 
to place each vertice and each line?

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  http://users.deltanet.com/~tegan/home.html  --


From s713221 at student.gu.edu.au  Mon Feb 12 09:38:04 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Mon, 12 Feb 2001 18:38:04 +1000
Subject: Graph (cyclic kind, not pretty picture kind) rendering algorithm?
References: <rw_g6.54$q04.78888@news.pacbell.net>
Message-ID: <3A87A0EC.BA802E3E@student.gu.edu.au>

Phlip wrote:
> 
> Python for Pyros:
> 
> Click here, then scoot down about 2 pages:
> 
>         http://www.tomsawyer.com/glt/index.html
> 
> On the left we see a Graph (a math object describing vertices connected by
> edges), and on the right's the same graph drawn with all the lengths of the
> lines balanced.
> 
> Is there a Python library out there that can draw a graph like that? Or
> which can take a graph as an input and return XY coordinates telling where
> to place each vertice and each line?

I don't know of any, but I'd be interested if anyone else could direct
us to one, I need exactly this for a chem editor I'm building at the
moment.

Joal Heagney/AncientHart


From robin at jessikat.fsnet.co.uk  Mon Feb 12 11:11:03 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 12 Feb 2001 10:11:03 +0000
Subject: Graph (cyclic kind, not pretty picture kind) rendering algorithm?
References: <rw_g6.54$q04.78888@news.pacbell.net>
 <3A87A0EC.BA802E3E@student.gu.edu.au>
Message-ID: <Jc1OOjA3a7h6EwpP@jessikat.fsnet.co.uk>

In article <3A87A0EC.BA802E3E at student.gu.edu.au>,
s713221 at student.gu.edu.au writes
>Phlip wrote:
>> 
>> Python for Pyros:
>> 
>> Click here, then scoot down about 2 pages:
>> 
>>         http://www.tomsawyer.com/glt/index.html
>> 
>> On the left we see a Graph (a math object describing vertices connected by
>> edges), and on the right's the same graph drawn with all the lengths of the
>> lines balanced.
>> 
>> Is there a Python library out there that can draw a graph like that? Or
>> which can take a graph as an input and return XY coordinates telling where
>> to place each vertice and each line?
>
>I don't know of any, but I'd be interested if anyone else could direct
>us to one, I need exactly this for a chem editor I'm building at the
>moment.
>
>Joal Heagney/AncientHart
Well you translate the java demo code to do this. There was also some
stuff in Dr Dobbs (in 1999 I think) on the spring/charge model for
displaying graphs.

A starting place for such efforts might be the demo/guido/tkinter/electr
ons.py file. Which shows how to put the nodes onto a canvas. 
-- 
Robin Becker


From dalke at acm.org  Tue Feb 13 03:26:00 2001
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 12 Feb 2001 19:26:00 -0700
Subject: Graph (cyclic kind, not pretty picture kind) rendering algorithm?
References: <rw_g6.54$q04.78888@news.pacbell.net> <3A87A0EC.BA802E3E@student.gu.edu.au>
Message-ID: <96a61a$9ln$1@slb1.atl.mindspring.net>

s713221 at student.gu.edu.au wrote:
>I don't know of any, but I'd be interested if anyone else could direct
>us to one, I need exactly this for a chem editor I'm building at the
>moment.
>
>Joal Heagney/AncientHart

Having been involved in several of these discussions,
don't use that approach.  There is a lot of chemistry
involved in making a chemical layout algorithm look
correct.

Your best bet is to look at how JMDraw
  http://vanilla.ice.mpg.de/~stein/projects/SmilesViewer/
or JChemPaint
  http://jchempaint.sourceforge.net/
do their layout.  The latter is LGPL'ed.  There isn't
much other publically available source code to look
at for development of new layout algorithms.

Even they won't be exactly what a chemist wants, because
chemists disagree with what they expect.  Eg, many
companies have guidelines on how different families
of compounds should be laid out, so the only real
solution involves a lot of substructure matching against
a set of templates.  Or you hire someone to draw
every molecule, which is done!

                    Andrew
                    dalke at acm.org





From -$Paul$- at verence.demon.co.uk  Sun Feb 18 23:27:59 2001
From: -$Paul$- at verence.demon.co.uk (Paul Wright)
Date: 18 Feb 2001 22:27:59 -0000
Subject: Graph (cyclic kind, not pretty picture kind) rendering algorithm?
References: <rw_g6.54$q04.78888@news.pacbell.net>
Message-ID: <96pi9f$3rg$1@verence.demon.co.uk>

In article <rw_g6.54$q04.78888 at news.pacbell.net>,
Phlip  <phlip_cpp at my-deja.com> wrote:
>Python for Pyros:
>
>Click here, then scoot down about 2 pages:
>
>        http://www.tomsawyer.com/glt/index.html
>
>On the left we see a Graph (a math object describing vertices connected by 
>edges), and on the right's the same graph drawn with all the lengths of the 
>lines balanced.
>
>Is there a Python library out there that can draw a graph like that? Or 
>which can take a graph as an input and return XY coordinates telling where 
>to place each vertice and each line?

AT&T's GraphViz tools will do directed and undirected graph drawing,
though I'm not sure what "balanced" means in this context so it may be
that they don't quite do what you want. The input format is something
like

digraph fred 
{
v1 -> v2;
v3 -> v2;
}

to put lines between v1 and v2 and between v3 and v2. Shapes, colours
and labels for vertices and edges can be controlled.

<http://www.research.att.com/sw/tools/graphviz/> is the URL I have, if
it's wrong, Google should find them. GraphViz is excellent: I wrote a
set of tools to produce data flow diagrams from C code using Exuberant
Ctags, Perl and GraphViz. 

-- 
----- Paul Wright ------| For my part, I travel not to go anywhere, but to
-paul.wright at pobox.com--| go. I travel for travel's sake. The great affair is
http://pobox.com/~pw201 | to move.  -Stevenson


From jeremy at alum.mit.edu  Sat Feb 10 01:28:12 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 9 Feb 2001 19:28:12 -0500 (EST)
Subject: Python 2.1 release schedule
Message-ID: <14980.35612.516421.741505@w221.z064000254.bwi-md.dsl.cnc.net>

I updated the Python 2.1 release schedule (PEP 226):
    http://python.sourceforge.net/peps/pep-0226.html

The schedule now has some realistic future release dates.  The plan is
to move to beta 1 before the Python conference, probably issue a
second beta in mid- to late-March, and aim for a final release
sometime in April.  The six-week period between first beta and final
release is about as long as the beta period for 2.0, which had many
more significant changes.

I have also added a section on open issues as we had in the 2.0
release schedule.  If you are responsible for any major changes or
fixes before the first beta, please add them to that section or send
me mail about them.  Remember that we are in feature freeze; only bug
fixes between now and beta 1.

Jeremy



From guido at digicool.com  Sat Feb 10 15:39:47 2001
From: guido at digicool.com (Guido van Rossum)
Date: Sat, 10 Feb 2001 09:39:47 -0500
Subject: [Python-Dev] Python 2.1 release schedule
In-Reply-To: Your message of "Fri, 09 Feb 2001 19:28:12 EST."
             <14980.35612.516421.741505@w221.z064000254.bwi-md.dsl.cnc.net> 
References: <14980.35612.516421.741505@w221.z064000254.bwi-md.dsl.cnc.net> 
Message-ID: <200102101439.JAA27319@cj20424-a.reston1.va.home.com>

> I updated the Python 2.1 release schedule (PEP 226):
>     http://python.sourceforge.net/peps/pep-0226.html

Thanks, Jeremy!

> The schedule now has some realistic future release dates.  The plan is
> to move to beta 1 before the Python conference, probably issue a
> second beta in mid- to late-March, and aim for a final release
> sometime in April.  The six-week period between first beta and final
> release is about as long as the beta period for 2.0, which had many
> more significant changes.

Feels good to me.

> I have also added a section on open issues as we had in the 2.0
> release schedule.  If you are responsible for any major changes or
> fixes before the first beta, please add them to that section or send
> me mail about them.  Remember that we are in feature freeze; only bug
> fixes between now and beta 1.

Here are a few issues that I wrote down recently.  I'm a bit out of
touch so some of these may already have been resolved...

- New schema for .pyc magic number?  (Eric, Tim)

- Call to C function without keyword args should pass NULL, not {}.
  (Jeremy)

- Reduce the errors for "from ... import *" to only those cases where
  it's a real problem for nested functions.  (Jeremy)

- Long ago, someone asked that 10**-15 should return a float rather
  than raise a ValueError.  I think this is an OK change, and unlikely
  to break code :-)  There may be a few other special cases like
  this, and of course ints and longs should act the same way.  (Tim?)

--Guido van Rossum (home page: http://www.python.org/~guido/)



From jeremy at alum.mit.edu  Sat Feb 10 05:57:51 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 9 Feb 2001 23:57:51 -0500 (EST)
Subject: [Python-Dev] Python 2.1 release schedule
In-Reply-To: <200102101439.JAA27319@cj20424-a.reston1.va.home.com>
References: <14980.35612.516421.741505@w221.z064000254.bwi-md.dsl.cnc.net>
	<200102101439.JAA27319@cj20424-a.reston1.va.home.com>
Message-ID: <14980.51791.171007.616771@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "GvR" == Guido van Rossum <guido at digicool.com> writes:

  >> I have also added a section on open issues as we had in the 2.0
  >> release schedule.  If you are responsible for any major changes
  >> or fixes before the first beta, please add them to that section
  >> or send me mail about them.  Remember that we are in feature
  >> freeze; only bug fixes between now and beta 1.

  GvR> Here are a few issues that I wrote down recently.  I'm a bit
  GvR> out of touch so some of these may already have been resolved...

  [...]

  GvR> - Call to C function without keyword args should pass NULL, not
  GvR>   {}.  (Jeremy)

  GvR> - Reduce the errors for "from ... import *" to only those cases
  GvR>   where it's a real problem for nested functions.  (Jeremy)

  [...]

These two are done and checked into CVS.

Jeremy



From esr at thyrsus.com  Sat Feb 10 16:43:42 2001
From: esr at thyrsus.com (Eric S. Raymond)
Date: Sat, 10 Feb 2001 10:43:42 -0500
Subject: [Python-Dev] Python 2.1 release schedule
In-Reply-To: <200102101439.JAA27319@cj20424-a.reston1.va.home.com>; from guido@digicool.com on Sat, Feb 10, 2001 at 09:39:47AM -0500
References: <14980.35612.516421.741505@w221.z064000254.bwi-md.dsl.cnc.net> <200102101439.JAA27319@cj20424-a.reston1.va.home.com>
Message-ID: <20010210104342.A20657@thyrsus.com>

Guido van Rossum <guido at digicool.com>:
> - New schema for .pyc magic number?  (Eric, Tim)

It looked to me like Tim had a good scheme, but he never specified the
latter (integrity-check) part of the header).
-- 
		<a href="http://www.tuxedo.org/~esr/">Eric S. Raymond</a>

Everything that is really great and inspiring is created by the
individual who can labor in freedom.
	-- Albert Einstein, in H. Eves Return to Mathematical Circles, 
		Boston: Prindle, Weber and Schmidt, 1988.



From loredo at spacenet.tn.cornell.edu  Mon Feb 12 22:33:34 2001
From: loredo at spacenet.tn.cornell.edu (Tom Loredo)
Date: Mon, 12 Feb 2001 16:33:34 -0500
Subject: Python 2.1 release schedule
References: <mailman.981764955.17485.python-list@python.org>
Message-ID: <3A8856AE.F49F5024@spacenet.tn.cornell.edu>

Jeremy et al.-

The 2.1 PEP mentions:

> Add a default unit testing framework to the standard library.

Is it yet settled which of the Python unit test frameworks this
will be?  Or is there at least a "short list" of candidates?
I looked for a PEP on this, but couldn't find one.  I ask because
I don't yet use unit testing, but was just starting to look into
it.  If there is a standard or default framework experts feel
is particularly suited to Python, I'd like to start there.

Thanks,
Tom Loredo


From fredrik at pythonware.com  Tue Feb 13 17:36:27 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 13 Feb 2001 16:36:27 GMT
Subject: Python 2.1 release schedule
References: <mailman.981764955.17485.python-list@python.org> <3A8856AE.F49F5024@spacenet.tn.cornell.edu>
Message-ID: <fodi6.13005$Qb7.2090549@newsb.telia.net>

Tom Loredo wrote:
> Is it yet settled which of the Python unit test frameworks this
> will be?

doctest.py (already in the repository, since it's cool)
PyUnit.py (by BDFL decree, since his employer was already using it)

> If there is a standard or default framework experts feel
> is particularly suited to Python, I'd like to start there.

doctest.py

Cheers /F




From tony at lsl.co.uk  Wed Feb 14 11:26:10 2001
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Wed, 14 Feb 2001 10:26:10 -0000
Subject: Python 2.1 release schedule
In-Reply-To: <fodi6.13005$Qb7.2090549@newsb.telia.net>
Message-ID: <008001c09670$8c5ad340$f05aa8c0@lslp7o.int.lsl.co.uk>

Tom Loredo wrote:
> Is it yet settled which of the Python unit test frameworks this
> will be?

Fredrik Lundh replied:
> doctest.py (already in the repository, since it's cool)
> PyUnit.py (by BDFL decree, since his employer was already using it)

Yeh! (you'll have to imagine me shouting that whilst raising my arms in
the air, as our two year old does when he gets good news - or perhaps
you'd rather not).

doctest is really neat (I've finally managed to get round to using it) -
having *just* changed the names of a handful of variables in some source
code, it's very nice to be able to run the module "at itself" and have
it tell me it's still working... (well, as much as it was before,
anyway!)

I guess now I'll have to bite the bullet and learn how to use PyUnit as
well.

Tibs
--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
Give a pedant an inch and they'll take 25.4mm
(once they've established you're talking a post-1959 inch, of course)
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From othello at javanet.com  Tue Feb 13 16:43:19 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 13 Feb 2001 10:43:19 -0500
Subject: Python 2.1 release schedule
References: <mailman.981764955.17485.python-list@python.org>
Message-ID: <3A895617.C7EDB5C8@javanet.com>

Jeremy Hylton wrote:

> I updated the Python 2.1 release schedule (PEP 226):
> <snip>
> second beta in mid- to late-March, and aim for a final release
> sometime in April.

Call me superstitious, but going final on Friday, April 13th
 in the first year of a new millineum, at tax time seems a
trifle risky.

RH



From jeremy at alum.mit.edu  Tue Feb 13 06:03:51 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Tue, 13 Feb 2001 00:03:51 -0500 (EST)
Subject: Python 2.1 release schedule
In-Reply-To: <3A895617.C7EDB5C8@javanet.com>
References: <mailman.981764955.17485.python-list@python.org>
	<3A895617.C7EDB5C8@javanet.com>
Message-ID: <14984.49207.794730.539580@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "RH" == Raymond Hettinger <othello at javanet.com> writes:

  RH> Jeremy Hylton wrote:
  >> I updated the Python 2.1 release schedule (PEP 226): <snip>
  >> second beta in mid- to late-March, and aim for a final release
  >> sometime in April.

  RH> Call me superstitious, but going final on Friday, April 13th
  RH> in the first year of a new millineum, at tax time seems a
  RH> trifle risky.

Because we're more likely to get audited or because we'll sit watching
the download statistics and forget to file our taxes?

Jeremy




From Bill.Scherer at VerizonWireless.com  Tue Feb 13 17:01:07 2001
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Tue, 13 Feb 2001 11:01:07 -0500
Subject: Python 2.1 release schedule
References: <mailman.981764955.17485.python-list@python.org> <3A895617.C7EDB5C8@javanet.com>
Message-ID: <3A895A43.6C2DA72@VerizonWireless.com>

Raymond Hettinger wrote:
> 
> Jeremy Hylton wrote:
> 
> > I updated the Python 2.1 release schedule (PEP 226):
> > <snip>
> > second beta in mid- to late-March, and aim for a final release
> > sometime in April.
> 
> Call me superstitious, but going final on Friday, April 13th
>  in the first year of a new millineum, at tax time seems a
> trifle risky.
> 
> RH

I disagree.  Could you imagine a better confluence of external events to
coincide with the release of a wonderful new Python?  There has to be a
balance between good and evil, right?  ;-)



-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From othello at javanet.com  Wed Feb 14 00:40:51 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 13 Feb 2001 18:40:51 -0500
Subject: Python 2.1 release schedule
References: <mailman.981764955.17485.python-list@python.org>
		<3A895617.C7EDB5C8@javanet.com> <mailman.982082474.11710.python-list@python.org>
Message-ID: <3A89C603.8A660552@javanet.com>


Jeremy Hylton wrote:

> >>>>> "RH" == Raymond Hettinger <othello at javanet.com> writes:
>
>   RH> Jeremy Hylton wrote:
>   >> I updated the Python 2.1 release schedule (PEP 226): <snip>
>   >> second beta in mid- to late-March, and aim for a final release
>   >> sometime in April.
>
>   RH> Call me superstitious, but going final on Friday, April 13th
>   RH> in the first year of a new millineum, at tax time seems a
>   RH> trifle risky.
>
> Because we're more likely to get audited or because we'll sit watching
> the download statistics and forget to file our taxes?
>
> Jeremy

File early, so you can watch the downloads without fear.
That is, unless, you get to thinking that the final bug fixes
are going in on April 10th during a full moon. ;)





From gillet at scripps.edu  Sat Feb 10 02:37:54 2001
From: gillet at scripps.edu (Alexandre Gillet)
Date: Fri, 09 Feb 2001 17:37:54 -0800
Subject: problem with numeric array on window.
Message-ID: <3A849B72.28AEAA22@scripps.edu>

Hi all,

We are having problems when creating numeric arrays in C extensions
under
windows. We narrowed the problem down to a very simple example (shown
below) where we simply allocate some memory, create a Numeric array
using
PyArrayFromDimsAndData. If we call that function as soon as we delete
the
returned array the python interpreter crashes as it tries to free
self->data. If we do not the the OWN_DATA flag in the array it works
fine
but we have a memory leak.
We tried this using both release and Debug versions of Python1.5.2, and
Numeric 17.1.1.
I have been using this mechanism under Unix for a long time and have not
had this problem before !
Using the same extension and test under unix works fine.
Has something changed ? Any help is welcome ...
Thank you

I join the two files we are using.
 bug.c : a C module that create an array numeric with the flags set to
own_data, so when the array is garbage collected, the memory is free.


#################
bug.c#####################################################
#ifdef WIN32
#include <stdlib.h>
#include <malloc.h>
#endif
#include "Python.h"
#include "arrayobject.h"

static PyObject* createArray(PyObject* self, PyObject* args)
{
    int *dims;
    float *data;
    PyArrayObject *out;

    dims = (int *)malloc(2 * sizeof(int));
    dims[0] = 500;
    dims[1] = 2;
    data = (float *)malloc(100000 * sizeof(float));
    out = (PyArrayObject *)PyArray_FromDimsAndData(2, dims,
                                                   PyArray_FLOAT,
                                                   (char *)data);
    if (!out) {
      PyErr_SetString(PyExc_RuntimeError,
                      "Failed to allocate memory for normals");
      return NULL;
    }
    out->flags |= OWN_DATA;   /*so we'll free this memory when this 
                                array will be garbage collected */
    return (PyObject *)out;
}

static PyMethodDef bug_methods[] = {
  {"createArray",             createArray,      1}, 
  {NULL,      NULL}        /* Sentinel */
};

static char bug_documentation[] = "No Doc";

#ifdef WIN32
extern __declspec(dllexport)
#endif
void initbug()
{
        PyObject *m, *d;
        m = Py_InitModule4("bug",
                           bug_methods,
                           bug_documentation,
                           (PyObject *)NULL,
                           PYTHON_API_VERSION);
        d = PyModule_GetDict(m);
        import_array();

        if (PyErr_Occurred()) 
                Py_FatalError("can't initialize module bug");
}
#####################################################################

###############testbug.py############################################
import bug
ar = bug.createArray()
del ar # fatal if ar owns the data member

#####################################################################




-- 
**********************************
Alexandre Gillet
The Scripps Research Institute,       tel: (858) 784-9557
Dept. Molecular Biology,  MB-5,       fax: (858) 784-2860
10550  North Torrey Pines Road,       email: gillet at scripps.edu
La Jolla,  CA 92037-1000,  USA.



From herne at orcon.net.nz  Sat Feb 10 04:55:19 2001
From: herne at orcon.net.nz (Grant)
Date: Sat, 10 Feb 2001 16:55:19 +1300
Subject: Should I use Python?
Message-ID: <3A84BBA7.74688DFD@orcon.net.nz>

I have been creating some useful utilities in java and now find other
people I work with also wish to use them. Problem is they are
applications and require the java runtime environment to be installed.

I would prefer  to be able to write the utilities on Linux as I do now.
The file created must be able to be run by others  on NT machines -
either from a server or from their own machines. It should not require
any additonal setup or libraries to install, and will  need to read from
remote servers on a local network, reading and parsing files.

Would Python be the best choice?Any other suggestions?







From msoulier at storm.ca  Sat Feb 10 05:58:01 2001
From: msoulier at storm.ca (Michael P. Soulier)
Date: Sat, 10 Feb 2001 04:58:01 GMT
Subject: Should I use Python?
References: <3A84BBA7.74688DFD@orcon.net.nz>
Message-ID: <tT3h6.119287$V22.24326302@news4.rdc1.on.home.com>

On Sat, 10 Feb 2001 16:55:19 +1300, Grant <herne at orcon.net.nz> wrote:
>I have been creating some useful utilities in java and now find other
>people I work with also wish to use them. Problem is they are
>applications and require the java runtime environment to be installed.
>
>I would prefer  to be able to write the utilities on Linux as I do now.
>The file created must be able to be run by others  on NT machines -
>either from a server or from their own machines. It should not require
>any additonal setup or libraries to install, and will  need to read from
>remote servers on a local network, reading and parsing files.
>
>Would Python be the best choice?Any other suggestions?

    The only way to do what you're talking about is to distribute binaries for
the target platform. Python is a good choice if you want to use the frozen
tool and compile the Python to native code. 

    Mike

-- 
Michael P. Soulier <msoulier at storm.ca>
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to UNIX
PGP Public Key: http://24.43.42.96/email.phtml


From erno-news at erno.iki.fi  Sat Feb 10 06:26:13 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 10 Feb 2001 07:26:13 +0200
Subject: Should I use Python?
References: <3A84BBA7.74688DFD@orcon.net.nz> <tT3h6.119287$V22.24326302@news4.rdc1.on.home.com>
Message-ID: <ku3ddn153e.fsf@lasipalatsi.fi>

In article <tT3h6.119287$V22.24326302 at news4.rdc1.on.home.com>,
msoulier at storm.ca (Michael P. Soulier) writes:

|     The only way to do what you're talking about is to distribute
| binaries for the target platform. Python is a good choice if you
| want to use the frozen tool and compile the Python to native code.

do you have a url for this "frozen"? or do you mean freeze?

from what i hear py2exe and installer have superseded freeze,
and freeze does not compile python to native code.

  -- erno


From erno-news at erno.iki.fi  Sat Feb 10 06:32:56 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 10 Feb 2001 07:32:56 +0200
Subject: Should I use Python?
References: <3A84BBA7.74688DFD@orcon.net.nz> <tT3h6.119287$V22.24326302@news4.rdc1.on.home.com>
Message-ID: <kuy9vfyuev.fsf@lasipalatsi.fi>

In article <tT3h6.119287$V22.24326302 at news4.rdc1.on.home.com>,
msoulier at storm.ca (Michael P. Soulier) writes:

|     The only way to do what you're talking about is to distribute
| binaries for the target platform. Python is a good choice if you
| want to use the frozen tool and compile the Python to native code.

do you have a url for this "frozen"? or do you mean freeze?

from what i hear py2exe and installer have superseded freeze,
and none of those compile python to native code...

  -- erno


From herne at orcon.net.nz  Sat Feb 10 23:45:21 2001
From: herne at orcon.net.nz (Grant)
Date: Sun, 11 Feb 2001 11:45:21 +1300
Subject: Should I use Python?
References: <3A84BBA7.74688DFD@orcon.net.nz> <tT3h6.119287$V22.24326302@news4.rdc1.on.home.com>
Message-ID: <3A85C481.7539DC3F@orcon.net.nz>

"Michael P. Soulier" wrote:

> On Sat, 10 Feb 2001 16:55:19 +1300, Grant <herne at orcon.net.nz> wrote:
> >I have been creating some useful utilities in java and now find other
> >people I work with also wish to use them. Problem is they are
> >applications and require the java runtime environment to be installed.
> >
> >I would prefer  to be able to write the utilities on Linux as I do now.
> >The file created must be able to be run by others  on NT machines -
> >either from a server or from their own machines. It should not require
> >any additonal setup or libraries to install, and will  need to read from
> >remote servers on a local network, reading and parsing files.
> >
> >Would Python be the best choice?Any other suggestions?
>
>     The only way to do what you're talking about is to distribute binaries for
> the target platform. Python is a good choice if you want to use the frozen
> tool and compile the Python to native code.
>
>     Mike

Thanks - I will keep it in mind. Learning a new language is always a big step - I
am still looking around at what best suits my needs.



From c941520 at alinga.newcastle.edu.au  Sat Feb 10 05:53:02 2001
From: c941520 at alinga.newcastle.edu.au (Ben de Luca)
Date: Fri, 9 Feb 2001 20:53:02 -0800
Subject: class scope questions
Message-ID: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>

if i do this

class a:
         dog='woof'

          def blah():
                    something



how do i call a
how do i get soemthing in blah to refernce dog? either to read it or write
to it






From stephen_purcell at yahoo.com  Fri Feb  9 11:29:54 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 9 Feb 2001 11:29:54 +0100
Subject: class scope questions
In-Reply-To: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>; from c941520@alinga.newcastle.edu.au on Fri, Feb 09, 2001 at 08:53:02PM -0800
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <20010209112954.B26384@freedom.puma-ag.com>

Ben de Luca wrote:
> if i do this
> 
> class a:
>          dog='woof'
> 
>           def blah():
>                     something
> how do i call a
> how do i get soemthing in blah to refernce dog? either to read it or write
> to it

Your 'blah' should have a 'self' parameter because it is a method of class 'a'.

Then the answer is;-

   class A:
      dog = 'woof'
      def bark(self):
         print self.dog


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From duncan at rcp.co.uk  Fri Feb  9 11:48:31 2001
From: duncan at rcp.co.uk (Duncan Booth)
Date: Fri, 9 Feb 2001 10:48:31 +0000 (UTC)
Subject: class scope questions
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <Xns90436DAD61713duncanrcpcouk@127.0.0.1>

"Ben de Luca" <c941520 at alinga.newcastle.edu.au> wrote in
<3a83bea8$0$16388$7f31c96c at news01.syd.optusnet.com.au>: 

>if i do this
>
>class a:
>         dog='woof'
>
>          def blah():
>                    something
>
>
>
>how do i call a
>how do i get soemthing in blah to refernce dog? either to read it or
>write to it
>
class a:
    dog='woof'
    def blah(self):
        print self.dog
    def update(self, value):
        self.dog = value
>>> a1 = a()
>>> a1.blah()
woof
>>> a1.update('bow wow')
>>> a2 = a()
>>> a2.blah()
woof
>>> a1.blah()
bow wow
>>> 

But, although this works fine, there is a problem with it. dog is
actually a variable in the class a until it is set by the update method
when it becomes  a variable in the instance a1. This works fine for
strings, but if you used a mutable object such as a list or dictionary
you would see that dog was shared between all instances of the class: 
class B: 	list = []
	def show(self):
		print self.list
	def append(self, value):
		self.list.append(value)

		
>>> b1 = B()
>>> b1.append('hello')
>>> b1.show()
['hello']
>>> b2 = B()
>>> b2.show()
['hello']
>>> b2.append('world')
>>> b1.show()
['hello', 'world']

To avoid this you should generally define an __init__ method in your
class and use that to initialise any instance variables. 


From sholden at holdenweb.com  Fri Feb  9 17:41:49 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 9 Feb 2001 11:41:49 -0500
Subject: class scope questions
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au> <Xns90436DAD61713duncanrcpcouk@127.0.0.1>
Message-ID: <GIVg6.27$t02.657@e420r-atl1.usenetserver.com>

"Duncan Booth" <duncan at rcp.co.uk> wrote in message
news:Xns90436DAD61713duncanrcpcouk at 127.0.0.1...
> "Ben de Luca" <c941520 at alinga.newcastle.edu.au> wrote in
> <3a83bea8$0$16388$7f31c96c at news01.syd.optusnet.com.au>:
>
> >if i do this
> >
> >class a:
> >         dog='woof'
> >
> >          def blah():
> >                    something
> >
> >
> >
> >how do i call a
> >how do i get soemthing in blah to refernce dog? either to read it or
> >write to it
> >
> class a:
>     dog='woof'
>     def blah(self):
>         print self.dog
>     def update(self, value):
>         self.dog = value
> >>> a1 = a()
> >>> a1.blah()
> woof
> >>> a1.update('bow wow')
> >>> a2 = a()
> >>> a2.blah()
> woof
> >>> a1.blah()
> bow wow
> >>>
>
> But, although this works fine, there is a problem with it. dog is
> actually a variable in the class a until it is set by the update method
> when it becomes  a variable in the instance a1. This works fine for
> strings, but if you used a mutable object such as a list or dictionary
> you would see that dog was shared between all instances of the class:
> class B: list = []
> def show(self):
> print self.list
> def append(self, value):
> self.list.append(value)
>
>
> >>> b1 = B()
> >>> b1.append('hello')
> >>> b1.show()
> ['hello']
> >>> b2 = B()
> >>> b2.show()
> ['hello']
> >>> b2.append('world')
> >>> b1.show()
> ['hello', 'world']
>
> To avoid this you should generally define an __init__ method in your
> class and use that to initialise any instance variables.
>
Note, however:

>>> a.dog
'woof'
>>>

If dog is really intended to be a class variable, shared among all
instances, then you are much safer treating it as an attribute of the class.

regards
 Steve




From aleaxit at yahoo.com  Fri Feb  9 12:58:13 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 9 Feb 2001 12:58:13 +0100
Subject: class scope questions
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <960m0p0d3r@news2.newsguy.com>

"Ben de Luca" <c941520 at alinga.newcastle.edu.au> wrote in message
news:3a83bea8$0$16388$7f31c96c at news01.syd.optusnet.com.au...
> if i do this
>
> class a:
>          dog='woof'
>
>           def blah():
>                     something

...then you're doing it wrong, because a method such as
blah should always have at least one argument.  The first
argument of a method is invariably called 'self', by
near-universal Python convention; it is bound to the
object on which the method is being called.

(Another popular convention is to name classes with a
name starting with an uppercase letter).

> how do i call a

You call a this way:

x = a()

this instantiates an instance of class a, and binds
variable x to said instance.


> how do i get soemthing in blah to refernce dog? either to read it or write
> to it

The normal approach to _reading_ a class-attribute is to
reference it as an attribute of self:

class A:
    dog = 'woof'
    def blah(self):
        print self.dog

x = A()
x.blah()

this will print 'woof'.  Since the instance does not directly
have an attribute named dog, the class's dictionary is looked
up for it, so the class-attribute is found.

However, if you *assign* self.dog, then that will bind a
(possibly new) attribute called 'dog' *in the instance*, and
NOT affect the _class_ attribute similarly named.  If you
do want to ensure you're accessing the class attribute,
you can use A.dog for reading, A.dog='buh' for writing;
or, getattr(A,'dog') for reading, setattr(A,'dog','buh')
for writing.


Alex





From neilt at bigfoot.com  Fri Feb  9 18:52:38 2001
From: neilt at bigfoot.com (Neil Timms)
Date: Fri, 9 Feb 2001 17:52:38 +0000
Subject: class scope questions
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au> <960m0p0d3r@news2.newsguy.com>
Message-ID: <961app$1ei1@imsp212.netvigator.com>

Not a real star for this sort of thing but...

in the post.. call a is incorrect. a  is a class so we don't call it but 
instaniate it - that is make an object described by a ...thus.. (and we'll 
call it A by convention because it is a class):

(presumes A is already imported)

        myObject = A()


use __init__ and put the variables in there. They are then private (but 
not totally, there is a way...) to instances of the class A. Attributes 
(variables of a class instance ) are best only altered by a setter or 
getter methods = encapsulation.. Now that we have loads of RAM and much 
speed from the CPUs we can use these conventions for more robust and esier 
to read/debug code.

so we can have:

        def setDog(anObject):
                self.dog = anObject

or:

        def getDog():
                return self.dog

or even more cunning (or not) ... the setter getter


        def dog(anObject=''):
                if anObject:
                        self.dog = anObject
                return self.dog

Using the setter/getter version of above - to alter the value of dog we 
would then:


        myObject.dog('puppy')

# the attribute dog of myObject (of class A) now has the value of 'puppy'

to get the value of the attribute dog we would:


        theValueOfDog = myObject.dog()


(sorry just back from finishing a Smalltalk exam - so I'm full of it.).

Basically we hide or variables away from direct access and only alter then 
via methods definind in the class. This restricts the extent of one part 
of the program interacting with others and causing unintended results 
(bugs, or in MS speak features), making implementation and maintenance 
easier - but making us think for the deisgn.

cheers

Neil


From paulp at ActiveState.com  Fri Feb  9 20:11:58 2001
From: paulp at ActiveState.com (Paul Prescod)
Date: Fri, 09 Feb 2001 11:11:58 -0800
Subject: class scope questions
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <3A8440FE.412C6EA4@ActiveState.com>

Ben de Luca wrote:
> 
> if i do this
> 
> class a:
>          dog='woof'
> 
>           def blah():
>                     something
> 
> how do i call a
> how do i get soemthing in blah to refernce dog? either to read it or write
> to it

class a:
    dog='woof'
    def blah(self ):
       print a.dog
       a.dog="woof woof"



From joconnor at cybermesa.com  Fri Feb  9 23:04:37 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Fri, 09 Feb 2001 22:04:37 +0000
Subject: class scope questions
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <3A846975.68708D86@cybermesa.com>

Ben de Luca wrote:
> 
> if i do this
> 
> class a:
>          dog='woof'
> 
>           def blah():
>                     something
> 
> how do i call a
> how do i get soemthing in blah to refernce dog? either to read it or write
> to it

You have to declare blah() to take 'self' as an argument (which will be
the object itself) and the reference dog with 'self.dog'

Such as

	class a:
		dog = 'woof'
		
		def blah (self):
			print self.dog


Take care,

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython


From phrxy at csv.warwick.ac.uk  Sat Feb 10 06:20:19 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Sat, 10 Feb 2001 05:20:19 +0000
Subject: win32 python path
Message-ID: <Pine.SOL.4.30.0102100456010.25873-100000@mimosa.csv.warwick.ac.uk>

I'm not sure if this is the same on unix as I'm not near a suitable
machine ATM, but I get this behaviour on win95:

C:\Python20\Lib\Local.pth
C:\Python20\Lib\Local\MyModuleDir.pth
C:\Python20\Lib\Local\MyModuleDir\mymodule.py

Where xxxx.pth points to the directory xxxx, of course.  Python doesn't
see mymodule.py with this set-up, but it does if you have instead:

C:\Python20\Lib\MyModuleDir.pth
C:\Python20\Lib\Local\MyModuleDir\mymodule.py

which is fine if you only have one locally-installed module, but annoying
if you have lots.  Why does Python (2.0) do this?  I am aware of the fact
that Guido is in favour (who knows why) of dumping all locally-installed
modules in Lib on windows.

I also don't like the way that .pth files are frowned upon.  For example,
tiny module 'TableIO' has a .so / .DLL _tableio.pyd and a pure python
module TableIO.py.  It would be nice to have them both in one directory,
because otherwise files like this will end up scattered over your module
directory and you won't know which ones belong together.  In this case I
suppose you could just make a directory TableIO and import with

from TableIO import TableIO

but many modules have a class with the same name as the module, so you'd
end up with

from TableIO import TableIO.TableIO

which is silly.

Thanks for any help


John

"Deprecating .pth's is silly, and does not at at all good things."



From phrxy at csv.warwick.ac.uk  Sun Feb 11 23:59:13 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Sun, 11 Feb 2001 22:59:13 +0000
Subject: win32 python path
In-Reply-To: <Pine.SOL.4.30.0102100456010.25873-100000@mimosa.csv.warwick.ac.uk>
References: <Pine.SOL.4.30.0102100456010.25873-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <Pine.SOL.4.30.0102112252390.16304-100000@mimosa.csv.warwick.ac.uk>

On Sat, 10 Feb 2001, John J. Lee wrote:

>
> I'm not sure if this is the same on unix as I'm not near a suitable
> machine ATM, but I get this behaviour on win95:
>
> C:\Python20\Lib\Local.pth
> C:\Python20\Lib\Local\MyModuleDir.pth
> C:\Python20\Lib\Local\MyModuleDir\mymodule.py

No, I mean:

C:\Python20\Local.pth
C:\Python20\Local\MyModuleDir.pth
C:\Python20\Local\MyModuleDir\mymodule.py

and, below, I mean:

C:\Python20\MyModuleDir.pth
C:\Python20\Local\MyModuleDir\mymodule.py

where MyModuleDir.pth contains the single line

Local/MyModuleDir

(Lib is where the _standard_ modules are)

> Where xxxx.pth points to the directory xxxx, of course.  Python doesn't
> see mymodule.py with this set-up, but it does if you have instead:
>
> C:\Python20\Lib\MyModuleDir.pth
> C:\Python20\Lib\Local\MyModuleDir\mymodule.py
>
> which is fine if you only have one locally-installed module, but annoying
> if you have lots.  Why does Python (2.0) do this?  I am aware of the fact
> that Guido is in favour (who knows why) of dumping all locally-installed
> modules in Lib on windows.
>
> I also don't like the way that .pth files are frowned upon.  For example,
> tiny module 'TableIO' has a .so / .DLL _tableio.pyd and a pure python
> module TableIO.py.  It would be nice to have them both in one directory,
> because otherwise files like this will end up scattered over your module
> directory and you won't know which ones belong together.  In this case I
> suppose you could just make a directory TableIO and import with
>
> from TableIO import TableIO
>
> but many modules have a class with the same name as the module, so you'd
> end up with
>
> from TableIO import TableIO.TableIO
>
> which is silly.
>
> Thanks for any help
>
>
> John
>
> "Deprecating .pth's is silly, and does not at at all good things."
>
>



From hatasaka at libertysw.com  Sat Feb 10 10:12:49 2001
From: hatasaka at libertysw.com (Mark Hatasaka)
Date: Sat, 10 Feb 2001 01:12:49 -0800
Subject: RedHat 7.0 LONG_BIT compilation error
Message-ID: <001a01c09341$b22f8570$3219a8c0@libertysw.com>

I'm trying to compile Python 2.0 on RedHat 7.0 i386, but keep getting the following LONG_BIT error.

gcc -g -O2 -Wall -Wstrict-prototypes -I./../Include -I.. -DHAVE_CONFIG_H   -c -o listnode.o listnode.c
gcc -g -O2 -Wall -Wstrict-prototypes -I./../Include -I.. -DHAVE_CONFIG_H   -c -o node.o node.c
In file included from ../Include/Python.h:54,
                 from node.c:3:
../Include/pyport.h:390:2: #error "LONG_BIT definition appears wrong for platform (bad gcc config?)."
make[1]: *** [node.o] Error 1
make[1]: Leaving directory `/usr/Python-2.0/Parser'
make: *** [Parser] Error 2
[root at Piglet Python-2.0]#

I've applied the latest RedHat gcc updates (gcc-2.96-69)

I've run the following commands

marke clean

rm config.cache

./configure

make



Any suggestions would be greatly appreciated.

Mark



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010210/db209a25/attachment.htm>

From avle at rogn.ii.uib.no  Sat Feb 10 10:49:28 2001
From: avle at rogn.ii.uib.no (Helge Avlesen)
Date: 10 Feb 2001 10:49:28 +0100
Subject: RedHat 7.0 LONG_BIT compilation error
References: <mailman.981795917.3251.python-list@python.org>
Message-ID: <72pugqx3yv.fsf@rogn.ii.uib.no>

Try update glibc as well, I think the problem was in the header
files, not the compiler. 

Helge



"Mark Hatasaka" <hatasaka at libertysw.com> writes:

| I'm trying to compile Python 2.0 on RedHat 7.0 i386, but keep getting the following LONG_BIT error.
| 
| gcc -g -O2 -Wall -Wstrict-prototypes -I./../Include -I.. -DHAVE_CONFIG_H   -c -o listnode.o listnode.c
| gcc -g -O2 -Wall -Wstrict-prototypes -I./../Include -I.. -DHAVE_CONFIG_H   -c -o node.o node.c
| In file included from ../Include/Python.h:54,
|                  from node.c:3:
| ../Include/pyport.h:390:2: #error "LONG_BIT definition appears wrong for platform (bad gcc config?)."
| make[1]: *** [node.o] Error 1
| make[1]: Leaving directory `/usr/Python-2.0/Parser'
| make: *** [Parser] Error 2
| [root at Piglet Python-2.0]#
| 
| I've applied the latest RedHat gcc updates (gcc-2.96-69)
| 
| I've run the following commands
| 
| marke clean
| 
| rm config.cache
| 
| ./configure
| 
| make
| 
| 
| 
| Any suggestions would be greatly appreciated.
| 
| Mark
| 
| 
| 

-- 
Helge Avlesen,
Para//ab, University of Bergen
WWW:   http://www.ii.uib.no/~avle
Phone: +47 55 58 40 35,  Fax: +47 55 58 41 99,  Office:  HIB 317H1


From etoffi at bigfoot.com  Sat Feb 10 11:15:39 2001
From: etoffi at bigfoot.com (etoffi at bigfoot.com)
Date: Sat, 10 Feb 2001 10:15:39 GMT
Subject: language switching
Message-ID: <9634c9$p2m$1@nnrp1.deja.com>

just a random thought that popped in my head:

what would be needed to debug code going across two languages, such as
python and eiffel, maybe with a little c in between (ie for gtk), maybe
over corba or xmlrpc?

is there a debugger that understands corba?
how about xmlrpc on a local machine?

how complicated would either (or both) of these be?

and without the rpc/ipc stuff, how difficult would that be?
i figure alot simpler than the other two...

background:

i like debugging code. not that much -- its better when it just works,
but it still is fun to see what goes on `inside'. within the year
(enough time for open source ;) i plan to have a working
editor/debugger that understands both python and eiffel (smalleiffel
compiles to c, which is a story in and f itself).

if i can step through both hard and soft layers of my programs, that
would be great.

--
etoffi


Sent via Deja.com
http://www.deja.com/


From manu_no_address at no_spam_svp.org  Sun Feb 11 09:08:25 2001
From: manu_no_address at no_spam_svp.org (Emmanuel STAPF)
Date: Sun, 11 Feb 2001 00:08:25 -0800
Subject: language switching
References: <9634c9$p2m$1@nnrp1.deja.com>
Message-ID: <B6AB8878.E2F%manu_no_address@no_spam_svp.org>

> what would be needed to debug code going across two languages, such as
> python and eiffel, maybe with a little c in between (ie for gtk), maybe
> over corba or xmlrpc?

If you take the new .NET technology from Microsoft, Eiffel# from ISE Eiffel
and Python from ActiveState. You don't have to do anything, it is simply
there for you to do cross language development and debugging.
 
Manu



From cbaumann at post.ch  Sun Feb 11 15:03:37 2001
From: cbaumann at post.ch (Carmen Baumann)
Date: Sun, 11 Feb 2001 15:03:37 +0100
Subject: language switching
References: <9634c9$p2m$1@nnrp1.deja.com> <B6AB8878.E2F%manu_no_address@no_spam_svp.org>
Message-ID: <3a869c62$1_1@news.datacomm.ch>

On OpenVMS, never a problem. During the debug session just tell the
debugger:
set lang c
set lang ada
set lang pascal
whatsoever

Jakob

Emmanuel STAPF <manu_no_address at no_spam_svp.org> schrieb in im Newsbeitrag:
B6AB8878.E2F%manu_no_address at no_spam_svp.org...
> > what would be needed to debug code going across two languages, such as
> > python and eiffel, maybe with a little c in between (ie for gtk), maybe
> > over corba or xmlrpc?
>
> If you take the new .NET technology from Microsoft, Eiffel# from ISE
Eiffel
> and Python from ActiveState. You don't have to do anything, it is simply
> there for you to do cross language development and debugging.
>
> Manu
>




From loewis at cs.tu-berlin.de  Sun Feb 18 11:24:57 2001
From: loewis at cs.tu-berlin.de (Martin v.Loewis)
Date: 18 Feb 2001 10:24:57 GMT
Subject: language switching
References: <9634c9$p2m$1@nnrp1.deja.com>
Message-ID: <96o7tp$put$1@news.cs.tu-berlin.de>

In article <9634c9$p2m$1 at nnrp1.deja.com>,  <etoffi at bigfoot.com> wrote:
>what would be needed to debug code going across two languages, such as
>python and eiffel, maybe with a little c in between (ie for gtk), maybe
>over corba or xmlrpc?

If the language runtimes all live in a single process, you need the 
debugger to tell what the language of the current function is. For example,
gdb supports multi-language debugging and will autodetect the language based
on the debug information.

>is there a debugger that understands corba?
>how about xmlrpc on a local machine?

In either case, the debugger would need to find out the process that runs
the service/object implementation, and attach to that process - which may
be difficult if the process is on a different machine. I'm not aware of
a debugger that does that for either CORBA or xmlrpc; I think MSVC supports
tracing COM calls over process boundaries.

>how complicated would either (or both) of these be?

You'd teach the debugger about the specifics of the runtime for the RPC
mechanism, e.g. how the language mapping works, how the IORs are represented
in memory, how to find out what sockets are used. With that, the debugger
could attach to the other end of the connection and trace the incoming call.

>and without the rpc/ipc stuff, how difficult would that be?
>i figure alot simpler than the other two...

Depends on the debugger, and the language. For gdb, telling it about another
"compiled" language is primarily a matter of defining debug information.
If the debugger already supports debugging that language, mixed-language
debugging should work out of the box.

In order to debug interpreted languages, you'd need to tell the debugger
details about the interpreter. E.g. for Python, you'd need to tell it
about the SET_LINENO instruction so it can do single-stepping, and you'd
need to inform it about the FrameObject and ThreadState so it can find
out what the local and global variables are.

Regards,
Martin



From zamurai at gmx.net  Sat Feb 10 11:20:12 2001
From: zamurai at gmx.net (Zamurai)
Date: Sat, 10 Feb 2001 11:20:12 +0100
Subject: Has anyone information on writing file compressing and decompressing programms?
Message-ID: <981800412.2025307082@news.t-online.de>

I want to write a file compressing/decompressing programm. Has anyone
inforamtions about how to compress data?

Zamurai


From erno-news at erno.iki.fi  Sat Feb 10 13:00:30 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 10 Feb 2001 14:00:30 +0200
Subject: Has anyone information on writing file compressing and decompressing programms?
References: <981800412.2025307082@news.t-online.de>
Message-ID: <kupugqzr1d.fsf@lasipalatsi.fi>

In article <981800412.2025307082 at news.t-online.de>, Zamurai
<zamurai at gmx.net> writes:

| I want to write a file compressing/decompressing programm. Has
| anyone inforamtions about how to compress data?

comp.compression might be a better place to ask. look in
the comp.compression faq, they probably list some introductory books.

i assume you are interested this in an academic sense, since
python is way too slow to write practical compression programs.

if you're just interested in how you can make stuff take up
less space in your python program, the zlib module is probably
what you want.

  -- erno


From ryu-jin at bigfoot.com  Sat Feb 10 12:33:16 2001
From: ryu-jin at bigfoot.com (Ryujin)
Date: Sat, 10 Feb 2001 11:33:16 GMT
Subject: XML Parser problem
Message-ID: <96395l$4e2$1@ryu.it>

I'm tryng to run some example from the Demos in the Python2.0 distro.  
When I try to run the xml parser example I get the following error:


bash-2.03$ python2.0 roundtrip.py /home/matte/PROGRAMMING/WORK/doc.xml 
Traceback (most recent call last):
  File "roundtrip.py", line 43, in ?
    parser = make_parser()
  File "/usr/local/lib/python2.0/xml/sax/__init__.py", line 88, in make_parser
    raise SAXReaderNotAvailable("No parsers found", None)  
xml.sax._exceptions.SAXReaderNotAvailable: No parsers found


And here is the content of the xml/sax directory in /usr/lib/python2.0:

bash-2.03$ ls /usr/local/lib/python2.0/xml/sax/
__init__.py   _exceptions.py   expatreader.py   handler.py   saxutils.py   xmlreader.py
__init__.pyc  _exceptions.pyc  expatreader.pyc  handler.pyc  saxutils.pyc  xmlreader.pyc
__init__.pyo  _exceptions.pyo  expatreader.pyo  handler.pyo  saxutils.pyo  xmlreader.pyo


Any suggestion? What am I missing?

thx in advance

matte







-- 
---------------------------------------
Memelli Matteo					
Could u imagine a world without Windoz?	 
ryu-jin at bigfoot.com			 
---------------------------------------



From zamurai at gmx.net  Sat Feb 10 12:48:16 2001
From: zamurai at gmx.net (Zamurai)
Date: Sat, 10 Feb 2001 12:48:16 +0100
Subject: Does an pdf create module exist?
Message-ID: <981805696.255883005@news.t-online.de>

?


From aahz at panix.com  Sat Feb 10 16:10:51 2001
From: aahz at panix.com (Aahz Maruch)
Date: 10 Feb 2001 07:10:51 -0800
Subject: Does an pdf create module exist?
References: <981805696.255883005@news.t-online.de>
Message-ID: <963llr$qjc$1@panix3.panix.com>

In article <981805696.255883005 at news.t-online.de>,
Zamurai  <zamurai at gmx.net> wrote:
>
>?

www.reportlab.com
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Actually, I've found that Usenet has helped a lot with my terror of
social stupidity.  Seeing people I really like and respect behave like
idiots in public has been very beneficial."  --Rachael Lininger


From f8dy at my-deja.com  Sat Feb 10 19:54:13 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Sat, 10 Feb 2001 18:54:13 GMT
Subject: Does an pdf create module exist?
References: <981805696.255883005@news.t-online.de>
Message-ID: <9642oi$eqo$1@nnrp1.deja.com>

In article <981805696.255883005 at news.t-online.de>,
  Zamurai <zamurai at gmx.net> wrote:
> ?
>

Depending on your needs, HTMLDoc may be enough for you.  It
will transform HTML documents into PDF (or PostScript).  It is free
software, licensed under the GPL, and is available for many
platforms.  I use it for my book, "Dive Into Python", and it works
fairly well.

http://www.easysw.com/htmldoc/

-M
--
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From teep at inet.co.th  Sun Feb 11 15:00:56 2001
From: teep at inet.co.th (Prateep Siamwalla)
Date: Sun, 11 Feb 2001 21:00:56 +0700
Subject: Does an pdf create module exist?
References: <981805696.255883005@news.t-online.de>
Message-ID: <9665p0$rgj$1@news.inet.co.th>

Definately check out report lab

www.reportlab.com

-teep








Zamurai <zamurai at gmx.net> wrote in message
news:981805696.255883005 at news.t-online.de...
?




From sholden at holdenweb.com  Sun Feb 11 18:00:50 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 11 Feb 2001 12:00:50 -0500
Subject: Does an pdf create module exist?
References: <981805696.255883005@news.t-online.de>
Message-ID: <TCzh6.3834$n4.27952@e420r-atl2.usenetserver.com>

The piddle library allows you to create PDF files, with a fair degree of
graphical complexity. Otherwise you might want to try reportlabs software,
which also creates PDFs.

regards
 Steve
--
Tools, training and technology to help you meet your information needs



"Zamurai" <zamurai at gmx.net> wrote in message
news:981805696.255883005 at news.t-online.de...
?




From clee at gnwy100.wuh.wustl.edu  Mon Feb 12 01:38:50 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 11 Feb 2001 18:38:50 -0600
Subject: Does an pdf create module exist?
References: <981805696.255883005@news.t-online.de> <TCzh6.3834$n4.27952@e420r-atl2.usenetserver.com>
Message-ID: <7un1bsybtx.fsf@gnwy100.wuh.wustl.edu>

>>>>> "Steve" == Steve Holden <sholden at holdenweb.com> writes:

    Steve> The piddle library allows you to create PDF files, with a fair
    Steve> degree of graphical complexity. Otherwise you might want to try
    Steve> reportlabs software, which also creates PDFs.

piddle would be a good choice if you wish to create cross-media graphics,
i.e., PDF and bitmaps and display in wxPython for the same drawing code.
It uses an early version of the same pdf creation module as reportlab.  If
you only need PDF creation, I would go directly to reportlab.com with it's
large, sophisticated and growing feature set.

Christopher Lee
Washington University Dept. of Anatomy and Neurobiology


From lakicsv at usa.net  Sat Feb 10 13:53:12 2001
From: lakicsv at usa.net (Viktor Lakics)
Date: Sat, 10 Feb 2001 12:53:12 +0000
Subject: Vim and python efficiently
Message-ID: <20010210125312.C2501@Diogenes.co.uk>

Hi All,

I would like to collect the tips, hints and macros, regarding how to edit python
sorce code in vim. I mean stuff like macros for commentify,
uncommentify, bracket and keyword autocomlpletion, python style
indenting,  etc. (I know that all this exist in IDLE, but I like vim
very much and I do not like to use different editors for different
languages... 

I have my ideas and I gradually try to write my own macros, but this
will take a long time, cause I just started to use vim. Besides, I
do not want to reinvent the wheel, if it is possible..:-))

This info could go into a Python-vim howto someday, similar to the C
editing with vim HOWTO from Siddharth Heroor, which is an excellent
one!

So if anyone knows a pointer  or an existing howto, or have an idea or
a macro which he or she is willing to contribute, please let me
know...

Thanks in advance:

Viktor 




From andi at opencan.cc  Sat Feb 10 22:06:12 2001
From: andi at opencan.cc (Andreas Poisel)
Date: 10 Feb 2001 21:06:12 GMT
Subject: Vim and python efficiently
References: <mailman.981809351.20580.python-list@python.org>
Message-ID: <964ag4$c4d$1@news.tuwien.ac.at>

Viktor Lakics <lakicsv at usa.net> wrote:

> Hi All,

> I would like to collect the tips, hints and macros, regarding how to edit python
> sorce code in vim. I mean stuff like macros for commentify,
> uncommentify, bracket and keyword autocomlpletion, python style
> indenting,  etc. (I know that all this exist in IDLE, but I like vim
> very much and I do not like to use different editors for different
> languages... 
[...]

Hi Viktor,

I really like your idea of a python + vim HOWTO. I'm not a vim expert, but I
have some small pieces of python code I use in vim, maybe it's useful for you:

I use this function to get a list of the python functions in the current buffer
with their line numbers:

def Functions(fname):
	print"----------------------------------------------------------------"
	cbuffer= vim.current.buffer
	zn= 0
	for line in cbuffer:
		zn= zn+1
		if re.compile(fname).search(line):
			print "%s %s" % (string.strip(line), zn)

I call it with 
map <F5> :py Functions("def ")<C-M>
from my .vimrc

Sometimes I work with files containing similar large functions. I use this one
to get the name of the function the cursor is in:

def thisFunc(str):
	cbuffer= vim.current.buffer
	cwindow= vim.current.window
	(line, column)= cwindow.cursor
	lnr= line-1
	while (lnr > 0):
		line= cbuffer[lnr]
		if string.find(line, str) != -1:
			print string.strip(line)
			break
		lnr= lnr-1

I call it with 
map <F6> :py thisFunc("def ")<C-M>
from my .vimrc

This one is a classbrowser using tkinter and Pmw. Its an ugly hack and I'm sure
it's buggy (I don't know, because I don't use this) but somehow it works.
You can navigate through buffers and classes by doubleclicking their names and
jump to a buffer/class/method by klicking the "go" button or doubleclicking on a
methods name.

class ClassBrowser:
	def __init__(self, kid, mid):
		sys.argv= ["/home/andi/vim/avim.py"]
		self.mid= mid
		self.kid= kid
		self.methDia= None
		self.claDia= None
		self.buffers= []
		self.getBuffer()
		self.root= Tk()
		Pmw.initialise()
		self.erzeugPanes()
		self.BufferBrowser()
		self.root.mainloop()

	def erzeugPanes(self):
		self.pane= Pmw.PanedWidget(self.root, hull_width= 650, 
									hull_height= 300, 
									orient= HORIZONTAL)
		self.pane.add("buffer", size= 200)
		self.pane.add("classes", size= 200)
		self.pane.add("methods", size= 200)
		self.pane.pack(expand= YES, fill= BOTH)

	def getClasses(self):
		self.classes= []
		self.cladict= {}
		self.lnrdict= {}
		lnr= 0
		for line in vim.current.buffer:
			lnr= lnr+1
			if re.compile(self.kid).search(line):
				kl= string.strip(line)
				self.classes.append(kl)
				self.cladict[kl]= []
				self.lnrdict[kl]= lnr
			elif re.compile(self.mid).search(line):
				me=  string.strip(line)
				try:
					self.cladict[kl].append(me)
				except NameError:
					ff= "Functions"
					try:
						self.cladict[ff].append(me)
					except KeyError:
						self.lnrdict[ff]= lnr
						self.classes.append(ff)
						self.cladict[ff]= []
						self.cladict[ff].append(me)
				self.lnrdict[me]= lnr

	def getBuffer(self):
		bnr= 1
		for buffer in vim.buffers:
			self.buffers.append("%s %s" % (bnr, os.path.basename(buffer.name)))
			bnr= bnr+1

	def goMethod(self):
		m= "%s" % self.methDia.getcurselection()
		lnr= "%s" % self.lnrdict[m]
		vim.command(lnr)
	
	def goClass(self):
		k= "%s" % self.claDia.getcurselection()
		lnr= "%s" % self.lnrdict[k]
		vim.command(lnr)

	def goBuffer(self):
		b= string.split("%s" % self.bufDia.getcurselection())
		vim.command("b%s" % b[0])

	def BufferBrowser(self):
		self.bufDia= Pmw.ScrolledListBox(self.pane.pane("buffer"),
listbox_selectmode= SINGLE,
										  labelpos= NW, usehullsize= 1,
										  hull_width= 200, hull_height= 200,
										  label_text= 'Buffers:',
										  items= self.buffers,
										  selectioncommand= self.goBuffer,
										  dblclickcommand= self.ClassBrowser)
		self.bufDia.pack(expand= 1, fill= BOTH, padx= 5, pady= 5, side= LEFT)
		Button(self.root, text= "go!", command= self.root.destroy).pack(side=
RIGHT, anchor= S)

	def ClassBrowser(self):
		self.getClasses()
		if self.methDia:
			self.methDia.destroy()
			self.methDia= None
		if self.claDia:
			self.claDia.destroy()
			self.KlaDia= None
		self.claDia= Pmw.ScrolledListBox(self.pane.pane("classes"), 
										  listbox_selectmode= SINGLE,
										  labelpos= NW, usehullsize= 1,
										  hull_width= 200, hull_height= 200,
										  label_text= 'Classes:',
										  items= self.classes,
										  selectioncommand= self.goClass,
										  dblclickcommand= self.MethodBrowser)
		self.claDia.pack(expand= 1, fill= BOTH, padx= 5, pady= 5, side= LEFT)

	def MethodBrowser(self):
		if self.methDia:
			self.methDia.destroy()
			self.methDia= None
		self.methDia= Pmw.ScrolledListBox(self.pane.pane("methods"), 
										  listbox_selectmode= SINGLE,
										  labelpos= NW, usehullsize= 1,
										  hull_width= 200, hull_height= 200,
										  label_text= 'Methods:',
										  items= self.cladict["%s" %
self.claDia.getcurselection()],
										  selectioncommand= self.goMethod,
										  dblclickcommand= self.root.destroy)
		self.methDia.pack(expand= 1, fill= BOTH, padx= 5, pady= 5, side=
LEFT)


I invoke this one in the .gvimrc:
map <F5> :py ClassBrowser("class ", "def ")<C-M>

Maybe you can use this somehow.
-- 
Andi


From andi at opencan.cc  Sun Feb 11 00:09:09 2001
From: andi at opencan.cc (Andreas Poisel)
Date: 10 Feb 2001 23:09:09 GMT
Subject: Vim and python efficiently
References: <mailman.981809351.20580.python-list@python.org> <964ag4$c4d$1@news.tuwien.ac.at>
Message-ID: <964hml$f36$1@news.tuwien.ac.at>

Andreas Poisel <andi at opencan.cc> wrote:
[...]

Sorry for the tabs...
-- 
Andi


From danielsatalandanielsdotcom  Sun Feb 11 04:47:12 2001
From: danielsatalandanielsdotcom (Alan Daniels)
Date: Sat, 10 Feb 2001 22:47:12 -0500
Subject: Vim and python efficiently
References: <mailman.981809351.20580.python-list@python.org>
Message-ID: <3n2c8tsol3b3rbh6uvkqm37tujvpk3d1rp@4ax.com>

On Sat, 10 Feb 2001 12:53:12 +0000, Viktor Lakics <lakicsv at usa.net>
wrote:

>I would like to collect the tips, hints and macros, regarding how to edit
>python sorce code in vim. 

I don't have any macros, but you may find my customized Pyrhon syntax
file for VIM to be handy...

http://www.alandaniels.com/vim/python.vim
==============================
Alan Daniels
daniels at alandaniels dot com


From none at none.com  Sat Feb 10 14:00:10 2001
From: none at none.com (knock knock)
Date: Sat, 10 Feb 2001 14:00:10 +0100
Subject: Tkinter event
Message-ID: <963ekd$7si$1@buty.wanadoo.nl>

got a canvas with items (lines). the mouseover event should give a specifc
desciption for every item in the canvas. so i did a mouseover bind to a
unique tag for every item! the idea is to make function that listens to the
event and look ups the description. how can this method find out which item
(tag) sent the event?? this must be possible, right?

- David




From fredrik at pythonware.com  Sat Feb 10 14:49:51 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 10 Feb 2001 13:49:51 GMT
Subject: Tkinter event
References: <963ekd$7si$1@buty.wanadoo.nl>
Message-ID: <3Gbh6.14482$AH6.2127525@newsc.telia.net>

"knock knock" <none at none.com> wrote:
> got a canvas with items (lines). the mouseover event should give a specifc
> desciption for every item in the canvas. so i did a mouseover bind to a
> unique tag for every item! the idea is to make function that listens to the
> event and look ups the description. how can this method find out which item
> (tag) sent the event?? this must be possible, right?

it's probably easier (for everyone, including Tk) to use a canvas-
level binding, and do the lookup in the event handler:

def eventhandler(event):
    canvas = event.widget
    x = canvas.canvasx(event.x)
    y = canvas.canvasx(event.y)
    items = canvas.find_overlapping(x, y, x+1, y+1)
    for item in items:
        tags = canvas.gettags(item)
        ...

Cheers /F




From cerutti at together.net  Mon Feb 12 20:02:55 2001
From: cerutti at together.net (Neil Cerutti)
Date: 12 Feb 2001 19:02:55 GMT
Subject: Tkinter event
References: <963ekd$7si$1@buty.wanadoo.nl>
Message-ID: <slrn98gd9b.7m.cerutti@fiad06.norwich.edu>

knock knock posted:
>got a canvas with items (lines). the mouseover event should give a specifc
>desciption for every item in the canvas. so i did a mouseover bind to a
>unique tag for every item! the idea is to make function that listens to the
>event and look ups the description. how can this method find out which item
>(tag) sent the event?? this must be possible, right?

Use the widget method of your event object.

-- 
Neil Cerutti <cerutti at together.net>


From zamurai at gmx.net  Sat Feb 10 14:50:18 2001
From: zamurai at gmx.net (Zamurai)
Date: Sat, 10 Feb 2001 14:50:18 +0100
Subject: Which command to use, to exit a programm?
Message-ID: <981813018.1461559148@news.t-online.de>

?


From sholden at holdenweb.com  Sat Feb 10 15:03:18 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 10 Feb 2001 09:03:18 -0500
Subject: Which command to use, to exit a programm?
References: <981813018.1461559148@news.t-online.de>
Message-ID: <rWbh6.3008$Tn6.81688@e420r-atl2.usenetserver.com>

It will usually be sys.exit(). In emergencies you might consider os._exit().

regards
 Steve

--
--
Tools, training and technology to help you meet your information needs



"Zamurai" <zamurai at gmx.net> wrote in message
news:981813018.1461559148 at news.t-online.de...
?




From digitig at cix.co.uk  Sat Feb 10 15:50:00 2001
From: digitig at cix.co.uk (Tim Rowe)
Date: Sat, 10 Feb 2001 14:50 +0000 (GMT Standard Time)
Subject: XML processing in Python -- changing horses in mid-stream
Message-ID: <memo.20010210145026.19785B@digitig.compulink.co.uk>

I have a Python program that parses a test script written in XML and puts 
the test results to stdout in XML (with a different DOCTYPE).

I also have a Python program that takes the test result XML and processes 
it for graphical presentation.

Here's my challenge. If the second program were given a test script 
instead of results, I would like it to recognise the different doctype, 
run the tests and process the results all in one.

When my result parser spots a test script I should be able to pass the 
script to the test program via a pipe, put the results of the test through 
a new result parser, and copy anything relevant from the new result parser 
to the old one. But can I stop the old result parser parsing the rest of 
the test script?

I plan to use either expat or pyxie, whichever is easiest.

TIA.




From sean at digitome.com  Sat Feb 10 17:24:49 2001
From: sean at digitome.com (Sean Mc Grath)
Date: Sat, 10 Feb 2001 16:24:49 GMT
Subject: XML processing in Python -- changing horses in mid-stream
References: <memo.20010210145026.19785B@digitig.compulink.co.uk>
Message-ID: <3a8563ae.3383090887@news.iol.ie>

Tim,

How big are the test scripts / test results? If the sizes are
manageable I would suggest looking at Pyxie Tree processing.
While processing a tree T, you can create another tree, T1 with
the results of running the second test script, you can then 
use the Paste functions to paste the new tree into the
existing one.

If size is a problem, you might like to look that the Sparse
Tree Processing stuff in Pyxie that allows you to switch
from event processing to tree processing as needed.

Sean

On Sat, 10 Feb 2001 14:50 +0000 (GMT Standard Time), digitig at cix.co.uk
(Tim Rowe) wrote:

>I have a Python program that parses a test script written in XML and puts 
>the test results to stdout in XML (with a different DOCTYPE).
>
>I also have a Python program that takes the test result XML and processes 
>it for graphical presentation.
>
>Here's my challenge. If the second program were given a test script 
>instead of results, I would like it to recognise the different doctype, 
>run the tests and process the results all in one.
>
>When my result parser spots a test script I should be able to pass the 
>script to the test program via a pipe, put the results of the test through 
>a new result parser, and copy anything relevant from the new result parser 
>to the old one. But can I stop the old result parser parsing the rest of 
>the test script?
>
>I plan to use either expat or pyxie, whichever is easiest.
>
>TIA.
>
>



From digitig at cix.co.uk  Tue Feb 13 19:29:00 2001
From: digitig at cix.co.uk (Tim Rowe)
Date: Tue, 13 Feb 2001 18:29 +0000 (GMT Standard Time)
Subject: XML processing in Python -- changing horses in mid-stream
References: <3a8563ae.3383090887@news.iol.ie>
Message-ID: <memo.20010213182932.64417A@digitig.compulink.co.uk>

The scripts are not very big. The result files could be. The sparse tree 
processing might be the thing.

Thanks,

Tim

In article <3a8563ae.3383090887 at news.iol.ie>, sean at digitome.com (Sean Mc 
Grath) wrote:

> Tim,
> 
> How big are the test scripts / test results? If the sizes are
> manageable I would suggest looking at Pyxie Tree processing.
> While processing a tree T, you can create another tree, T1 with
> the results of running the second test script, you can then 
> use the Paste functions to paste the new tree into the
> existing one.
> 
> If size is a problem, you might like to look that the Sparse
> Tree Processing stuff in Pyxie that allows you to switch
> from event processing to tree processing as needed.
> 
> Sean
> 
> On Sat, 10 Feb 2001 14:50 +0000 (GMT Standard Time), digitig at cix.co.uk
> (Tim Rowe) wrote:
> 
> >I have a Python program that parses a test script written in XML and 
> puts >the test results to stdout in XML (with a different DOCTYPE).
> >
> >I also have a Python program that takes the test result XML and 
> processes >it for graphical presentation.
> >
> >Here's my challenge. If the second program were given a test script 
> >instead of results, I would like it to recognise the different 
> doctype, >run the tests and process the results all in one.
> >
> >When my result parser spots a test script I should be able to pass the 
> >script to the test program via a pipe, put the results of the test 
> through >a new result parser, and copy anything relevant from the new 
> result parser >to the old one. But can I stop the old result parser 
> parsing the rest of >the test script?
> >
> >I plan to use either expat or pyxie, whichever is easiest.
> >
> >TIA.
> >
> >
> 
> 



From hungjunglu at yahoo.com  Sat Feb 10 17:50:35 2001
From: hungjunglu at yahoo.com (hungjunglu at yahoo.com)
Date: Sat, 10 Feb 2001 16:50:35 -0000
Subject: ODBC: creating DSN in Windows
Message-ID: <963rgr+6tns@eGroups.com>

Hi,

Is there a way of creating DSN from within Python (or Win32 
extensions)?

Creating DSN manually is no problem, but it'll be nice to be able to 
add one from an application.

How is it done usually? Via InstallShield?

regards,

Hung Jung





From dsavitsk at e-coli.net  Sat Feb 10 20:03:05 2001
From: dsavitsk at e-coli.net (douglas savitsky)
Date: Sat, 10 Feb 2001 13:03:05 -0600
Subject: ODBC: creating DSN in Windows
References: <mailman.981823871.8867.python-list@python.org>
Message-ID: <3A859069.69D8CC7B@e-coli.net>

hungjunglu at yahoo.com wrote:
> 
> Hi,
> 
> Is there a way of creating DSN from within Python (or Win32
> extensions)?
> 
> Creating DSN manually is no problem, but it'll be nice to be able to
> add one from an application.
> 
> How is it done usually? Via InstallShield?
> 
> regards,
> 
> Hung Jung

see the discussion from a few days ago (Wed) regarding DSN-Less
connection strings (ADO/ODBC/IIS etc.)

ds


From hungjunglu at yahoo.com  Sun Feb 11 08:11:35 2001
From: hungjunglu at yahoo.com (hungjunglu at yahoo.com)
Date: Sun, 11 Feb 2001 07:11:35 -0000
Subject: ODBC: creating DSN in Windows
In-Reply-To: <3A859069.69D8CC7B@e-coli.net>
Message-ID: <965dv7+4jnc@eGroups.com>

--- In python-list at y..., douglas savitsky <dsavitsk at e...> wrote:

> see the discussion from a few days ago (Wed) regarding DSN-Less
> connection strings (ADO/ODBC/IIS etc.)

Thanks. I have tried the DSN-less ADO.

I found another message by Scott Hathaway (with the info source 
ultimately from Alex Martelli). It only deals with registry tweaking. 
I tried it and it worked. (Though I had to be careful with line 
breaking in Yahoo egroups.) This is definitely what I was looking for.

http://groups.yahoo.com/group/python-list/message/115714

ADO is fun. But after all it's another layer on top of traditional 
SQL. I'd be more inclined if something equivalent in Python exists 
(meaning not Microsoft.)

regards,

Hung Jung





From slhath at home.com  Sun Feb 11 22:31:44 2001
From: slhath at home.com (Scott Hathaway)
Date: Sun, 11 Feb 2001 21:31:44 GMT
Subject: creating DSN in Windows
References: <mailman.981823871.8867.python-list@python.org>
Message-ID: <4xDh6.66249$bb.1038647@news1.rdc1.tx.home.com>

I posted this a few weeks ago:

I asked how to create an odbc link from python a couple of days ago.  Thanks
to help from Alex Martelli, I have solved my problem.  If anyone else is
interested, here is the code below.

Since I am just beginning with python, I am sure that vast improvements can
be made.  This code adds a MS Access System dsn named "tri2".

import _winreg as w

parentKey =
w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI',0,w.KEY_WRITE)
newKey = w.CreateKey(parentKey,'tri2')
w.SetValueEx(newKey,'DBQ',0,w.REG_SZ,r'c:\apache\htdocs\t2\mdb\Trinity.mdb')
w.SetValueEx(newKey,'Driver',0,w.REG_SZ,r'c:\windows\system\odbcjt32.dll')
w.SetValueEx(newKey,'DriverId',0,w.REG_DWORD,25)
w.SetValueEx(newKey,'FIL',0,w.REG_SZ,'MS Access;')
w.SetValueEx(newKey,'SafeTransactions',0,w.REG_DWORD,0)
w.SetValueEx(newKey,'UID',0,w.REG_SZ,'')

parentKey =
w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI\tri2',0,w.KEY_WRITE)
newKey = w.CreateKey(parentKey,'Engines')

parentKey =
w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI\tri2\Engines',0,w.KE
Y_WRITE)
newKey = w.CreateKey(parentKey,'Jet')
w.SetValueEx(newKey,'ImplicitCommitSync',0,w.REG_SZ,'')
w.SetValueEx(newKey,'MaxBufferSize',0,w.REG_DWORD,2048)
w.SetValueEx(newKey,'PageTimeout',0,w.REG_DWORD,5)
w.SetValueEx(newKey,'Threads',0,w.REG_DWORD,3)
w.SetValueEx(newKey,'UserCommitSync',0,w.REG_SZ,'Yes')

parentKey = w.OpenKey(w.HKEY_LOCAL_MACHINE,r'Software\ODBC\ODBC.INI\ODBC
Data Sources',0,w.KEY_WRITE)
w.SetValueEx(parentKey,'tri2',0,w.REG_SZ,'Microsoft Access Driver (*.mdb)')

Hope this helps someone else!

Scott

<hungjunglu at yahoo.com> wrote in message
news:mailman.981823871.8867.python-list at python.org...
> Hi,
>
> Is there a way of creating DSN from within Python (or Win32
> extensions)?
>
> Creating DSN manually is no problem, but it'll be nice to be able to
> add one from an application.
>
> How is it done usually? Via InstallShield?
>
> regards,
>
> Hung Jung
>
>
>




From hungjunglu at yahoo.com  Mon Feb 12 03:45:32 2001
From: hungjunglu at yahoo.com (hungjunglu at yahoo.com)
Date: Mon, 12 Feb 2001 02:45:32 -0000
Subject: creating DSN in Windows
In-Reply-To: <4xDh6.66249$bb.1038647@news1.rdc1.tx.home.com>
Message-ID: <967ioc+9p7t@eGroups.com>

--- In python-list at y..., "Scott Hathaway" <slhath at h...> wrote:
> I posted this a few weeks ago:
> ...
> Hope this helps someone else!
> 
> Scott

It surely does. Thanks so much! (See my other posting, where I 
already found your posted code.) I have done the code now in Python 
and Visual Basic, for Windows 95 and NT. And I think I'll just do it 
in C directly next week, just to make the executable as small as 
possible. I almost did in in Java, too, but somehow found out Java 
community is not very keen towards Windows registry... some of them 
charge money for some very simple api to talk to Windows dlls. Arrghh.

thanks again,

Hung Jung





From clickron at webtv.net  Sat Feb 10 17:56:18 2001
From: clickron at webtv.net (clickron at webtv.net)
Date: Sat, 10 Feb 2001 11:56:18 -0500 (EST)
Subject: def problem
Message-ID: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net>

I write this, save it. and run it.
def hello():
      print "hello"
      print "bye bye"

hello()

and I get this message

hello
Traceback innermost last):
File "<pyshell27>",line 1, in ?
hello
NameError: There is no variable named 'hello'

What am I doing wrong?

Ron





From emile at fenx.com  Sat Feb 10 19:34:35 2001
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 10 Feb 2001 10:34:35 -0800
Subject: def problem
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net>
Message-ID: <9641mn$iu28d$1@ID-11957.news.dfncis.de>

It must be in the way you're running it.  When I run it
here, I get pretty much what you'd expect:

hello
bye bye

If you give us some info on how you're trying to do this, we
should be able to get you going.

--

Emile van Sebille
emile at fenx.com
-------------------


<clickron at webtv.net> wrote in message
news:10844-3A8572B2-38 at storefull-161.iap.bryant.webtv.net...
> I write this, save it. and run it.
> def hello():
>       print "hello"
>       print "bye bye"
>
> hello()
>
> and I get this message
>
> hello
> Traceback innermost last):
> File "<pyshell27>",line 1, in ?
> hello
> NameError: There is no variable named 'hello'
>
> What am I doing wrong?
>
> Ron
>
>
>




From gregj at pobox.com  Sat Feb 10 19:53:49 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Sat, 10 Feb 2001 18:53:49 GMT
Subject: def problem
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net>
Message-ID: <9642nr$eqg$1@nnrp1.deja.com>

In article <10844-3A8572B2-38 at storefull-161.iap.bryant.webtv.net>,
  clickron at webtv.net wrote:
> I write this, save it. and run it.
> def hello():
>       print "hello"
>       print "bye bye"
>
> hello()
>
> and I get this message
>
> hello
> Traceback innermost last):
> File "<pyshell27>",line 1, in ?
> hello
> NameError: There is no variable named 'hello'
>
> What am I doing wrong?

It works for me. My guess is that you are putting the function in a
file, saving it, running it, and then typing "hello()" into IDLE (or
PyWin). That would cause the error you report. Is the hello() function
call in the same file as the function?

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From clickron at webtv.net  Sat Feb 10 22:13:04 2001
From: clickron at webtv.net (clickron at webtv.net)
Date: Sat, 10 Feb 2001 16:13:04 -0500 (EST)
Subject: def problem
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net>
Message-ID: <10844-3A85AEE0-152@storefull-161.iap.bryant.webtv.net>

I write this, save it. and run it.
def hello(): 
??????????????print "hello"
???????print "bye bye" 
hello() 
and I get this message 
hello
Traceback innermost last):
File "<pyshell27>",line 1, in ?
hello 
NameError: There is no variable named 'hello' 
What am I doing wrong? 
It works for me. My guess is that you are putting the function in a
file, saving it, running it, and then typing "hello()" into IDLE (or
PyWin). That would cause the error you report. Is the hello() function
call in the same file as the function? 
_________________________________

That's exactly what I'm doing. I go to IDLE, open a window, write the
function, save it, and run it. Then  I type in hello()
and get the error message. 

Now I retyped it, both with the function call  in the same file as the
function and without it. Did everything else the same and it works both
ways.

I don't know what's going on. How should I do it?

Ron



From ljz at asfast.com  Sat Feb 10 23:16:39 2001
From: ljz at asfast.com (Lloyd Zusman)
Date: 10 Feb 2001 17:16:39 -0500
Subject: def problem
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net> <10844-3A85AEE0-152@storefull-161.iap.bryant.webtv.net>
Message-ID: <ltd7cqi3p4.fsf@asfast.com>

clickron at webtv.net writes:

> I write this, save it. and run it.
> def hello(): 
> ??????????????print "hello"
> ???????print "bye bye" 
> hello() 
> and I get this message 
> hello
> Traceback innermost last):
> File "<pyshell27>",line 1, in ?
> hello 
> NameError: There is no variable named 'hello' 
> What am I doing wrong? 

Python is a language for which proper indentation is essential.
Try this:

def hello():
        print "hello"
        print "bye bye"

hello()


[Also, note the blank line after the function definition]

Does this work any better for you?


> [ ... ]
>
> Ron

-- 
 Lloyd Zusman
 ljz at asfast.com


From gregj at pobox.com  Sun Feb 11 04:55:44 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 11 Feb 2001 03:55:44 GMT
Subject: def problem
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net> <10844-3A85AEE0-152@storefull-161.iap.bryant.webtv.net>
Message-ID: <9652fv$7og$1@nnrp1.deja.com>

In article <10844-3A85AEE0-152 at storefull-161.iap.bryant.webtv.net>,
clickron at webtv.net wrote:

> It works for me. My guess is that you are putting the function in a
> file, saving it, running it, and then typing "hello()" into IDLE (or
> PyWin). That would cause the error you report. Is the hello() function
> call in the same file as the function?
> _________________________________
>
> That's exactly what I'm doing. I go to IDLE, open a window, write the
> function, save it, and run it. Then  I type in hello()
> and get the error message.
>
> Now I retyped it, both with the function call  in the same file as the
> function and without it. Did everything else the same and it works
both
> ways.

I think you had a typo or indentation problem somewhere.

What you may be trying to do is import the hello module:

---
# hello.py
def hello():
    print "hello, world"
---

>>> import hello.py
>>> hello()
hello, world!

This is explained in the Python tutorial at http://python.org.

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From m.faassen at vet.uu.nl  Mon Feb 12 01:09:25 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 12 Feb 2001 00:09:25 GMT
Subject: def problem
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net> <10844-3A85AEE0-152@storefull-161.iap.bryant.webtv.net> <9652fv$7og$1@nnrp1.deja.com>
Message-ID: <9679jl$c70$1@newshost.accu.uu.nl>

Greg Jorgensen <gregj at pobox.com> wrote:
> What you may be trying to do is import the hello module:

> ---
> # hello.py
> def hello():
>     print "hello, world"
> ---

>>>> import hello.py
>>>> hello()
> hello, world!

This won't work. Instead try this (keep hello.py the same):

>>> import hello
>>> hello.hello()

or alternatively:

>>> from hello import hello
>>> hello()

There's also the 'from hello import *' example which I won't demonstrate. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From clickron at webtv.net  Sun Feb 11 16:41:51 2001
From: clickron at webtv.net (clickron at webtv.net)
Date: Sun, 11 Feb 2001 10:41:51 -0500 (EST)
Subject: def problem (Got it!)
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net>
Message-ID: <25253-3A86B2BF-12@storefull-163.iap.bryant.webtv.net>

As a newbie you have to know I'm going to make some stupid mistakes, and
this was one of them.

What I was doing was sometimes clicking "run script" and mostly hitting
the "F5 key" which I just realized is not "run script" but "import
module."

I might have another question, but I think I'll go and hang my head in
shame for while. 

Thanks again for taking the time to answer my question.

Ron



From sholden at holdenweb.com  Sun Feb 11 18:26:27 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 11 Feb 2001 12:26:27 -0500
Subject: def problem (Got it!)
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net> <25253-3A86B2BF-12@storefull-163.iap.bryant.webtv.net>
Message-ID: <S_zh6.3841$n4.29052@e420r-atl2.usenetserver.com>

<clickron at webtv.net> wrote in message
news:25253-3A86B2BF-12 at storefull-163.iap.bryant.webtv.net...
> As a newbie you have to know I'm going to make some stupid mistakes, and
> this was one of them.
>
Don't worry, we all do it (except Tim Peters).

> What I was doing was sometimes clicking "run script" and mostly hitting
> the "F5 key" which I just realized is not "run script" but "import
> module."
>
Cool, so you have found ot what was wrong.

> I might have another question, but I think I'll go and hang my head in
> shame for while.
>
Please don't feel you need to do that.

> Thanks again for taking the time to answer my question.
>
> Ron
>
>



From carifio.nospam at nospam.usys.com  Sat Feb 10 17:59:26 2001
From: carifio.nospam at nospam.usys.com (Mike Carifio)
Date: Sat, 10 Feb 2001 11:59:26 -0500
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
Message-ID: <t8c7a69cqerq7c@corp.supernews.com>

I have a network of Linux and Windows machines. On a Windows box, I have
several python scripts which I'd also like to run on the Linux machines.
I've
smbmounted the Windows share to see the scripts, but having been created
on the Windows box, the line termination in those files is \r\n. When I try
to execute the script, say 'start', I get:

bash: ./start.py: No such file or directory

When I "clean up" start.py with tr - tr -d '\r' < start.py > reallystart.py,
it
runs. So it seems like python 2 can't compensate for wierd Windows line
terminators. Any workarounds, other than 'tr'? Pl. advise. Tx.




From porter at et.byu.edu  Sun Feb 11 07:29:41 2001
From: porter at et.byu.edu (C. Porter Bassett)
Date: Sat, 10 Feb 2001 23:29:41 -0700 (MST)
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
In-Reply-To: <t8c7a69cqerq7c@corp.supernews.com>
References: <t8c7a69cqerq7c@corp.supernews.com>
Message-ID: <Pine.GHP.4.21.0102102328170.26046-100000@don.et.byu.edu>

I have noticed the same thing.  One easy work-around is to run 
"python start.py" instead of "./start.py".  It appears that it is bash
that has problems with the \r\n, not python.

--------------------------------------------------------------------------
C. Porter Bassett     porter at et.byu.edu   http://www.et.byu.edu/~porter
--------------------------------------------------------------------------
"Pretend like this is a really witty saying." - Anonymous
--------------------------------------------------------------------------

On Sat, 10 Feb 2001, Mike Carifio wrote:

> I have a network of Linux and Windows machines. On a Windows box, I have
> several python scripts which I'd also like to run on the Linux machines.
> I've
> smbmounted the Windows share to see the scripts, but having been created
> on the Windows box, the line termination in those files is \r\n. When I try
> to execute the script, say 'start', I get:
> 
> bash: ./start.py: No such file or directory
> 
> When I "clean up" start.py with tr - tr -d '\r' < start.py > reallystart.py,
> it
> runs. So it seems like python 2 can't compensate for wierd Windows line
> terminators. Any workarounds, other than 'tr'? Pl. advise. Tx.
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From sholden at holdenweb.com  Sun Feb 11 18:10:52 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 11 Feb 2001 12:10:52 -0500
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com>
Message-ID: <qMzh6.1817$o4.18669@e420r-atl1.usenetserver.com>

"Mike Carifio" <carifio.nospam at nospam.usys.com> wrote in message
news:t8c7a69cqerq7c at corp.supernews.com...
> I have a network of Linux and Windows machines. On a Windows box, I have
> several python scripts which I'd also like to run on the Linux machines.
> I've
> smbmounted the Windows share to see the scripts, but having been created
> on the Windows box, the line termination in those files is \r\n. When I
try
> to execute the script, say 'start', I get:
>
> bash: ./start.py: No such file or directory
>
> When I "clean up" start.py with tr - tr -d '\r' < start.py >
reallystart.py,
> it
> runs. So it seems like python 2 can't compensate for wierd Windows line
> terminators. Any workarounds, other than 'tr'? Pl. advise. Tx.
>
I tend to use Programmer's File Editor, and keep my Python source code in
Unix format even on Windows machines.

regards
 Steve
--
Tools, training and technology to help you meet your information needs







From clarence at netlojix.com  Sun Feb 11 23:25:48 2001
From: clarence at netlojix.com (Clarence Gardner)
Date: Sun, 11 Feb 2001 14:25:48 -0800
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com> <qMzh6.1817$o4.18669@e420r-atl1.usenetserver.com>
Message-ID: <981930999.1446113299@news.silcom.com>

>"Mike Carifio" <carifio.nospam at nospam.usys.com> wrote in message
>news:t8c7a69cqerq7c at corp.supernews.com...
>> I have a network of Linux and Windows machines. On a Windows box, I have
>> several python scripts which I'd also like to run on the Linux machines.
>> I've
>> smbmounted the Windows share to see the scripts, but having been created
>> on the Windows box, the line termination in those files is \r\n. When I
>try
>> to execute the script, say 'start', I get:
>>
>> bash: ./start.py: No such file or directory
>>
>> When I "clean up" start.py with tr - tr -d '\r' < start.py >
>reallystart.py,
>> it
>> runs. So it seems like python 2 can't compensate for wierd Windows line
>> terminators. Any workarounds, other than 'tr'? Pl. advise. Tx.

Most likely, this is not a Python 'problem'; the 'problem' is with Linux.
The first line of your script has a '#!<path to Python>' and there is a
carriage return at the end of the word 'python'.  Since you don't have
a program on your system called 'python\r', it won't run.

The best thing to do is simply use newlines only, since it will work on
both.  I was going to say that you could also just put a space after the
word python on your shebang line (#!python \r), but then Python tries
to open '\r' as a script, because Linux supplies the \r as a command
line argument.  Sometimes, you just can't win....



From cfelling at iae.nl  Mon Feb 12 01:54:03 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 12 Feb 2001 01:54:03 +0100
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com> <qMzh6.1817$o4.18669@e420r-atl1.usenetserver.com> <981930999.1446113299@news.silcom.com>
Message-ID: <967c7b$hub$1@animus.fel.iae.nl>

...
> Most likely, this is not a Python 'problem'; the 'problem' is with Linux.
> The first line of your script has a '#!<path to Python>' and there is a
> carriage return at the end of the word 'python'.  Since you don't have
> a program on your system called 'python\r', it won't run.

> The best thing to do is simply use newlines only, since it will work on
> both.  I was going to say that you could also just put a space after the
> word python on your shebang line (#!python \r), but then Python tries
> to open '\r' as a script, because Linux supplies the \r as a command
> line argument.  Sometimes, you just can't win....

Ne, one space wont do, try around 112 and to your delight it will work 
-- 
groetjes, carel


From csrabak at uol.com.br  Mon Feb 12 03:47:52 2001
From: csrabak at uol.com.br (Cesar Rabak)
Date: Sun, 11 Feb 2001 23:47:52 -0300
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com> <qMzh6.1817$o4.18669@e420r-atl1.usenetserver.com> <981930999.1446113299@news.silcom.com>
Message-ID: <3A874ED8.75C9A92@uol.com.br>

Clarence Gardner wrote:
> 
> >"Mike Carifio" <carifio.nospam at nospam.usys.com> wrote in message
> >news:t8c7a69cqerq7c at corp.supernews.com...
> >> I have a network of Linux and Windows machines. On a Windows box, I have
> >> several python scripts which I'd also like to run on the Linux machines.
> >> I've
> >> smbmounted the Windows share to see the scripts, but having been created
> >> on the Windows box, the line termination in those files is \r\n. When I
> >try
> >> to execute the script, say 'start', I get:
> >>
> >> bash: ./start.py: No such file or directory
> >>
> >> When I "clean up" start.py with tr - tr -d '\r' < start.py >
> >reallystart.py,
> >> it
> >> runs. So it seems like python 2 can't compensate for wierd Windows line
> >> terminators. Any workarounds, other than 'tr'? Pl. advise. Tx.

I'm not a samba guru, but get a look (perhaps with a local expert) if
samba has the same functionality that mtools have to automagically
convert DOS (CRLF) to Unix (LF) text files. May be the final touch
you're looking for...

HTH

Cesar


From mbel44 at dial.pipex.net  Tue Feb 13 14:59:58 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Tue, 13 Feb 2001 13:59:58 +0000
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com> <qMzh6.1817$o4.18669@e420r-atl1.usenetserver.com> <981930999.1446113299@news.silcom.com> <3A874ED8.75C9A92@uol.com.br>
Message-ID: <cdfi8tsll6htt2gtjs8omckjgb146lui8p@4ax.com>

I use cvs to distribute scripts, which will happily take care of of
changing between \n and \r\n.


Toby Dickenson
tdickenson at geminidataloggers.com


From timr at probo.com  Mon Feb 12 07:16:12 2001
From: timr at probo.com (Tim Roberts)
Date: Sun, 11 Feb 2001 22:16:12 -0800
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com>
Message-ID: <nkve8tgpdq6s3103l562j09v5lsokskb2b@4ax.com>

"Mike Carifio" <carifio.nospam at nospam.usys.com> wrote:
>
>I have a network of Linux and Windows machines. On a Windows box, I have
>several python scripts which I'd also like to run on the Linux machines.
>I've
>smbmounted the Windows share to see the scripts, but having been created
>on the Windows box, the line termination in those files is \r\n. When I try
>to execute the script, say 'start', I get:
>
>bash: ./start.py: No such file or directory
>
>When I "clean up" start.py with tr - tr -d '\r' < start.py > reallystart.py,
>it runs. So it seems like python 2 can't compensate for wierd Windows line
>terminators. Any workarounds, other than 'tr'? Pl. advise. Tx.

Your diagnosis is incorrect, as you should have seen by the error message.
Python handles either line terminator on either operating system.  The
problem here is not Pyhton, it is the shell.  The first line of your script
probably says:
  #! /usr/bin/python

When bash tries to figure out how to run this file, it looks at the first
two characters.  When it finds #!, it assumes the next token on the line is
the name of a file to be executed to handle the script.  In your case, it
reads the file name as /usr/bin/python\r, and there is no such file or
directory.  Hence, the error.

The script will run correctly if you say:
  python start.py

Admittedly, this is not ideal.  You also might be able to work around this
by adding something to the #! line, but this might require experimentation.
Adding a single space might be an interesting option.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From carifio.nospam at nospam.usys.com  Mon Feb 12 17:32:54 2001
From: carifio.nospam at nospam.usys.com (Mike Carifio)
Date: Mon, 12 Feb 2001 11:32:54 -0500
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com>
Message-ID: <t8hegl11hkm465@corp.supernews.com>

Insightful responses.

Mr. Rabak suggested convincing Samba to convert the end of line
characteristics.
A good suggestion, but one that the Samba team decided against for several
versions.
They contend that they shouldn't fool with the insides of files. I tend to
agree with them.
In any event, its not in version 2.0.7.

So here's my solution, crufty though it is. I created a link to the python
executable
named 'python^M', in other words I added the carriage return to the end of
the file name
so bash would find it. (In bash, Ctrl-V inserts the next character as a
literal.) This way, I
don't have to worry about whether I created the file on Linux or NT. Umm,
that's not really
true, but I don't have to worry about line termination for invocation.
That's a start.

This "solution" unfortunately doesn't handle options, for example:

    #!/usr/local/bin/python -i
    print "some stuff"

doesn't work.

Thanks for help.





From stephen_purcell at yahoo.com  Tue Feb 13 09:46:23 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 13 Feb 2001 09:46:23 +0100
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
In-Reply-To: <t8hegl11hkm465@corp.supernews.com>; from carifio.nospam@nospam.usys.com on Mon, Feb 12, 2001 at 11:32:54AM -0500
References: <t8c7a69cqerq7c@corp.supernews.com> <t8hegl11hkm465@corp.supernews.com>
Message-ID: <20010213094623.D5558@freedom.puma-ag.com>

Mike Carifio wrote:
> So here's my solution, crufty though it is. I created a link to the python
> executable named 'python^M', in other words I added the carriage return to
> the end of the file name so bash would find it.

Ay-ay-ay...

Ingenious lateral thinking, but...

I can't imagine why people freak out when they can't write UNIX scripts that
don't have UNIX line endings! If the first line starts '#!' it should finish
'\n', not '\r\n'. If it doesn't, fix the file, not the OS.

-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From csrabak at uol.com.br  Tue Feb 13 15:45:23 2001
From: csrabak at uol.com.br (Cesar Rabak)
Date: Tue, 13 Feb 2001 11:45:23 -0300
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com> <t8hegl11hkm465@corp.supernews.com> <mailman.982053978.29454.python-list@python.org>
Message-ID: <3A894883.61C64508@uol.com.br>

Steve Purcell wrote:
> 
> Mike Carifio wrote:
> > So here's my solution, crufty though it is. I created a link to the python
> > executable named 'python^M', in other words I added the carriage return to
> > the end of the file name so bash would find it.
> 
> Ay-ay-ay...
> 
> Ingenious lateral thinking, but...
> 
> I can't imagine why people freak out when they can't write UNIX scripts that
> don't have UNIX line endings! If the first line starts '#!' it should finish
> '\n', not '\r\n'. If it doesn't, fix the file, not the OS.
> 

So elaborating, on your reasoning, Steve, I think the 'sollution' would
be the OP to have a kind of "header" or macro in his favorite editor
which inserts the correct line in whatever OS be the script being
written.

just my .01999...


From stephen_purcell at yahoo.com  Tue Feb 13 15:59:47 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 13 Feb 2001 15:59:47 +0100
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
In-Reply-To: <3A894883.61C64508@uol.com.br>; from csrabak@uol.com.br on Tue, Feb 13, 2001 at 11:45:23AM -0300
References: <t8c7a69cqerq7c@corp.supernews.com> <t8hegl11hkm465@corp.supernews.com> <mailman.982053978.29454.python-list@python.org> <3A894883.61C64508@uol.com.br>
Message-ID: <20010213155947.B759@freedom.puma-ag.com>

Cesar Rabak wrote:
> 
> So elaborating, on your reasoning, Steve, I think the 'sollution' would
> be the OP to have a kind of "header" or macro in his favorite editor
> which inserts the correct line in whatever OS be the script being
> written.

Perhaps, although what's the simplest thing? Save the file in UNIX text
format (there must be Windows editors that support this).

As Toby Dickenson suggests, you can use a technique like CVS' filters to
transparently correct line endings according to the system on which files are
checked out from source code control. In this case, files are usually saved
in UNIX format in the source repository, and Windowsified as necessary on
checkout.

I think there's a utility that ships with Python for changing the #! line of
a number of scripts -- perhaps that would help. Anybody remember the name?

-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From carifio.nospam at nospam.usys.com  Tue Feb 13 14:11:47 2001
From: carifio.nospam at nospam.usys.com (Mike Carifio)
Date: Tue, 13 Feb 2001 08:11:47 -0500
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com> <t8hegl11hkm465@corp.supernews.com> <mailman.982053978.29454.python-list@python.org> <3A894883.61C64508@uol.com.br> <"from csrabak"@uol.com.br> <20010213155947.B759@freedom.puma-ag.com> <mailman.982078635.28038.python-list@python.org>
Message-ID: <t8jn3l8im23v52@corp.supernews.com>

I don't want to provoke an argument. And I certainly won't try to defend the
'^M' hack in the file name itself.
But other technologies that started on Unix have made some accomodations to
the pragmatics of other
platforms (typically starting with 'W'). Python itself, for example, doesn't
choke on \r\n or \n lines. I have a hetrogeneous environment, not all my
scripts are under CVS control and I don't want to have to remember if I'm
creating a Unix file with a Windows utility or a Windows file with a Unix
utility and so forth. If bash accomodated
both \r\n and \n at the end of the first line, I wouldn't be asking the
question because it wouldn't matter. It
would just work.

In any event, the original question was "are there work arounds?" To
summarize the answers:

1) Yes, but the workarounds are crufty. Obey the \n rule. In some cases, you
can get your favorite editor
     or IDE to do it for you regardless of platform.
2) You can graft \r onto the python filename. Pro: you can create scripts
from Windows or Linux on the native
     file system or over Samba. Cons: You can't pass additional arguments,
it leaves a foul taste in your mouth.
3) If you put your scripts under source code control, you can convert end of
line on checkout.

All this was helpful. Thanks.




From scarblac at pino.selwerd.nl  Thu Feb 15 01:19:29 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 15 Feb 2001 00:19:29 GMT
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
References: <t8c7a69cqerq7c@corp.supernews.com> <t8hegl11hkm465@corp.supernews.com> <mailman.982053978.29454.python-list@python.org> <3A894883.61C64508@uol.com.br> <"from csrabak"@uol.com.br> <20010213155947.B759@freedom.puma-ag.com> <mailman.982078635.28038.python-list@python.org> <t8jn3l8im23v52@corp.supernews.com>
Message-ID: <slrn98m80u.67u.scarblac@pino.selwerd.nl>

Mike Carifio <carifio.nospam at nospam.usys.com> wrote in comp.lang.python:
> If bash accomodated both \r\n and \n at the end of the first line, I 
> wouldn't be asking the question because it wouldn't matter. It would 
> just work.

It's not bash doing that, it's the Linux kernel. And for all it knows,
you're just asking for a file that ends with \r, which is perfectly legal.
Actually, files with \n in them are perfectly legal as well but they won't
work with a #! line I think, so maybe you have a point... But it has to work
this way to be compatible with Unix.

>From a distance I'd guess the chance of the Linux kernel understanding
Windows text files for #! in the near future is extremely close to 0...

> In any event, the original question was "are there work arounds?" To
> summarize the answers:
> 
> 1) Yes, but the workarounds are crufty. Obey the \n rule. In some cases, you
> can get your favorite editor
>      or IDE to do it for you regardless of platform.
> 2) You can graft \r onto the python filename. Pro: you can create scripts
> from Windows or Linux on the native
>      file system or over Samba. Cons: You can't pass additional arguments,
> it leaves a foul taste in your mouth.
> 3) If you put your scripts under source code control, you can convert end of
> line on checkout.

I'd say run a utility that mass converts your files after you move them to
the linux server. Also, if you move them by ftp in ASCII mode, it will
convert the files.

-- 
Remco Gerlich


From dsh8290 at rit.edu  Tue Feb 13 16:35:56 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 13 Feb 2001 10:35:56 -0500
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
In-Reply-To: <20010213155947.B759@freedom.puma-ag.com>; from stephen_purcell@yahoo.com on Tue, Feb 13, 2001 at 03:59:47PM +0100
References: <t8c7a69cqerq7c@corp.supernews.com> <t8hegl11hkm465@corp.supernews.com> <mailman.982053978.29454.python-list@python.org> <3A894883.61C64508@uol.com.br> <"from csrabak"@uol.com.br> <20010213155947.B759@freedom.puma-ag.com>
Message-ID: <20010213103555.B10906@harmony.cs.rit.edu>

On Tue, Feb 13, 2001 at 03:59:47PM +0100, Steve Purcell wrote:
| Cesar Rabak wrote:
| > 
| > So elaborating, on your reasoning, Steve, I think the 'sollution' would
| > be the OP to have a kind of "header" or macro in his favorite editor
| > which inserts the correct line in whatever OS be the script being
| > written.
| 
| Perhaps, although what's the simplest thing? Save the file in UNIX text
| format (there must be Windows editors that support this).

There are -- I use gvim :-).

:set fileformat=unix


| 
| As Toby Dickenson suggests, you can use a technique like CVS' filters to
| transparently correct line endings according to the system on which files are
| checked out from source code control. In this case, files are usually saved
| in UNIX format in the source repository, and Windowsified as necessary on
| checkout.
| 
| I think there's a utility that ships with Python for changing the #! line of
| a number of scripts -- perhaps that would help. Anybody remember the name?
| 

I don't know if one came with it, but it shouldn't be hard to make
such a thing.  (Just be sure to open the file in _binary_ not text
mode) It would be simplest to use a RandomAccessFile type of interface
(there is a Java class by that name) and simply change the first \r\n
to \n\n.

In (g)vim you could type the following commands :

:1 
:set binary
:s/^V^M//


where ^V means Ctrl-V and ^M means Ctrl-M (or enter, in this case)

-D




From olivier.deckmyn at mail.dotcom.fr  Sat Feb 10 20:03:54 2001
From: olivier.deckmyn at mail.dotcom.fr (Olivier DECKMYN)
Date: Sat, 10 Feb 2001 20:03:54 +0100
Subject: Getting "current" module file ?
Message-ID: <964317$1hpm$1@feed.teaser.fr>

Hi all,

I am looking for a magic function that would tell me where is stored the
file that is currently running.
A good example is better that thousands of words :

here is /home/olivier/python/test1.py :
"""
import string
...

print "The current module test1 is stored in", magic_get_location()
"""

would get :
The current module test1 is stored in /home/olivier/python/test1.py

Thanx for your support.

ps : Python is real nice piece of code...







From emile at fenx.com  Sat Feb 10 20:53:09 2001
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 10 Feb 2001 11:53:09 -0800
Subject: Getting "current" module file ?
References: <964317$1hpm$1@feed.teaser.fr>
Message-ID: <96469u$jkgg1$1@ID-11957.news.dfncis.de>

For imported modules, a __file__ attribute exists that tells
where it came from.

If the name of the current module is __main__, you'll need
to check argv for the file name, but unless it's fully
qualified, you'll need to guess at the directory.  pwd may
help, but if it has been changed, may not.

#-----test.py----
import bisect
print bisect.__file__ # --> d:\python20\lib\bisect.pyc
print __name__ # --> __main__
print __main__.__file__ # --> error
#----end test.py----


d:\python20\lib\bisect.pyc
__main__
Traceback (most recent call last):
  File "C:\My Documents\test1.py", line 4, in ?
    print __main__.__file__
NameError: There is no variable named '__main__'

--

Emile van Sebille
emile at fenx.com
-------------------


"Olivier DECKMYN" <olivier.deckmyn at mail.dotcom.fr> wrote in
message news:964317$1hpm$1 at feed.teaser.fr...
> Hi all,
>
> I am looking for a magic function that would tell me where
is stored the
> file that is currently running.
> A good example is better that thousands of words :
>
> here is /home/olivier/python/test1.py :
> """
> import string
> ...
>
> print "The current module test1 is stored in",
magic_get_location()
> """
>
> would get :
> The current module test1 is stored in
/home/olivier/python/test1.py
>
> Thanx for your support.
>
> ps : Python is real nice piece of code...
>
>
>
>
>




From tawee199 at ksc.th.com  Sat Feb 10 20:41:47 2001
From: tawee199 at ksc.th.com (Tawee Laoitichote)
Date: Sun, 11 Feb 2001 02:41:47 +0700
Subject: How to enable native language
Message-ID: <96469l$i5j$1@news.ksc.co.th>

Hello, I have python 2.0 installed on my win'98 environment. How can I
enable my native language on it. My native language is "Thai"

Thanks
Tawee Language




From deragon at aqiii.org  Sat Feb 10 21:01:31 2001
From: deragon at aqiii.org (deragon at aqiii.org)
Date: Sat, 10 Feb 2001 20:01:31 -0000
Subject: Compiling Python 2.0 on RedHat 7.0. -> XML stuff missing.
Message-ID: <9646mr+7cfv@eGroups.com>

Greetings.


  I am trying to compile Python 2.0 on RedHat 7.0, but I get
the following problem:

gcc -fpic  -I/usr/local/include/xmlparse -g -O2 -Wall
-Wstrict-prototypes -I./../Include -I.. -DHAVE_CONFIG_H -c ./pyexpat.c
./pyexpat.c: In function `newxmlparseobject':
./pyexpat.c:639: warning: implicit declaration of function
`XML_ParserCreateNS'
./pyexpat.c:639: warning: assignment makes pointer from integer
without a cast
./pyexpat.c: In function `pyxml_SetStartNamespaceDeclHandler':
./pyexpat.c:1024: `XML_SetNamespaceDeclHandler' undeclared (first use
in this function)
./pyexpat.c:1024: (Each undeclared identifier is reported only once
./pyexpat.c:1024: for each function it appears in.)
./pyexpat.c: In function `pyxml_SetEndNamespaceDeclHandler':
./pyexpat.c:1032: `XML_SetNamespaceDeclHandler' undeclared (first use
in this function)
./pyexpat.c: In function `pyxml_SetStartCdataSection':
./pyexpat.c:1040: `XML_SetCdataSectionHandler' undeclared (first use
in this function)
./pyexpat.c: In function `pyxml_SetEndCdataSection':
./pyexpat.c:1048: `XML_SetCdataSectionHandler' undeclared (first use
in this function)
./pyexpat.c: At top level:
./pyexpat.c:1077: `XML_SetCommentHandler' undeclared here (not in a
function)
./pyexpat.c:1077: initializer element is not constant
./pyexpat.c:1077: (near initialization for `handler_info[8].setter')
./pyexpat.c:1089: `XML_SetDefaultHandlerExpand' undeclared here (not
in a function)
./pyexpat.c:1089: initializer element is not constant
./pyexpat.c:1089: (near initialization for `handler_info[12].setter')
./pyexpat.c:1092: `XML_SetNotStandaloneHandler' undeclared here (not
in a function)
./pyexpat.c:1092: initializer element is not constant
./pyexpat.c:1092: (near initialization for `handler_info[13].setter')
make[1]: *** [pyexpat.o] Error 1
make[1]: Leaving directory `/usr/src/redhat/BUILD/Python-2.0/Modules'
make: *** [sharedmods] Error 2


  I do have expat-1.95.0-1 installed.  I am building the
  python-2.0-2tummy.src.rpm.  Anybody have a clue how to get passed
  this problem?

  Or if there are any other ways to compile Python 2.0 on RedHat 7.0,
  please provide the "receipe".


Sincerely,
Hans Deragon




From TRoeschmann at ohltec.de  Sat Feb 10 22:54:41 2001
From: TRoeschmann at ohltec.de (=?iso-8859-1?Q?=22R=F6schmann=2C_Thies=22?=)
Date: Sat, 10 Feb 2001 22:54:41 +0100
Subject: CGIHTTPServer 
Message-ID: <2153422094D9D311A58B00508B9B199533EA57@HZ_KOMM>

Hello,

How can I configure my own CGIHTTPServer from Python1.5.2-Lib to run my
Python-scripts on WinNT4.0?
Or,
how can I set the world-executable and world-readable bits like this: 
        "chmod ugo+rx myfile.py"

Thanks for your help,
Thies



From emile at fenx.com  Sat Feb 10 23:56:49 2001
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 10 Feb 2001 14:56:49 -0800
Subject: CGIHTTPServer
References: <mailman.981842112.10482.python-list@python.org>
Message-ID: <964h29$j9l7o$1@ID-11957.news.dfncis.de>

I find Zope an easy to deploy alternative for simply running
scripts under NT.  Download and install the Zope software.
I generally install it to run as a service under NT.  Browse
to the manage interface, set up a new user with Owner and
Manager roles selected (highlight them in the list). Put a
sample.py in the extensions directory, something like:

def sample(self, REQUEST=REQUEST):
    "function doc string required by zope"
    return "There's no such thing as the P.S"

Then, sign back in to manage Zope as the new user.  Add a
new external method and specify sample for each of the
fields.

Now you should be able to browse to http://localhost/sample
and see the return value.  Changes to your function are
reflected immediately by Zope without further intervention.
I sometimes need to go into zope and update the external
method, but I'm unclear as to what kind of changes require
this (cache/reload?).

I'm not sure if this is what you were after, but perhaps it
helps.

--

Emile van Sebille
emile at fenx.com
-------------------


"R?schmann, Thies" <TRoeschmann at ohltec.de> wrote in message
news:mailman.981842112.10482.python-list at python.org...
> Hello,
>
> How can I configure my own CGIHTTPServer from
Python1.5.2-Lib to run my
> Python-scripts on WinNT4.0?
> Or,
> how can I set the world-executable and world-readable bits
like this:
>         "chmod ugo+rx myfile.py"
>
> Thanks for your help,
> Thies
>




From aleaxit at yahoo.com  Sun Feb 11 08:40:32 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 11 Feb 2001 08:40:32 +0100
Subject: CGIHTTPServer
References: <mailman.981842112.10482.python-list@python.org>
Message-ID: <965fr202v6o@news2.newsguy.com>

"R?schmann, Thies" <TRoeschmann at ohltec.de> wrote in message
news:mailman.981842112.10482.python-list at python.org...
> Hello,
>
> How can I configure my own CGIHTTPServer from Python1.5.2-Lib to run my
> Python-scripts on WinNT4.0?

Get the CGIHTTPServer from Python 2 -- it's been enhanced so it
does run well under NT too, so it will show you what changes you
have to make to your older version for the same purpose.


Alex





From rhyde99 at email.msn.com  Sat Feb 10 23:06:24 2001
From: rhyde99 at email.msn.com (Roger H)
Date: Sat, 10 Feb 2001 14:06:24 -0800
Subject: Python equivalent of static member data?
Message-ID: <#USeY36kAHA.359@cpmsnbbsa07>

I've just gotten done digging through the docs for 1.5.2,
and did not find an answer to my question:  Is there a
Python equivalent to the C++ concept of static member
data?  I'd like to make a class, and have each instance
of that class share a variable and its associated value.
Each instance would have the ability to modify that value,
and the new value would be updated through all instances.

Make sense?

~Roger the Shrubber~




From emile at fenx.com  Sat Feb 10 23:32:24 2001
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 10 Feb 2001 14:32:24 -0800
Subject: Python equivalent of static member data?
References: <#USeY36kAHA.359@cpmsnbbsa07>
Message-ID: <964fkh$j21aa$1@ID-11957.news.dfncis.de>

I'm not familiar with C++ static members, but it sounds like
you are looking for something like:

class Test:
    instances = 0
    def __init__(self, incBy=1):
        Test.instances = Test.instances + incBy
        print Test.instances

a = Test()
b = Test(2)


HTH,

--

Emile van Sebille
emile at fenx.com
-------------------


"Roger H" <rhyde99 at email.msn.com> wrote in message
news:#USeY36kAHA.359 at cpmsnbbsa07...
> I've just gotten done digging through the docs for 1.5.2,
> and did not find an answer to my question:  Is there a
> Python equivalent to the C++ concept of static member
> data?  I'd like to make a class, and have each instance
> of that class share a variable and its associated value.
> Each instance would have the ability to modify that value,
> and the new value would be updated through all instances.
>
> Make sense?
>
> ~Roger the Shrubber~
>
>




From f8dy at my-deja.com  Sun Feb 11 02:01:39 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Sun, 11 Feb 2001 01:01:39 GMT
Subject: Python equivalent of static member data?
References: <#USeY36kAHA.359@cpmsnbbsa07>
Message-ID: <964o9g$45$1@nnrp1.deja.com>

In article <#USeY36kAHA.359 at cpmsnbbsa07>,
  "Roger H" <rhyde99 at email.msn.com> wrote:
> I've just gotten done digging through the docs for 1.5.2,
> and did not find an answer to my question:  Is there a
> Python equivalent to the C++ concept of static member
> data?  I'd like to make a class, and have each instance
> of that class share a variable and its associated value.
> Each instance would have the ability to modify that value,
> and the new value would be updated through all instances.

In Python, these are called class attributes.  You can define them at design
time in your class by putting the attribute definition inside your class but
outside any methods.  You can also access/add/modify them at run time through
the built-in __class__ attribute in each instance of the class.

This is explained in more detail in my book, with an example of exactly what
you're trying to do here. 
http://diveintopython.org/fileinfo_classattributes.html

-M
--
You're smart; why haven't you learned Python yet?  http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From gregj at pobox.com  Sun Feb 11 04:39:29 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 11 Feb 2001 03:39:29 GMT
Subject: Python equivalent of static member data?
References: <#USeY36kAHA.359@cpmsnbbsa07>
Message-ID: <9651hh$6tp$1@nnrp1.deja.com>

In article <#USeY36kAHA.359 at cpmsnbbsa07>,
  "Roger H" <rhyde99 at email.msn.com> wrote:
> I've just gotten done digging through the docs for 1.5.2,
> and did not find an answer to my question:  Is there a
> Python equivalent to the C++ concept of static member
> data?  I'd like to make a class, and have each instance
> of that class share a variable and its associated value.
> Each instance would have the ability to modify that value,
> and the new value would be updated through all instances.

>From the Python Reference Manual, section 7.6:

Programmer's note: variables defined in the class definition are class
variables; they are shared by all instances. To define instance
variables, they must be given a value in the the __init__() method or
in another method. Both class and instance variables are accessible
through the notation ```codeself.name'', and an instance variable hides
a class variable with the same name when accessed in this way. Class
variables with immutable values can be used as defaults for instance
variables.

You can find this and much more at http://python.org.

If you are used to C++ or Java, Python's syntax for defining class
variables may be a bit confusing. For example, in C++:

  class foo {
      static int x = 33;
      int z;
      ...
  }

x is a class variable, shared by all instances of foo, and z is an
instance variable, unique to each instance of foo. The heavily
overloaded "static" keyword is what makes the difference.

The Python equivalent is:

  class foo:
      x = 33

      def __init__(self):
          self.z = 99

One big difference is that Python will let you easily hide the class
variable behind an instance variable:

>>> foo.x
33
>>> bar = foo()
>>> bar.x
33
>>> bar.z
99
>>> bar.x = 66  # creates a new instance variable for bar
>>> foo.x
33
>>> bar.x
66
>>> bar.__class__.x  # refer to class variable through instance
33
>>> del bar.x
>>> bar.x
33


There's a thread going on about the built-in name None that discusses
scopes and name hiding at greater length.

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From e_d_k at my-deja.com  Sat Feb 10 23:26:01 2001
From: e_d_k at my-deja.com (EdK)
Date: Sat, 10 Feb 2001 22:26:01 GMT
Subject: Whn to read a pipe
Message-ID: <964f5i$oqr$1@nnrp1.deja.com>

I'm having a problem. I use popen3 to run I program I want to controle from a
Python Program. My problem is that I want to read data from the progams stdio
pipe if there is data to read. If I try to read date when there is no data
the progam hangs until there is data present, I do not know how to tell if
there is data to read, and only read it if it is there. Can anyone help me
with this?

   -EdK


Sent via Deja.com
http://www.deja.com/


From bowman at montana.com  Sun Feb 11 05:05:02 2001
From: bowman at montana.com (bowman)
Date: Sat, 10 Feb 2001 21:05:02 -0700
Subject: Whn to read a pipe
References: <964f5i$oqr$1@nnrp1.deja.com>
Message-ID: <m8oh6.649$hF5.4895@newsfeed.slurp.net>

EdK <e_d_k at my-deja.com> wrote in message news:964f5i$oqr$1 at nnrp1.deja.com...
> I'm having a problem. I use popen3 to run I program I want to controle
from a
> Python Program. My problem is that I want to read data from the progams
stdio
> pipe if there is data to read. If I try to read date when there is no data
> the progam hangs until there is data present, I do not know how to tell if
> there is data to read, and only read it if it is there. Can anyone help me
> with this?

Depending on your OS, select may work for you.





From sl0op at my-deja.com  Sat Feb 10 23:26:06 2001
From: sl0op at my-deja.com (sl0op at my-deja.com)
Date: Sat, 10 Feb 2001 22:26:06 GMT
Subject: Python compiler/installer?
Message-ID: <964f5n$oqs$1@nnrp1.deja.com>

Hi. I'm new to Python (which seems like a very nice language to me) and
have a few questions. First, is there any way I can compile it so I
don't have to send out any source code I don't want to? Are there any
installers for either compiled-Python (if that exists) or simple .py
files? And where can I find some examples of Python programs ?(some
that will help me learn Tkinter.)

Any help would be greatly appreciated.

Thank you,
Jason


Sent via Deja.com
http://www.deja.com/


From bwilk_97 at yahoo.com  Sun Feb 11 02:42:35 2001
From: bwilk_97 at yahoo.com (Bill Wilkinson)
Date: Sun, 11 Feb 2001 01:42:35 GMT
Subject: Python compiler/installer?
References: <964f5n$oqs$1@nnrp1.deja.com>
Message-ID: <f6mh6.12452$lk6.897687@typhoon.midsouth.rr.com>


| Hi. I'm new to Python (which seems like a very nice language to me) and
| have a few questions. First, is there any way I can compile it so I
| don't have to send out any source code I don't want to?

It compiles to byte code. But the persistant can get to it if they want.
It decompiles quite nicely. :)

|Are there any installers for either compiled-Python (if that exists) or
simple .py
| files?

For windows only:
http://starship.python.net/crew/theller/   (Thanks Thomas!)

For windows and Linux:
http://starship.python.net/crew/gmcm/distribute.html


|And where can I find some examples of Python programs ?(some
| that will help me learn Tkinter.)

The best way to learn Tkinter IMHO is to read /F's tutorial, It's righteous.
http://www.secretlabs.com/library/tkinter/introduction/index.htm




From zamurai at gmx.net  Sat Feb 10 23:45:04 2001
From: zamurai at gmx.net (Zamurai)
Date: Sat, 10 Feb 2001 23:45:04 +0100
Subject: How to label loops?
Message-ID: <981845104.329821352@news.t-online.de>



From sholden at holdenweb.com  Sun Feb 11 18:20:32 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 11 Feb 2001 12:20:32 -0500
Subject: How to label loops?
References: <981845104.329821352@news.t-online.de>
Message-ID: <jVzh6.3840$n4.28565@e420r-atl2.usenetserver.com>

First, convince me that loops need labelling!

regards
 Steve
--
--
Tools, training and technology to help you meet your information needs



"Zamurai" <zamurai at gmx.net> wrote in message
news:981845104.329821352 at news.t-online.de...




From zamurai at gmx.net  Sun Feb 11 19:31:20 2001
From: zamurai at gmx.net (Zamurai)
Date: Sun, 11 Feb 2001 19:31:20 +0100
Subject: How to label loops?
References: <981845104.329821352@news.t-online.de> <jVzh6.3840$n4.28565@e420r-atl2.usenetserver.com>
Message-ID: <981916280.1465025599@news.t-online.de>

"Steve Holden" <sholden at holdenweb.com> wrote:
> First, convince me that loops need labelling!
> 
> regards
>  Steve
> --
> --
> Tools, training and technology to help you meet your information needs
> 
> 
> 
> "Zamurai" <zamurai at gmx.net> wrote in message
> news:981845104.329821352 at news.t-online.de...


for example:

while 1:                # first loop
    while 1:            # second loop 
        break


this break breaks only out of the first loop, but what, if i want to break out
of the first loop, while i?m in the second one?


From emile at fenx.com  Sun Feb 11 20:28:03 2001
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 11 Feb 2001 11:28:03 -0800
Subject: How to label loops?
References: <981845104.329821352@news.t-online.de> <jVzh6.3840$n4.28565@e420r-atl2.usenetserver.com> <981916280.1465025599@news.t-online.de>
Message-ID: <966p6r$jrm94$1@ID-11957.news.dfncis.de>

for i in range(10):
    for j in range(10):
        breakOuter = 0
        if i == 3 and j == 3:
            breakOuter = 1
            break
    if breakOuter:
        break
print i , j

HTH,

--

Emile van Sebille
emile at fenx.com
-------------------


"Zamurai" <zamurai at gmx.net> wrote in message
news:981916280.1465025599 at news.t-online.de...
"Steve Holden" <sholden at holdenweb.com> wrote:
> First, convince me that loops need labelling!
>
> regards
>  Steve
> --
> --
> Tools, training and technology to help you meet your
information needs
>
>
>
> "Zamurai" <zamurai at gmx.net> wrote in message
> news:981845104.329821352 at news.t-online.de...


for example:

while 1:                # first loop
    while 1:            # second loop
        break


this break breaks only out of the first loop, but what, if i
want to break out
of the first loop, while i?m in the second one?




From steffen.ries at sympatico.ca  Sun Feb 11 20:35:53 2001
From: steffen.ries at sympatico.ca (Steffen Ries)
Date: Sun, 11 Feb 2001 19:35:53 GMT
Subject: How to label loops?
References: <981845104.329821352@news.t-online.de> <jVzh6.3840$n4.28565@e420r-atl2.usenetserver.com> <981916280.1465025599@news.t-online.de>
Message-ID: <m3ofw9hv1k.fsf@gondolin.beleriand>

Zamurai <zamurai at gmx.net> writes:

> "Steve Holden" <sholden at holdenweb.com> wrote:
> > First, convince me that loops need labelling! 
... 
> 
> for example:
> 
> while 1:                # first loop
>     while 1:            # second loop 
>         break
> 
> 
> this break breaks only out of the first loop, but what, if i want to
> break out of the first loop, while i?m in the second one?

one way is to use control variables, e.g.:

outer_active = 1
inner_active = 1
while outer_active:
    while inner_active:
       outer_active = 0
       break

another way would be to use exceptions, e.g.:

try:
    while 1:
        while 1:
            raise Exception, "broken loop"
except Exception:
    pass

/steffen
-- 
steffen.ries at sympatico.ca	<> Gravity is a myth -- the Earth sucks!


From quinn at lira.ugcs.caltech.edu  Mon Feb 12 01:07:42 2001
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 12 Feb 2001 00:07:42 GMT
Subject: How to label loops?
References: <981845104.329821352@news.t-online.de> <jVzh6.3840$n4.28565@e420r-atl2.usenetserver.com> <981916280.1465025599@news.t-online.de> <m3ofw9hv1k.fsf@gondolin.beleriand>
Message-ID: <slrn98eaae.qne.quinn@lira.ugcs.caltech.edu>

On Sun, 11 Feb 2001 19:35:53 GMT, Steffen Ries <steffen.ries at sympatico.ca>
wrote:
>> while 1:                # first loop
>>     while 1:            # second loop 
>>         break
>> 
>> 
>> this break breaks only out of the first loop, but what, if i want to
>> break out of the first loop, while i?m in the second one?
>
>one way is to use control variables, e.g.:
>
>outer_active = 1
>inner_active = 1
>while outer_active:
>    while inner_active:
>       outer_active = 0
>       break
>
>another way would be to use exceptions, e.g.:
>
>try:
>    while 1:
>        while 1:
>            raise Exception, "broken loop"
>except Exception:
>    pass

That will catch and ignore all sorts of things you don't want to ignore, like
OSError and MemoryError.  If you must (ab)use an exception for control flow,
you should probably use a string:

try:
    while 1:
        while 1:
            raise 'outer loop'
except 'outer loop': # just don't misspell this :)
    pass

I've found that I rarely need to break from two levels at once, and when I
think I do, often python's (unique) 'while x: ... else:' construct will do the
trick.  But if you find yourself doing complicated 'while x: xxx else: break'
to break from the outer loop if the inner one isn't, and then break from the
inner loop if you want to *not* break the outer loop, which is only a nominal
loop since it has 'break' at the end and you use continue... well, what you
really want here is 'goto' :)

As in "go back to fortran" :)


From dvdavins at aol.com  Sat Feb 10 23:52:16 2001
From: dvdavins at aol.com (DvdAvins)
Date: 10 Feb 2001 22:52:16 GMT
Subject: Is Python for me?
Message-ID: <20010210175216.26827.00000474@ng-mh1.aol.com>

I'm planning to write a program for myself which, if I'm happy with the way it
turns out, I will likely distribute as shareware. I'm looking for as many of
the following traits as possible:

1:a functional language. I've been intrigued by functional languages for a
while, but have had no occasion to use them at work.
2: automatic garbage collection and other programmers' convenience features (I
still have to take care of my day job)
3: support for objects (not necessary, just a convenience)
4: it MUST be distributable as a stand-alone program for Windows. I'd like to
be able to distribute to Mac and Linux as well.
5. there has to a good book on the language available in English or at least an
IDE with and excdellent help environment.

I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common Lisp.
Which would you use? Why?


From j.spies at hccnet.nl  Sun Feb 11 00:07:02 2001
From: j.spies at hccnet.nl (Jaap Spies)
Date: Sun, 11 Feb 2001 00:07:02 +0100
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <3A85C996.9000504@hccnet.nl>

DvdAvins wrote:

> I'm planning to write a program for myself which, if I'm happy with the way it
> turns out, I will likely distribute as shareware. I'm looking for as many of
> the following traits as possible:
> 
> 1:a functional language. I've been intrigued by functional languages for a
> while, but have had no occasion to use them at work.
> 2: automatic garbage collection and other programmers' convenience features (I
> still have to take care of my day job)
> 3: support for objects (not necessary, just a convenience)
> 4: it MUST be distributable as a stand-alone program for Windows. I'd like to
> be able to distribute to Mac and Linux as well.
> 5. there has to a good book on the language available in English or at least an
> IDE with and excdellent help environment.
> 
> I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common Lisp.
> Which would you use? Why?

Grow up! Miracles are of another World! Are you joking?

Jaap Spies



From dvdavins at aol.com  Sun Feb 11 00:50:33 2001
From: dvdavins at aol.com (DvdAvins)
Date: 10 Feb 2001 23:50:33 GMT
Subject: Is Python for me?
References: <3A85C996.9000504@hccnet.nl>
Message-ID: <20010210185033.09372.00000461@ng-mm1.aol.com>

Huh? No, I'm not joking. Looking over my post, maybe something was unclear. I'm
not saying the language has to redistributable, jsut that is has to produce
redistributable compiled code. (Or, as in a few languages, produce shipable
interpertable code with a stripped-downl interperter included)

It seems Lisp, for one, meets all the criteria. Why shouldn't I get opinions on
what else might be interesting?


From m.faassen at vet.uu.nl  Mon Feb 12 00:18:26 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 11 Feb 2001 23:18:26 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <3A85C996.9000504@hccnet.nl>
Message-ID: <9676k2$brc$1@newshost.accu.uu.nl>

Jaap Spies <j.spies at hccnet.nl> wrote:
[snip]
> Grow up! Miracles are of another World! Are you joking?

What's the reason for such a weird response? I'm not used to such
behavior on the Python newsgroup, so please adjust. There are many languages
that more or less fulfill his criterions, and even if there weren't,
you could respond in a more polite fashion. The PSU thanks you.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From j.spies at hccnet.nl  Mon Feb 12 16:26:12 2001
From: j.spies at hccnet.nl (Jaap Spies)
Date: Mon, 12 Feb 2001 16:26:12 +0100
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <3A85C996.9000504@hccnet.nl> <9676k2$brc$1@newshost.accu.uu.nl>
Message-ID: <3A880094.6060906@hccnet.nl>

Martijn Faassen wrote:

> Jaap Spies <j.spies at hccnet.nl> wrote:
> [snip]
> 
>> Grow up! Miracles are of another World! Are you joking?
> 
> 
> What's the reason for such a weird response? I'm not used to such
> behavior on the Python newsgroup, so please adjust. There are many languages
> that more or less fulfill his criterions, and even if there weren't,
> you could respond in a more polite fashion. The PSU thanks you.
> 
> Regards,
> 
> Martijn

You'r right! I was a bit short (tempered).

But I don't agree that there are many languages (and environments)
that fulfill the stated criterions. If any. Python will be close.

I followed the thread and I admire all the kind answers.
Please accept my apologies.

Jaap Spies



From m.faassen at vet.uu.nl  Tue Feb 13 07:41:45 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 13 Feb 2001 06:41:45 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <3A85C996.9000504@hccnet.nl> <9676k2$brc$1@newshost.accu.uu.nl> <3A880094.6060906@hccnet.nl>
Message-ID: <96akv9$dik$2@newshost.accu.uu.nl>

Jaap Spies <j.spies at hccnet.nl> wrote:
[snip]
> You'r right! I was a bit short (tempered).

> But I don't agree that there are many languages (and environments)
> that fulfill the stated criterions. If any. Python will be close.

Well, I used 'more or less' fulfill, a conveniently vague expression. :)

> I followed the thread and I admire all the kind answers.
> Please accept my apologies.

And accept my thanks; that was gracious of you.

Thanks again,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From paul at fxtech.com  Sun Feb 11 02:06:35 2001
From: paul at fxtech.com (Paul Miller)
Date: Sat, 10 Feb 2001 19:06:35 -0600
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <3A85E59B.C58F86E8@fxtech.com>

> I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common Lisp.
> Which would you use? Why?

Python will handle all of that. It is a great language.

--
Paul Miller - stele at fxtech.com


From m.faassen at vet.uu.nl  Mon Feb 12 00:20:43 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 11 Feb 2001 23:20:43 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <3A85E59B.C58F86E8@fxtech.com>
Message-ID: <9676ob$brc$2@newshost.accu.uu.nl>

Paul Miller <paul at fxtech.com> wrote:
>> I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common Lisp.
>> Which would you use? Why?

> Python will handle all of that. It is a great language.

While Python handles all the other criteria listed it is not a 
functional programming language (though it does have some functional
programming features). Python's a great language, but it's not entirely
correct to say it'll handle it all. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From f8dy at my-deja.com  Sun Feb 11 02:27:32 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Sun, 11 Feb 2001 01:27:32 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <964ppu$156$1@nnrp1.deja.com>

In article <20010210175216.26827.00000474 at ng-mh1.aol.com>,
  dvdavins at aol.com (DvdAvins) wrote:
> I'm planning to write a program for myself which, if I'm happy with the way it
> turns out, I will likely distribute as shareware. I'm looking for as many of
> the following traits as possible:

Python has all of these traits.

> 1:a functional language. I've been intrigued by functional languages for a
> while, but have had no occasion to use them at work.

Python has many features of functional programming, including lambda, map,
filter, and reduce.  http://diveintopython.org/apihelper_lambda.html 
http://python.org/doc/current/tut/node7.html

> 2: automatic garbage collection and other programmers' convenience features (I
> still have to take care of my day job)

Python has automatic garbage collection based on a combination of reference
counting and mark-and-sweep collection. 
http://diveintopython.org/fileinfo_instantiation.html

> 3: support for objects (not necessary, just a convenience)

Python is fully object-oriented.
  http://diveintopython.org/fileinfo_divein.html
  http://python.org/doc/current/tut/node11.html

> 4: it MUST be distributable as a stand-alone program for Windows. I'd like to
> be able to distribute to Mac and Linux as well.

Although I have never done it personally, Python comes with utilities to
"freeze" a program and distribute it in object (not source) form.  Python is
cross-platform, and you can freeze for any supported platform (including the
three you mention).

> 5. there has to a good book on the language available in English or at least an
> IDE with and excdellent help environment.

O'Reilly just did a review of several free Python books available online.
  http://www.oreillynet.com/pub/a/python/2001/02/07/pythonnews.html

If you are on Windows, I would recommend ActiveState's IDE, which has
method/attribute completion and popup help (similar to Visual Studio), syntax
highlighting, integrated debugging, and a built-in method/class browser. 
http://www.activestate.com/Products/ActivePython/

Hope this helps.

-M
--
You're smart; why haven't you learned Python yet?  http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From bwilk_97 at yahoo.com  Sun Feb 11 02:42:35 2001
From: bwilk_97 at yahoo.com (Bill Wilkinson)
Date: Sun, 11 Feb 2001 01:42:35 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <f6mh6.12451$lk6.897622@typhoon.midsouth.rr.com>

| I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common
Lisp.
| Which would you use? Why?

An older comparison by Guido on python.org.
http://www.python.org/doc/essays/comparisons.html

A cool comparison by Norvig of Lisp and Python.
http://www.norvig.com/python-lisp.html






From gregj at pobox.com  Sun Feb 11 04:15:16 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 11 Feb 2001 03:15:16 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <965041$618$1@nnrp1.deja.com>

In article <20010210175216.26827.00000474 at ng-mh1.aol.com>,
dvdavins at aol.com (DvdAvins) wrote:
> I'm planning to write a program for myself which, if I'm happy with
> the way it turns out, I will likely distribute as shareware.
> ...
> I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and
Common Lisp. Which would you use? Why?

You don't say anything about the program you are planning, so no one
can give a meaningful answer to your question. The language I would
choose for a cross-platform GUI application may not be the language I'd
choose for writing web server applications, or querying a database.
There is no one best programming language for everything.

I would use the language and tools I am most productive with. For the
vast majority of programmers, that means using one of the handful (or
fewer) programming languages they know well. I can't tell for sure, but
your post gives the impression that you don't know any of the
programming languages you are asking about. Do you have any programming
experience at all? If yes, what language or languages are you most
experienced and comfortable with?

Imagine if someone asked you: "I'm planning to take some photographs,
and if they turn out I may sell them. Should I use a Nikon, Minolta,
Canon, Hasselblad, or Olympus camera? Why?" What kind of sensible
answer could you give them?

If you don't have any significant programming experience, I'd say
Python is a great choice to start with. The language is easy to learn,
tools are widely available, several good books are in bookstores, and
the support community is first-rate. But the hard part is learning how
to program, not learning a language. I can teach my kids how to use a
camera, but that won't make them as good as Ansel Adams.

For more about learning programming and why you should be prepared to
spend a lot of time at it, have a look at:

http://www.norvig.com/21-days.html

--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From dsavitsk at e-coli.net  Sun Feb 11 05:14:27 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sat, 10 Feb 2001 22:14:27 -0600
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <965041$618$1@nnrp1.deja.com>
Message-ID: <Hkoh6.87168$Ch.14259401@newsrump.sjc.telocity.net>


> Imagine if someone asked you: "I'm planning to take some photographs,
> and if they turn out I may sell them. Should I use a Nikon, Minolta,
> Canon, Hasselblad, or Olympus camera? Why?" What kind of sensible
> answer could you give them?

i would say they should use a leica ;-)

ds




From dvdavins at aol.com  Sun Feb 11 06:22:36 2001
From: dvdavins at aol.com (DvdAvins)
Date: 11 Feb 2001 05:22:36 GMT
Subject: Is Python for me?
References: <965041$618$1@nnrp1.deja.com>
Message-ID: <20010211002236.20899.00000455@ng-fr1.aol.com>

>You don't say anything about the program you are planning, so no one
>can give a meaningful answer to your question. The language I would
>choose for a cross-platform GUI application may not be the language I'd
>choose for writing web server applications, or querying a database.
>There is no one best programming language for everything.
>
>I would use the language and tools I am most productive with. For the
>vast majority of programmers, that means using one of the handful (or
>fewer) programming languages they know well. I can't tell for sure, but
>your post gives the impression that you don't know any of the
>programming languages you are asking about. Do you have any programming
>experience at all? If yes, what language or languages are you most
>experienced and comfortable with?

I do have programming experience, but the vast majority of my work has been
LotusScript (an OO variant of VB) or Domino/Notes's @Formula language
(which has lovely list operations and can *almost* be made to behave like a
real
language, despite its total lack of a looping/iteration construct). I've also
written
production code in Javascript and Sybase's stored procedure language (T-SQL
is the name, if I remember right). On top of that, I fiddled around a little in
Java
recently and C, Fortran, and Pascal in the distant past.

But the point here is to use something I'm *not* familiar with and something
that
will help me think in new ways.

The project is a non-animated game. To a less technical audience, I'd say
"non-graphical", but I will use a GUI. The game will rely on an extensive
database of actual historical information, but that database will be updatable
and
even replacable by the user. Therefor I imagine I'll use CSVs rather than write
a custom import/export interface. I don't think I'll need the performance of an
RDBMS.

Incidentally, there are similar programs out there now, but they rely on
proprietary data formats. My innovation would be to sell people the game
without making them have to pay me to update the data.

The first iteration of the game will be, as you say, a cross-platform GUI
application, but there is also a market for a web-based version.


From m.faassen at vet.uu.nl  Mon Feb 12 00:33:35 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 11 Feb 2001 23:33:35 GMT
Subject: Is Python for me?
References: <965041$618$1@nnrp1.deja.com> <20010211002236.20899.00000455@ng-fr1.aol.com>
Message-ID: <9677gf$brc$3@newshost.accu.uu.nl>

DvdAvins <dvdavins at aol.com> wrote:
[snip]
> I do have programming experience, but the vast majority of my work has been
> LotusScript (an OO variant of VB) or Domino/Notes's @Formula language
> (which has lovely list operations and can *almost* be made to behave like a
> real language, despite its total lack of a looping/iteration construct).
> I've also written production code in Javascript and Sybase's stored
> procedure language (T-SQL is the name, if I remember right). On top of that,
> I fiddled around a little in Java recently and C, Fortran, and Pascal
> in the distant past.

Given this background, Python would be among the languages easiest to
learn for you. It's a breath of fresh air to step away from application
specific scripting languages to a general one, I imagine, and it's still
far higher level than Java, and definitely C, Fortran or Pascal. If you're
interested in object oriented programming techniques Python is a very
nice language to help learn about them as well. Given your background 
Python may also be intereresting if you want to learn about how to
scale your applications up to larger sizes, while you still don't lose
easy integration abilities with all kinds of external systems (databases,
GUI toolkits, etc).

> But the point here is to use something I'm *not* familiar with and something
> that will help me think in new ways.

I imagine there will be enough to learn that you're not familiar with in
Python. But if this is really a big goal of yours then a functional language
may be more suitable. Python has some properties of functional languages
but is not really one itself. 

> The project is a non-animated game. To a less technical audience, I'd say
> "non-graphical", but I will use a GUI. The game will rely on an extensive
> database of actual historical information, but that database will be
> updatable and even replacable by the user.

> Therefor I imagine I'll use CSVs rather than write
> a custom import/export interface. I don't think I'll need the performance of an
> RDBMS.

If you're interested in learning new things, you could also look into
the Zope Object Database that you can use with Python, or just the
plain 'pickle' module that comes with the Python standard library.

> Incidentally, there are similar programs out there now, but they rely on
> proprietary data formats. My innovation would be to sell people the game
> without making them have to pay me to update the data.

Makes me curious about what the game could be. Historical information?

> The first iteration of the game will be, as you say, a cross-platform GUI
> application, but there is also a market for a web-based version.

Python is good at glueing all kinds of components together, probably
better than most of the other languages you're looking at. So if this is
a criterion you should look at Python.

Regards,

Martijn 
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From tim_one at email.msn.com  Sun Feb 11 06:45:05 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Sun, 11 Feb 2001 00:45:05 -0500
Subject: Is Python for me?
In-Reply-To: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEKJIOAA.tim_one@email.msn.com>

[DvdAvins]
> ...
> I'm looking for as many of the following traits as possible:

I bet you're not married <wink>.

> 1:a functional language. I've been intrigued by functional
> languages for a while, but have had no occasion to use them at work.
> 2: automatic garbage collection and other programmers'
> convenience features (I still have to take care of my day job)
> 3: support for objects (not necessary, just a convenience)
> 4: it MUST be distributable as a stand-alone program for Windows.

You have to understand that this one implies large executables, because the
language's entire runtime system and libraries have to be built into the
standalone.  Standalones for, e.g., C programs can be much smaller, because
they can *assume* every system already has the standard C runtime libraries
installed (or a JVM, in the case of Java).

> I'd like to be able to distribute to Mac and Linux as well.

This is at odds with desiring standalones, unless you want to build a
different standalone for every platform.

> 5. there has to a good book on the language available in English
> or at least an IDE with and excdellent help environment.
>
> I'm investigating at M,

Don't know it.

> Python,

Not functional at heart; making standalones is possible with enough pain but
not how the language was intended to be used (i.e., you'll fight it in this
area).

> BETA,

Could be cool.  Unsure about license restrictions (it's a commercial
product).

> Scheme,

Many, many implementations available, and generally compact.  Does it
support objects?  Sure, provided you build them yourself -- which is a
common answer to many "how do I xyz?" questions in Scheme.  Nevertheless,
it's a beautiful little language.

> Haskell,

Lots of fun.  See also Clean (which was designed to make standalones):

    http://www.cs.kun.nl/~clean/

> Icon,

Not functional; difficult to make standalones; no native support for
objects; but unique in many respects and well worth learning anyway.

> and Common Lisp.

Huge.  Really huge.

> Which would you use?

Python.

> Why?

This is a Python newsgroup.  Stick around and you'll get a few thousand
answers to that every month without even asking <wink>.

python-scores-highest-on-"plays-well-with-others"-ly y'rs  - tim




From aleaxit at yahoo.com  Sun Feb 11 08:37:56 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 11 Feb 2001 08:37:56 +0100
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <965fhd02uv1@news2.newsguy.com>

"DvdAvins" <dvdavins at aol.com> wrote in message
news:20010210175216.26827.00000474 at ng-mh1.aol.com...
> I'm planning to write a program for myself which, if I'm happy with the
way it
> turns out, I will likely distribute as shareware. I'm looking for as many
of
> the following traits as possible:
>
> 1:a functional language. I've been intrigued by functional languages for a
> while, but have had no occasion to use them at work.
> 2: automatic garbage collection and other programmers' convenience
features (I
> still have to take care of my day job)
> 3: support for objects (not necessary, just a convenience)
> 4: it MUST be distributable as a stand-alone program for Windows. I'd like
to
> be able to distribute to Mac and Linux as well.
> 5. there has to a good book on the language available in English or at
least an
> IDE with and excdellent help environment.
>
> I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common
Lisp.
> Which would you use? Why?

I see you got several replies, but I don't think anybody mentioned that
none of the languages you list [save possibly M -- I never heard of it]
really meets your constraints [1] and [3] at once.  In Scheme, Lisp,
Python, you can do some functional programming, but they aren't
functional languages -- one normally does use assignment in each
of them.  And similarly, Haskell, the only real functional language out
of the ones you list, does not really support objects.

If I wanted to meet your constraints 1-5, I'd be torn between Haskell
(no objects, but...) and ML variants -- OCaml has objects, other ML
versions may have better books in English (but the O'Reilly OCaml
book should get translated by a volunteer project any time soon, and
the docs that come with OCaml as well as Cousineau's earlier book
"Functional Approach to Programming" do a pretty good job in the
meantime).  Haskell might be considered 'more' functional in some
respects (lazy by default, no 'ref' for assignment but rather the
elegant encapsulation of monads) and the English books about it
are nothing short of superb.

Mostly, given constraint number 1, I'd be asking on comp.lang.functional --
people who hang out there are likeliest to know what books are out, or
brewing, about various languages, as well as implementation details
such as whether a DLL needs to accompany a distributed program for
a certain compiler or a really standalone EXE may be built, etc.


Alex





From phrxy at csv.warwick.ac.uk  Mon Feb 12 00:23:24 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Sun, 11 Feb 2001 23:23:24 +0000
Subject: Is Python for me?
In-Reply-To: <965fhd02uv1@news2.newsguy.com>
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <965fhd02uv1@news2.newsguy.com>
Message-ID: <Pine.SOL.4.30.0102112321160.16304-100000@mimosa.csv.warwick.ac.uk>

On Sun, 11 Feb 2001, Alex Martelli wrote:
[...about Haskell...]
> elegant encapsulation of monads) and the English books about it
> are nothing short of superb.
[...]

Which ones?  Or just all of them?


John



From kendall at monkeyfist.com  Tue Feb 20 11:49:46 2001
From: kendall at monkeyfist.com (Kendall G. Clark)
Date: 20 Feb 2001 10:49:46 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <965fhd02uv1@news2.newsguy.com>
Message-ID: <slrn994ieu.utn.kclark@cmpu.net>

On Sun, 11 Feb 2001 08:37:56 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:

>If I wanted to meet your constraints 1-5, I'd be torn between Haskell
>(no objects, but...) 

Well, there's always O'Haskell

	http://www.cs.chalmers.se/~nordland/ohaskell/

and Mondrian 

	http://www.mondrian-script.org/

(But beware the Mondrian syntax; my Python-as-first-language eyes were
nearly burned out at the site of the unholy mixture of Haskell and
Java syntaxes. Yeek!)

So lemme see if I've got this right:

1. There's something Alex Martelli doesn't know; and

2. the Timbot is (still) spreading about the hoary 'Common Lisp is
*really* huge' myth[1]?

Martelli omissions and a Timbot imprecision in one night's
news-reading session? c.l.py is slipping! :>

Best, 
Kendall Clark

[1] See the recent c.l.lisp thread wherein Kent Pitman (the Timbot of
c.l.l.  Or is our Timbot the Kentbot of c.l.py? -- those CL hackers
have all that tricky AI, doncha know!) said,

   Add to this the fact that Lisp has a Large Footprint(TM). That is,
   it takes up MANY megabytes.  I say "(TM)" to underscore the fact
   that there are some factoids (or false-oids) that get enshrined as
   lemmas and not re-evaluated with time.  Lisp WAS big back in the
   early 90's but it so panicked customers that Lisp images have
   hardly grown in size for the last decade.  During the same
   interval, every release of every other language has grown by leaps
   and bounds.  A Lisp image is now competitive with, and often tiny,
   compared to images of other languages.  The difference is often a
   distribution of .dll's that make it hard to see the size of the
   other things.  And even there, the commercial vendors have dll
   solutions that do similar tricks.  So this is just a myth
   perpetuated by people not going back to find out what the real
   truth is.

(I haven't sorted out the Google refurb of Deja well enough to have a
chance in hell of finding a proper URL. My SLRN newspool says
"Message-ID: <sfwg0hc26dy.fsf at world.std.com>". It's the thread called
"Re: Help me convince skeptics! Lisp success stories wanted...".)

-- 
THIS MACHINE KILLS FASCISTS		http://monkeyfist.com/


From aleaxit at yahoo.com  Tue Feb 20 13:37:40 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 13:37:40 +0100
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <965fhd02uv1@news2.newsguy.com> <slrn994ieu.utn.kclark@cmpu.net>
Message-ID: <96toeh012hd@news2.newsguy.com>

"Kendall G. Clark" <kendall at monkeyfist.com> wrote in message
news:slrn994ieu.utn.kclark at cmpu.net...
> On Sun, 11 Feb 2001 08:37:56 +0100, Alex Martelli <aleaxit at yahoo.com>
wrote:
>
> >If I wanted to meet your constraints 1-5, I'd be torn between Haskell
> >(no objects, but...)
>
> Well, there's always O'Haskell
> http://www.cs.chalmers.se/~nordland/ohaskell/
    [snip]
> 1. There's something Alex Martelli doesn't know; and

Yes (such as: what is the best C-coded ANOVA package for Python...
see my latest post!), but the (non-empty) set does not include
O'Haskell (nor, alas, Mondrian -- which is a hack of which
respondent sayeth no further).  O'Haskell looks cool, but I just
would not recommend it within the given constraints 1-5 (which
I invite you to review) -- it may mature into something *really*
great, but it's still at "rabid early-adopters" stage now (IMHO).


> 2. the Timbot is (still) spreading about the hoary 'Common Lisp is
> *really* huge' myth[1]?

Wasn't that in the context of a comparison with Scheme?  In which
case, it seems to me that it's no myth, but simple fact.  It's not
an issue of dimension of executable images, etc -- it's an issue
of how big the *language* (with its libraries) is.  CLisp is right
up there with C++, Ada, etc, in the "really huge languages" class
(Java is an interesting anomaly -- middling *language*, but the
'standard libraries' are *SO* deucedly MANY, that overall it may
be considered big again:-).

I don't particularly care if some clever compiler is able to turn:
    #include <iostream>
    int main() { std::cout<<"Hello world!\n"; }
into a "completely standalone" executable of 392 bytes or less --
that doesn't make *C++* any smaller.  (How many alleged C++ users
KNOW that the 'int' here is mandatory _but_ it's OK to leave the
"return 0;" out because it's IMPLIED if function ::main -- and
IT only -- "falls off the end"...?  just a tiny example of the
kind of things that make a _language_ "big" -- lots of "convenient"
ad-hoc little special-cases and such...:-).


Alex





From kclark at ntlug.org  Wed Feb 21 11:05:25 2001
From: kclark at ntlug.org (Kendall G. Clark)
Date: 21 Feb 2001 10:05:25 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <965fhd02uv1@news2.newsguy.com> <slrn994ieu.utn.kclark@cmpu.net> <96toeh012hd@news2.newsguy.com>
Message-ID: <slrn996jpl.3tg.kclark@cmpu.net>

On Tue, 20 Feb 2001 13:37:40 +0100, Alex Martelli <aleaxit at yahoo.com> wrote:

>Yes (such as: what is the best C-coded ANOVA package for Python...
>see my latest post!), but the (non-empty) set does not include
>O'Haskell (nor, alas, Mondrian -- which is a hack of which
>respondent sayeth no further).  O'Haskell looks cool, but I just
>would not recommend it within the given constraints 1-5 (which
>I invite you to review) -- it may mature into something *really*
>great, but it's still at "rabid early-adopters" stage now (IMHO).

Re: Mondrian, I didn't find it very compelling either, but if
someone is just really aching for objects and Haskell, it's at
least in the neighborhood. (I don't claim it satisfies the
5 criteria of the original poster.)

(Perhaps the only real claim to widespread interest Mondrian
has is that it seems one of the many languages, variants,
or alternative implementations springing up around MS's .NET
project. As an assiduous avoider of all things MS, and a booster
of all things open source or free software, I have some concerns
about .NET. But I'm also finding the kind of ancillary initiatives
it's either causing or funding to be of some interest when I put
on my 'social informatics' hat. Off the top of my head, I can think
of several languages that are working to accomodate .NET in some way:
Mondrian, Mercury, Smalltalk, Python, and so on. There may yet be
some interesting things of an analytic sort to be said about all
this. I just don't know what they are yet!)

Of course if one is *really* aching for objects and Haskell,
Ocaml may be the best practical choice (even given the differ-
ences between Haskell and the ML family).

Er, and it was also a decent chance to give a pointer to O'Haskell,
an effort that deserves to find some success.

>> 2. the Timbot is (still) spreading about the hoary 'Common Lisp is
>> *really* huge' myth[1]?

>Wasn't that in the context of a comparison with Scheme?  In which
>case, it seems to me that it's no myth, but simple fact.  It's not
>an issue of dimension of executable images, etc -- it's an issue
>of how big the *language* (with its libraries) is.  

Hmm, no, I didn't read it like that. But the original point was
just a throwaway line, I suspect, so it's not worth quibbling about,
at least not as matter of Timbotean exegesis. :>.

Your distinction between library and language conceptual size and 
the size of *all* the parts and pieces required to make working
programs in a given language is exactly right. CL isn't a large system
in the latter sense; but for many cases -- except perhaps all those
where Scheme *really* flourishes (like pedagogical uses and embed-and-
extend uses) -- library "largeness" is a very good thing. It's just
about the only good thing *I* can say about Perl, for example, though
Perl and CL libraries are rather different in the terrain they cover
(and cover best).

I did take Timbot to be making a claim about CL's largeness re:
development infrastructure, runtime requirements, etc. Maybe I was
just misreading him. At any rate, that *is* one of the hoary myths --
together with 'it's slow' and 'it's interpreted, so slow' and 'it's
just for AI' and 'it's just for academic wankers' -- that, together
with other factors, tends to hold CL back from the more widespread
commercial acceptance it rightly, in my view, deserves.

My *real* point was that one has to look far and wide, as I think I've
proven (!), to find stuff to complain about on c.l.python, which is
consistently excellent and often fun. ;>

Best,
Kendall Clark
-- 
In this world there are only two tragedies. One is not getting what one
wants, and the other is getting it. --Oscar Wilde


From tim_one at email.msn.com  Wed Feb 21 07:07:03 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Feb 2001 01:07:03 -0500
Subject: Is Python for me?
In-Reply-To: <slrn994ieu.utn.kclark@cmpu.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEODJAAA.tim_one@email.msn.com>

[Kendall G. Clark]
> ...
> So lemme see if I've got this right:
>
> 1. There's something Alex Martelli doesn't know; and

Na, that's an illusion.  Alex is working hard to keep his posts under 50Kb
these days, so he can't afford to divulge more than the tiniest fraction of
what he knows.

> 2. the Timbot is (still) spreading about the hoary 'Common Lisp is
> *really* huge' myth[1]?

In terms of language size, yes.  CL is a big language -- it's a *huge*
language.  The original questioner was attracted to functional languages,
and they tend to be on the minimal side (notwithstanding that in recent
years they've grudgingly allowed that some people need to do I/O <0.9
wink>).

> Martelli omissions and a Timbot imprecision in one night's
> news-reading session? c.l.py is slipping! :>

The question is whether that precedes or follows Python ballooning out of
control.  Newsgroups and languages reflect each other, but it's hard to say
which is more cause and which effect <wink>.

BTW, I suspect the biggest contributor to the myth that Lisp
*implementations* are huge may still be Emacs.  More than a decade ago that
stood for Eighty Megs And Constantly Swapping, and I doubt it's gone on a
diet since then.  Doesn't matter that Lisp has nothing to do with that; it's
like the equally bizarre myth that Python's treatment of whitespace was
taken from Fortran.

myths-are-too-comforting-to-fight-ly y'rs  - tim




From scarblac at pino.selwerd.nl  Sun Feb 11 13:19:49 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 11 Feb 2001 12:19:49 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <slrn98d0au.452.scarblac@pino.selwerd.nl>

DvdAvins <dvdavins at aol.com> wrote in comp.lang.python:
> I'm planning to write a program for myself which, if I'm happy with the way it
> turns out, I will likely distribute as shareware. I'm looking for as many of
> the following traits as possible:
> 
> 1:a functional language. I've been intrigued by functional languages for a
> while, but have had no occasion to use them at work.

Although Python has some features from functional languages, it is not
itself a functional language. It's more procedural/OO.

> 2: automatic garbage collection and other programmers' convenience features (I
> still have to take care of my day job)

Python is absolutely great at this.

> 3: support for objects (not necessary, just a convenience)

Check.

> 4: it MUST be distributable as a stand-alone program for Windows. I'd like to
> be able to distribute to Mac and Linux as well.

What counts as a stand-alone program? There are installers that will install
Python if it's not there yet and then your program; and there's a program
that will bundle your program and the interpreter into one big executable
that you can distribute. In the Mac case it'll probably be more work but I
don't know, for Linux you might assume a Python version to be around (but
then it'd be Python 1.5.2, I think).

I think most candidate languages have problems like this, several will have
worse problems.

> 5. there has to a good book on the language available in English or at least an
> IDE with and excdellent help environment.

There are IDEs, there is a good library reference (but I don't know how well
the IDEs integrate it), and there's this newsgroup and the Tutor list!
Excellent help environment includes a good newsgroup :).

> I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common Lisp.
> Which would you use? Why?

Common Lisp *should* be good at most of the things you list, but I don't
know. Python is very good too, then there's a gap.

Of course, you never said what you were going to do, you also need good
libraries available. Python is rather all-round :).

-- 
Remco Gerlich


From cjc26 at nospam.cornell.edu  Sun Feb 11 17:43:40 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Sun, 11 Feb 2001 16:43:40 GMT
Subject: Is Python for me?
References: <mailman.981870373.15031.python-list@python.org>
Message-ID: <0jzh6.36317$v.1980202@typhoon.nyroc.rr.com>

* Tim Peters <tim_one at email.msn.com> menulis:
| [DvdAvins]
| >
| > I'm investigating at M,
| 
| Don't know it.

It's a weird database language developed at a Boston-area hospital in
the late '60s.  It used to be called MUMPS; I guess they changed the
name because they felt silly saying "I program in MUMPS" ;)

I-program-with-snakes-and-gemstones-ly y'rs, Cliff


-- 
Cliff Crawford               http://www.people.cornell.edu/pages/cjc26/
                             print "Just another Python hacker"


From m.faassen at vet.uu.nl  Mon Feb 12 00:40:18 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 11 Feb 2001 23:40:18 GMT
Subject: Is Python for me?
References: <mailman.981870373.15031.python-list@python.org> <0jzh6.36317$v.1980202@typhoon.nyroc.rr.com>
Message-ID: <9677t2$brc$4@newshost.accu.uu.nl>

Cliff Crawford <cjc26 at nospam.cornell.edu> wrote:
> * Tim Peters <tim_one at email.msn.com> menulis:
> | [DvdAvins]
> | >
> | > I'm investigating at M,
> | 
> | Don't know it.

> It's a weird database language developed at a Boston-area hospital in
> the late '60s.  It used to be called MUMPS; I guess they changed the
> name because they felt silly saying "I program in MUMPS" ;)

Woah, if that's really the M he's talking about.. at my previous workplace
MUMP ('M systems') was still used. It's not really a modern language. 
Though it has a verbose mode, its default representation looks more like
line noise than Perl does, and it's not a functional language. It does
have some weird database system integrated with it; not really relational
but it comes closest to that.

I believe there are a few modern open source implementations of it
these days, but I'm not sure how well they work. Anyway, it's still
being used, though my ex-coworker is happily playing with Zope and Python
these days as well. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From garry at sage.att.com  Tue Feb 13 20:06:55 2001
From: garry at sage.att.com (Garry Hodgson)
Date: Tue, 13 Feb 2001 19:06:55 GMT
Subject: Is Python for me?
References: <mailman.981870373.15031.python-list@python.org>
Message-ID: <3A8985CF.33A67363@sage.att.com>

Tim Peters wrote:

> Lots of fun.  See also Clean (which was designed to make standalones):
> 
>     http://www.cs.kun.nl/~clean/

while it appears rather nice, it varies widely with respect to what
libraries
are supported on what platforms.  building something truly cross
platform
might be tricky, restricting one to a small subset.

-- 
Garry Hodgson                   Once in a while
Senior Hacker                   you can get shown the light
Software Innovation Services    in the strangest of places
AT&T Labs                       if you look at it right


From claird at starbase.neosoft.com  Tue Feb 13 22:30:17 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 13 Feb 2001 15:30:17 -0600
Subject: Is Python for me?
References: <mailman.981870373.15031.python-list@python.org>
Message-ID: <2A9F6176035467BF.087B976324B429F1.FD59F3E2FF0B9542@lp.airnews.net>

In article <mailman.981870373.15031.python-list at python.org>,
Tim Peters <tim_one at email.msn.com> wrote:
			.
			.
			.
>> I'd like to be able to distribute to Mac and Linux as well.
>
>This is at odds with desiring standalones, unless you want to build a
>different standalone for every platform.
			.
			.
			.
There *is* a device for coming as close as possible to dissolving that conflict:
<URL:http://starbase.neosoft.com/~claird/comp.software-eng/scripted_documents.html>
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From aleaxit at yahoo.com  Wed Feb 21 17:29:41 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 17:29:41 +0100
Subject: Is Python for me?
References: <mailman.982766607.21100.python-list@python.org>
Message-ID: <970qdj0t60@news1.newsguy.com>

"Tim Peters" <tim_one at email.msn.com> wrote in message
news:mailman.982766607.21100.python-list at python.org...
> [Kendall G. Clark]
> > ...
> > So lemme see if I've got this right:
> >
> > 1. There's something Alex Martelli doesn't know; and
>
> Na, that's an illusion.  Alex is working hard to keep his posts under 50Kb
> these days, so he can't afford to divulge more than the tiniest fraction
of
> what he knows.

Actually, my target-function is still the sum of squares of
my posts' sizes, just as it always was -- except that a little
programming bug was recently fixed (seems I must _minimize_
this target under certain constraints, rather than _maximize_
it -- don't you wish bot programmers tested more carefully...!).

But I still don't know exactly what C-coded-for-Python module
is best to do some logistic regression, ANOVA, and a few other
things (quite a few suggestions have come in but I still have
not had the time to try them...!).  So, there IS something I
don't know (at least until I do find the time to try out the
various suggestions I received).


> language.  The original questioner was attracted to functional languages,
> and they tend to be on the minimal side (notwithstanding that in recent

O'Caml appears to work rather hard to buck that tendency, IMHO -- still
not as big a language as C++, CLisp, or Ada, but to me it seems on that
route (which is part of why O'Haskell IS so interesting -- it seems to
be managing to be adding a LOT of deep functionality *without* all that
many "convenient" [!] doodads -- but I have to admit my opinion is based
strictly on reading _about_ it... haven't coded so much as a helloworld
in O'Haskell, yet).


Alex





From barron at bitment.com  Sun Feb 11 01:14:36 2001
From: barron at bitment.com (Barron)
Date: Sat, 10 Feb 2001 18:14:36 -0600
Subject: Fonts in Pmw.ScrolledListBox?
Message-ID: <B6AB358C.D8C%barron@bitment.com>

How does one set the font in a Pmw.ScrolledListBox. The "font" option
doesn't  seem to work like it does in a regular list box and I didn't see
anything about fonts in the Megawidget documentation for the
ScrolledListBox.

Thanks,
Barron



From matt at mondoinfo.com  Sun Feb 11 01:29:46 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Sun, 11 Feb 2001 00:29:46 GMT
Subject: Fonts in Pmw.ScrolledListBox?
References: <B6AB358C.D8C%barron@bitment.com>
Message-ID: <slrn98bn7p.qe.matt@happy-hour.mondoinfo.com>

On Sat, 10 Feb 2001 18:14:36 -0600, Barron <barron at bitment.com> wrote:

>How does one set the font in a Pmw.ScrolledListBox. The "font" option
>doesn't  seem to work like it does in a regular list box and I didn't see
>anything about fonts in the Megawidget documentation for the
>ScrolledListBox.

Barron,
You want something like:

mylistbox.component("listbox").configure(font="7x13")

Since Pmw commonly wraps up standard Tkinter widgets, configuring the
ordinary widget is usually just a matter of asking the megawidget for
it and doing the configuration normally.

Regards,
Matt


From martin.franklin at westerngeco.com  Mon Feb 12 09:17:27 2001
From: martin.franklin at westerngeco.com (Martin Franklin)
Date: Mon, 12 Feb 2001 08:17:27 +0000
Subject: Fonts in Pmw.ScrolledListBox?
Message-ID: <3A879C17.C46AE6F5@westerngeco.com>

I think you can also do it when you create the widget...(untested)

lb=Pmw.ScrolledListBox(root, listbox_font='7x15', listbox_height=8,
labelpos='n' ..... etc)

Martin

Matthew Dixon Cowles wrote:
>
> On Sat, 10 Feb 2001 18:14:36 -0600, Barron <barron at bitment.com> wrote:
>
> >How does one set the font in a Pmw.ScrolledListBox. The "font" option
> >doesn't  seem to work like it does in a regular list box and I didn't see
> >anything about fonts in the Megawidget documentation for the
> >ScrolledListBox.
>
> Barron,
> You want something like:
>
> mylistbox.component("listbox").configure(font="7x13")
>
> Since Pmw commonly wraps up standard Tkinter widgets, configuring the
> ordinary widget is usually just a matter of asking the megawidget for
> it and doing the configuration normally.
>
> Regards,
> Matt
> --
> http://mail.python.org/mailman/listinfo/python-list



From rpm1deleteme at frontiernet.net  Sun Feb 11 01:17:37 2001
From: rpm1deleteme at frontiernet.net (RPM1)
Date: Sat, 10 Feb 2001 19:17:37 -0500
Subject: Grab window contents in Python?
Message-ID: <964m55$oee$1@node17.cwnet.frontiernet.net>

Hi,
    Is there a way I can grab the contents of a text window of a commercial
program, using Python 2.0?  I've seen a VB program that could tell you the
window handle of any window you pointed at with the mouse.  I'm thinking
that if I could do something like that in Python 2.0, that would be a start.
Any thoughts?

    Thanks,
        Patrick M.

P.S. I don't want to use VB.





From gst at sysfrog.org  Sun Feb 11 01:26:59 2001
From: gst at sysfrog.org (Guenther Starnberger)
Date: 11 Feb 2001 01:26:59 +0100
Subject: nassi-shneiderman generator
Message-ID: <87bssa6p4c.fsf@nirvana.sysfrog.org>

i'm searching a nassi-shneiderman generator for python code (or some
other tool which generates nice diagrams, etc. stuff from my python
code (i need it for a school project)).

tnx
/gst

-- 
sysfrog.org


From dima at xenon.spb.ru  Sun Feb 11 01:40:02 2001
From: dima at xenon.spb.ru (Dmitry Rozmanov)
Date: Sun, 11 Feb 2001 03:40:02 +0300
Subject: test2
Message-ID: <964n12$6c7$1@news.wplus.spb.ru>

test. direct test.



From billy_bill_1 at my-deja.com  Sun Feb 11 10:28:36 2001
From: billy_bill_1 at my-deja.com (billy_bill_1 at my-deja.com)
Date: Sun, 11 Feb 2001 09:28:36 GMT
Subject: Using Python for Web development
Message-ID: <965m01$lr7$1@nnrp1.deja.com>

Hi all,

I'm new to Python and so far I like what I see!  I would really like to
use it for web development as I'm currently using PHP and ASP and I'm a
bit sick of thier lack of good OO support and the c style syntax etc
etc (especially php).  PHP can be used as an apache module, can python
be used in this way too?  I don't want to have to deal with any of the
header information stuff as well like in PERL cgi scripts.  Any links
or information anyone can give me to get me on my way?

Thanks,
Billy


Sent via Deja.com
http://www.deja.com/


From stephen_purcell at yahoo.com  Sun Feb 11 13:28:22 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sun, 11 Feb 2001 13:28:22 +0100
Subject: Using Python for Web development
In-Reply-To: <965m01$lr7$1@nnrp1.deja.com>; from billy_bill_1@my-deja.com on Sun, Feb 11, 2001 at 09:28:36AM +0000
References: <965m01$lr7$1@nnrp1.deja.com>
Message-ID: <20010211132822.B22574@freedom.puma-ag.com>

billy_bill_1 at my-deja.com wrote:
> I'm new to Python and so far I like what I see!

Then you are a wise and discerning soul.

> I would really like to
> use it for web development as I'm currently using PHP and ASP and I'm a
> bit sick of thier lack of good OO support and the c style syntax etc
> etc (especially php).  PHP can be used as an apache module, can python
> be used in this way too?  I don't want to have to deal with any of the
> header information stuff as well like in PERL cgi scripts.  Any links
> or information anyone can give me to get me on my way?

Some things that spring to mind:

 * Zope
      An entire web development and content management platform written in
      and customisable with Python.
      See http://www.zope.org/

 * 'cgi' module:
      Undoubtedly nicer than the Perl equivalent.
      See http://www.python.org/doc/lib/module-cgi.html

 * WebWare:
      Server-side utilities, framework and templating tools for Python
      See http://webware.sourceforge.net/

 * PSP (Python server pages)
      ASP equivalent for Python
      See http://www.ciobriefings.com/psp/

 * Quixote
      "Yet Another framework for developing Web applications in Python"
      See http://www.mems-exchange.org/software/python/quixote/


Cameron Laird has an excellent and more complete collection of alternatives
catalogued at

  http://starbase.neosoft.com/~claird/comp.lang.python/web_python.html

Hope that helps,

-Steve

-- 
Steve Purcell, Pythangelist
PythonTesting  = urlopen('http://pyunit.sourceforge.net/')
PythonServlets = urlopen('http://pyserv.sourceforge.net/')
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From doughellmann at bigfoot.com  Sun Feb 11 15:06:43 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Sun, 11 Feb 2001 14:06:43 GMT
Subject: Using Python for Web development
References: <965m01$lr7$1@nnrp1.deja.com>
Message-ID: <T%wh6.62614$m5.3421201@news1.rdc1.ga.home.com>

You can get mod_python from http://www.modpython.org.

You'll also want to check out http://www.zope.org, though, since Zope
does a lot of the hard work for you up front.

Doug

In article <965m01$lr7$1 at nnrp1.deja.com>, billy_bill_1 at my-deja.com wrote:

> Hi all,
> 
> I'm new to Python and so far I like what I see!  I would really like to
> use it for web development as I'm currently using PHP and ASP and I'm a
> bit sick of thier lack of good OO support and the c style syntax etc etc
> (especially php).  PHP can be used as an apache module, can python be
> used in this way too?  I don't want to have to deal with any of the
> header information stuff as well like in PERL cgi scripts.  Any links or
> information anyone can give me to get me on my way?
> 
> Thanks, Billy
> 
> 
> Sent via Deja.com http://www.deja.com/


From serdars at ixir.com  Sun Feb 11 18:04:21 2001
From: serdars at ixir.com (SHiBuMi)
Date: Sun, 11 Feb 2001 19:04:21 +0200
Subject: Using Python for Web development
In-Reply-To: <965m01$lr7$1@nnrp1.deja.com>
Message-ID: <BPEKKLEMEILKBMAKMBKOIEMLCKAA.serdars@ixir.com>

Yes, you can use Python as an apache module, see www.modpython.org and
search mod_snake project at www.sourceforge.net.

If you use CGI module, that will me some more familiar to you as you are
using PHP & ASP. But mod_python & mod_snake both have different approaches.
mod_snake has an advantage, you can embed your python code in HTML (as PHP &
ASP). But both of this modules needs much more customization on the web
server than PHP Apache module. And, with mod_python, you should deal with
headers...



-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of
billy_bill_1 at my-deja.com
Sent: Sunday, February 11, 2001 11:29 AM
To: python-list at python.org
Subject: Using Python for Web development


Hi all,

I'm new to Python and so far I like what I see!  I would really like to
use it for web development as I'm currently using PHP and ASP and I'm a
bit sick of thier lack of good OO support and the c style syntax etc
etc (especially php).  PHP can be used as an apache module, can python
be used in this way too?  I don't want to have to deal with any of the
header information stuff as well like in PERL cgi scripts.  Any links
or information anyone can give me to get me on my way?

Thanks,
Billy


Sent via Deja.com
http://www.deja.com/




From brian_zhou at agilentNOSPAM.com  Mon Feb 12 04:29:55 2001
From: brian_zhou at agilentNOSPAM.com (Brian Zhou)
Date: Sun, 11 Feb 2001 19:29:55 -0800
Subject: Using Python for Web development
References: <965m01$lr7$1@nnrp1.deja.com>
Message-ID: <981948593.79078@emperor.labs.agilent.com>

One more:

Java servlet in jython - It rocks!

/Brian

<billy_bill_1 at my-deja.com> wrote in message
news:965m01$lr7$1 at nnrp1.deja.com...
> Hi all,
>
> I'm new to Python and so far I like what I see!  I would really like to
> use it for web development as I'm currently using PHP and ASP and I'm a
> bit sick of thier lack of good OO support and the c style syntax etc
> etc (especially php).  PHP can be used as an apache module, can python
> be used in this way too?  I don't want to have to deal with any of the
> header information stuff as well like in PERL cgi scripts.  Any links
> or information anyone can give me to get me on my way?
>
> Thanks,
> Billy
>
>
> Sent via Deja.com
> http://www.deja.com/




From billy_bill_1 at my-deja.com  Mon Feb 12 05:23:37 2001
From: billy_bill_1 at my-deja.com (billy_bill_1 at my-deja.com)
Date: Mon, 12 Feb 2001 04:23:37 GMT
Subject: Using Python for Web development
References: <965m01$lr7$1@nnrp1.deja.com>
Message-ID: <967og3$7hl$1@nnrp1.deja.com>

Thanks heaps for all the links and help, I think I'll start with
mod_python.  One more thing I just thought of, how fast is python
compared to php, asp perl etc?  PHP is damn fast for a scripting
language, any benchmarks around?  I just need to see if its in the same
sort of league as php and asp.  I have to know this (for client and
boss) if I'm going to get use it in a commercial environment... here's
hoping!  Oh and that Servlet in Jpython, would that be like JSP and
servlets with no type checking crap?

Thanks!
Billy

In article <965m01$lr7$1 at nnrp1.deja.com>,
  billy_bill_1 at my-deja.com wrote:
> Hi all,
>
> I'm new to Python and so far I like what I see!  I would really like
to
> use it for web development as I'm currently using PHP and ASP and I'm
a
> bit sick of thier lack of good OO support and the c style syntax etc
> etc (especially php).  PHP can be used as an apache module, can python
> be used in this way too?  I don't want to have to deal with any of the
> header information stuff as well like in PERL cgi scripts.  Any links
> or information anyone can give me to get me on my way?
>
> Thanks,
> Billy
>
> Sent via Deja.com
> http://www.deja.com/
>


Sent via Deja.com
http://www.deja.com/


From timr at probo.com  Mon Feb 12 07:21:11 2001
From: timr at probo.com (Tim Roberts)
Date: Sun, 11 Feb 2001 22:21:11 -0800
Subject: Using Python for Web development
References: <965m01$lr7$1@nnrp1.deja.com> <967og3$7hl$1@nnrp1.deja.com>
Message-ID: <avve8ts7n40vc3bnl42osvs69dgi71ma37@4ax.com>

billy_bill_1 at my-deja.com wrote:

>Thanks heaps for all the links and help, I think I'll start with
>mod_python.

It might be easier for you to forget about an Apache module to start with,
and just do plain CGI scripts.  The python interpreter isn't all that big.

>One more thing I just thought of, how fast is python
>compared to php, asp perl etc?  PHP is damn fast for a scripting
>language, any benchmarks around?

Does it really matter?  CGI script execution time is negligible compared to
network overhead time.  The cost savings in development is significant
enough that I'll pay a performance penalty that cannot be measured, anyway.

If you program Python like you program Perl, you will find Python to be
slower.  For example, Perl programs often become a morass of regular
expression tweaks; Python doesn't do as well for that.  But once you figure
out the "Python" way to do things, the difference becomes slight.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From claird at starbase.neosoft.com  Mon Feb 12 15:41:38 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 12 Feb 2001 08:41:38 -0600
Subject: Performance considerations (was: Using Python for Web development)
References: <965m01$lr7$1@nnrp1.deja.com> <967og3$7hl$1@nnrp1.deja.com> <avve8ts7n40vc3bnl42osvs69dgi71ma37@4ax.com>
Message-ID: <73858481A1811258.15D110EC0C18A11D.E00F2C1B9EE10D58@lp.airnews.net>

In article <avve8ts7n40vc3bnl42osvs69dgi71ma37 at 4ax.com>,
Tim Roberts  <timr at probo.com> wrote:
>billy_bill_1 at my-deja.com wrote:
			.
			.
			.
>>One more thing I just thought of, how fast is python
>>compared to php, asp perl etc?  PHP is damn fast for a scripting
>>language, any benchmarks around?
>
>Does it really matter?  CGI script execution time is negligible compared to
>network overhead time.  The cost savings in development is significant
>enough that I'll pay a performance penalty that cannot be measured, anyway.
>
>If you program Python like you program Perl, you will find Python to be
>slower.  For example, Perl programs often become a morass of regular
>expression tweaks; Python doesn't do as well for that.  But once you figure
>out the "Python" way to do things, the difference becomes slight.
			.
			.
			.
This answer deserves reiteration.

Yes, there are benchmarks around, and occasionally I even
publish a few myself.  They are of EXTREMELY limited value,
though, value utterly swamped by the moral hazard that they
might fall into the hands of uneducated managers.  In 2001,
we're in the happy situation that Perl, PHP, Python, ...
execution times are all essentially equivalent, at least in
comparison with the challenges that fill most developers'
days.

Python's faster than PHP for many things, slower for others.
While it's right to ask about such differences, which have
the potential to matter, the answer for performance is, you
won't be able to tell 'em apart.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From brian_zhouNOSPAM at agilent.com  Mon Feb 12 18:45:36 2001
From: brian_zhouNOSPAM at agilent.com (Brian Zhou)
Date: Mon, 12 Feb 2001 09:45:36 -0800
Subject: Using Python for Web development
References: <965m01$lr7$1@nnrp1.deja.com> <967og3$7hl$1@nnrp1.deja.com>
Message-ID: <981999949.368015@emperor.labs.agilent.com>

<billy_bill_1 at my-deja.com> wrote in message
news:967og3$7hl$1 at nnrp1.deja.com...

> hoping!  Oh and that Servlet in Jpython, would that be like JSP and
> servlets with no type checking crap?
>

Here's a "Hello world!" servlet in jython:

########### ><8
import sys, java, javax

class hello(javax.servlet.http.HttpServlet):
    def doGet(self, request, response):
        response.setContentType("text/html")
        out = response.getOutputStream()

        out.println("""\
<html>
<head><title>Hello World</title></head>
<body>Hello World!
<P>
Current server time is: %s<br>
</P>
</body>
</html>
""" % (java.util.Date()))

        return
########### ><8

It's pretty much like mod_python, only using servlet API instead of CGI.
With java stuff like JDBC all available, it gets best of both worlds.
Personally I don't like embed HTML markup in python code, there're template
systems out there that you can use with mod_python or jython.

Regards,

/Brian




From fuess at att.net  Mon Feb 12 05:28:31 2001
From: fuess at att.net (David Fuess)
Date: Mon, 12 Feb 2001 04:28:31 GMT
Subject: Using Python for Web development
References: <965m01$lr7$1@nnrp1.deja.com> <mailman.981894494.8608.python-list@python.org>
Message-ID: <atoe8t4lgfje10e388n5a8gpolpid2oi54@4ax.com>

On Sun, 11 Feb 2001 13:28:22 +0100, Steve Purcell
<stephen_purcell at yahoo.com> wrote:

>billy_bill_1 at my-deja.com wrote:
>> I'm new to Python and so far I like what I see!
>
>Then you are a wise and discerning soul.
>
>> I would really like to
>> use it for web development as I'm currently using PHP and ASP and I'm a
>> bit sick of thier lack of good OO support and the c style syntax etc
>> etc (especially php).  PHP can be used as an apache module, can python
>> be used in this way too?  I don't want to have to deal with any of the
>> header information stuff as well like in PERL cgi scripts.  Any links
>> or information anyone can give me to get me on my way?
[snip]

I particularly like the Python web scripting that installs with
Win32Com extensions. Then, you can execute Python from within an ASP
file like this ...

<html>
<head>
<title>About Me</title>
</head>
<body>
<%@ Language=Python %>
<% url = Request.ServerVariables("PATH_INFO") %>
<h2>Current Document</h2>
<%
for key in Request.ServerVariables:
	print("%s=%s<br>"%(key,Request.ServerVariables(key))
%>
</body>
</html>




From jdukarm at hydracen.com  Mon Feb 12 20:33:57 2001
From: jdukarm at hydracen.com (J.J. Dukarm)
Date: Mon, 12 Feb 2001 11:33:57 -0800
Subject: Using Python for Web development
Message-ID: <3.0.1.32.20010212113357.00d60390@pop.islandnet.com>

There are some very nice things happening at webware.sourceforge.net.
Webware for Python is a toolkit for developing python-driven Web applications.

Jim Dukarm
DELTA-X RESEARCH
jdukarm at hydracen.com




From marco at crepido.com  Sun Feb 11 13:02:54 2001
From: marco at crepido.com (marco at crepido.com)
Date: Sun, 11 Feb 2001 13:02:54 +0100
Subject: PyQT problems...
Message-ID: <3A867F6E.FD5C7C3F@crepido.com>

Hi all,

I hope this is the right place to kindly ask for help. 

I have a problem with the Python bindings for QT. But I'm not
sure if it is a QT or Python problem...

When I run a pyQT program that uses signals and slots (about 99.99%
do...) 
Python complains about the signal being of the wrong type to match
the slot. 

I tried several PyQT programs and they all behave the same.

Example error message running Eric...

Traceback (most recent call last):
  File "/usr/local/lib/python2.1/site-packages/eric/eric.py", line 23,
in ?
    mw = UserInterface(dbs,prog)
  File "/usr/local/lib/python2.1/site-packages/eric/UserInterface.py",
line 53, in __init__
    self.connect(openAct,SIGNAL('activated()'),self.handleOpen)
RuntimeError: Signal has wrong argument types for slot


Have anyone seen this error and knows what to do about it??


My world:
- Python 2.1a2
- QT 2.2.4
- PyQT 2.4
- sip 2.3


TIA
/Marco


From dan at eevolved.com  Sun Feb 11 21:26:25 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 20:26:25 GMT
Subject: PyQT problems...
References: <3A867F6E.FD5C7C3F@crepido.com>
Message-ID: <RzCh6.101327$Pm2.1931128@news20.bellglobal.com>

marco at crepido.com wrote:

> Hi all,
> 
> I hope this is the right place to kindly ask for help.

Sure. I can direct you towards better help though

Here's docs on signals and slots
http://www.thekompany.com/projects/pykde/pyqt_docs/x125.html

Here's the mailing list (full of other PyQt developers that can answer your 
questions)
http://mats.gmd.de/mailman/listinfo/pykde

> Have anyone seen this error and knows what to do about it??

Wow. No I haven't. Very interesting though. I suggest joining the mailing 
list and asking this question there (not that you won't receive help here, 
it's just that the people with the knowledge are there)

peace,
Dan



From boud at rempt.xs4all.nl  Sun Feb 11 22:08:32 2001
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 11 Feb 2001 21:08:32 GMT
Subject: PyQT problems...
References: <3A867F6E.FD5C7C3F@crepido.com>
Message-ID: <966v0g$f20$1@news1.xs4all.nl>

marco at crepido.com wrote:
> Hi all,

> I hope this is the right place to kindly ask for help. 

> I have a problem with the Python bindings for QT. But I'm not
> sure if it is a QT or Python problem...

> When I run a pyQT program that uses signals and slots (about 99.99%
> do...) 
> Python complains about the signal being of the wrong type to match
> the slot. 

> I tried several PyQT programs and they all behave the same.

> Example error message running Eric...

> Traceback (most recent call last):
>   File "/usr/local/lib/python2.1/site-packages/eric/eric.py", line 23,
> in ?
>     mw = UserInterface(dbs,prog)
>   File "/usr/local/lib/python2.1/site-packages/eric/UserInterface.py",
> line 53, in __init__
>     self.connect(openAct,SIGNAL('activated()'),self.handleOpen)
> RuntimeError: Signal has wrong argument types for slot


> Have anyone seen this error and knows what to do about it??

Most likely self.handleOpen isn't defined as:

class ...
  def handleOpen(self)

but has zero or more than one arguments. There are certainly no
problems at all using signals and slots in PyQt: look at the example
scripts, or my work (http://www.valdyas.org/python), or my article
on the subject in January's Dr Dobbs.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From phrxy at csv.warwick.ac.uk  Mon Feb 12 00:15:29 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Sun, 11 Feb 2001 23:15:29 +0000
Subject: PyQT problems...
In-Reply-To: <3A867F6E.FD5C7C3F@crepido.com>
References: <3A867F6E.FD5C7C3F@crepido.com>
Message-ID: <Pine.SOL.4.30.0102112311170.16304-100000@mimosa.csv.warwick.ac.uk>

On Sun, 11 Feb 2001 marco at crepido.com wrote:

> I hope this is the right place to kindly ask for help.
[snipped PyQt question]

Yes, but there's a PyKDE mailing list too (covers both PyQt and PyKDE).

http://mats.gmd.de/mailman/listinfo/pykde


John



From Marco.Seirio at crepido_dot_.com  Mon Feb 12 14:22:56 2001
From: Marco.Seirio at crepido_dot_.com (Marco =?iso-8859-1?Q?Seiri=F6?=)
Date: Mon, 12 Feb 2001 14:22:56 +0100
Subject: PyQT problems... [2.1a2+PyQT broken?]
References: <3A867F6E.FD5C7C3F@crepido.com>
Message-ID: <968o3h$sjr@fullstop.carmen.se>

I tried on another machine with Python 2.0 and it works
ok!! Could it be that 2.1a2 is broken when used together 
with PyQT? (Or PyQT could be the one that is broken... well
they don't mix anyway...)

Anybody involved in PyQT that cares to comment on this one?

/Marco


marco at crepido.com wrote:
> 
> Hi all,
> 
> I hope this is the right place to kindly ask for help.
> 
> I have a problem with the Python bindings for QT. But I'm not
> sure if it is a QT or Python problem...
> 
> When I run a pyQT program that uses signals and slots (about 99.99%
> do...)
> Python complains about the signal being of the wrong type to match
> the slot.
> 
> I tried several PyQT programs and they all behave the same.
> 
> Example error message running Eric...
> 
> Traceback (most recent call last):
>   File "/usr/local/lib/python2.1/site-packages/eric/eric.py", line 23,
> in ?
>     mw = UserInterface(dbs,prog)
>   File "/usr/local/lib/python2.1/site-packages/eric/UserInterface.py",
> line 53, in __init__
>     self.connect(openAct,SIGNAL('activated()'),self.handleOpen)
> RuntimeError: Signal has wrong argument types for slot
> 
> Have anyone seen this error and knows what to do about it??
> 
> My world:
> - Python 2.1a2
> - QT 2.2.4
> - PyQT 2.4
> - sip 2.3
> 
> TIA
> /Marco


From ssthapa at cs.uchicago.edu  Sun Feb 11 15:06:37 2001
From: ssthapa at cs.uchicago.edu (ssthapa at cs.uchicago.edu)
Date: Sun, 11 Feb 2001 14:06:37 GMT
Subject: CPAN functionality for python
Message-ID: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>

  I've created a sort of proof of concept that implements something 
similar to perl's CPAN.  Right now, it allows basic user interaction 
and should be able to download and install modules.  
  At this point, I want to know whether I should continue on this. Also
and possibly more importantly, I would like to know if there is any place
to discuss formats for holding information about packages and other similar
information.  I have some ideas floating around but would like to get other
opinons and it seems like the Catalog-SIG is inactive.  If there is enough
positive feedback, I'll open up a project on Sourceforge and put my
code up there.  Any comments, suggestions, or help on where to go from here 
would also be appreciated.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at alumni.uchicago.edu        | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From sholden at holdenweb.com  Sun Feb 11 18:29:43 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 11 Feb 2001 12:29:43 -0500
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <T1Ah6.3842$n4.29046@e420r-atl2.usenetserver.com>

<ssthapa at cs.uchicago.edu> wrote in message
news:slrn98c5fc.2ah.ssthapa at ntcs-ip45.uchicago.edu...
>   I've created a sort of proof of concept that implements something
> similar to perl's CPAN.  Right now, it allows basic user interaction
> and should be able to download and install modules.

Sounds like what catalog-sig should have been doing.

>   At this point, I want to know whether I should continue on this. Also
> and possibly more importantly, I would like to know if there is any place
> to discuss formats for holding information about packages and other
similar
> information.  I have some ideas floating around but would like to get
other
> opinons and it seems like the Catalog-SIG is inactive.  If there is enough
> positive feedback, I'll open up a project on Sourceforge and put my
> code up there.  Any comments, suggestions, or help on where to go from
here
> would also be appreciated.
>
>
I think the most important thing is maintaining dependencies, which becomes
trickier the more you change things. Assuming nobody ever breaks with
backward compaitibility, then at least a module should be able to specifiy
the minimum facility-level it requires in other modules. But then I don't
want to change Python into Perl!

Just my 2c-worth.

Steve



From ssthapa at classes.cs.uchicago.edu  Mon Feb 12 20:15:17 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Mon, 12 Feb 2001 19:15:17 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <T1Ah6.3842$n4.29046@e420r-atl2.usenetserver.com>
Message-ID: <slrn98g6r6.3eo.ssthapa@ntcs-ip45.uchicago.edu>

Steve Holden <sholden at holdenweb.com> wrote:
>I think the most important thing is maintaining dependencies, which becomes
>trickier the more you change things. Assuming nobody ever breaks with
>backward compaitibility, then at least a module should be able to specifiy
>the minimum facility-level it requires in other modules. But then I don't
>want to change Python into Perl!

  I agree on this point, the format I had in mind is something a list for
each package formated something like this 
  [name, version, size, author, note, [[depend1, version], ...]]]
the size, author, and note would be optional, although the size and author
fields would be strongly encouraged.  With this format, the user could be
told what packages would need to be downloaded and the size of the total 
download  before starting the download.  CPAN has dependency information but
this is done in the course of installing the packages so users don't know
how long the install/downloads will take and how far along they are in 
the process.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From aahz at panix.com  Mon Feb 12 22:36:37 2001
From: aahz at panix.com (Aahz Maruch)
Date: 12 Feb 2001 13:36:37 -0800
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <T1Ah6.3842$n4.29046@e420r-atl2.usenetserver.com> <slrn98g6r6.3eo.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <969l15$8cm$1@panix6.panix.com>

In article <slrn98g6r6.3eo.ssthapa at ntcs-ip45.uchicago.edu>,
 <s-thapa-11 at NOSPAM.alumni.uchicago.edu> wrote:
>Steve Holden <sholden at holdenweb.com> wrote:
>
>  I agree on this point, the format I had in mind is something a list for
>each package formated something like this 
>  [name, version, size, author, note, [[depend1, version], ...]]]

I suggest:

    ... [[depend1, minVersion, maxVersion], ...

with None a valid value for minVersion and maxVersion.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

'Gender' isn't a definition, it's a heuristic.  --Aahz


From ssthapa at classes.cs.uchicago.edu  Tue Feb 13 04:39:34 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Tue, 13 Feb 2001 03:39:34 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <T1Ah6.3842$n4.29046@e420r-atl2.usenetserver.com> <slrn98g6r6.3eo.ssthapa@ntcs-ip45.uchicago.edu> <969l15$8cm$1@panix6.panix.com>
Message-ID: <slrn98h9sj.523.ssthapa@ntcs-ip45.uchicago.edu>

Aahz Maruch <aahz at panix.com> wrote:
>I suggest:
>
>    ... [[depend1, minVersion, maxVersion], ...
>
>with None a valid value for minVersion and maxVersion.

  That sounds like a great idea.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From dan at eevolved.com  Sun Feb 11 21:28:28 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 20:28:28 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <MBCh6.101338$Pm2.1931128@news20.bellglobal.com>

ssthapa at cs.uchicago.edu wrote:

> If there is enough
> positive feedback, I'll open up a project on Sourceforge and put my
> code up there.  Any comments, suggestions, or help on where to go from
> here would also be appreciated.

POSITIVE FEEDBACK!!
POSITIVE FEEDBACK!!
POSITIVE FEEDBACK!!

is that enough? ;)

I'd be glad to pitch in as well!

peace,
Dan


From doughellmann at bigfoot.com  Sun Feb 11 23:01:58 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Sun, 11 Feb 2001 22:01:58 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <qZDh6.62857$m5.3456640@news1.rdc1.ga.home.com>

Yea!!  

It would be interesting to see how to integrate this type of
functionality with the distutils project.

Doug

In article <slrn98c5fc.2ah.ssthapa at ntcs-ip45.uchicago.edu>,
ssthapa at cs.uchicago.edu  wrote:

>   I've created a sort of proof of concept that implements something 
> similar to perl's CPAN.  Right now, it allows basic user interaction 
> and should be able to download and install modules.  
>   At this point, I want to know whether I should continue on this. Also
> and possibly more importantly, I would like to know if there is any
> place to discuss formats for holding information about packages and
> other similar information.  I have some ideas floating around but would
> like to get other opinons and it seems like the Catalog-SIG is inactive.
>  If there is enough positive feedback, I'll open up a project on
> Sourceforge and put my code up there.  Any comments, suggestions, or
> help on where to go from here  would also be appreciated.
> 
>


From ssthapa at classes.cs.uchicago.edu  Mon Feb 12 20:15:49 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Mon, 12 Feb 2001 19:15:49 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <qZDh6.62857$m5.3456640@news1.rdc1.ga.home.com>
Message-ID: <slrn98g740.3eo.ssthapa@ntcs-ip45.uchicago.edu>

Doug Hellmann <doughellmann at bigfoot.com> wrote:
>Yea!!  
>
>It would be interesting to see how to integrate this type of
>functionality with the distutils project.

  Well, the code I have depends on distutils to build and install packages.
I actually see this as being somewhat independent of distutils since distutils
doesn't need it.  I guess the relationship is more like that between debian's
dpkg and dselect.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From amk at mira.erols.com  Mon Feb 12 03:13:58 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 12 Feb 2001 02:13:58 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <slrn98eii0.h5.amk@mira.erols.com>

On Sun, 11 Feb 2001 14:06:37 GMT, 
	ssthapa at cs.uchicago.edu <ssthapa at cs.uchicago.edu> wrote:
>code up there.  Any comments, suggestions, or help on where to go from here 
>would also be appreciated.

How about getting a Starship account?  Then you can run CGI scripts to
allow searching and using your proposed index, as a way of getting
user feedback.

--amk



From ssthapa at classes.cs.uchicago.edu  Mon Feb 12 20:15:54 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Mon, 12 Feb 2001 19:15:54 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <slrn98eii0.h5.amk@mira.erols.com>
Message-ID: <slrn98g7ed.3eo.ssthapa@ntcs-ip45.uchicago.edu>

A.M. Kuchling <amk at mira.erols.com> wrote:
>On Sun, 11 Feb 2001 14:06:37 GMT, 
>	ssthapa at cs.uchicago.edu <ssthapa at cs.uchicago.edu> wrote:
>>code up there.  Any comments, suggestions, or help on where to go from here 
>>would also be appreciated.
>
>How about getting a Starship account?  Then you can run CGI scripts to
>allow searching and using your proposed index, as a way of getting
>user feedback.

  I'll set up an account on starship early this week.  The biggest problems
I see are getting authors to add their modules to the package index and 
getting sites to host the archive.  But before that I would like to work out
a format for package entries so that it wouldn't need to be changed later on.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From akuchlin at mems-exchange.org  Mon Feb 12 23:53:12 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 12 Feb 2001 17:53:12 -0500
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <slrn98eii0.h5.amk@mira.erols.com> <slrn98g7ed.3eo.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <3d4rxz8qef.fsf@ute.cnri.reston.va.us>

ssthapa at classes.cs.uchicago.edu () writes:
>   I'll set up an account on starship early this week.  The biggest problems
> I see are getting authors to add their modules to the package index and 
> getting sites to host the archive.  But before that I would like to work out
> a format for package entries so that it wouldn't need to be changed later on.

I posted a list of fields
(http://mail.python.org/pipermail/catalog-sig/2000-November/000016.html)
that no one had problems with.  

--amk



From doughellmann at bigfoot.com  Tue Feb 13 12:50:05 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Tue, 13 Feb 2001 11:50:05 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <slrn98eii0.h5.amk@mira.erols.com>
Message-ID: <Nb9i6.64779$m5.3618739@news1.rdc1.ga.home.com>

In article <slrn98eii0.h5.amk at mira.erols.com>, amk at mira.erols.com  wrote:

> On Sun, 11 Feb 2001 14:06:37 GMT, 
> 	ssthapa at cs.uchicago.edu <ssthapa at cs.uchicago.edu> wrote:
>>code up there.  Any comments, suggestions, or help on where to go from
>>here  would also be appreciated.
> 
> How about getting a Starship account?  Then you can run CGI scripts to
> allow searching and using your proposed index, as a way of getting user
> feedback.

Why use standard CGI? This is a perfect application for Zope, the ZODB,
and the Catalog. There could be a standard browsable interface which could
be used with a web browser, and the special-purpose client could use
XML-RPC to communicate with the server.

I would be willing to help implement a Zope-based solution, if
experienced Zope developers are needed.

Doug


From phd at phd.pp.ru  Tue Feb 13 13:14:33 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 13 Feb 2001 15:14:33 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <Nb9i6.64779$m5.3618739@news1.rdc1.ga.home.com>
Message-ID: <Pine.LNX.4.30.0102131513460.348-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Doug Hellmann wrote:
> Why use standard CGI? This is a perfect application for Zope, the ZODB,

   The big advatnage of CPAn is that it's distributed and has many mirrors.
Zope-based solution will be hard to mirror.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From dsh8290 at rit.edu  Mon Feb 12 17:02:04 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 12 Feb 2001 11:02:04 -0500
Subject: CPAN functionality for python
In-Reply-To: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>; from ssthapa@cs.uchicago.edu on Sun, Feb 11, 2001 at 02:06:37PM +0000
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <20010212110204.A4625@harmony.cs.rit.edu>

On Sun, Feb 11, 2001 at 02:06:37PM +0000, ssthapa at cs.uchicago.edu wrote:
|   I've created a sort of proof of concept that implements something 
| similar to perl's CPAN.  Right now, it allows basic user interaction 
| and should be able to download and install modules.  
|   At this point, I want to know whether I should continue on this. Also
| and possibly more importantly, I would like to know if there is any place
| to discuss formats for holding information about packages and other similar
| information.  I have some ideas floating around but would like to get other
| opinons and it seems like the Catalog-SIG is inactive.  If there is enough
| positive feedback, I'll open up a project on Sourceforge and put my
| code up there.  Any comments, suggestions, or help on where to go from here 
| would also be appreciated.
| 

Debian has a great system for package management and dependency
resolving.  Maybe we should port dpkg/apt to all other OSes? ;-)
Certainly installing (python) modules on a Debian system isn't too
bad.  (I've never used CPAN so I don't know how it compares)

-D




From jafo at tummy.com  Mon Feb 12 19:51:18 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 12 Feb 2001 11:51:18 -0700
Subject: CPAN functionality for python
In-Reply-To: <20010212110204.A4625@harmony.cs.rit.edu>; from dsh8290@rit.edu on Mon, Feb 12, 2001 at 11:02:04AM -0500
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu>
Message-ID: <20010212115117.A7546@tummy.com>

On Mon, Feb 12, 2001 at 11:02:04AM -0500, D-Man wrote:
>On Sun, Feb 11, 2001 at 02:06:37PM +0000, ssthapa at cs.uchicago.edu wrote:
>|   I've created a sort of proof of concept that implements something 
>| similar to perl's CPAN.  Right now, it allows basic user interaction 
>| and should be able to download and install modules.  
>
>Debian has a great system for package management and dependency
>resolving.  Maybe we should port dpkg/apt to all other OSes? ;-)

The first question I would ask is: is there any way we can leverage
the work done by the Trove project?  I really don't know the status
of that project though.  Eric?  Any way to leverage off the work done
by CPAN?  Trying to leverage off existing art is a good idea, but I
don't think that dpkg is the right sort of tool for it.  If the
system could integrate with dpkg, rpm, swinstall, great!

The other thing I'd say is exactly what I said to ESR about Trove:
it would be ideal if the system could deal with mirror functionality.
The central site(s) wouldn't necessarily have the content, but you'd
ask them and they'd tell you where you can get the content.

For example, you may ask about "filterinput", and it would report
that there are versions X and Y and Z.  Then you'd ask where to find
"filterinput" version X and it would give you a list of URLs that
have it (or the like).  Ideally, you'd then be able to send a light-weight
packet to the servers listed asking about their willingness to provide
that resource, but that could be optional.  It would be nice to be
able to easily prune out the mirrors that were already at capacity,
no longer had that file or that version, etc...

ESR dismissed it because it wasn't really possible to do with a regular
web or ftp client.  However, if we're probably going to have our own
client anyway (a browser isn't likely to be able to install the software
once it's downloaded).

The problem of finding a mirror that's still got the file and isn't
overloaded is certainly not getting any better though.

It's probably for the best that I didn't implement those ideas back in
'97 when I first discussed them with ESR...  I'd be in court and Metallica
would hate me.  ;-)

Sean
-- 
 Windows NT: From the people who brought you 640K and EDLIN
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From dsh8290 at rit.edu  Mon Feb 12 21:15:45 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 12 Feb 2001 15:15:45 -0500
Subject: CPAN functionality for python
In-Reply-To: <20010212115117.A7546@tummy.com>; from jafo@tummy.com on Mon, Feb 12, 2001 at 11:51:18AM -0700
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu> <"from dsh8290"@rit.edu> <20010212115117.A7546@tummy.com>
Message-ID: <20010212151545.D5667@harmony.cs.rit.edu>

On Mon, Feb 12, 2001 at 11:51:18AM -0700, Sean Reifschneider wrote:
| On Mon, Feb 12, 2001 at 11:02:04AM -0500, D-Man wrote:
| >On Sun, Feb 11, 2001 at 02:06:37PM +0000, ssthapa at cs.uchicago.edu wrote:
| >|   I've created a sort of proof of concept that implements something 
| >| similar to perl's CPAN.  Right now, it allows basic user interaction 
| >| and should be able to download and install modules.  
| >
| >Debian has a great system for package management and dependency
| >resolving.  Maybe we should port dpkg/apt to all other OSes? ;-)
| 
| The first question I would ask is: is there any way we can leverage
| the work done by the Trove project?  I really don't know the status
| of that project though.  Eric?  Any way to leverage off the work done

I'm not familiar with Trove.  What is it and where is it used?

| by CPAN?  Trying to leverage off existing art is a good idea, but I
| don't think that dpkg is the right sort of tool for it.  If the

I don't know.  It would be cool to have dpkg and apt-get for Windo~1
though (Or at least rpm).  I was over at my friend's house yesterday
trying to install drivers for an ethernet card on his Win98 system.
It was a mess!  To finally get the install to work, we had to manually
remove the old (from earlier failed installs) driver files from the
C:\Windows\System directory.  My friend is computer literate, and
familiar with RH Linux as well.  We were both wishing we could locate
files (such as with a rpm/dpkg query) and not reboot several dozen
times!

Probably just wishful thinking ...

| system could integrate with dpkg, rpm, swinstall, great!
| 
| The other thing I'd say is exactly what I said to ESR about Trove:
| it would be ideal if the system could deal with mirror functionality.
| The central site(s) wouldn't necessarily have the content, but you'd
| ask them and they'd tell you where you can get the content.
| 
| For example, you may ask about "filterinput", and it would report
| that there are versions X and Y and Z.  Then you'd ask where to find
| "filterinput" version X and it would give you a list of URLs that
| have it (or the like).  Ideally, you'd then be able to send a light-weight
| packet to the servers listed asking about their willingness to provide
| that resource, but that could be optional.  It would be nice to be
| able to easily prune out the mirrors that were already at capacity,
| no longer had that file or that version, etc...
| 

That would be a cool feature.

| ESR dismissed it because it wasn't really possible to do with a regular
| web or ftp client.  However, if we're probably going to have our own
| client anyway (a browser isn't likely to be able to install the software
| once it's downloaded).

Something like dselect or apt right?  I don't have a problem with
that.

-D




From esr at thyrsus.com  Tue Feb 13 07:59:24 2001
From: esr at thyrsus.com (Eric S. Raymond)
Date: Tue, 13 Feb 2001 01:59:24 -0500
Subject: CPAN functionality for python
In-Reply-To: <20010212115117.A7546@tummy.com>; from jafo@tummy.com on Mon, Feb 12, 2001 at 11:51:18AM -0700
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu> <20010212115117.A7546@tummy.com>
Message-ID: <20010213015924.Q1764@thyrsus.com>

Sean Reifschneider <jafo at tummy.com>:
> The first question I would ask is: is there any way we can leverage
> the work done by the Trove project?  I really don't know the status
> of that project though.  Eric?

Trove is dead.  Long live SourceForge.

The SourceForge guys picked up many of Trove's core ideas.  SourceForge II,
which we're doing a planning meeting for on Wedneday, will pick up more
of them.
-- 
		<a href="http://www.tuxedo.org/~esr/">Eric S. Raymond</a>

Morality is always the product of terror; its chains and
strait-waistcoats are fashioned by those who dare not trust others,
because they dare not trust themselves, to walk in liberty.
	-- Aldous Huxley 



From jafo at tummy.com  Tue Feb 13 18:27:31 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 13 Feb 2001 10:27:31 -0700
Subject: CPAN functionality for python
In-Reply-To: <20010213015924.Q1764@thyrsus.com>; from esr@thyrsus.com on Tue, Feb 13, 2001 at 01:59:24AM -0500
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu> <20010212115117.A7546@tummy.com> <20010213015924.Q1764@thyrsus.com>
Message-ID: <20010213102731.B24074@tummy.com>

On Tue, Feb 13, 2001 at 01:59:24AM -0500, Eric S. Raymond wrote:
>Trove is dead.  Long live SourceForge.
>
>The SourceForge guys picked up many of Trove's core ideas.  SourceForge II,
>which we're doing a planning meeting for on Wedneday, will pick up more
>of them.

No lessons that can be learned from it?

Sean
-- 
 The "PEANUTS" gang finds their first root-kit in "YOU'RE AN 3L33T H4CK3R
 NOW, CHARLIE BROWN".
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From esr at thyrsus.com  Tue Feb 13 19:41:47 2001
From: esr at thyrsus.com (Eric S. Raymond)
Date: Tue, 13 Feb 2001 13:41:47 -0500
Subject: CPAN functionality for python
In-Reply-To: <20010213102731.B24074@tummy.com>; from jafo@tummy.com on Tue, Feb 13, 2001 at 10:27:31AM -0700
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu> <20010212115117.A7546@tummy.com> <20010213015924.Q1764@thyrsus.com> <20010213102731.B24074@tummy.com>
Message-ID: <20010213134147.B15155@thyrsus.com>

Sean Reifschneider <jafo at tummy.com>:
> On Tue, Feb 13, 2001 at 01:59:24AM -0500, Eric S. Raymond wrote:
> >Trove is dead.  Long live SourceForge.
> >
> >The SourceForge guys picked up many of Trove's core ideas.  SourceForge II,
> >which we're doing a planning meeting for on Wedneday, will pick up more
> >of them.
> 
> No lessons that can be learned from it?

Oh, believe you me, I am carrying forward *lots* of lessons from Trove.
-- 
		<a href="http://www.tuxedo.org/~esr/">Eric S. Raymond</a>

The direct use of physical force is so poor a solution to the problem of
limited resources that it is commonly employed only by small children and
great nations.
	-- David Friedman



From ssthapa at classes.cs.uchicago.edu  Mon Feb 12 22:03:39 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Mon, 12 Feb 2001 21:03:39 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu> <mailman.982005089.16180.python-list@python.org>
Message-ID: <slrn98gkd6.40n.ssthapa@ntcs-ip45.uchicago.edu>

Sean Reifschneider <jafo at tummy.com> wrote:
>The first question I would ask is: is there any way we can leverage
>the work done by the Trove project?  I really don't know the status
>of that project though.  Eric?  Any way to leverage off the work done
>by CPAN?  Trying to leverage off existing art is a good idea, but I
>don't think that dpkg is the right sort of tool for it.  If the
>system could integrate with dpkg, rpm, swinstall, great!

  I'm not familiar with Trove, what is it? It would be difficult 
to integrate this system with dkpg, or rpm.  At the very least, it 
would involve creating dummy packages.  
  I've taken some of CPAN's ideas as inspiration while writing my code
but I'm not sure how much more help CPAN can be outside as being an example
of a similar system.

>For example, you may ask about "filterinput", and it would report
>that there are versions X and Y and Z.  Then you'd ask where to find
>"filterinput" version X and it would give you a list of URLs that
>have it (or the like).  Ideally, you'd then be able to send a light-weight
>packet to the servers listed asking about their willingness to provide
>that resource, but that could be optional.  It would be nice to be
>able to easily prune out the mirrors that were already at capacity,
>no longer had that file or that version, etc...
  
  Parts of this are already implemented, e.g. the config file lists 
a group of mirrors and the script will try to grab the file from one and will
move on to the next if its connection is refused.  I guess an extension to 
this would be randomize which server is choosen or to allow the mirrors
to have preferences associated with them.  However, at the moment the 
program uses ftp to grab the files but a new module implementing a different
method wouldn't be hard to add.

>ESR dismissed it because it wasn't really possible to do with a regular
>web or ftp client.  However, if we're probably going to have our own
>client anyway (a browser isn't likely to be able to install the software
>once it's downloaded).
>
>The problem of finding a mirror that's still got the file and isn't
>overloaded is certainly not getting any better though.

  With a conventional ftp client,  this is sort of doable.  If you let
administrators place a dummy file in the archive that indicates that this
mirror is overloaded, then you can get a crude way of skipping loaded servers.

-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From jafo-pythonlist at tummy.com  Fri Feb 16 00:38:25 2001
From: jafo-pythonlist at tummy.com (Sean Reifschneider)
Date: Thu, 15 Feb 2001 16:38:25 -0700
Subject: CPAN functionality for python
In-Reply-To: <slrn98gkd6.40n.ssthapa@ntcs-ip45.uchicago.edu>; from ssthapa@classes.cs.uchicago.edu on Mon, Feb 12, 2001 at 09:03:39PM +0000
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu> <mailman.982005089.16180.python-list@python.org> <slrn98gkd6.40n.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <20010215163825.H23577@tummy.com>

On Mon, Feb 12, 2001 at 09:03:39PM +0000, ssthapa at classes.cs.uchicago.edu wrote:
>  I'm not familiar with Trove, what is it? It would be difficult 

Seems based on Eric's response that you need to use "was" instead of "is".
It was meant to be a large software catalog, originally meant to replace
the mirrors of Sunsite.

>to integrate this system with dkpg, or rpm.  At the very least, it 

Being a catalog means that it doesn't have to interfere with dpkg or
rpm.

>  I've taken some of CPAN's ideas as inspiration while writing my code
>but I'm not sure how much more help CPAN can be outside as being an example
>of a similar system.

The architecture I see is (simplified):

	A package has one or more versions.
	A version has one or more flavors.
	A flavor has one or more locations.

A package identifies the name and other attributes of the software.  The
version identifies the revisions.  The flavor would identify:

	Packaging format (tar, SRPM, dpkg, Distutils)
	Architecture (any, i386, alpha)
	Platform (any, RedHat, Debian, Windows)
	PlatformVersion (any, 7.0, 2.2, 3.11)

So that way a user can set preferences:

	I'd prefer to get an SRPM to have built on my box.  If there isn't one of
	those, snag a binary RPM for RedHat 7.0.  Ok, then I'd try installing from
	Distutils, and if all else fails give me the tar file and I'll try building
	from that.

It doesn't preclude dpkg, it embraces it.  IMHO, a requirement.

>  Parts of this are already implemented, e.g. the config file lists 
>a group of mirrors and the script will try to grab the file from one and will
>move on to the next if its connection is refused.  I guess an extension to 
>this would be randomize which server is choosen or to allow the mirrors
>to have preferences associated with them.  However, at the moment the 

Personally, I think that's more a mirror decision than something that should
be put in the catalog.  If I've got a list of 10 mirrors for a file, it would
be nice to send them a UDP packet asking if how willing they are to serve
a file.  Their willingness depends highly on the time of day and other
factors that you can't predict ahead of time (going to a host that has a
RedHat mirror is going to suck the day that they announce 7.1, eh).

Really, most of the mirror selection should probably be done by the client
software.  It'd be especially cool if the client could ask it's squid proxy
if it had any of the URLs cached...

>  With a conventional ftp client,  this is sort of doable.  If you let
>administrators place a dummy file in the archive that indicates that this
>mirror is overloaded, then you can get a crude way of skipping loaded servers.

Well, they usually do that by spitting out a file that says "I'm busy, go
away", and then you're off to search for another mirror.  Not very
effective, and rather heavy-weight.  Ideally, the client would send a UDP
packet to all the mirrors saying "Hey, I want this file", the mirrors
would receive it, maybe delay responding if they were loaded, drop it if
they were heavily loaded, and respond with a willingness level if they
have it and are up to sending it.  The first reply the client gets is
likely to be the best choice for trying to get a response from, especially
if servers make an effort to delay the response an appropriate period of
time based on the mirror load.

Sean
-- 
 "All I'm saying is that when I'm around you I find myself showing off,
 which is the idiots version of being interesting."  -- _LA_Story_
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From doughellmann at bigfoot.com  Tue Feb 13 13:49:35 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Tue, 13 Feb 2001 12:49:35 GMT
Subject: CPAN functionality for python
References: <mailman.982066514.11480.python-list@python.org>
Message-ID: <z3ai6.64782$m5.3620394@news1.rdc1.ga.home.com>

In article <mailman.982066514.11480.python-list at python.org>, "Oleg
Broytmann" <phd at phd.pp.ru> wrote:

> On Tue, 13 Feb 2001, Doug Hellmann wrote:
>> Why use standard CGI? This is a perfect application for Zope, the ZODB,
> 
>    The big advatnage of CPAn is that it's distributed and has many
>    mirrors.
> Zope-based solution will be hard to mirror.
> 

I don't follow that at all.  To support mirrors, the data needs to be
replicated from one server to another. That problem exists 
independant of any tool.

Doug


From phd at phd.pp.ru  Tue Feb 13 14:04:15 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 13 Feb 2001 16:04:15 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <z3ai6.64782$m5.3620394@news1.rdc1.ga.home.com>
Message-ID: <Pine.LNX.4.30.0102131603550.348-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Doug Hellmann wrote:
> >> Why use standard CGI? This is a perfect application for Zope, the ZODB,
> >
> >    The big advatnage of CPAn is that it's distributed and has many
> >    mirrors.
> > Zope-based solution will be hard to mirror.
> >
>
> I don't follow that at all.  To support mirrors, the data needs to be
> replicated from one server to another. That problem exists
> independant of any tool.

   What is "data" in this case?

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From m.faassen at vet.uu.nl  Tue Feb 13 16:21:30 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 13 Feb 2001 15:21:30 GMT
Subject: CPAN functionality for python
References: <mailman.982066514.11480.python-list@python.org> <z3ai6.64782$m5.3620394@news1.rdc1.ga.home.com>
Message-ID: <96bjdq$ico$1@newshost.accu.uu.nl>

Doug Hellmann <doughellmann at bigfoot.com> wrote:
> In article <mailman.982066514.11480.python-list at python.org>, "Oleg
> Broytmann" <phd at phd.pp.ru> wrote:

>> On Tue, 13 Feb 2001, Doug Hellmann wrote:
>>> Why use standard CGI? This is a perfect application for Zope, the ZODB,
>> 
>>    The big advatnage of CPAn is that it's distributed and has many
>>    mirrors.
>> Zope-based solution will be hard to mirror.
>> 

> I don't follow that at all.  To support mirrors, the data needs to be
> replicated from one server to another. That problem exists 
> independant of any tool.

Plus in the case of Zope, you may be able to leverage ZEO, though if
you're going to upload very large modules that may not be the best
idea; don't know how ZEO behaves with large objects across a possibly
flakey network.

Regards,

Martijn
--
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From jafo at tummy.com  Fri Feb 16 00:52:34 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Thu, 15 Feb 2001 16:52:34 -0700
Subject: CPAN functionality for python
In-Reply-To: <z3ai6.64782$m5.3620394@news1.rdc1.ga.home.com>; from doughellmann@bigfoot.com on Tue, Feb 13, 2001 at 12:49:35PM +0000
References: <mailman.982066514.11480.python-list@python.org> <z3ai6.64782$m5.3620394@news1.rdc1.ga.home.com>
Message-ID: <20010215165234.J23577@tummy.com>

On Tue, Feb 13, 2001 at 12:49:35PM +0000, Doug Hellmann wrote:
>I don't follow that at all.  To support mirrors, the data needs to be
>replicated from one server to another. That problem exists 
>independant of any tool.

I'd recommend that the catalog be exported to the primary mirror so that other
mirrors can pick it up using the normal FTP mirroring (possibly from a mirror
that doesn't have the cataloging software running on it.

What about dealing with partial mirrors?  It would be nice to be able to
take advantage of places that, say, only mirror the *LATEST* version of
the packages becuase of space constraints.  I see the catalog and the
actual file contents being two very distinct things, and that they don't
necessarily need to be placed on the same machine.

What about storing the md5sum of the file to help ensure that you've gotten
the whole file and it (hopefully) hasn't been tampered with?

Sean
-- 
 I'd rather write programs to write programs than write programs. 
                 -- Dick Sites
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From dsh8290 at rit.edu  Fri Feb 16 01:58:27 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 15 Feb 2001 19:58:27 -0500
Subject: CPAN functionality for python
In-Reply-To: <20010215165234.J23577@tummy.com>; from jafo@tummy.com on Thu, Feb 15, 2001 at 04:52:34PM -0700
References: <mailman.982066514.11480.python-list@python.org> <z3ai6.64782$m5.3620394@news1.rdc1.ga.home.com> <"from doughellmann"@bigfoot.com> <20010215165234.J23577@tummy.com>
Message-ID: <20010215195827.B2150@harmony.cs.rit.edu>

On Thu, Feb 15, 2001 at 04:52:34PM -0700, Sean Reifschneider wrote:
| I see the catalog and the actual file contents being two very
| distinct things, and that they don't necessarily need to be placed
| on the same machine.

I agree with this.


Something I sometimes wish for is a single site that has a list of all
the software goodies I want.  I can easily forget where I found a
certain neat proggy, and google, etc, usually give a lot of extraneous
results.  Sounds like this is what Trove tried to do, but it must not
have succeeded.  I guess I need to keep my own list ;-).

-D




From doughellmann at bigfoot.com  Tue Feb 13 14:17:59 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Tue, 13 Feb 2001 13:17:59 GMT
Subject: CPAN functionality for python
References: <mailman.982069514.23859.python-list@python.org>
Message-ID: <buai6.64783$m5.3620776@news1.rdc1.ga.home.com>

In article <mailman.982069514.23859.python-list at python.org>, "Oleg
Broytmann" <phd at phd.pp.ru> wrote:

> On Tue, 13 Feb 2001, Doug Hellmann wrote:
>> >> Why use standard CGI? This is a perfect application for Zope, the
>> >> ZODB,
>> >
>> >    The big advatnage of CPAn is that it's distributed and has many
>> >    mirrors.
>> > Zope-based solution will be hard to mirror.
>> >
>>
>> I don't follow that at all.  To support mirrors, the data needs to be
>> replicated from one server to another. That problem exists independant
>> of any tool.
> 
>    What is "data" in this case?

The server is going to have to keep up with what modules are available,
and all the meta data about those modules (version, dependencies,
description, location, etc.).

That information needs to be available on all mirror sites in order for
the mirror sites to be useful.  It is not enough to mirror some web
pages, the database needs to be mirrored to support the custom client
search mechanism.

Doug


From doughellmann at bigfoot.com  Tue Feb 13 14:21:36 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Tue, 13 Feb 2001 13:21:36 GMT
Subject: CPAN functionality for python
References: <mailman.982069514.23859.python-list@python.org> <buai6.64783$m5.3620776@news1.rdc1.ga.home.com>
Message-ID: <Axai6.64784$m5.3621015@news1.rdc1.ga.home.com>

In article <buai6.64783$m5.3620776 at news1.rdc1.ga.home.com>, "Doug
Hellmann" <doughellmann at bigfoot.com> wrote:

> The server is going to have to keep up with what modules are available,
> and all the meta data about those modules (version, dependencies,
> description, location, etc.).
> 
> That information needs to be available on all mirror sites in order for
> the mirror sites to be useful.  It is not enough to mirror some web
> pages, the database needs to be mirrored to support the custom client
> search mechanism.

Hmm.  Maybe that's not true.

Can we learn something from the P2P networks that are springing up all
over the place?  What if the server software knew how to talk other
servers running the same software as though they were clients. Dunno if
that would be more/less efficient than mirroring the data somehow.

Doug


From phd at phd.pp.ru  Tue Feb 13 15:29:15 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 13 Feb 2001 17:29:15 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <buai6.64783$m5.3620776@news1.rdc1.ga.home.com>
Message-ID: <Pine.LNX.4.30.0102131727520.2534-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Doug Hellmann wrote:
> The server is going to have to keep up with what modules are available,
> and all the meta data about those modules (version, dependencies,
> description, location, etc.).
>
> That information needs to be available on all mirror sites in order for
> the mirror sites to be useful.  It is not enough to mirror some web
> pages, the database needs to be mirrored to support the custom client
> search mechanism.

   This metadata can be put in static files. Thus, it can be easily
mirrored/reached by www/ftp/rsync/etc. In case of Zope, mirror site must
run Zope, so mirroring will be limited to probably 2-3 sites in the world.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From jeremy at alum.mit.edu  Tue Feb 13 05:46:59 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Mon, 12 Feb 2001 23:46:59 -0500 (EST)
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.30.0102131727520.2534-100000@fep132.fep.ru>
References: <buai6.64783$m5.3620776@news1.rdc1.ga.home.com>
	<Pine.LNX.4.30.0102131727520.2534-100000@fep132.fep.ru>
Message-ID: <14984.48195.455785.172253@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "OB" == Oleg Broytmann <phd at phd.pp.ru> writes:

  OB> On Tue, 13 Feb 2001, Doug Hellmann wrote:
  >> The server is going to have to keep up with what modules are
  >> available, and all the meta data about those modules (version,
  >> dependencies, description, location, etc.).
  >>
  >> That information needs to be available on all mirror sites in
  >> order for the mirror sites to be useful.  It is not enough to
  >> mirror some web pages, the database needs to be mirrored to
  >> support the custom client search mechanism.

  OB> This metadata can be put in static files. Thus, it can be easily
  OB> mirrored/reached by www/ftp/rsync/etc. In case of Zope, mirror
  OB> site must run Zope, so mirroring will be limited to probably 2-3
  OB> sites in the world.

I think this would be a fine project for Zope, in part because we're
going to be spinning up zope.python.org in the next few weeks and
could probably provide a home for it.  (Yeah, I work for Digital
Creations, too, but only because I think it's a great company.)

A couple of possibilities come to mind.  One is that the Zope service
has an export feature that produces a snapshot of the database in a
simple format that can be exchanged with other non-Zope servers.  I
imagine XML is well-suited for this kind of project.

Another possibility, though not ready right now, is to use ZEO to
mirror the database contents.  There are two possibilities, both of
which are being worked on:

    - Each mirror runs a ZEO client program and its own Zope server.
      The client will probably end up with the entire database in its
      cache and can serve clients from the cache most of the time.

    - Use a ZEO replicated storage where each replica runs a ZEO
      server that manages a copy of the data.  This provides greater
      fault tolerance, because the database is available for reads and
      writes as along as a majority of the servers are available.

Jeremy



From phd at phd.pp.ru  Tue Feb 13 17:34:38 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 13 Feb 2001 19:34:38 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <14984.48195.455785.172253@w221.z064000254.bwi-md.dsl.cnc.net>
Message-ID: <Pine.LNX.4.30.0102131928550.2534-100000@fep132.fep.ru>

On Mon, 12 Feb 2001, Jeremy Hylton wrote:
>   OB> This metadata can be put in static files. Thus, it can be easily
>   OB> mirrored/reached by www/ftp/rsync/etc. In case of Zope, mirror
>   OB> site must run Zope, so mirroring will be limited to probably 2-3
>   OB> sites in the world.
>
> I think this would be a fine project for Zope, in part because we're
> going to be spinning up zope.python.org in the next few weeks and
> could probably provide a home for it.  (Yeah, I work for Digital
> Creations, too, but only because I think it's a great company.)

   Ok, so we've found one of these 2 or 3 mirrors.

>     - Each mirror runs a ZEO client program and its own Zope server.

   What will be these other 1 or 2 mirror sites that will run Zope?

   I completeley agree that this task *can* be solved using Zope. We can
develop data formats, protocols, etc. And I completely disagree that the
task *should* be solved with Zope. It shouldn't.
    Please count current mirrors of CPAN. The figure looks good, isn't it?
This is because there are sites that provide FTP mirrors.
   Who will agree to run Zope only to mirror CPyAN?

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From jeremy at alum.mit.edu  Tue Feb 13 06:02:31 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Tue, 13 Feb 2001 00:02:31 -0500 (EST)
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.30.0102131928550.2534-100000@fep132.fep.ru>
References: <14984.48195.455785.172253@w221.z064000254.bwi-md.dsl.cnc.net>
	<Pine.LNX.4.30.0102131928550.2534-100000@fep132.fep.ru>
Message-ID: <14984.49127.853606.286789@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "OB" == Oleg Broytmann <phd at phd.pp.ru> writes:

  OB> I completeley agree that this task *can* be solved using Zope. We can
  OB> develop data formats, protocols, etc. And I completely disagree
  OB> that the task *should* be solved with Zope. It shouldn't.

I think you skipped my very first suggestion -- to use some simple
data format to support exchange between a Zope-based solution and
other solutions.

  OB> Please count current mirrors of CPAN. The figure looks good,
  OB> isn't it? This is because there are sites that provide FTP
  OB> mirrors. 

The CPAN mirrors provide access to the actual libraries, right?  That
is, you can download mumble.pm from any CPAN mirror.  I'm not clear on
whether the application being proposed for Python has only the catalog
information or also the actual data.  If the latter, then an FTP
mirror doesn't buy you much.

  OB> Who will agree to run Zope only to mirror CPyAN?

I don't know.  I presume we'd need to ask places that are already
running Zope.  There are at least a couple sites...

Jeremy



From phd at phd.pp.ru  Tue Feb 13 17:50:25 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 13 Feb 2001 19:50:25 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <14984.49127.853606.286789@w221.z064000254.bwi-md.dsl.cnc.net>
Message-ID: <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Jeremy Hylton wrote:
> I think you skipped my very first suggestion -- to use some simple
> data format to support exchange between a Zope-based solution and
> other solutions.

   Ok, so we should create formats and protocols and reference
implementations. Zope can be use for one of such reference implementations,
but we must have other implementations as well.

>   OB> Please count current mirrors of CPAN. The figure looks good,
>   OB> isn't it? This is because there are sites that provide FTP
>   OB> mirrors.
>
> The CPAN mirrors provide access to the actual libraries, right?  That
> is, you can download mumble.pm from any CPAN mirror.  I'm not clear on
> whether the application being proposed for Python has only the catalog
> information or also the actual data.  If the latter, then an FTP
> mirror doesn't buy you much.

   But I certainly think of real FTP mirrors, not only catalogs. I hate
perl, and avoid it as much as possible. But even for me there is one good
thing about perl - and it is exactly CPAN. The process of
   perl -MCPAN -e "install mod_perl;"
is simply the best! Python really need such thing. With all goodies about
local/nearest mirror, "make", "make test" and "make install"!

PS. When I first ran "perl -MCPAN" it started to configure itself, and it
asked lots of questions to find nearest mirror. The first question was
"Please select your nearest continent or a list of continetens" :)))
Tthanks goodnes, not "a list of nearest planets" or "galaxies" :)))

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From doughellmann at home.com  Wed Feb 14 01:31:47 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Tue, 13 Feb 2001 19:31:47 -0500
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru>
References: <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru>
Message-ID: <01021319355901.05138@branagan>

On Tue, 13 Feb 2001, Oleg Broytmann wrote:
> On Tue, 13 Feb 2001, Jeremy Hylton wrote:
> > I think you skipped my very first suggestion -- to use some simple
> > data format to support exchange between a Zope-based solution and
> > other solutions.
> 
>    Ok, so we should create formats and protocols and reference
> implementations. Zope can be use for one of such reference implementations,
> but we must have other implementations as well.

Sure, why not.  We could build the initial application very quickly using Zope,
though, so I suggest starting there.

We need a name for this project, and we need to start working out actual
requirements.

What meta data do we need to store for a module?  

What formats are going to be supported (tgz, rpm, zip, etc.)?  

Who hosts the modules?  Is it sufficient to provide a URL to the module when
registering the module, or does it have to be uploaded to a server
participating in the module-sharing-network?

Doug



From phd at phd.pp.ru  Wed Feb 14 10:33:21 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 14 Feb 2001 12:33:21 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <01021319355901.05138@branagan>
Message-ID: <Pine.LNX.4.30.0102141225570.8189-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Doug Hellmann wrote:
> What meta data do we need to store for a module?

   I don't care about metadata (others will do). I want:
configure/autoconf or Distutil-based "make"
test suite ("make test")
install script ("make install"); again, it's possible to do this with
   Distutil

> What formats are going to be supported (tgz, rpm, zip, etc.)?

   I vote for tgz.

> Who hosts the modules?  Is it sufficient to provide a URL to the module when
> registering the module, or does it have to be uploaded to a server
> participating in the module-sharing-network?

   Certainly, just an URL is not enough. How much you will be disappointed
if you were trying to install a module (and you are in need and in haste)
and the server is not responding? A list of mirrors is a must; and I hope
every our mirror soon will mirror all the modules, not only links to
author's homepages.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From jafo-pythonlist at tummy.com  Fri Feb 16 01:02:27 2001
From: jafo-pythonlist at tummy.com (Sean Reifschneider)
Date: Thu, 15 Feb 2001 17:02:27 -0700
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru>; from phd@phd.pp.ru on Tue, Feb 13, 2001 at 07:50:25PM +0300
References: <14984.49127.853606.286789@w221.z064000254.bwi-md.dsl.cnc.net> <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru>
Message-ID: <20010215170227.K23577@tummy.com>

On Tue, Feb 13, 2001 at 07:50:25PM +0300, Oleg Broytmann wrote:
>   Ok, so we should create formats and protocols and reference
>implementations. Zope can be use for one of such reference implementations,
>but we must have other implementations as well.

Ok, so from a client side we'll need the following commands available:

   Search:
      Takes:
         Module name
         Module Version*
         Platform*
         Platform Version*
         Architecture*
         Package Format*
      Returns:
         List of matching entities:
            Module name
            Module Version
            Platform
            Platform Version
            Architecture
            Package Format
   
   Find:
      Takes:
         Module name
         Module Version
         Platform
         Platform Version
         Architecture
         Package Format
      Returns:
         List of:
            URL

For the protocol, I'd recommend sockets using DJB's netstring encoding (which
I just happen to have a module for).  Netstrings allow embedded spaces and
nulls, and send the size of the messages before the messages themselves, so
that space can be allocated (preventing buffer overruns) and overly large
messages can be dropped right up front.  They can also be embedded within
other netstrings to represent lists or dictionaries:

   >>> print netstring.stons([ 'a', 'b', 'c' ])
   12:1:a,1:b,1:c,,
   >>> print netstring.dicttons({ 'a' : 1, 'b' : 2, 'c' : 'foo' })
   26:1:a,1:1,1:b,1:2,1:c,3:foo,,
   >>> print netstring.nstodict('26:1:a,1:1,1:b,1:2,1:c,3:foo,,')
   {'b': '2', 'c': 'foo', 'a': '1'}

You get the idea.  The module is available from
ftp://ftp.tummy.com/pub/tummy/Python/netstring/

The schema required for the database is pretty obvious, but the questions
become wether we want to try to track descriptions of the packages (and how
to search them), groupings of packages, etc.

Sean
-- 
 Actual error message:  "Keyboard not connected.  Press F1 to continue."
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From jafo at tummy.com  Mon Feb 26 09:39:03 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 01:39:03 -0700
Subject: CPAN functionality for python
In-Reply-To: <20010215170227.K23577@tummy.com>; from jafo-pythonlist@tummy.com on Thu, Feb 15, 2001 at 05:02:27PM -0700
References: <14984.49127.853606.286789@w221.z064000254.bwi-md.dsl.cnc.net> <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru> <20010215170227.K23577@tummy.com>
Message-ID: <20010226013903.B3221@tummy.com>

Unfortunately, this week has been pretty busy.  I've been giving the idea
some thought again (it's been something I've wanted to work on for quite
a while), and just finally today had the chance to look at pythonsiphon.

I'm not really clear (from the code what) as to what it's job is to be.
It smells like it's wanting to be something that understands, given a
distutils package, how to install it.

My thoughts on that are that it's a job that's probably going to be
something subject largely to being handed off to appropriate scripts
based on the platform and some user input (for example, a user may
prefer to have packages they download installed in ~/lib/python1.5,
instead of /usr/lib/python1.5).

Ideally though, the tool should be able to deal with allowing the user
to select their preferred distribution media.  I'd prefer to see an ix86
RedHat 7.0 RPM, a SRPM, and then would fall back to a distutils file or
tar file.

As far as the server side for allowing users to query packages,
dependencies, and locations...

I had actually hoped to pound out some code on this, but that just
didn't happen.  I have built up a schema which I think encompasses
everything I see such a system wanting to track.  Initial versions
probably won't track all that, but I believe they should be considered.

The schema is attached below.

A quick overview of the tables.  I hate the name "items", but it's
the best I could come up with.  That's what I'm calling the general
module (things that can be imported) or package (for example, programs
like Sketch -- why not have them in here as well) description.  In
this case, let's use "sockserv".  The URL would point to the main
location you can find "sockserv" module information.

"Packages" are actually meant to describe what you can download.
This database will track multiple versions, include a checksum
and information on what platform it's meant for (for binary
distributions, for example).

Mirrors are handled by a special table.  The idea is that one should
be able to add mirrors without having to list all the items in that
particular mirror (again).  New entries should be able to be listed
without having to list all the mirrors they might reside on.  So,
I've come up with the idea that a mirror simply specifies a "prefix"
for the URL listed in the "locations" table.

To be honest, my plan is to have a special-case "null" mirror which
non-mirror members are listed under -- that would have an empty
prefix.

So, given a package, version, format, architecture, etc...  The
"locations" table will produce list of all locations which that
file can be found at.

Note that there is a dependency table which lists that one "item"
relies on another "item".

As far as the network interface goes, at the current moment I'm thinking
that queries will be submitted to a web server CGI as a POST, and the
results will be returned as the body as text/plain...  I'll probably use
netstrings (as discussed before) to encode things for transport.

So, any comments?

Sean
=========================
#ISSUES:
#  This doesn't handle partial mirrors.  Should it?  How?

CREATE TABLE items (
   id serial primary key,
   lastUpdated datetime default 'now',

   name text,
   description text,

   type text,     #  module, package
   homepageURL text,
   )

CREATE TABLE packages (
   id serial primary key,
   itemID int4,
   version text,
   md5sum text,
	filesize text,
   insertedOn datetime default 'now',

   description text,       #  description about this specific file
   platformName text,
   platformVersion text,
   architecture text,
   packageFormat text,
   )

#  a collection of URLs that all contain the same set of files (within some
#  precision).  A mirror can be added without having to add all the URLs it
#  contains.  Not useful for handling partial-mirrors though.
CREATE TABLE mirrors (
   id serial primary key,
   name text,
   baseURL text,
   )

CREATE TABLE locations (
   id serial primary key,
   packageID int4,
   mirrorID int4,    #  if found on a mirror, list it
   URL text,            #  append this text to the mirror baseURL
   )

CREATE TABLE depends (
   id serial primary key,
   itemID int4,
   requiresItemID int4
   )

CREATE TABLE users (
   id serial primary key,
   name text,
   password text,
   lastLogin datetime default NULL,
   loginCookie text
   )

CREATE TABLE maintainers (
   userID int4,
   type text,        #  package, item
   piID int4
   )

CREATE TABLE ranking (
   id serial primary key,
   rank int2,
   itemID int4,
   userID int4
   date datetime default 'now',
   )
-- 
 Millions long for immortality who don't know what to do with themselves
 on a rainy Sunday afternoon.                -- Heinlein
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From bsass at freenet.edmonton.ab.ca  Mon Feb 26 10:56:51 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Mon, 26 Feb 2001 02:56:51 -0700 (MST)
Subject: CPAN functionality for python
In-Reply-To: <20010226013903.B3221@tummy.com>
Message-ID: <Pine.LNX.4.33.0102260150060.28098-100000@bms>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:
<...>
> My thoughts on that are that it's a job that's probably going to be
> something subject largely to being handed off to appropriate scripts
> based on the platform and some user input (for example, a user may
> prefer to have packages they download installed in ~/lib/python1.5,
> instead of /usr/lib/python1.5).
>
> Ideally though, the tool should be able to deal with allowing the user
> to select their preferred distribution media.  I'd prefer to see an ix86
> RedHat 7.0 RPM, a SRPM, and then would fall back to a distutils file or
> tar file.

Do you mean:

A tool that would download a python pkg, then pass it onto scripts
that could turn it into a .deb, .rpm (bunch of flavours), tarball,
etc., then install it.

or

A tool that would download pre-packaged distro specific pkgs.


Regarding the first... I like this because users will always get an
installable package that fits into their system, and every system can
have packages in their native format - something you can't expect if
developers are doing the distro specific packaging and arbitrary
individuals are filling in the rest.  I realize the distros will
probably start auto-building once python modules are available in a
standard format, but users should also have the ability to do the
packaging with a minimum of hoops to jump through.


As for the second... it seems like an awful waste of archive space to
have the same code in many different formats, just because some want
to put stuff in /usr/share and others in /usr/lib, etc.


Maybe I've just been spoiled by the Debian packaging system and tools,
but tarballs/zips and third-party packages are definately second
class, I would much rather have the packages created on the fly then
installed by a script (probably provided by a Debian developer).

I suppose you could replace "Debian" with RH, Mandrake, Microsoft, ...
whatever your fav distro is.


- Bruce




From doughellmann at home.com  Mon Feb 26 12:45:33 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Mon, 26 Feb 2001 06:45:33 -0500
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102260150060.28098-100000@bms>
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms>
Message-ID: <01022606501904.15046@branagan>

On Mon, 26 Feb 2001, Bruce Sass wrote:
> On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> <...>
> > My thoughts on that are that it's a job that's probably going to be
> > something subject largely to being handed off to appropriate scripts
> > based on the platform and some user input (for example, a user may
> > prefer to have packages they download installed in ~/lib/python1.5,
> > instead of /usr/lib/python1.5).
> >
> > Ideally though, the tool should be able to deal with allowing the user
> > to select their preferred distribution media.  I'd prefer to see an ix86
> > RedHat 7.0 RPM, a SRPM, and then would fall back to a distutils file or
> > tar file.
> 
> Do you mean:
> 
> A tool that would download a python pkg, then pass it onto scripts
> that could turn it into a .deb, .rpm (bunch of flavours), tarball,
> etc., then install it.
> 
> or
> 
> A tool that would download pre-packaged distro specific pkgs.
> 
> 
> Regarding the first... I like this because users will always get an
> installable package that fits into their system, and every system can
> have packages in their native format - something you can't expect if
> developers are doing the distro specific packaging and arbitrary
> individuals are filling in the rest.  I realize the distros will
> probably start auto-building once python modules are available in a
> standard format, but users should also have the ability to do the
> packaging with a minimum of hoops to jump through.

Why is this a desirable feature?  Is it worth the complexity?  

What if Python kept up with its own packages and modules like XEmacs does?  We
could then use the same format on all platforms -- it might even be something
we roll ourself to make sure we can unpack it with *only* Python installed on a
platform.

Doug



From bsass at freenet.edmonton.ab.ca  Mon Feb 26 19:14:35 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Mon, 26 Feb 2001 11:14:35 -0700 (MST)
Subject: CPAN functionality for python
In-Reply-To: <01022606501904.15046@branagan>
Message-ID: <Pine.LNX.4.33.0102261043060.28411-100000@bms>


On Mon, 26 Feb 2001, Doug Hellmann wrote:

> On Mon, 26 Feb 2001, Bruce Sass wrote:
> > Do you mean:
> >
> > A tool that would download a python pkg, then pass it onto scripts
> > that could turn it into a .deb, .rpm (bunch of flavours), tarball,
> > etc., then install it.
> >
> > or
> >
> > A tool that would download pre-packaged distro specific pkgs.
> >
> >
> > Regarding the first... I like this because users will always get an
> > installable package that fits into their system, and every system can
> > have packages in their native format - something you can't expect if
> > developers are doing the distro specific packaging and arbitrary
> > individuals are filling in the rest.  I realize the distros will
> > probably start auto-building once python modules are available in a
> > standard format, but users should also have the ability to do the
> > packaging with a minimum of hoops to jump through.
>
> Why is this a desirable feature?  Is it worth the complexity?

Because distros are not always building what users are running.

It should be no more complex than defining the inter-dependencies
between the modules in the archive, and declaring what is required
from outside the archive.  The translation from a generic pkg to a
distro specific pkg would be a problem for the disto to solve
(probably a matter of repacking info about the pkg, then using
whatever packaging tools they have).  A second hook would be needed
for mapping the python pkg name to the distro specific name.

> What if Python kept up with its own packages and modules like XEmacs does?  We
> could then use the same format on all platforms -- it might even be something
> we roll ourself to make sure we can unpack it with *only* Python installed on a
> platform.

There are enough different packaging schemes around that it seems
unnecessarily heavy-handed to pick one (or two or three) over the
others - Python should not appear to favor any paticular system.  The
only way I can see to do that is to not package to anyones spec, but
provide enough information so that anyone can package to their
fav spec.


- Bruce




From guido at digicool.com  Mon Feb 26 19:22:45 2001
From: guido at digicool.com (Guido van Rossum)
Date: Mon, 26 Feb 2001 13:22:45 -0500
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: Your message of "Mon, 26 Feb 2001 11:14:35 MST."
             <Pine.LNX.4.33.0102261043060.28411-100000@bms> 
References: <Pine.LNX.4.33.0102261043060.28411-100000@bms> 
Message-ID: <200102261822.NAA26272@cj20424-a.reston1.va.home.com>

> There are enough different packaging schemes around that it seems
> unnecessarily heavy-handed to pick one (or two or three) over the
> others - Python should not appear to favor any paticular system.  The
> only way I can see to do that is to not package to anyones spec, but
> provide enough information so that anyone can package to their
> fav spec.

But practicality beats purity: if a large fraction of users,
especially newbie users, are using a specific platform, it makes a lot
of sense to provide pre-packaged distributions for them!  E.g. Windows
installers, Red Hat RPMs.  (So far my experience with Windows
installers is much more positive than with RPMs though -- RPMs forever
seem to depend on some version of some other RPM that you don't have.)

--Guido van Rossum (home page: http://www.python.org/~guido/)



From phrxy at csv.warwick.ac.uk  Mon Feb 26 19:56:37 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Mon, 26 Feb 2001 18:56:37 +0000 (GMT)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <200102261822.NAA26272@cj20424-a.reston1.va.home.com>
Message-ID: <Pine.SOL.4.30.0102261852520.26805-100000@mimosa.csv.warwick.ac.uk>

On Mon, 26 Feb 2001, Guido van Rossum wrote:
[...]
> installers, Red Hat RPMs.  (So far my experience with Windows
> installers is much more positive than with RPMs though -- RPMs forever
> seem to depend on some version of some other RPM that you don't have.)

Surely this is a function of free vs. non-free software as opposed to
which particular packaging format you are using?


John




From tim.one at home.com  Mon Feb 26 20:14:27 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 26 Feb 2001 14:14:27 -0500
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <Pine.SOL.4.30.0102261852520.26805-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEBGJCAA.tim.one@home.com>

[Guido]
> (So far my experience with Windows installers is much more
> positive than with RPMs though -- RPMs forever seem to depend
> on some version of some other RPM that you don't have.)

[John J. Lee]
> Surely this is a function of free vs. non-free software as opposed to
> which particular packaging format you are using?

I doubt it has much to do with either.  It's more that, over very painful
years of installers that routinely crashed, Windows apps that rely on
anything external have learned to ship with *everything* they need.  Heck,
the Windows Python installer even ships with a copy of Microsoft's C runtime
libraries, just in case the platform installed on doesn't have a reasonably
up-to-date copy.  If we did something similar on Linux, we'd ship Python
there with its own glibc <0.7 wink>.

shudder-if-you-like-but-it-would-have-stopped-the-platform-
    LONG_BIT-screwups-ly y'rs  - tim




From phrxy at csv.warwick.ac.uk  Tue Feb 27 19:06:23 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Tue, 27 Feb 2001 18:06:23 +0000 (GMT)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <LNBBLJKPBEHFEDALKOLCEEBGJCAA.tim.one@home.com>
Message-ID: <Pine.SOL.4.30.0102270947140.3918-100000@mimosa.csv.warwick.ac.uk>

On Mon, 26 Feb 2001, Tim Peters wrote:

> [Guido]
> > (So far my experience with Windows installers is much more
> > positive than with RPMs though -- RPMs forever seem to depend
> > on some version of some other RPM that you don't have.)
>
> [John J. Lee]
> > Surely this is a function of free vs. non-free software as opposed to
> > which particular packaging format you are using?
>
> I doubt it has much to do with either.  It's more that, over very painful
> years of installers that routinely crashed, Windows apps that rely on
> anything external have learned to ship with *everything* they need.  Heck,
[...]

Well yes, that was one of the things I was thinking of (I don't remember
what the others were now :-).  But you are right in saying that 'bundling
everything' does not really divide along the free / non-free line.


John




From bsass at freenet.edmonton.ab.ca  Mon Feb 26 20:22:47 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Mon, 26 Feb 2001 12:22:47 -0700 (MST)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <Pine.SOL.4.30.0102261852520.26805-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <Pine.LNX.4.33.0102261214430.28411-100000@bms>

On Mon, 26 Feb 2001, John J. Lee wrote:
> On Mon, 26 Feb 2001, Guido van Rossum wrote:
> [...]
> > installers, Red Hat RPMs.  (So far my experience with Windows
> > installers is much more positive than with RPMs though -- RPMs forever
> > seem to depend on some version of some other RPM that you don't have.)
>
> Surely this is a function of free vs. non-free software as opposed to
> which particular packaging format you are using?

<subject to correction by someone who knows more about this than I>
I think the problem with RPMs is that they depend on a specific
package, instead of depending on the contents of a package.

e.g., If you have code depending on libfoo-1.x, depending on the
package that provides libfoo-1.1 will work, but is no good to people
with libfoo-1.2, the dependency should be on any package that provides
libfoo-1.
</subject to correction by someone who knows more about this than I>


- Bruce




From jafo at tummy.com  Tue Feb 27 02:38:57 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 18:38:57 -0700
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102261214430.28411-100000@bms>; from bsass@freenet.edmonton.ab.ca on Mon, Feb 26, 2001 at 12:22:47PM -0700
References: <Pine.SOL.4.30.0102261852520.26805-100000@mimosa.csv.warwick.ac.uk> <Pine.LNX.4.33.0102261214430.28411-100000@bms>
Message-ID: <20010226183857.G7531@tummy.com>

On Mon, Feb 26, 2001 at 12:22:47PM -0700, Bruce Sass wrote:
>I think the problem with RPMs is that they depend on a specific
>package, instead of depending on the contents of a package.

In RPM, you can let the packaging system decide what revisions of
various packages are needed (by looking at the results, it will
find out what libraries it links against, what's in the she-bang
(#!) lines, etc).  That mechanism is usually the most paranoid
about versions (if we linked against glibc 2.1.7.9.pi, then that's
exactly what we expect).

On the other hand, if you know better, as an application developer,
you can include something like:

	Requires: python >= 1.5.2

in the .spec file which says "I know that 1.5.1 won't work, but I hope
that 1.6 won't break this package".

Sean
-- 
 Some girl with psychic power, She said "T-bone, what's your sign?"
 I blinked and answered "Neon", I thought I'd blow her mind.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From bsass at freenet.edmonton.ab.ca  Tue Feb 27 10:54:29 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 02:54:29 -0700 (MST)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <20010226183857.G7531@tummy.com>
Message-ID: <Pine.LNX.4.33.0102270250030.28512-100000@bms>

This confirms what I got out of another list member in p-email...
all the problems surrounding RPMs I hear about are most likely the
result of inexperienced packagers.

Debian can have the same problems, but the packaging tools are
thorough enough that things usually work, even if the result is really
ugly (e.g., the Active State py-2.0 .debs <shudder>).

--
On Mon, 26 Feb 2001, Sean Reifschneider wrote:

> On Mon, Feb 26, 2001 at 12:22:47PM -0700, Bruce Sass wrote:
> >I think the problem with RPMs is that they depend on a specific
> >package, instead of depending on the contents of a package.
>
> In RPM, you can let the packaging system decide what revisions of
> various packages are needed (by looking at the results, it will
> find out what libraries it links against, what's in the she-bang
> (#!) lines, etc).  That mechanism is usually the most paranoid
> about versions (if we linked against glibc 2.1.7.9.pi, then that's
> exactly what we expect).
>
> On the other hand, if you know better, as an application developer,
> you can include something like:
>
> 	Requires: python >= 1.5.2
>
> in the .spec file which says "I know that 1.5.1 won't work, but I hope
> that 1.6 won't break this package".
>
> Sean
>




From guido at digicool.com  Tue Feb 27 03:10:48 2001
From: guido at digicool.com (Guido van Rossum)
Date: Mon, 26 Feb 2001 21:10:48 -0500
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: Your message of "Tue, 27 Feb 2001 02:54:29 MST."
             <Pine.LNX.4.33.0102270250030.28512-100000@bms> 
References: <Pine.LNX.4.33.0102270250030.28512-100000@bms> 
Message-ID: <200102270210.VAA01457@cj20424-a.reston1.va.home.com>

> This confirms what I got out of another list member in p-email...
> all the problems surrounding RPMs I hear about are most likely the
> result of inexperienced packagers.

That's easy to say.

If lots of people contributed Python modules that were coded in a
non-portable way, would we blame those users?  No!  We'd look for a
way to encourage newbies to do the right thing!

I still blame RPM for making it too damn hard to do the right thing.
RPM mist be the Perl of packaging. :-)

--Guido van Rossum (home page: http://www.python.org/~guido/)



From bsass at freenet.edmonton.ab.ca  Mon Feb 26 19:57:36 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Mon, 26 Feb 2001 11:57:36 -0700 (MST)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <200102261822.NAA26272@cj20424-a.reston1.va.home.com>
Message-ID: <Pine.LNX.4.33.0102261132560.28411-100000@bms>

On Mon, 26 Feb 2001, Guido van Rossum wrote:

> > There are enough different packaging schemes around that it seems
> > unnecessarily heavy-handed to pick one (or two or three) over the
> > others - Python should not appear to favor any paticular system.  The
> > only way I can see to do that is to not package to anyones spec, but
> > provide enough information so that anyone can package to their
> > fav spec.
>
> But practicality beats purity: if a large fraction of users,
> especially newbie users, are using a specific platform, it makes a lot
> of sense to provide pre-packaged distributions for them!  E.g. Windows
> installers, Red Hat RPMs.  (So far my experience with Windows
> installers is much more positive than with RPMs though -- RPMs forever
> seem to depend on some version of some other RPM that you don't have.)

Sure, but if you can build packages on the fly, transparently to the
user... it is a "six of one, half dozen of the other" situation.  I
would expect there to be situation where autobuilding would fail or
not be practical (C based modules), supporting "binary" packages would
be a necessity.

Ya, RPMs are not known for doing a good job of dependencies.  Debian
would be a better model to look at with respect to what dependency
info is required and how it should be handled.  I rarely (can't
remember the last time) have dependency problems with DEBs, that
includes third party DEBs (they usually have problems with "Policy",
but that's a whole other story ;).


- Bruce




From jafo at tummy.com  Tue Feb 27 02:31:38 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 18:31:38 -0700
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102261132560.28411-100000@bms>; from bsass@freenet.edmonton.ab.ca on Mon, Feb 26, 2001 at 11:57:36AM -0700
References: <200102261822.NAA26272@cj20424-a.reston1.va.home.com> <Pine.LNX.4.33.0102261132560.28411-100000@bms>
Message-ID: <20010226183138.E7531@tummy.com>

On Mon, Feb 26, 2001 at 11:57:36AM -0700, Bruce Sass wrote:
>Ya, RPMs are not known for doing a good job of dependencies.  Debian

They aren't?  The example given was a situation where one could say that
RPM was doing *TOO* good a job of dependencies -- it was failing to
install because you didn't have what was needed.

For quite a long time I was runing a 5.2 machine running the newest versions
of GTK, but older versions of glibc.  Practicly none of the stock binary RPMs
would install on it because of that.

However, if you pick up an SRPM, and do "rpm --rebuild package.src.rpm", it
will build the RPM against *EXACTLY* the set of libraries and files you
have.  It works brilliantly.

Sean
-- 
 "No early worm is giving ME the bird!" -- Bullwinkle J. Moose
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From bsass at freenet.edmonton.ab.ca  Tue Feb 27 09:35:56 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 01:35:56 -0700 (MST)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <20010226183138.E7531@tummy.com>
Message-ID: <Pine.LNX.4.33.0102270123170.28512-100000@bms>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:

> On Mon, Feb 26, 2001 at 11:57:36AM -0700, Bruce Sass wrote:
> >Ya, RPMs are not known for doing a good job of dependencies.  Debian
>
> They aren't?  The example given was a situation where one could say that
> RPM was doing *TOO* good a job of dependencies -- it was failing to
> install because you didn't have what was needed.

s/*TOO* good/overboard/  ;)

> For quite a long time I was runing a 5.2 machine running the newest versions
> of GTK, but older versions of glibc.  Practicly none of the stock binary RPMs
> would install on it because of that.
>
> However, if you pick up an SRPM, and do "rpm --rebuild package.src.rpm", it
> will build the RPM against *EXACTLY* the set of libraries and files you
> have.  It works brilliantly.

Agreed, so supply something that can be turned into an SRPM, or a
Debian .dsc+.diff+.orig.tar.gz source package, or a MS whatever.

I suppose one could provide SRPMs only - let everyone else do a
conversion - but that is favoring a specific packaging system and
could easily be misconstrued... which is why I suggested a neutral
format.


- Bruce




From jafo at tummy.com  Tue Feb 27 09:41:16 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 01:41:16 -0700
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102270123170.28512-100000@bms>; from bsass@freenet.edmonton.ab.ca on Tue, Feb 27, 2001 at 01:35:56AM -0700
References: <20010226183138.E7531@tummy.com> <Pine.LNX.4.33.0102270123170.28512-100000@bms>
Message-ID: <20010227014116.V1781@tummy.com>

On Tue, Feb 27, 2001 at 01:35:56AM -0700, Bruce Sass wrote:
>I suppose one could provide SRPMs only - let everyone else do a
>conversion - but that is favoring a specific packaging system and
>could easily be misconstrued... which is why I suggested a neutral
>format.

There is no such thing as a neutral format.  If it uses .tar.gz files,
then Slackware users will be winging that we swiped theirs.  ;-)

Sean
-- 
 It usually takes more than three weeks to prepare a good impromptu speech.
                 -- Mark Twain
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From skip at mojam.com  Mon Feb 26 21:04:05 2001
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 26 Feb 2001 14:04:05 -0600 (CST)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <200102261822.NAA26272@cj20424-a.reston1.va.home.com>
References: <Pine.LNX.4.33.0102261043060.28411-100000@bms>
	<200102261822.NAA26272@cj20424-a.reston1.va.home.com>
Message-ID: <15002.46773.642780.730206@beluga.mojam.com>

    Guido> (So far my experience with Windows installers is much more
    Guido> positive than with RPMs though -- RPMs forever seem to depend on
    Guido> some version of some other RPM that you don't have.)

Guido,

Can you elaborate?  I rarely have any problem installing stuff that's
packaged as an RPM.  If I understand your statement correctly, the Windows
Python installer had the same sort of problem with Tcl/Tk until the decision
was made to bite the bullet and simply include it as part of the installer.

-- 
Skip Montanaro (skip at mojam.com)
Support Mojam & Musi-Cal: http://www.musi-cal.com/sponsor.shtml
(847)971-7098



From tim.one at home.com  Mon Feb 26 21:21:21 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 26 Feb 2001 15:21:21 -0500
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <15002.46773.642780.730206@beluga.mojam.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEBOJCAA.tim.one@home.com>

[Skip Montanaro, to Guido]
> Can you elaborate?  I rarely have any problem installing stuff that's
> packaged as an RPM.  If I understand your statement correctly, the
> Windows Python installer had the same sort of problem with Tcl/Tk until
> the decision was made to bite the bullet and simply include it as part
> of the installer.

Actually, the Python Windows installer included the Tcl/Tk installer for a
long time without solving anything.  It wasn't until Python *stopped*
running the Tcl/Tk installer, and took the *pieces* of Tcl/Tk Python needs
and installed them itself, that the endless Windows problems went away.
Python no longer tries to cooperate with any pre-existing Tcl/Tk
installation on Windows, and indeed does all it can to ensure that the
Tcl/Tk it installs neither affects nor is affected by any other Tcl/Tk
installation on the box, or that may exist in the future, or that may get
uninstalled in the future.

removing-potential-cooperation-served-in-reality-to-remove-potential-
    conflict-ly y'rs  - tim




From mwa at gate.net  Mon Feb 26 21:14:24 2001
From: mwa at gate.net (Mark W. Alexander)
Date: Mon, 26 Feb 2001 15:14:24 -0500 (EST)
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <200102261822.NAA26272@cj20424-a.reston1.va.home.com>
Message-ID: <Pine.A41.4.21.0102261511490.44014-100000@dakota.gate.net>

On Mon, 26 Feb 2001, Guido van Rossum wrote:

> Date: Mon, 26 Feb 2001 13:22:45 -0500
> From: Guido van Rossum <guido at digicool.com>
> To: Bruce Sass <bsass at freenet.edmonton.ab.ca>
> Cc: python-list at python.org, distutils-sig at python.org
> Subject: Re: [Distutils] Re: CPAN functionality for python
>
> But practicality beats purity: if a large fraction of users,
> especially newbie users, are using a specific platform, it makes a lot
> of sense to provide pre-packaged distributions for them!  E.g. Windows
> installers, Red Hat RPMs.  (So far my experience with Windows
> installers is much more positive than with RPMs though -- RPMs forever
> seem to depend on some version of some other RPM that you don't have.)
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)

Isn't this the point of distutils --bdist functionality? The 
catalog should provide download capability of any binary format
supported byu distutils. It can either be done on the fly at
request time, or batched periodically depending on the space
availability of the catalog providing site.

Mark
mwa at gate.net





From robin at alldunn.com  Mon Feb 26 21:36:46 2001
From: robin at alldunn.com (Robin Dunn)
Date: Mon, 26 Feb 2001 12:36:46 -0800
Subject: [Distutils] Re: CPAN functionality for python
References: <Pine.A41.4.21.0102261511490.44014-100000@dakota.gate.net>
Message-ID: <02f501c0a033$d7bedd00$0100a8c0@Rogue>

> >
> > But practicality beats purity: if a large fraction of users,
> > especially newbie users, are using a specific platform, it makes a lot
> > of sense to provide pre-packaged distributions for them!  E.g. Windows
> > installers, Red Hat RPMs.  (So far my experience with Windows
> > installers is much more positive than with RPMs though -- RPMs forever
> > seem to depend on some version of some other RPM that you don't have.)
> >
> > --Guido van Rossum (home page: http://www.python.org/~guido/)
>
> Isn't this the point of distutils --bdist functionality? The
> catalog should provide download capability of any binary format
> supported byu distutils. It can either be done on the fly at
> request time, or batched periodically depending on the space
> availability of the catalog providing site.
>

That would work for pure python packages, but not for any that contains an
extension module.  They would still have to be built on the target platform
and then the taget package format can be built with --bdist.

--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!







From wware at world.std.com  Tue Feb 27 00:49:32 2001
From: wware at world.std.com (Will Ware)
Date: Mon, 26 Feb 2001 23:49:32 GMT
Subject: [Distutils] Re: CPAN functionality for python
References: <mailman.983213887.18394.python-list@python.org>
Message-ID: <G9E26K.su@world.std.com>

On Mon, 26 Feb 2001, Guido van Rossum wrote:
> So far my experience with Windows
> installers is much more positive than with RPMs though -- RPMs forever
> seem to depend on some version of some other RPM that you don't have.

John J. Lee (phrxy at csv.warwick.ac.uk) wrote:
> Surely this is a function of free vs. non-free software as opposed to
> which particular packaging format you are using?

Mebbe not. I read something recently, referenced in a Slashdot article
IIRC, which claims that apt-get (the Debian distribution's equivalent to
Red Hat's RPM installer) handles this much better. Where the RPM system
will inform you about dependencies but insist that the burden of resolving
them falls to you (and heaven forbid that _those_ packages have unmet
dependencies; exponentially-nightmarish upgrading ensues), apt-get will
seek out the required packages for you, and whatever packages they may
require, in a painless manner like "make". In order to do this, apt-get
needs an internet connection and a list of places to look for stuff.

-- 
-----------------------------------+---------------------
 22nd century: Esperanto, geodesic | Will Ware
 domes, hovercrafts, metric system | wware at world.std.com


From jafo at tummy.com  Tue Feb 27 09:53:38 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 01:53:38 -0700
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <G9E26K.su@world.std.com>; from wware@world.std.com on Mon, Feb 26, 2001 at 11:49:32PM +0000
References: <mailman.983213887.18394.python-list@python.org> <G9E26K.su@world.std.com>
Message-ID: <20010227015338.B13115@tummy.com>

On Mon, Feb 26, 2001 at 11:49:32PM +0000, Will Ware wrote:
>Red Hat's RPM installer) handles this much better. Where the RPM system
>will inform you about dependencies but insist that the burden of resolving
>them falls to you (and heaven forbid that _those_ packages have unmet

My impression from most RedHat users is that they really aren't interested
in having their system pick up files unless they ask for them.  The "rpm"
tool can be given a URL to pick the files up from, so implementing something
that follows dependencies and picks them up isn't really a hard thing to
do...

It has been implemented in things like rpmfind and autorpm, but they just
don't seem to be that popular...

The question that makes this relavent to the discussion at hand is: is this
more or less true of Python modules than of your OS software?

Correct me if I'm wrong, but CPAN doesn't do this...  Or at least doesn't
seem to.  Just last night I was installing one package and the test
for it failed.  I tried installing other packages listed which the software
needed, and then went back to the "trouble child" and found it now worked
fine.

The thing is to give the users the option of doing it manually.  What about
if the download and build is happening because of an
"import module_i_dont_already_have"?

Sean
-- 
 Home is where your source is.  -- Sean Reifschneider, 1999
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From jafo at tummy.com  Tue Feb 27 02:09:35 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 18:09:35 -0700
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102261043060.28411-100000@bms>; from bsass@freenet.edmonton.ab.ca on Mon, Feb 26, 2001 at 11:14:35AM -0700
References: <01022606501904.15046@branagan> <Pine.LNX.4.33.0102261043060.28411-100000@bms>
Message-ID: <20010226180935.B7531@tummy.com>

On Mon, Feb 26, 2001 at 11:14:35AM -0700, Bruce Sass wrote:
>from outside the archive.  The translation from a generic pkg to a
>distro specific pkg would be a problem for the disto to solve
>(probably a matter of repacking info about the pkg, then using

The only way having automaticly-generated packages be possible is
if we really force a conformance from the authors of the packages.
I've packaged quite a number of RPMs in my day, and you can get pretty
close to just wrapping a tar file with some meta-information in
some cases, but these are the minority.

In most cases you have to provide patches which fix incorrect paths
in the tar file (for example, the Python tar file uses
"#!/usr/local/bin/python" all over the place).

If we can force the distributions to be more compatible with being
packaged, it may be very easy.  Otherwise, we may have to rely on
people contributing their time to building RPMs if we want to have
RPMs.

>only way I can see to do that is to not package to anyones spec, but
>provide enough information so that anyone can package to their
>fav spec.

I don't agree...  I believe that we should embrace the different packaging
formats, not act like they don't exist.  If the distutils package can be
made to generate native RPMs, that's great.  The *CLIENT* program should
dictate that.  If it can handle turning a distutil package into an RPM,
it'll have it's preference be a distutil package.

I don't believe having the archive not able to deal with different package
formats is acceptable.  If that were the case, then if the package didn't
provide enough information to build an RPM, the user would have no option
within the archive of working around it.

Sean
-- 
 When in doubt, use brute force. 
                 -- Ken Thompson
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From bsass at freenet.edmonton.ab.ca  Tue Feb 27 08:53:45 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 00:53:45 -0700 (MST)
Subject: CPAN functionality for python
In-Reply-To: <20010226180935.B7531@tummy.com>
Message-ID: <Pine.LNX.4.33.0102270016190.28512-100000@bms>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> On Mon, Feb 26, 2001 at 11:14:35AM -0700, Bruce Sass wrote:
> >from outside the archive.  The translation from a generic pkg to a
> >distro specific pkg would be a problem for the disto to solve
> >(probably a matter of repacking info about the pkg, then using
>
> The only way having automaticly-generated packages be possible is
> if we really force a conformance from the authors of the packages.
> I've packaged quite a number of RPMs in my day, and you can get pretty
> close to just wrapping a tar file with some meta-information in
> some cases, but these are the minority.

I have assumed that submitted packages would need to meet a
standard of some sort - was that incorrect?

> In most cases you have to provide patches which fix incorrect paths
> in the tar file (for example, the Python tar file uses
> "#!/usr/local/bin/python" all over the place).

That is to be expected... and only the distro is really equipted to
handle these issues.

> If we can force the distributions to be more compatible with being
> packaged, it may be very easy.  Otherwise, we may have to rely on
> people contributing their time to building RPMs if we want to have
> RPMs.

I'm not sure what you mean here.  The distributions are the ones doing
the packaging, the Python version of CPAN should supply all the bits
needed for the distros to perform that function.

> >only way I can see to do that is to not package to anyones spec, but
> >provide enough information so that anyone can package to their
> >fav spec.
>
> I don't agree...  I believe that we should embrace the different packaging
> formats, not act like they don't exist.

Maybe we see "embracing" them a little differently.  I believe
providing everything they need to turn a Python module/package into
their standard format, using their packaging and QC tools, does them
more of a service than providing pre-built packages created by a third
party.  I'm not saying binary style packages shouldn't be supported,
just that the emphasis should be on portable (in that they can easily
be turned into a system's native pkg format) source packages.

> If the distutils package can be
> made to generate native RPMs, that's great.  The *CLIENT* program should
> dictate that.  If it can handle turning a distutil package into an RPM,
> it'll have it's preference be a distutil package.

That sounds like exactly what I was enquiring about - as long as the
distutils pkg provides enough info for every distro to create their
own pkg.

> I don't believe having the archive not able to deal with different package
> formats is acceptable.  If that were the case, then if the package didn't
> provide enough information to build an RPM, the user would have no option
> within the archive of working around it.

Agreed.  The submission should not be accepted if it does not contain
enough information to be packaged.


- Bruce




From chris at voodooland.net  Tue Feb 27 09:00:40 2001
From: chris at voodooland.net (Chris Watson)
Date: Tue, 27 Feb 2001 02:00:40 -0600 (CST)
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102270016190.28512-100000@bms>
Message-ID: <Pine.BSF.4.21.0102270157430.56975-100000@open-systems.net>

> > I don't believe having the archive not able to deal with different package
> > formats is acceptable.  If that were the case, then if the package didn't
> > provide enough information to build an RPM, the user would have no option
> > within the archive of working around it.
> 
> Agreed.  The submission should not be accepted if it does not contain
> enough information to be packaged.

	Not to toss a pebble in the still waters here. But it *might* be
worth the effort to work with or at least get idea's from the openpackages
people. Who are now wrestling with certain issues discussed here for
handling a common package system across several platforms. Just a thought.
Info can be found at http://www.openpackages.net.

--
=============================================================================
-Chris Watson         (316) 326-3862 | FreeBSD Consultant, FreeBSD Geek 
Work:              scanner at jurai.net | Open Systems Inc., Wellington, Kansas
Home:  scanner at deceptively.shady.org | http://open-systems.net
=============================================================================
WINDOWS: "Where do you want to go today?"
LINUX: "Where do you want to go tomorrow?"
BSD: "Are you guys coming or what?"
=============================================================================
irc.openprojects.net #FreeBSD -Join the revolution!
ICQ: 20016186




From jafo at tummy.com  Tue Feb 27 09:30:19 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 01:30:19 -0700
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102270016190.28512-100000@bms>; from bsass@freenet.edmonton.ab.ca on Tue, Feb 27, 2001 at 12:53:45AM -0700
References: <20010226180935.B7531@tummy.com> <Pine.LNX.4.33.0102270016190.28512-100000@bms>
Message-ID: <20010227013019.T1781@tummy.com>

(I'm going to remove the non-list entities in further messages -- if
you want to follow this discussion you can find it on the lists.  Not
entirely sure how the CC list got so big in the first place)

On Tue, Feb 27, 2001 at 12:53:45AM -0700, Bruce Sass wrote:
>I have assumed that submitted packages would need to meet a
>standard of some sort - was that incorrect?

Obviously there will be some level of standards.  I don't know that the
tool should govern a policy decision though...  It seems that you're
pushing for a standard that's more than "just dump a .tar.gz here"
and less than "schlep up tars and RPMs and .debs, and don't forget about
HP-UX swinstall".  Maybe something like those (defunct?) Linux Software
Map entries?

I don't know the answers to what the standards will be and how they'll
be enforced...  I do suspect that they'll change over time, so the tool
should probably be flexible.

>That is to be expected... and only the distro is really equipted to
>handle these issues.

I don't agree with that.  Historicly, it's been fairly common for the
package to say "Hey, I've got some paths I need in the various files.
Here's the macros that represent them."

>I'm not sure what you mean here.  The distributions are the ones doing
>the packaging, the Python version of CPAN should supply all the bits
>needed for the distros to perform that function.

What do you mean by "the distributions"?  How will they do the packaging?
Are you saying that RedHat should be creating the RPMs?  That's just not
how it's actually done in the real world...  We just can't wait for
the distribution vendors to decide to include them on their disc.
It's not that sort of world.

If tools can be provided so that a user can turn a .tar.gz file into a
.RPM file easily, I'm all for it.  What about systems where that isn't
possible?

As far as I can tell, my system doesn't *REQUIRE* the inclusion of
RPMs, it just *ALLOWS* it.  You seem to be arguing for the exclusion
of that capability.  Why?

Sean
-- 
 The early bird may get the worm, but the second mouse gets the cheese.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From jafo at tummy.com  Tue Feb 27 02:36:06 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 18:36:06 -0700
Subject: CPAN functionality for python
In-Reply-To: <01022606501904.15046@branagan>; from doughellmann@home.com on Mon, Feb 26, 2001 at 06:45:33AM -0500
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022606501904.15046@branagan>
Message-ID: <20010226183606.F7531@tummy.com>

On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
>What if Python kept up with its own packages and modules like XEmacs does?  We
>could then use the same format on all platforms -- it might even be something
>we roll ourself to make sure we can unpack it with *only* Python installed on a
>platform.

My understanding is that that's what distutils is meant to do, to an extent.
The problem is that it's not entirely unreasonable to expect that an
"rpm -qa" produce a list of all the software that's installed on your
box.  If you don't use an RPM, you can't really make use of that, and
packages WILL get dropped after a re-install.

Sean
-- 
 ISA isn't dead, it's just that people wish it were.  The correct term for
 this condition is "legacy"...  --  Sean Reifschneider, 1999
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From doughellmann at home.com  Tue Feb 27 04:45:38 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Mon, 26 Feb 2001 22:45:38 -0500
Subject: CPAN functionality for python - requirements
In-Reply-To: <20010226183606.F7531@tummy.com>
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022606501904.15046@branagan> <20010226183606.F7531@tummy.com>
Message-ID: <01022623140707.15046@branagan>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
> >What if Python kept up with its own packages and modules like XEmacs does?  We
> >could then use the same format on all platforms -- it might even be something
> >we roll ourself to make sure we can unpack it with *only* Python installed on a
> >platform.
> 
> My understanding is that that's what distutils is meant to do, to an extent.
> The problem is that it's not entirely unreasonable to expect that an
> "rpm -qa" produce a list of all the software that's installed on your
> box.  If you don't use an RPM, you can't really make use of that, and
> packages WILL get dropped after a re-install.

The command "rpm -qa" only produces meaningful results on a system where RPM is
the default and primary package installation system.  There are more OSes where
that is *not* the case (and where RPM isn't installed at all) than where it is.

It still isn't clear to me why we would want to use more than one distribution
package format.  It appears that one argument is so that sysadmins can use
those tools to track what is installed on the system.  But isn't that part of
what this new thing (cyphon?) is supposed to do?  Do we want Python to require
an external package management system on each platform where this tool is
supported, or do we want a Python-esq tool which just works the way it is
supposed to, in the same way, everywhere Python is available?

But I think we're arguing over a design before we've worked out requirements. 
The basic tasks of this new tool, as I see them, are:

1. Locate Python packages you want on the net.
2. Resolve dependencies between what you want, what you have, and what you
   need.
3. Download the appropriate packages to give you what you want.
4. Install those packages.

Do we agree these are the primary features?

I sense a consensus that the "install" part should be handled by distutils.  Is
that right?  

Other requirements we might lay down up front:

1. Should run on all platforms where Python runs.
2. Must support mirrors on the server side.
3. Need to include documentation along with source for packages.

The features related to dependencies and downloads could be handled by a
platform-specific packaging system, but integrating with all of the different
options on all of the different platforms where Python runs increases the
complexity of the new tool.  (How do we handle RPMs on MacOS?  How do we handle
HQX files on Win32?  What about for Pippy, where files as such might not even
be appropriate?  Does Jython have any special requirements?)  Do we deal with
this complexity by deciding on a per-platform basis what format to use, or do
we force the user to specify the format they want downloaded?

If we're modeling what we're doing based on CPAN, maybe we should look at
that design.  What format does CPAN use when downloading Perl packages?  What
features does CPAN have that we want?  What does it not have that we want?

Doug



From jafo at tummy.com  Tue Feb 27 05:49:09 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 21:49:09 -0700
Subject: CPAN functionality for python - requirements
In-Reply-To: <01022623140707.15046@branagan>; from doughellmann@home.com on Mon, Feb 26, 2001 at 10:45:38PM -0500
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022606501904.15046@branagan> <20010226183606.F7531@tummy.com> <01022623140707.15046@branagan>
Message-ID: <20010226214909.A10411@tummy.com>

On Mon, Feb 26, 2001 at 10:45:38PM -0500, Doug Hellmann wrote:
>The command "rpm -qa" only produces meaningful results on a system where RPM is
>the default and primary package installation system.  There are more OSes where
>that is *not* the case (and where RPM isn't installed at all) than where it is.

Apaprently I wasn't canonical enough...  On an rpm-based system, it's
bogus if "rpm -qa" doesn't list packages installed by CPAN.  Just as it's
bogus for the similar to happen on Debian, HP, Sun, Windows, etc...

>It still isn't clear to me why we would want to use more than one distribution
>package format.  It appears that one argument is so that sysadmins can use
>those tools to track what is installed on the system.  But isn't that part of

That's one argument, and I'd say it's a good one at that...  Other reasons
are:

	The admin doesn't have to learn new packaging formats to deal with the
	Python packages.

	If we don't leverage off the platform's package management tool, we have
	to write our own.  Who's volunteering?

	If the packages aren't listed in the native packaging system, they are
	much more likely to get overlooked if admins search for things to
	update and the like.

>1. Locate Python packages you want on the net.

By "packages" do you mean modules only?  I envision it also handling
distribution of Python applications and even Python itself.

>2. Resolve dependencies between what you want, what you have, and what you
>   need.
>3. Download the appropriate packages to give you what you want.
>4. Install those packages.
>
>Do we agree these are the primary features?

Yes.

>I sense a consensus that the "install" part should be handled by distutils.  Is
>that right?  

That seems to be where we diverge...  If somone has made an RPM of it, I'd
rather have that than some files winding up hanging around my file-system.

>3. Need to include documentation along with source for packages.

Not gonna happen...  Until there's tools and standards for such documentation,
it's not really possible to deal with them...  In the cases where docstrings
are used, installing the module produces the documentation, but that seems
to be the exception...

>complexity of the new tool.  (How do we handle RPMs on MacOS?  How do we handle

My schema lists the package format, architecture and platform.  If you
search for a package for MacOS, and are returned an RPM, then apparently
somone has built an RPM for MacOS.  Your client should list the package
formats that it desires and prefers, which means that if you don't have
RPM on your Mac, you will pull down something else -- without penalizing
those who *DO* have RPM running on their Mac...

>HQX files on Win32?  What about for Pippy, where files as such might not even
>be appropriate?  Does Jython have any special requirements?)  Do we deal with

My setup handles that -- set the platform to "pilot" and you get access to
only the things that will work there (in theory).

Again, there are two distinct things here.  The archive network and the
tools which make use of it...

>this complexity by deciding on a per-platform basis what format to use, or do
>we force the user to specify the format they want downloaded?

You say "we force the user to specify".  I say "we allow the user the choose".

>If we're modeling what we're doing based on CPAN, maybe we should look at
>that design.  What format does CPAN use when downloading Perl packages?  What
>features does CPAN have that we want?  What does it not have that we want?

CPAN is a Unix-like directory structure, files are downloaded as .tar bundles
which are extracted and a "perl Makefile.pl; make; make test; make install"
is run.  Does Pippy have "make"?

Sean
-- 
 Gone Postal Sort: Iterate over elements, any element that is out of order you
 blow away.  -- Evelyn, Kevin, and Sean, watching Monty Python and reading DDJ
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From bsass at freenet.edmonton.ab.ca  Tue Feb 27 10:49:40 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 02:49:40 -0700 (MST)
Subject: CPAN functionality for python - requirements
In-Reply-To: <20010226214909.A10411@tummy.com>
Message-ID: <Pine.LNX.4.33.0102270234520.28512-100000@bms>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> On Mon, Feb 26, 2001 at 10:45:38PM -0500, Doug Hellmann wrote:
<...>
> >I sense a consensus that the "install" part should be handled by distutils.  Is
> >that right?
>
> That seems to be where we diverge...  If somone has made an RPM of it, I'd
> rather have that than some files winding up hanging around my file-system.

Even if the RPM was made by someone who didn't have a clue?
Wouldn't it be better to have a standard way to convert a Python
package into an RPM, that way any packaging problems will be seen by
everyone and the fix will be fast.

> >3. Need to include documentation along with source for packages.
>
> Not gonna happen...  Until there's tools and standards for such documentation,
> it's not really possible to deal with them...  In the cases where docstrings
> are used, installing the module produces the documentation, but that seems
> to be the exception...

Huh, there are tools and standards.
Isn't /usr/shar/doc/package the FHS place for general package specific
docs, and before that it was /usr/doc/package (FSSTND), MS users just put
them into the installation dir.  Special formats usually have a place
to reside, but I wouldn't expect an archive network or its tools to
know or care what and where they are.

> CPAN is a Unix-like directory structure, files are downloaded as .tar bundles
> which are extracted and a "perl Makefile.pl; make; make test; make install"
> is run.  Does Pippy have "make"?

The above is what Debian appears to have replaced - *ifiact*, because
it was not versatile enough, it thinks it knows how things should be
done and is wrong.


- Bruce




From jafo at tummy.com  Tue Feb 27 21:20:28 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 13:20:28 -0700
Subject: CPAN functionality for python - requirements
In-Reply-To: <Pine.LNX.4.33.0102270234520.28512-100000@bms>; from bsass@freenet.edmonton.ab.ca on Tue, Feb 27, 2001 at 02:49:40AM -0700
References: <20010226214909.A10411@tummy.com> <Pine.LNX.4.33.0102270234520.28512-100000@bms>
Message-ID: <20010227132028.L1781@tummy.com>

On Tue, Feb 27, 2001 at 02:49:40AM -0700, Bruce Sass wrote:
>Even if the RPM was made by someone who didn't have a clue?

Yes, because it's easier to remove an RPM from somone who doesn't have a
clue than a .tar file from somone who doesn't have a clue.

>Wouldn't it be better to have a standard way to convert a Python
>package into an RPM, that way any packaging problems will be seen by
>everyone and the fix will be fast.

The thing about automated systems is that they often fail in particularly
spetacular ways.  If that happens an human intervention is required, then
the archive network should be able to handle it, without requiring that
distutils be fixed, then handed off to the author of the package to make
a new release from.

>Huh, there are tools and standards.
>Isn't /usr/shar/doc/package the FHS place for general package specific

Where are the tools for turning doc-strings into HTML, and how are the
links between doc packages handled?  What is the markup format for the
doc strings?

We spent quite a while discussing the sad state of the documentation at our
last pythoneers meeting.  Mark Lutz was shocked that people would actually
use javadoc, but they do seem to simply becuase it was available and was
"good enough".

We are missing similar functionality.

Sean
-- 
 Give me immortality or give me death!
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From akuchlin at mems-exchange.org  Tue Feb 27 21:25:10 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: Tue, 27 Feb 2001 15:25:10 -0500
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <20010227132028.L1781@tummy.com>; from jafo@tummy.com on Tue, Feb 27, 2001 at 01:20:28PM -0700
References: <20010226214909.A10411@tummy.com> <Pine.LNX.4.33.0102270234520.28512-100000@bms> <20010227132028.L1781@tummy.com>
Message-ID: <20010227152510.B14307@ute.cnri.reston.va.us>

On Tue, Feb 27, 2001 at 01:20:28PM -0700, Sean Reifschneider wrote:
>On Tue, Feb 27, 2001 at 02:49:40AM -0700, Bruce Sass wrote:
>>Wouldn't it be better to have a standard way to convert a Python
>>package into an RPM, that way any packaging problems will be seen by
>>everyone and the fix will be fast.

There is such a way: python setup.py bdist_rpm.  (Or build
--build-format=rpm, if you prefer.)

>We spent quite a while discussing the sad state of the documentation at our
>last pythoneers meeting.  Mark Lutz was shocked that people would actually
>use javadoc, but they do seem to simply becuase it was available and was
>"good enough".
>We are missing similar functionality.

Indeed; no one ever pronounced on a format, so work has been stagnant.
That will probably change with the addition of pydoc in 2.1beta1.

--amk




From aahz at panix.com  Wed Feb 28 03:19:15 2001
From: aahz at panix.com (Aahz Maruch)
Date: 27 Feb 2001 18:19:15 -0800
Subject: [Distutils] Re: CPAN functionality for python - requirements
References: <mailman.983298609.17548.python-list@python.org>
Message-ID: <97hn73$iar$1@panix6.panix.com>

In article <mailman.983298609.17548.python-list at python.org>,
John J. Lee <phrxy at csv.warwick.ac.uk> wrote:
>On Tue, 27 Feb 2001, Sean Reifschneider wrote:
>> On Tue, Feb 27, 2001 at 09:30:13AM -0700, Evelyn Mitchell wrote:
>>>
>>>But it will also discover and resolve dependences in your perl
>>>site-packages, and automatically fetch them from your closest
>>>CPAN archive.
>>
>> Not according to my tests the night before last.  I did a test CPAN
>> install of "News::Newsrc", which failed because the "make test" was
>> failing.  I then installed the "Set::BitSet" (?  Something like that)
>> module and tried News::Newsrc again and it worked...
>
>Did you use CPAN.pm?  CPAN the archive doesn't do dependencies.
>Dependencies are listed in the Makefile.PL (BTW, I have no idea why the
>capitalised PL), and after the module CPAN.pm downloads what you asked
>for, it checks that to see if there are dependencies, then downloads them
>if you don't already have them installed.

It may depend on the version of CPAN.pm; last time I used Perl, it
didn't have this functionality (about three years ago).
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Nostalgia just ain't what it used to be


From bsass at freenet.edmonton.ab.ca  Wed Feb 28 00:10:32 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 16:10:32 -0700 (MST)
Subject: CPAN functionality for python - requirements
In-Reply-To: <20010227132028.L1781@tummy.com>
Message-ID: <Pine.LNX.4.33.0102271542110.30946-100000@bms>

On Tue, 27 Feb 2001, Sean Reifschneider wrote:

> On Tue, Feb 27, 2001 at 02:49:40AM -0700, Bruce Sass wrote:
> >Even if the RPM was made by someone who didn't have a clue?
>
> Yes, because it's easier to remove an RPM from somone who doesn't have a
> clue than a .tar file from somone who doesn't have a clue.

I'm a little more careful.  I only install DEBs from Debian or a
trusted packager (usually unofficial pkgs from a Debian developer),
third party DEBs and tarballs get manually unpacked and inspected.

Installing a third-party DEB or RPM is worse than doing
"cd / ; tar xzf..." with that tarball you found at
mess-with-your-system.com - at least the tarball is not going to run
scripts as root on you system!

> >Wouldn't it be better to have a standard way to convert a Python
> >package into an RPM, that way any packaging problems will be seen by
> >everyone and the fix will be fast.
>
> The thing about automated systems is that they often fail in particularly
> spetacular ways.  If that happens an human intervention is required, then
> the archive network should be able to handle it, without requiring that
> distutils be fixed, then handed off to the author of the package to make
> a new release from.


That sounds much better than the alternatives to me; everything is
likely to fail at some point, when it comes to the dull task of
packaging software... I'll go with the well defined automated process
'cause it doesn't get tired and make mistakes.

The archive network is not going to fix anything, it may dump packages
that are getting complaints, but that should - must - require human
intervention.

> >Huh, there are tools and standards.
> >Isn't /usr/shar/doc/package the FHS place for general package specific
>
> Where are the tools for turning doc-strings into HTML, and how are the
> links between doc packages handled?  What is the markup format for the
> doc strings?
>
> We spent quite a while discussing the sad state of the documentation at our
> last pythoneers meeting.  Mark Lutz was shocked that people would actually
> use javadoc, but they do seem to simply becuase it was available and was
> "good enough".
>
> We are missing similar functionality.

Oh, ok.  I though the reference was to READMEs, .html files, etc.

Doc strings are a great replacement for bunches of "rem" statements,
but as program documentation...
I don't want my programs cluttered up with user level documentation.

Literate programming is interesting.


- Bruce




From doughellmann at home.com  Tue Feb 27 12:50:52 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Tue, 27 Feb 2001 06:50:52 -0500
Subject: CPAN functionality for python - requirements
In-Reply-To: <20010226214909.A10411@tummy.com>
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022623140707.15046@branagan> <20010226214909.A10411@tummy.com>
Message-ID: <01022706564001.20069@branagan>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> On Mon, Feb 26, 2001 at 10:45:38PM -0500, Doug Hellmann wrote:
> >The command "rpm -qa" only produces meaningful results on a system where RPM is
> >the default and primary package installation system.  There are more OSes where
> >that is *not* the case (and where RPM isn't installed at all) than where it is.
> 
> Apaprently I wasn't canonical enough...  On an rpm-based system, it's
> bogus if "rpm -qa" doesn't list packages installed by CPAN.  Just as it's
> bogus for the similar to happen on Debian, HP, Sun, Windows, etc...
> 
> >It still isn't clear to me why we would want to use more than one distribution
> >package format.  It appears that one argument is so that sysadmins can use
> >those tools to track what is installed on the system.  But isn't that part of
> 
> That's one argument, and I'd say it's a good one at that...  Other reasons
> are:
> 
> 	The admin doesn't have to learn new packaging formats to deal with the
> 	Python packages.
> 
> 	If we don't leverage off the platform's package management tool, we have
> 	to write our own.  Who's volunteering?
> 
> 	If the packages aren't listed in the native packaging system, they are
> 	much more likely to get overlooked if admins search for things to
> 	update and the like.
> 
> >1. Locate Python packages you want on the net.
> 
> By "packages" do you mean modules only?  I envision it also handling
> distribution of Python applications and even Python itself.

Sure, all of that.  Anything which uses distutils for installation, right?

> >2. Resolve dependencies between what you want, what you have, and what you
> >   need.
> >3. Download the appropriate packages to give you what you want.
> >4. Install those packages.
> >
> >Do we agree these are the primary features?
> 
> Yes.
> 
> >I sense a consensus that the "install" part should be handled by distutils.  Is
> >that right?  
> 
> That seems to be where we diverge...  If somone has made an RPM of it, I'd
> rather have that than some files winding up hanging around my file-system.

So why did distutils make it into the Python core?

> >3. Need to include documentation along with source for packages.
> 
> Not gonna happen...  Until there's tools and standards for such documentation,
> it's not really possible to deal with them...  In the cases where docstrings
> are used, installing the module produces the documentation, but that seems
> to be the exception...

Ok, so make that:

3. Need to support including documentation along with source for packages when
documentation is available.

> >complexity of the new tool.  (How do we handle RPMs on MacOS?  How do we handle
> 
> My schema lists the package format, architecture and platform.  If you
> search for a package for MacOS, and are returned an RPM, then apparently
> somone has built an RPM for MacOS.  Your client should list the package
> formats that it desires and prefers, which means that if you don't have
> RPM on your Mac, you will pull down something else -- without penalizing
> those who *DO* have RPM running on their Mac...

But that leaves it up to the person posting the package to make distributions
in formats for any platforms where folks might want to install their stuff, or
to have someone *else* create those packages.  That may lead to a situation
where someone believes that they cannot install my package on MacOS because I
didn't make an HQX file (or whatever) when I posted HappyDoc.

> >HQX files on Win32?  What about for Pippy, where files as such might not even
> >be appropriate?  Does Jython have any special requirements?)  Do we deal with
> 
> My setup handles that -- set the platform to "pilot" and you get access to
> only the things that will work there (in theory).
> 
> Again, there are two distinct things here.  The archive network and the
> tools which make use of it...

Right.  We need to make it easy for the user to do the right thing.

> >this complexity by deciding on a per-platform basis what format to use, or do
> >we force the user to specify the format they want downloaded?
> 
> You say "we force the user to specify".  I say "we allow the user the choose".

I can go with "allow" but if we do not help then the user must specify in order
to get useful results from the tool, and that becomes "force."

> >If we're modeling what we're doing based on CPAN, maybe we should look at
> >that design.  What format does CPAN use when downloading Perl packages?  What
> >features does CPAN have that we want?  What does it not have that we want?
> 
> CPAN is a Unix-like directory structure, files are downloaded as .tar bundles
> which are extracted and a "perl Makefile.pl; make; make test; make install"

So, it sounds like their tool just does the download and then they use the
equivalent of distutils to do the installation.  Is that all we want?

> is run.  Does Pippy have "make"?

I don't know about that.  Maybe it isn't even appropriate to support Pippy, I'm
not sure.

Doug



From jafo at tummy.com  Tue Feb 27 20:29:03 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 12:29:03 -0700
Subject: CPAN functionality for python - requirements
In-Reply-To: <01022706564001.20069@branagan>; from doughellmann@home.com on Tue, Feb 27, 2001 at 06:50:52AM -0500
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022623140707.15046@branagan> <20010226214909.A10411@tummy.com> <01022706564001.20069@branagan>
Message-ID: <20010227122903.F1781@tummy.com>

On Tue, Feb 27, 2001 at 06:50:52AM -0500, Doug Hellmann wrote:
>> That seems to be where we diverge...  If somone has made an RPM of it, I'd
>> rather have that than some files winding up hanging around my file-system.
>
>So why did distutils make it into the Python core?

How does distutils making it into the Python distribution rule out that
somone may want to prefer RPMs?

>3. Need to support including documentation along with source for packages when
>documentation is available.

The practice in many RPMs is that the documentation is a "sub package" (see
my Python SRPM), which can be installed on your development machines, but
bypassed on the production machines...

Don't get me wrong, I'm all for documentation.  The lack of "javadoc" for
Python has really hurt though...

>But that leaves it up to the person posting the package to make distributions
>in formats for any platforms where folks might want to install their stuff, or
>to have someone *else* create those packages.  That may lead to a situation

What's wrong with allowing people to select different packaging formats
if they're available?

>where someone believes that they cannot install my package on MacOS because I
>didn't make an HQX file (or whatever) when I posted HappyDoc.

Again, my tool isn't concerned with such policies except that it allows the
user to define it.  If the user has defined a policy of "I won't accept
any package that's not in .hqx format", then they should indeed not be shown
HappyDoc if it doesn't have one.  However, it would also allow them to say
"I prefer .hqx files, but will take a distutils package if it's my only
choice".  That user then *WOULD* see HappyDoc if it had a distutils
package.

>I can go with "allow" but if we do not help then the user must specify in order
>to get useful results from the tool, and that becomes "force."

I don't follow that logic.

>So, it sounds like their tool just does the download and then they use the
>equivalent of distutils to do the installation.  Is that all we want?

For the archive server, I think it is...  That doesn't preclude the
archive client from being overly clever though...  That's what I'm going
to work on next, since I have the server in a state where it can handle
giving out such information.

Sean
-- 
 He who wonders discovers that this in itself is wonder.
                 -- M. C. Escher
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From doughellmann at bigfoot.com  Wed Feb 28 12:53:11 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Wed, 28 Feb 2001 11:53:11 GMT
Subject: CPAN functionality for python - requirements
References: <20010226214909.A10411@tummy.com> <mailman.983305271.1946.python-list@python.org>
Message-ID: <HE5n6.77745$m5.4881814@news1.rdc1.ga.home.com>

In article <mailman.983305271.1946.python-list at python.org>, "Sean
Reifschneider" <jafo at tummy.com> wrote:

> On Tue, Feb 27, 2001 at 02:49:40AM -0700, Bruce Sass wrote:
> 
>>Huh, there are tools and standards. Isn't /usr/shar/doc/package the FHS
>>place for general package specific
> 
> Where are the tools for turning doc-strings into HTML, and how are the

http://happydoc.sourceforge.net

> links between doc packages handled? 

Hrm, not sure about that one.  HappyDoc can write output to a
subdirectory inside where the source code is...

> What is the markup format for the
> doc strings?

StructuredText.

> We spent quite a while discussing the sad state of the documentation at
> our last pythoneers meeting.  Mark Lutz was shocked that people would
> actually use javadoc, but they do seem to simply becuase it was
> available and was
> "good enough".
> 
> We are missing similar functionality.

Working on it...

Doug


From doughellmann at bigfoot.com  Wed Feb 28 12:53:14 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Wed, 28 Feb 2001 11:53:14 GMT
Subject: CPAN functionality for python - requirements
References: <20010226214909.A10411@tummy.com> <mailman.983302211.26581.python-list@python.org>
Message-ID: <KE5n6.77747$m5.4881862@news1.rdc1.ga.home.com>

In article <mailman.983302211.26581.python-list at python.org>, "Sean
Reifschneider" <jafo at tummy.com> wrote:

> On Tue, Feb 27, 2001 at 06:50:52AM -0500, Doug Hellmann wrote:
>>> That seems to be where we diverge...  If somone has made an RPM of it,
>>> I'd rather have that than some files winding up hanging around my
>>> file-system.
>>
>>So why did distutils make it into the Python core?
> 
> How does distutils making it into the Python distribution rule out that
> somone may want to prefer RPMs?

It doesn't, now that I realize distutils will make RPMs.

>>3. Need to support including documentation along with source for
>>packages when
>>documentation is available.
> 
> The practice in many RPMs is that the documentation is a "sub package"
> (see my Python SRPM), which can be installed on your development
> machines, but bypassed on the production machines...
> 
> Don't get me wrong, I'm all for documentation.  The lack of "javadoc"
> for Python has really hurt though...

See http://happydoc.sourceforge.net.

>>But that leaves it up to the person posting the package to make
>>distributions in formats for any platforms where folks might want to
>>install their stuff, or to have someone *else* create those packages. 
>>That may lead to a situation
> 
> What's wrong with allowing people to select different packaging formats
> if they're available?

We still have to support *some* form of installation if the package
format you want is not available.  It sounds like the best way to handle
that is to enhance distutils to have it generate more package types as
output.

>>where someone believes that they cannot install my package on MacOS
>>because I didn't make an HQX file (or whatever) when I posted HappyDoc.
> 
> Again, my tool isn't concerned with such policies except that it allows
> the user to define it.  If the user has defined a policy of "I won't
> accept any package that's not in .hqx format", then they should indeed
> not be shown HappyDoc if it doesn't have one.  However, it would also
> allow them to say
> "I prefer .hqx files, but will take a distutils package if it's my only
> choice".  That user then *WOULD* see HappyDoc if it had a distutils
> package.
> 
>>I can go with "allow" but if we do not help then the user must specify
>>in order to get useful results from the tool, and that becomes "force."
> 
> I don't follow that logic.

If the tool says, "I'm running on a Mac therefore I want HQX files." that
may result in an artificial limit to the list of packages available,
forcing the user to install HQX files.  If the tools says, "I was told to search
for HQX files." the user has made a choice and has more responsibility to
be aware that they are only going to get HQX files and that there may be
other options for them.

Doug


From fredrik at pythonware.com  Wed Feb 28 15:21:18 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 14:21:18 GMT
Subject: CPAN functionality for python - requirements
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022623140707.15046@branagan> <20010226214909.A10411@tummy.com> <01022706564001.20069@branagan> <mailman.983302211.26581.python-list@python.org>
Message-ID: <yP7n6.15965$Qb7.2607352@newsb.telia.net>

Sean Reifschneider wrote:
> Don't get me wrong, I'm all for documentation.  The lack of "javadoc" for
> Python has really hurt though...

current CVS version:

>>> import pydoc
>>> print pydoc.__doc__
Generate Python documentation in HTML or as text for interactive use.

At the shell command line outside of Python, run "pydoc <name>" to show
documentation on something.  <name> may be the name of a Python function,
module, package, or a dotted reference to a class or function within a
module or module in a package.  Alternatively, the argument can be the
path to a Python source file.

Or, at the shell prompt, run "pydoc -k <keyword>" to search for a keyword
in the one-line descriptions of modules.

Or, at the shell prompt, run "pydoc -p <port>" to start an HTTP server
on a given port on the local machine to generate documentation web pages.

Or, at the shell prompt, run "pydoc -w <name>" to write out the HTML
documentation for a module to a file named "<name>.html".

In the Python interpreter, do "from pydoc import help" to provide online
help.  Calling help(thing) on a Python object documents the object.

Cheers /F




From bsass at freenet.edmonton.ab.ca  Tue Feb 27 10:33:29 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 02:33:29 -0700 (MST)
Subject: CPAN functionality for python - requirements
In-Reply-To: <01022623140707.15046@branagan>
Message-ID: <Pine.LNX.4.33.0102270137070.28512-100000@bms>

On Mon, 26 Feb 2001, Doug Hellmann wrote:

> On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> > On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
<...>
> It still isn't clear to me why we would want to use more than one distribution
> package format.  It appears that one argument is so that sysadmins can use
> those tools to track what is installed on the system.  But isn't that part of
> what this new thing (cyphon?) is supposed to do?  Do we want Python to require
> an external package management system on each platform where this tool is
> supported, or do we want a Python-esq tool which just works the way it is
> supposed to, in the same way, everywhere Python is available?

I want users to be able to find and install Python stuff, preferably
in a manner consistent with their native package management system.

> But I think we're arguing over a design before we've worked out requirements.
> The basic tasks of this new tool, as I see them, are:
>
> 1. Locate Python packages you want on the net.
> 2. Resolve dependencies between what you want, what you have, and what you
>    need.

There is no way an archive network can resolve dependencies because
ther is no way for the archive network to know what is installed on a
users machine (at least there had better not be!), or what packages
are available to fulfil those dependencies.  The best that can
be done is to tell the target system what is required by the pkg being
looked at.  The dependencies get resolved when the native system
decides on where it should fetch the other stuff from.

Dependencies are a big problem.  I've spoken with a local LUG member
who was complaining that Debian does not know about the stuff he gets from
CPAN, which makes mixing CPAN and debianized CPAN stuff a pain.
Debian's solution is a user level program to debianize arbitrary CPAN
modules... I want the same solution for Python modules, except I want
it done transparently.

> 3. Download the appropriate packages to give you what you want.
> 4. Install those packages.

Doesn't that mean the program needs to know about all the distros, and
all the version of those distros, and all the supported ways a user
can tweak the system.  If you don't do that, and keep it up, then your
installer is going to be breaking often (or always broken, depending
on the user or distro).

> Do we agree these are the primary features?

Both 2. and 4. are the domain of the target system.  2. can be
modified (or maybe just reworded).  4...

> I sense a consensus that the "install" part should be handled by distutils.  Is
> that right?

As long as distutils does not presume to know how to do the
installation.

> Other requirements we might lay down up front:
>
> 1. Should run on all platforms where Python runs.
> 2. Must support mirrors on the server side.
> 3. Need to include documentation along with source for packages.
>
> The features related to dependencies and downloads could be handled by a
> platform-specific packaging system, but integrating with all of the different
> options on all of the different platforms where Python runs increases the
> complexity of the new tool.

Downloading is independent of the native package management system.
Complexity is only increased if the new tool thinks it should handle
platform/OS/distro specific issues.

> (How do we handle RPMs on MacOS?  How do we handle
> HQX files on Win32?  What about for Pippy, where files as such might not even
> be appropriate?  Does Jython have any special requirements?)  Do we deal with
> this complexity by deciding on a per-platform basis what format to use, or do
> we force the user to specify the format they want downloaded?

You don't.  You give the data to the target system and let it worry
about the complexity of packaging.

There is no reason for the archive network to know anything except
where to send the data!

> If we're modeling what we're doing based on CPAN, maybe we should look at
> that design.  What format does CPAN use when downloading Perl packages?  What
> features does CPAN have that we want?  What does it not have that we want?

Pros:
don't know (long strings of punctuation and symbols give me a headache ;).

Cons:
Dependencies problems when mixed with modules fetched by other methods.


- Bruce




From doughellmann at home.com  Tue Feb 27 12:28:32 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Tue, 27 Feb 2001 06:28:32 -0500
Subject: CPAN functionality for python - requirements
In-Reply-To: <Pine.LNX.4.33.0102270137070.28512-100000@bms>
References: <Pine.LNX.4.33.0102270137070.28512-100000@bms>
Message-ID: <01022706484400.20069@branagan>

On Tue, 27 Feb 2001, Bruce Sass wrote:
> On Mon, 26 Feb 2001, Doug Hellmann wrote:
> 
> > On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> > > On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
> <...>
> > It still isn't clear to me why we would want to use more than one distribution
> > package format.  It appears that one argument is so that sysadmins can use
> > those tools to track what is installed on the system.  But isn't that part of
> > what this new thing (cyphon?) is supposed to do?  Do we want Python to require
> > an external package management system on each platform where this tool is
> > supported, or do we want a Python-esq tool which just works the way it is
> > supposed to, in the same way, everywhere Python is available?
> 
> I want users to be able to find and install Python stuff, preferably
> in a manner consistent with their native package management system.
> 
> > But I think we're arguing over a design before we've worked out requirements.
> > The basic tasks of this new tool, as I see them, are:
> >
> > 1. Locate Python packages you want on the net.
> > 2. Resolve dependencies between what you want, what you have, and what you
> >    need.
> 
> There is no way an archive network can resolve dependencies because

No, the network doesn't.  But the network by itself isn't going to do most of
these things.  The client app would need to resolve the dependencies.

If I tell the client to find and download PmwContribD, it needs to figure out
that I also need Pmw (and perhaps a minimum version) and either suggest that I
go ahead and install it or fail with a message telling me to install that first
depending on the command line arguments given.

> ther is no way for the archive network to know what is installed on a
> users machine (at least there had better not be!), or what packages
> are available to fulfil those dependencies.  The best that can
> be done is to tell the target system what is required by the pkg being
> looked at.  The dependencies get resolved when the native system
> decides on where it should fetch the other stuff from.

It sounds like the knows how to find a particular package, but getting it
installed is your problem.  We can make it smarter than that.

> Dependencies are a big problem.  I've spoken with a local LUG member
> who was complaining that Debian does not know about the stuff he gets from
> CPAN, which makes mixing CPAN and debianized CPAN stuff a pain.
> Debian's solution is a user level program to debianize arbitrary CPAN
> modules... I want the same solution for Python modules, except I want
> it done transparently.

I can see that as somewhat desirable, and it could be implemented, but do
you make the cyphon client download the right package format or convert
a standard distro into the right format?  Doesn't the user have to specify a
package type?

> > 3. Download the appropriate packages to give you what you want.
> > 4. Install those packages.
> 
> Doesn't that mean the program needs to know about all the distros, and
> all the version of those distros, and all the supported ways a user
> can tweak the system.  If you don't do that, and keep it up, then your
> installer is going to be breaking often (or always broken, depending
> on the user or distro).
> 
> > Do we agree these are the primary features?
> 
> Both 2. and 4. are the domain of the target system.  2. can be
> modified (or maybe just reworded).  4...

What I want is a single app I can use that will do all of those things.  I want
to type one command and have it find and install software on my system that
will allow me to use some Python package that I don't have yet.  That app would
have a client side and server side, but as a user I don't care about the server
side unless I want to specify a specific mirror to use by default.

> > I sense a consensus that the "install" part should be handled by distutils.  Is
> > that right?
> 
> As long as distutils does not presume to know how to do the
> installation.

How can it do the install if it doesn't know how?

It sounds like you do *not* want distutils to do the install, and that you
cyphon to download/create a platform-specific package to be installed.  Since we
*have* distutils as an ostensibly standard installation system for Python
modules, shouldn't we use it?

> > Other requirements we might lay down up front:
> >
> > 1. Should run on all platforms where Python runs.
> > 2. Must support mirrors on the server side.
> > 3. Need to include documentation along with source for packages.
> >
> > The features related to dependencies and downloads could be handled by a
> > platform-specific packaging system, but integrating with all of the different
> > options on all of the different platforms where Python runs increases the
> > complexity of the new tool.
> 
> Downloading is independent of the native package management system.

No, it isn't.  Even if *all* the tool does is download things I told it to find
and write them to the current directory, it has to know what format I like or
it may download (for example) debian packages for my Mac -- and then I won't
get useful results from the tool.

> Complexity is only increased if the new tool thinks it should handle
> platform/OS/distro specific issues.

It *must* handle those issues to be useful.

> > (How do we handle RPMs on MacOS?  How do we handle
> > HQX files on Win32?  What about for Pippy, where files as such might not even
> > be appropriate?  Does Jython have any special requirements?)  Do we deal with
> > this complexity by deciding on a per-platform basis what format to use, or do
> > we force the user to specify the format they want downloaded?
> 
> You don't.  You give the data to the target system and let it worry
> about the complexity of packaging.

What is the "target system," the OS?
 
> There is no reason for the archive network to know anything except
> where to send the data!

Right, but we do need a client so that users have a way to interact with the
network.

Wait -- maybe that's where the disconnect is!  I'm describing features of the
system from the client-side view, assuming the network would provide services
to support those features.  It sounds like you are describing features of the
network.  We should determine the requirements from the user perspective first,
and design something to meet those requirements.  Designs for the network
services will follow from the requirements.

> > If we're modeling what we're doing based on CPAN, maybe we should look at
> > that design.  What format does CPAN use when downloading Perl packages?  What
> > features does CPAN have that we want?  What does it not have that we want?
> 
> Pros:
> don't know (long strings of punctuation and symbols give me a headache ;).

Heh, tell me about it.

I can't remember now whether CPAN did installations or just downloaded stuff
and made you figure out how to install it.

> Cons:
> Dependencies problems when mixed with modules fetched by other methods.

Right, so we need a single way of determining, maintaining, and
resolving dependencies.

Doug



From mwa at gate.net  Tue Feb 27 15:23:33 2001
From: mwa at gate.net (Mark W. Alexander)
Date: Tue, 27 Feb 2001 09:23:33 -0500 (EST)
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <01022706484400.20069@branagan>
Message-ID: <Pine.A41.4.21.0102270918130.53710-100000@dakota.gate.net>

On Tue, 27 Feb 2001, Doug Hellmann wrote:
> Subject: [Distutils] Re: CPAN functionality for python - requirements
> 
> I can't remember now whether CPAN did installations or just downloaded stuff
> and made you figure out how to install it.

(My understanding of) CPAN is that it's a repository of things that
install a standard way. You download what you want and

	perl Makefile.pl
	make install

Mark





From efm at tummy.com  Tue Feb 27 17:30:13 2001
From: efm at tummy.com (Evelyn Mitchell)
Date: Tue, 27 Feb 2001 09:30:13 -0700
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <Pine.A41.4.21.0102270918130.53710-100000@dakota.gate.net>; from mwa@gate.net on Tue, Feb 27, 2001 at 09:23:33AM -0500
References: <01022706484400.20069@branagan> <Pine.A41.4.21.0102270918130.53710-100000@dakota.gate.net>
Message-ID: <20010227093012.C4181@tummy.com>

On Tue, Feb 27, 2001 at 09:23:33AM -0500, Mark W. Alexander wrote:
> (My understanding of) CPAN is that it's a repository of things that
> install a standard way. You download what you want and
> 
> 	perl Makefile.pl
> 	make install

CPAN is a lot closer to appget with debian than that.
Yes, there is a standard perl Makefile, which looks through your
system for quirks, and creates a makefile which can work.

But it will also discover and resolve dependences in your perl
site-packages, and automatically fetch them from your closest
CPAN archive.

Regards,
Evelyn Mitchell
efm at tummy.com



From akuchlin at mems-exchange.org  Tue Feb 27 17:42:58 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: Tue, 27 Feb 2001 11:42:58 -0500
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <20010227093012.C4181@tummy.com>; from efm@tummy.com on Tue, Feb 27, 2001 at 09:30:13AM -0700
References: <01022706484400.20069@branagan> <Pine.A41.4.21.0102270918130.53710-100000@dakota.gate.net> <20010227093012.C4181@tummy.com>
Message-ID: <20010227114258.E17362@ute.cnri.reston.va.us>

On Tue, Feb 27, 2001 at 09:30:13AM -0700, Evelyn Mitchell wrote:
>On Tue, Feb 27, 2001 at 09:23:33AM -0500, Mark W. Alexander wrote:
>> (My understanding of) CPAN is that it's a repository of things that
>> install a standard way. You download what you want and
>> 
>> 	perl Makefile.pl
>> 	make install
>
>CPAN is a lot closer to appget with debian than that.
>Yes, there is a standard perl Makefile, which looks through your
>system for quirks, and creates a makefile which can work.

Distutils aims to fix the installation problem, providing a consistent
way to build a module once you've downloaded it.  Discussion of
indexing, dependencies, finding modules, &c., is the Catalog-SIG's
job, and should be discussed there; they're off-topic for the
Distutils SIG.  

--amk




From jafo at tummy.com  Tue Feb 27 19:08:03 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 11:08:03 -0700
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <20010227093012.C4181@tummy.com>; from efm@tummy.com on Tue, Feb 27, 2001 at 09:30:13AM -0700
References: <01022706484400.20069@branagan> <Pine.A41.4.21.0102270918130.53710-100000@dakota.gate.net> <20010227093012.C4181@tummy.com>
Message-ID: <20010227110803.Z1781@tummy.com>

On Tue, Feb 27, 2001 at 09:30:13AM -0700, Evelyn Mitchell wrote:
>But it will also discover and resolve dependences in your perl
>site-packages, and automatically fetch them from your closest
>CPAN archive.

Not according to my tests the night before last.  I did a test CPAN
install of "News::Newsrc", which failed because the "make test" was
failing.  I then installed the "Set::BitSet" (?  Something like that)
module and tried News::Newsrc again and it worked...

Maybe this was just a fluke and News::Newsrc is the exception and/or
isn't used enough that people have gotten the prereqs right yet.  If
anyone knows for sure, I'm curious.

Sean
-- 
 The dude abides.
                 -- _The_Big_Lebowski_
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From phrxy at csv.warwick.ac.uk  Tue Feb 27 19:29:18 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Tue, 27 Feb 2001 18:29:18 +0000 (GMT)
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <20010227110803.Z1781@tummy.com>
Message-ID: <Pine.SOL.4.30.0102271823190.22762-100000@mimosa.csv.warwick.ac.uk>

Followup to Catalog-sig only (pine doesn't seem to let me set that
in the headers...)?  I don't think this has much to do with Distutils.

On Tue, 27 Feb 2001, Sean Reifschneider wrote:

> On Tue, Feb 27, 2001 at 09:30:13AM -0700, Evelyn Mitchell wrote:
> >But it will also discover and resolve dependences in your perl
> >site-packages, and automatically fetch them from your closest
> >CPAN archive.
>
> Not according to my tests the night before last.  I did a test CPAN
> install of "News::Newsrc", which failed because the "make test" was
> failing.  I then installed the "Set::BitSet" (?  Something like that)
> module and tried News::Newsrc again and it worked...
[...]

Did you use CPAN.pm?  CPAN the archive doesn't do dependencies.
Dependencies are listed in the Makefile.PL (BTW, I have no idea why the
capitalised PL), and after the module CPAN.pm downloads what you asked
for, it checks that to see if there are dependencies, then downloads them
if you don't already have them installed.


John




From jafo at tummy.com  Wed Feb 28 05:17:51 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 21:17:51 -0700
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <Pine.SOL.4.30.0102271823190.22762-100000@mimosa.csv.warwick.ac.uk>; from phrxy@csv.warwick.ac.uk on Tue, Feb 27, 2001 at 06:29:18PM +0000
References: <20010227110803.Z1781@tummy.com> <Pine.SOL.4.30.0102271823190.22762-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <20010227211751.C24378@tummy.com>

On Tue, Feb 27, 2001 at 06:29:18PM +0000, John J. Lee wrote:
>Did you use CPAN.pm?  CPAN the archive doesn't do dependencies.

Yes, the latest version obtaied from CPAN as I understand it...

Sean
-- 
 People who interview themselves shouldn't criticize writing styles.
                 -- John Bentley, Programming Pearls
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From phrxy at csv.warwick.ac.uk  Tue Feb 27 19:16:39 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Tue, 27 Feb 2001 18:16:39 +0000 (GMT)
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <Pine.A41.4.21.0102270918130.53710-100000@dakota.gate.net>
Message-ID: <Pine.SOL.4.30.0102271436480.3740-100000@mimosa.csv.warwick.ac.uk>

On Tue, 27 Feb 2001, Mark W. Alexander wrote:

> On Tue, 27 Feb 2001, Doug Hellmann wrote:
[...]
> > I can't remember now whether CPAN did installations or just downloaded stuff
> > and made you figure out how to install it.
>
> (My understanding of) CPAN is that it's a repository of things that
> install a standard way. You download what you want and
>
> 	perl Makefile.pl
> 	make install
[...]

There are two different 'CPAN's:

1. The archive itself -- a bunch of files in .tar.gz format, READMEs, etc.
It uses a mirroring system that sets itself up on your first visit to
point to a nearby mirror, and subsequently takes you there automatically.

2. The CPAN module, CPAN.pm, which can be used interactively or through
its API.  It lets you search for, download, find dependencies (and
download again if required), build and install modules.  [Activestate's
windows distribution has its own binary package system, PPM, but CPAN (the
module) works on windows as long as you have libwww-perl.  I don't know
what fraction of modules including C code compile without trouble on
windows.]


John




From bsass at freenet.edmonton.ab.ca  Tue Feb 27 15:35:57 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 07:35:57 -0700 (MST)
Subject: CPAN functionality for python - requirements
In-Reply-To: <01022706484400.20069@branagan>
Message-ID: <Pine.LNX.4.33.0102270530470.28590-100000@bms>

On Tue, 27 Feb 2001, Doug Hellmann wrote:
> On Tue, 27 Feb 2001, Bruce Sass wrote:
> > On Mon, 26 Feb 2001, Doug Hellmann wrote:
> > > On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> > > > On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
> > <...>
> > There is no way an archive network can resolve dependencies because
>
> No, the network doesn't.  But the network by itself isn't going to do most of
> these things.  The client app would need to resolve the dependencies.
>
> If I tell the client to find and download PmwContribD, it needs to figure out
> that I also need Pmw (and perhaps a minimum version) and either suggest that I
> go ahead and install it or fail with a message telling me to install that first
> depending on the command line arguments given.

You are easily pleased.  :)

> > ther is no way for the archive network to know what is installed on a
> > users machine (at least there had better not be!), or what packages
> > are available to fulfil those dependencies.  The best that can
> > be done is to tell the target system what is required by the pkg being
> > looked at.  The dependencies get resolved when the native system
> > decides on where it should fetch the other stuff from.
>
> It sounds like the knows how to find a particular package, but getting it
> installed is your problem.  We can make it smarter than that.

Not "your", the native package management system's (and they have
been doing it a relatively long time).

> > Dependencies are a big problem.  I've spoken with a local LUG member
> > who was complaining that Debian does not know about the stuff he gets from
> > CPAN, which makes mixing CPAN and debianized CPAN stuff a pain.
> > Debian's solution is a user level program to debianize arbitrary CPAN
> > modules... I want the same solution for Python modules, except I want
> > it done transparently.
>
> I can see that as somewhat desirable, and it could be implemented, but do
> you make the cyphon client download the right package format or convert
> a standard distro into the right format?  Doesn't the user have to specify a
> package type?

I prefer building from a generic package to a system specific package
(QC is reduced to checking the contents of the generic source package
and having a good package making script, hopefully provided by the
same group of people who control the format of the package), but there
is no reason for not supporting both (except maybe room to store it
all), and in some cases it would be desirable.

Sure the user has to specify a package type, _if_ they want a specific
type of package, that info can be included in the name of the file
being requested (built up from the name of the python piece, a client
id, etc.)

> > > 3. Download the appropriate packages to give you what you want.
> > > 4. Install those packages.
> >
> > Doesn't that mean the program needs to know about all the distros, and
> > all the version of those distros, and all the supported ways a user
> > can tweak the system.  If you don't do that, and keep it up, then your
> > installer is going to be breaking often (or always broken, depending
> > on the user or distro).
> >
> > > Do we agree these are the primary features?
> >
> > Both 2. and 4. are the domain of the target system.  2. can be
> > modified (or maybe just reworded).  4...
>
> What I want is a single app I can use that will do all of those things.  I want
> to type one command and have it find and install software on my system that
> will allow me to use some Python package that I don't have yet.  That app would
> have a client side and server side, but as a user I don't care about the server
> side unless I want to specify a specific mirror to use by default.
>
> > > I sense a consensus that the "install" part should be handled by distutils.  Is
> > > that right?
> >
> > As long as distutils does not presume to know how to do the
> > installation.
>
> How can it do the install if it doesn't know how?
>
> It sounds like you do *not* want distutils to do the install, and that you
> cyphon to download/create a platform-specific package to be installed.  Since we
> *have* distutils as an ostensibly standard installation system for Python
> modules, shouldn't we use it?

I would not use it if it means bypassing the native package management
system, and would be willing to wait for "official" packages rather
than use third-party packages.

Best of all scenarios would be if all the Python programmers used the
same system I did and spit out packages in my favorite format, next
best is having a package built by someone who knows my system just as
well or better than I, then their proxy in the form of a scripted
build, followed by someone I don't know building the package, a
tarball or some other package format I can handle.

The first is a dream, the second could happen now and then if I was
lucky, so I'll try for my third choice (and it sounds like distutils
can handle it).

> > > Other requirements we might lay down up front:
> > >
> > > 1. Should run on all platforms where Python runs.
> > > 2. Must support mirrors on the server side.
> > > 3. Need to include documentation along with source for packages.
> > >
> > > The features related to dependencies and downloads could be handled by a
> > > platform-specific packaging system, but integrating with all of the different
> > > options on all of the different platforms where Python runs increases the
> > > complexity of the new tool.
> >
> > Downloading is independent of the native package management system.
>
> No, it isn't.  Even if *all* the tool does is download things I told it to find
> and write them to the current directory, it has to know what format I like or
> it may download (for example) debian packages for my Mac -- and then I won't
> get useful results from the tool.

Build up the name of the download from the name of the package and
system specific information, just like is done with the names of .deb
and .rpm packages.

> > Complexity is only increased if the new tool thinks it should handle
> > platform/OS/distro specific issues.
>
> It *must* handle those issues to be useful.

Isn't that is a lot of duplicated work.

> > > (How do we handle RPMs on MacOS?  How do we handle
> > > HQX files on Win32?  What about for Pippy, where files as such might not even
> > > be appropriate?  Does Jython have any special requirements?)  Do we deal with
> > > this complexity by deciding on a per-platform basis what format to use, or do
> > > we force the user to specify the format they want downloaded?
> >
> > You don't.  You give the data to the target system and let it worry
> > about the complexity of packaging.
>
> What is the "target system," the OS?

The system the python package is being downloaded to.

> > There is no reason for the archive network to know anything except
> > where to send the data!
>
> Right, but we do need a client so that users have a way to interact with the
> network.
>
> Wait -- maybe that's where the disconnect is!  I'm describing features of the
> system from the client-side view, assuming the network would provide services
> to support those features.  It sounds like you are describing features of the
> network.  We should determine the requirements from the user perspective first,
> and design something to meet those requirements.  Designs for the network
> services will follow from the requirements.

I see them as two pieces of the same puzzle.  The third, and most
important, is the data structure used to represent the python package.

> > > If we're modeling what we're doing based on CPAN, maybe we should look at
> > > that design.  What format does CPAN use when downloading Perl packages?  What
> > > features does CPAN have that we want?  What does it not have that we want?
> >
> > Pros:
> > don't know (long strings of punctuation and symbols give me a headache ;).
>
> Heh, tell me about it.
>
> I can't remember now whether CPAN did installations or just downloaded stuff
> and made you figure out how to install it.
>
> > Cons:
> > Dependencies problems when mixed with modules fetched by other methods.
>
> Right, so we need a single way of determining, maintaining, and
> resolving dependencies.

Or a way of enabling the native tools so they can do it.  I would not
want to accept the job of keeping up with changing packaging standards
for <how many?> different systems, and don't expect that anyone else
would want to (or be able to keep it up).


- Bruce





From jafo at tummy.com  Tue Feb 27 20:53:51 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 12:53:51 -0700
Subject: CPAN functionality for python - requirements
In-Reply-To: <01022706484400.20069@branagan>; from doughellmann@home.com on Tue, Feb 27, 2001 at 06:28:32AM -0500
References: <Pine.LNX.4.33.0102270137070.28512-100000@bms> <01022706484400.20069@branagan>
Message-ID: <20010227125351.H1781@tummy.com>

On Tue, Feb 27, 2001 at 06:28:32AM -0500, Doug Hellmann wrote:
>It sounds like the knows how to find a particular package, but getting it
>installed is your problem.  We can make it smarter than that.

That's more a distutils problem than an archive network problem.  My
vision is that the archive network simply gives the client package
information, which the client then can deal with as they see fit.
Ideally, once downloaded, the archive network client would then just
tell the package "Install thyself", but there will probably have to
be some additional smarts in the client than that.

I expect these additional smarts to be little more than scripts:

	mkdir tmpdirectory
	cd tmpdirectory
	tar xpvfz archivefile
	cd archivename-with-.tar.gz-removed
	python setup.py

That sort of thing...

>you make the cyphon client download the right package format or convert
>a standard distro into the right format?  Doesn't the user have to specify a
>package type?

That decision is up to the client.  If the client has the smarts to turn
a distutils package into an RPM, then the client would list it's preferred
format as a distutils package and it would handle the rest.  If it can't,
you can either select an RPM, or fall back to a distutils package.

>No, it isn't.  Even if *all* the tool does is download things I told it to find
>and write them to the current directory, it has to know what format I like or
>it may download (for example) debian packages for my Mac -- and then I won't
>get useful results from the tool.

That's exactly my target.

>> Complexity is only increased if the new tool thinks it should handle
>> platform/OS/distro specific issues.
>
>It *must* handle those issues to be useful.

The schema lists platform version, name, architecture, and format.  So, it
can handle "RedHat 7.0 ix86 in RPM format" if we wish.  Any of those
fields can be "*", which will probably be the case in source distutils
package cases.

>Wait -- maybe that's where the disconnect is!  I'm describing features of the
>system from the client-side view, assuming the network would provide services
>to support those features.  It sounds like you are describing features of the

The archive network client I believe should have a step which is "install
this", but I believe that largely needs to be abstracted from the client.
Scripts that are used to control the "I've got this thing, here's how to
install it".  In the case of a tar file (not a distutils package) we
probably want to extract it then drop the user into a shell saying
"Check out the README and INSTALL files, do the install, then type
'exit' when done".  Automating it to the point that we can...

>network.  We should determine the requirements from the user perspective first,
>and design something to meet those requirements.  Designs for the network
>services will follow from the requirements.

That's why I've been posting my schema and trying to decide what my vision
of the client interaction is.  If it doesn't meet the user needs, then
it needs to be re-worked.  However, I believe it has the flexability to
handle the policy decisions that have been discussed here.

>I can't remember now whether CPAN did installations or just downloaded stuff
>and made you figure out how to install it.

It does a "make test" and a "make install".  There is a convention that CPAN
packages will have a "make test", even if it doesn't do much.  It's a great
idea...

With Python modules we could almost make some of the dependencies automatic.
If we try to import the module and it throws an exception on an import, we
know we've hit a dependency...  That'd be cute...

>Right, so we need a single way of determining, maintaining, and
>resolving dependencies.

One could say that the appropriate way to find out if a required module
is available is to try importing it.

I firmly believe that we have to allow not only system-wide installs,
but also user installs (into their home directory, for example).

Sean
-- 
 Come see the violence inherent in the system!
 Help!  Help!  I'm being repressed!  -- Monty Python and the Holy Grail
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From bsass at freenet.edmonton.ab.ca  Tue Feb 27 23:37:21 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 15:37:21 -0700 (MST)
Subject: CPAN functionality for python - requirements
In-Reply-To: <20010227125351.H1781@tummy.com>
Message-ID: <Pine.LNX.4.33.0102271450220.30946-100000@bms>

I'm a little concerned about the privacy aspect of this...

> That decision is up to the client.  If the client has the smarts to turn
> a distutils package into an RPM, then the client would list it's preferred
> format as a distutils package and it would handle the rest.  If it can't,
> you can either select an RPM, or fall back to a distutils package.

Why should the client need to "list" anything.

client: what do you have?
server: this(deb,rh-5.rpm,rh-6.2.rpm,rh-7.rpm) that() other(hqx)
client: send me this(deb), that(mandrake.rpm), other()
server: sorry don't have that(mandrake.rpm)
client: send me that(deb)
server: sorry don't have that(deb)
client: send me that()

This is how I would handle a curious server...
client: what do you have?
server: what format do you want?
<click>

In other words... the server is just that - a server - it does not ask
questions, it does not keep track of any preferences, it serves up
files.

This could work...
client: only show me (deb) and ()
client: what do you have?
...

The server should NOT be usable as a tool to track Python users and
their habits, and making it do so should require a conscious effort on
the operators part (so there is no opportunity for the operator to
say, "I don't track you, that's just how the software works").


- Bruce




From doughellmann at home.com  Wed Feb 28 03:03:43 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Tue, 27 Feb 2001 21:03:43 -0500
Subject: CPAN functionality for python - requirements
In-Reply-To: <Pine.LNX.4.33.0102271450220.30946-100000@bms>
References: <Pine.LNX.4.33.0102271450220.30946-100000@bms>
Message-ID: <01022721090603.20069@branagan>

On Tue, 27 Feb 2001, Bruce Sass wrote:
> I'm a little concerned about the privacy aspect of this...
> 
> > That decision is up to the client.  If the client has the smarts to turn
> > a distutils package into an RPM, then the client would list it's preferred
> > format as a distutils package and it would handle the rest.  If it can't,
> > you can either select an RPM, or fall back to a distutils package.
> 
> Why should the client need to "list" anything.
> 
> client: what do you have?
> server: this(deb,rh-5.rpm,rh-6.2.rpm,rh-7.rpm) that() other(hqx)
> client: send me this(deb), that(mandrake.rpm), other()
> server: sorry don't have that(mandrake.rpm)
> client: send me that(deb)
> server: sorry don't have that(deb)
> client: send me that()

I'm not sure why the "what do you have" question is needed.  The "send me
that(mandrake.rpm)" interaction is what we want.

The client requests packages in the formats prefered.  A standard source format
should be available for all packages so that everybody can get every package in
some form.  If the standard source format contains distutil-enabled scripts,
platform specific distribution files could be generated from the source.

> In other words... the server is just that - a server - it does not ask
> questions, it does not keep track of any preferences, it serves up
> files.

No, of course not.

> This could work...
> client: only show me (deb) and ()
> client: what do you have?

Right, that would be a bit more efficient than specifying a format for every
package for every request.

> The server should NOT be usable as a tool to track Python users and
> their habits, and making it do so should require a conscious effort on
> the operators part (so there is no opportunity for the operator to
> say, "I don't track you, that's just how the software works").

The server is likely to be a cgi, which by its nature means requests may be
written to a log file.  Should that be disallowed?

Doug



From jafo at tummy.com  Wed Feb 28 04:41:38 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 20:41:38 -0700
Subject: CPAN functionality for python - requirements
In-Reply-To: <01022721090603.20069@branagan>; from doughellmann@home.com on Tue, Feb 27, 2001 at 09:03:43PM -0500
References: <Pine.LNX.4.33.0102271450220.30946-100000@bms> <01022721090603.20069@branagan>
Message-ID: <20010227204138.C1781@tummy.com>

On Tue, Feb 27, 2001 at 09:03:43PM -0500, Doug Hellmann wrote:
>> > That decision is up to the client.  If the client has the smarts to turn
>> > a distutils package into an RPM, then the client would list it's preferred
>> > format as a distutils package and it would handle the rest.  If it can't,
>> > you can either select an RPM, or fall back to a distutils package.
>> 
>> Why should the client need to "list" anything.

If you mean "why would it need to send the list of preferred package
formats to the server", it would do that so that the server could return
the list of specific entries that the client wants.  Currently, the
implementation is that the client can ask for a specific package format,
and can ask again if there isn't one available, or it can ask the server
to list all of them that it has, and then discard the ones it does't want.

If this is a privacy concern, don't use it, eh?  We can only do so much
if you want to be tight-lipped:  "I want a package, but I don't want to
say what it is."  "Ok."

>I'm not sure why the "what do you have" question is needed.  The "send me
>that(mandrake.rpm)" interaction is what we want.

If you know exactly the versions, package formats, etc which are available?
One could say that if you expect the client to know all this, then you might
as well expect them to know where to get the thing...

>The server is likely to be a cgi, which by its nature means requests may be
>written to a log file.  Should that be disallowed?

A user is more than welcome to use an anonimizer for connecting to the
archive.

Sean
-- 
 A computer is like an Old Testament god, with a lot of rules and no mercy.
                -- Joseph Campbell
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From jack at oratrix.nl  Wed Feb 28 10:13:51 2001
From: jack at oratrix.nl (Jack Jansen)
Date: Wed, 28 Feb 2001 10:13:51 +0100
Subject: [Distutils] Re: CPAN functionality for python - requirements 
In-Reply-To: Message by Doug Hellmann <doughellmann@home.com> ,
	     Tue, 27 Feb 2001 21:03:43 -0500 , <01022721090603.20069@branagan> 
Message-ID: <20010228091351.C4A41371690@snelboot.oratrix.nl>

> I'm not sure why the "what do you have" question is needed.  The "send me
> that(mandrake.rpm)" interaction is what we want.

Well, one of the things I'd like to do is integrate all this in MacPython IDE, 
and the same probably holds for PythonWin and Idle. Then the "what do you 
have" is pretty useful for building menus and such.
--
Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
Jack.Jansen at oratrix.com | ++++ if you agree copy these lines to your sig ++++
www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 





From Paul.Moore at uk.origin-it.com  Wed Feb 28 10:59:38 2001
From: Paul.Moore at uk.origin-it.com (Moore, Paul)
Date: Wed, 28 Feb 2001 09:59:38 -0000
Subject: [Distutils] Re: CPAN functionality for python - requirements
Message-ID: <714DFA46B9BBD0119CD000805FC1F53B01B5AD44@UKRUX002.rundc.uk.origin-it.com>

From: Sean Reifschneider [mailto:jafo at tummy.com]
> On Tue, Feb 27, 2001 at 09:30:13AM -0700, Evelyn Mitchell wrote:
> >But it will also discover and resolve dependences in your perl
> >site-packages, and automatically fetch them from your closest
> >CPAN archive.
> 
> Not according to my tests the night before last.  I did a test CPAN
> install of "News::Newsrc", which failed because the "make test" was
> failing.  I then installed the "Set::BitSet" (?  Something like that)
> module and tried News::Newsrc again and it worked...
> 
> Maybe this was just a fluke and News::Newsrc is the exception and/or
> isn't used enough that people have gotten the prereqs right yet.  If
> anyone knows for sure, I'm curious.

There are basically a number of aspects to "CPAN", which need separating
out.

MakeMaker
---------

This is a perl module which implements a build process. You write a
Makefile.PL, which calls MakeMaker defining the module's structure and
metadata. Thanks to MakeMaker, the process for building a Perl module is
(nearly always) simply

    perl Makefile.PL
    make
    make test      <-- optional, but pretty much standard - runs module unit
tests
    make install   <-- installs the module

This is, in both concept and functionality, almost identical to Distutils.
There are some areas in which distutils is better (building of
platform-specific installers, for instance) and some where MakeMaker is
better (it hooks into a very standard structure for modules, generated by
thye h2xs program, which practically forces module authors to write test
suites and documentation in a standard format), but these are details.

We can consider this covered. (Although the distutils-sig could still
usefully learn from MakeMaker).

The system of FTP sites and mirrors
-----------------------------------

Frankly, this is nothing special. It has some nice features (automated
uploads for module authors, plus quite nice indexing and server multiplexing
features), but it isn't rocket science as far as I know. We could quite
happily start with a simple FTP site for this (that's what CPAN did - the
mirroring came later as popularity grew).

CPAN.pm
-------

This is a Perl module, which automates the process of downloading and
installing modules. I don't use this personally, for a number of reasons.
Frankly, I find that manual downloading and running the 4 lines noted above
is fine.

It relies for things like dependency tracking on metadata added into the
Makefile.PL which is not necessary for the straight 4-line build above. As
such, the level to which that metadata is added by module authors is
variable (to be polite). In practice, I wouldn't rely on it - accurate
dependency data seems to be the exception rather than the rule.

I *don't* regard CPAN.pm to be important to the overall CPAN "phenomenon".
But some people like it. Writing something "at least as good as" CPAN.pm
shouldn't be hard in Python - not least because the standard library is rich
enough that things like FTP client support is available out of the box
(whereas CPAN.pm won't work until you manually install libnet, and possibly
some other modules, I forget which...)

But writing a "perfect" utility for automated download-and-install, with
dependency tracking, etc etc, is going to be VERY HARD. Don't get misled -
Perl doesn't have such a beast. And We won't even have what Perl has if we
focus on perfection rather than practicality.

The h2xs program
----------------

This is VERY important. The idea is that when you write a Perl module,
either pure perl or a C (XS) extension, you run h2xs first, to generate a
template build directory. It automatically includes

    * The perl module, with some basic template code and embedded POD
      documentation
    * The XS extension, with template code (if requested)
    * A Makefile.PL shell
    * A basic test script - all it does is test that the module loads,
      but it includes a placeholder for your own tests

Essentially, h2xs forces a standard structure on all Perl modules. This is
important for Perl, where modules have to conform to some standards in order
to work at all. However, it brings HUGE benefits in standardisation of all
the "other" parts of the process (documentation, tests, etc).

Python is at a disadvantage here, precisely because writing a Python module
involves so little in the way of a specific structure. So people will likely
rebel against having a structure "imposed"...

A social structure
------------------

This is a bit of a chicken and egg issue, but Perl developers expect to
write modules using h2xs and MakeMaker, they expect to write tests (even if
they are minimal), they expect to fill in the sections in the POD
documentation, and they expect to submit their modules to CPAN. So this all
"just works".

Interestingly, developers probably don't "expect" to have to include
dependency information, and hence many don't - resulting in the problems you
hit. But then again, Perl users don't "expect" to be totally shielded from
dependency issues.

Python is VERY far behind here. This is a maturity issue - distutils is
still (relatively) new, and so there are LOTS of packages which don't come
with a setup.py yet. Often, adding one isn't hard, but it is still to
happen. And when you are distributing a pure python module, as a single .py
file, it's hard to see the benefit of changing that into a .tar.gz file
containing the module, plus a setup.py. (Once you start adding test suites
and documentation, the point of the whole archive bit is clearer, but we're
not even close to that stage yet).

Things are looking better with Python 2.1, though. Included with 2.1, it
looks like there will be a standard unit testing module, and a new "pydoc"
package, which will extract documentation from module docstrings. So the
infrastructure will be there, it just becomes a case of getting developers
to use it consistently. Distutils can help with this, by "just working" if
people follow a standard structure.

Sorry, this wasn't meant to get so long...
Paul.



From robin at alldunn.com  Wed Feb 28 17:50:03 2001
From: robin at alldunn.com (Robin Dunn)
Date: Wed, 28 Feb 2001 08:50:03 -0800
Subject: The KISS principle [ Re: CPAN functionality for python - requirements ]
References: <714DFA46B9BBD0119CD000805FC1F53B01B5AD44@UKRUX002.rundc.uk.origin-it.com>
Message-ID: <08af01c0a1a6$800b2130$0100a8c0@Rogue>

>
> But writing a "perfect" utility for automated download-and-install, with
> dependency tracking, etc etc, is going to be VERY HARD. Don't get misled -
> Perl doesn't have such a beast. And We won't even have what Perl has if we
> focus on perfection rather than practicality.

Agreed.  There is an old saying about shooting for the stars to reach the
moon.  Some of the ideas and proposals for all this seem to be shooting for
distant galaxies and are unable to even get off the ground.

IMHO we should start with something simple that allows for some of the
ultimate goals, and that can be built upon later to get to more of the
goals.

Here is what I think the simple version needs:

1. A standard place where a file containing a list of mirrors and maybe a
bit of mirror meta-data can be fetched from.

2. A python module to parse that mirror data into a meaningful data
structure such as a dictionary or a list of instance objects of some type.

3. At every site in the mirrors file there will be a file containing
meta-data about all the packages in the archive.

4. A python module to parse the package meta-data into something the client
code can easily use.  (The meta-data format is probably XML, but hiding it
like this means the tool developer doesn't need to know or care what format
the file is in.  Ditto for the mirror list.)

5. An automated way to extract at least most if not all of the package
meta-data from a Distutils based package and upload the meta-data, the
sdist, and any bdists that the developer has made to the archive.  This
could be a new command added to Distutils, with a bit of functionality on
the server side for receiving the files, moving them into the "right place"
in the archive, and updating the package meta-data file.

6. Define a file format and supporting Python code for tracking which
packages and versions have been fetched from the network and installed.
Again, this should probably be a function of Distutils so it will also catch
the cases where you downloaded some package not in the archive network and
ran "python setup.py install" yourself.

That's it.  With just that bit in place intelligent clients could be written
as command-line or GUI tools, or even a web-based interface.  They simply
fetch the list of mirrors and fetch the package meta-data from a desired
mirror and cache this info locally.  Then the client tools can do things
without having to talk to a server like list available packages, query
dependencies, list packages you have installed, etc.  When you want to fetch
and install a package then the client tool can use the package meta-data and
urllib to fetch the sdist and/or a desired bdist and then optionally install
it either using the package's setup.py or by whatever command is neccessary
for the bdist.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!







From guido at digicool.com  Wed Feb 28 16:26:20 2001
From: guido at digicool.com (Guido van Rossum)
Date: Wed, 28 Feb 2001 10:26:20 -0500
Subject: Please disable your mail mirror, whoever it is!
In-Reply-To: Your message of "28 Feb 2001 01:17:57 PST."
             <20010228091351.C4A41371690@snelboot.oratrix.nl> 
References: <20010228091351.C4A41371690@snelboot.oratrix.nl> 
Message-ID: <200102281526.KAA28444@cj20424-a.reston1.va.home.com>

Somebody or something is echoing all mail in this thread.  Below is a
sample message -- look at all the repeated headers with an X- in front
of them.  Can this please stop?

--Guido van Rossum (home page: http://www.python.org/~guido/)

> Return-Path: distutils-sig-admin at python.org
> Delivery-Date: Wed Feb 28 04:25:40 2001
> Return-Path: <distutils-sig-admin at python.org>
> Received: from localhost (IDENT:guido at localhost.localdomain [127.0.0.1])
> 	by cj20424-a.reston1.va.home.com (8.9.3/8.9.3) with ESMTP id EAA23220
> 	for <guido at localhost>; Wed, 28 Feb 2001 04:25:40 -0500
> Received: from pop.digicool.com
> 	by localhost with POP3 (fetchmail-5.3.1)
> 	for guido at localhost (single-drop); Wed, 28 Feb 2001 04:25:40 -0500 (EST)
> Received: from ns2.digicool.com ([216.164.72.2] verified)
>   by digicool.com (CommuniGate Pro SMTP 3.4)
>   with ESMTP id 1628614 for guido at mail.digicool.com; Wed, 28 Feb 2001 04:24:25 -0500
> Received: from mail.python.org (mail.python.org [63.102.49.29])
> 	by ns2.digicool.com (8.9.3/8.9.3) with ESMTP id EAA06781
> 	for <guido at digicool.com>; Wed, 28 Feb 2001 04:22:01 -0500
> Received: from localhost.localdomain ([127.0.0.1] helo=mail.python.org)
> 	by mail.python.org with esmtp (Exim 3.21 #1)
> 	id 14Y2mB-0003d5-00; Wed, 28 Feb 2001 04:20:03 -0500
> Received: from [204.123.2.56] (helo=mail2.digital.com)
> 	by mail.python.org with esmtp (Exim 3.21 #1)
> 	id 14Y2lY-0003Yz-00; Wed, 28 Feb 2001 04:19:25 -0500
> Received: from src-news.pa.dec.com (src-news.pa.dec.com [16.4.16.111])
> 	by mail2.digital.com (8.9.2/8.9.3/WV2.0h) with ESMTP id BAA25142;
> 	Wed, 28 Feb 2001 01:19:13 -0800 (PST)
> Received: by src-news.pa.dec.com; id BAA15660; Wed, 28 Feb 2001 01:17:57 -0800 (PST)
> To: dec-mail-lists-python-list at usenet.pa.dec.com
> Path: pa.dec.com!owner-python-list-redist
> From: jack at oratrix.nl (Jack Jansen)
> Newsgroups: dec.mail.lists.python-list
> Subject: Re: [Distutils] Re: CPAN functionality for python - requirements
> Organization: Systems Research Center, Compaq Computer Corporation
> Lines: 16
> Distribution: dec
> Message-ID: <20010228091351.C4A41371690 at snelboot.oratrix.nl>
> To: Doug Hellmann <doughellmann at home.com>
> Cc: Bruce Sass <bsass at freenet.edmonton.ab.ca>,
>         Sean Reifschneider <jafo at tummy.com>, python-list at python.org,
>         distutils-sig at python.org
> In-Reply-To: Message by Doug Hellmann <doughellmann at home.com> ,
> 	     Tue, 27 Feb 2001 21:03:43 -0500 , <01022721090603.20069 at branagan> 
> Mime-Version: 1.0
> Content-Type: text/plain; charset=us-ascii
> X-List-Id: General discussion list for the Python programming language <python-list.python.org>
> X-Beenthere: python-list at python.org
> X-Phone: +31 20 6795452(work), +31 20 6795309 (fax), +31 20 6160335(home)
> X-List-Subscribe: <http://mail.python.org/mailman/listinfo/python-list>,
> 	<mailto:python-list-request at python.org?subject=subscribe>
> X-Errors-To: python-list-admin at python.org
> X-List-Unsubscribe: <http://mail.python.org/mailman/listinfo/python-list>,
> 	<mailto:python-list-request at python.org?subject=unsubscribe>
> X-List-Post: <mailto:python-list at python.org>
> X-Mini-Review: Sixties punkrock with an ATTITUDE!
> X-Mailman-Version: 2.0.1 (101270)
> X-Mailer: exmh version 2.1.0 04/14/1999
> X-Received: by src-news.pa.dec.com; id BAA13907; Wed, 28 Feb 2001 01:17:55 -0800 (PST)
> X-Received: by pobox1.pa.dec.com (5.65v3.2/1.1.10.5/07Nov97-1157AM)
> 	id AA08971; Wed, 28 Feb 2001 01:17:55 -0800
> X-Received: from zmamail01.nz-tay.cpqcorp.net by pobox1.pa.dec.com (5.65v3.2/1.1.10.5/07Nov97-1157AM)
> 	id AA14595; Wed, 28 Feb 2001 01:17:54 -0800
> X-Received: by zmamail01.zma.compaq.com (Postfix, from userid 12345)
> 	id A497011C86; Wed, 28 Feb 2001 04:17:30 -0500 (EST)
> X-List-Archive: <http://mail.python.org/pipermail/python-list/>
> X-Received: from mail.python.org (mail.python.org [63.102.49.29])
> 	by zmamail01.zma.compaq.com (Postfix) with ESMTP id 7DFCAC1AB
> 	for <python-list-redist at pa.dec.com>; Wed, 28 Feb 2001 04:17:27 -0500 (EST)
> X-Received: from localhost.localdomain ([127.0.0.1] helo=mail.python.org)
> 	by mail.python.org with esmtp (Exim 3.21 #1)
> 	id 14Y2hV-0003IL-00; Wed, 28 Feb 2001 04:15:13 -0500
> X-Received: from [212.136.221.130] (helo=oratrix.oratrix.nl)
> 	by mail.python.org with esmtp (Exim 3.21 #1)
> 	id 14Y2gO-00038k-00; Wed, 28 Feb 2001 04:14:04 -0500
> X-Organisation: Oratrix Development, Valeriusplein 30, 1075 BJ Amsterdam, NL
> X-Received: from snelboot.oratrix.nl (snelboot.oratrix.nl [212.136.221.131])
> 	by oratrix.oratrix.nl (Postfix) with ESMTP
> 	id E61C211CA2A; Wed, 28 Feb 2001 10:14:03 +0100 (MET)
> X-Precedence: bulk
> X-Received: from snelboot.oratrix.nl (localhost [127.0.0.1])
> 	by snelboot.oratrix.nl (Postfix) with ESMTP
> 	id C4A41371690; Wed, 28 Feb 2001 10:13:51 +0100 (MET)
> X-List-Help: <mailto:python-list-request at python.org?subject=help>
> X-Last-Band-Seen: Hives (Melkweg, 25-2)
> Sender: distutils-sig-admin at python.org
> Errors-To: distutils-sig-admin at python.org
> X-BeenThere: distutils-sig at python.org
> Precedence: bulk
> List-Help: <mailto:distutils-sig-request at python.org?subject=help>
> List-Post: <mailto:distutils-sig at python.org>
> List-Subscribe: <http://mail.python.org/mailman/listinfo/distutils-sig>,
> 	<mailto:distutils-sig-request at python.org?subject=subscribe>
> List-Id: Python Distribution Utilities <distutils-sig.python.org>
> List-Unsubscribe: <http://mail.python.org/mailman/listinfo/distutils-sig>,
> 	<mailto:distutils-sig-request at python.org?subject=unsubscribe>
> List-Archive: <http://mail.python.org/pipermail/distutils-sig/>
> X-Original-Date: 28 Feb 2001 01:17:57 -0800
> Date: 28 Feb 2001 01:17:57 -0800
> 
> 
> > I'm not sure why the "what do you have" question is needed.  The "send me
> > that(mandrake.rpm)" interaction is what we want.
> 
> Well, one of the things I'd like to do is integrate all this in MacPython IDE, 
> and the same probably holds for PythonWin and Idle. Then the "what do you 
> have" is pretty useful for building menus and such.
> --
> Jack Jansen             | ++++ stop the execution of Mumia Abu-Jamal ++++
> Jack.Jansen at oratrix.com | ++++ if you agree copy these lines to your sig ++++
> www.oratrix.nl/~jack    | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
> _______________________________________________
> Distutils-SIG maillist  -  Distutils-SIG at python.org
> http://mail.python.org/mailman/listinfo/distutils-sig



From mwa at gate.net  Tue Feb 27 15:50:52 2001
From: mwa at gate.net (Mark W. Alexander)
Date: Tue, 27 Feb 2001 09:50:52 -0500 (EST)
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <Pine.LNX.4.33.0102270137070.28512-100000@bms>
Message-ID: <Pine.A41.4.21.0102270924480.53710-100000@dakota.gate.net>

I have no idea how this cc: list came to be. I'm adding catalog-sig
and will drop individuals on subsequent posts....

On Tue, 27 Feb 2001, Bruce Sass wrote:
> Subject: [Distutils] Re: CPAN functionality for python - requirements
>
> > I sense a consensus that the "install" part should be handled by distutils.  Is
> > that right?
> 
> As long as distutils does not presume to know how to do the
> installation.
> 
> > Other requirements we might lay down up front:
> >
> > 1. Should run on all platforms where Python runs.
> > 2. Must support mirrors on the server side.
> > 3. Need to include documentation along with source for packages.
> >
> > The features related to dependencies and downloads could be handled by a
> > platform-specific packaging system, but integrating with all of the different
> > options on all of the different platforms where Python runs increases the
> > complexity of the new tool.

It seems there's a misunderstanding of what distutils does. It provides
a standard mechanism for producing installable python packages. IOW,
exactly what's under discussion here. It does so by the package
author providing code and meta-data. Distutils uses the code to
produce the python/C modules, and the meta-data to produce binary
packages in _any_supported_format. Currently, RPM, tar.gz, and 
Windows installers are included. Solaris and HP-UX package support
is available in horrendously hacked format, but it's been demonstrated
to be possible. I imagine dpkg to be of minimal difficulty for 
someone who understands debian packaging. Note that, theoretically
at least, the author needs no knowledge of packaging for distutils
to do it's work. All they need to know is what meta-data is
required by distutils. (I think there are still some issues regarding
a "standard" meta-data base that is necessary to support "all" packagers,
but that's not in insurmountable problem.)

So distutils can be used to do the installation. The problem,
however, is in the capabilities of the target system. Is there
a compiler available for compiled modules? If not, the source
is worthless, and the user is forced to rely on binary packages
(a significant number of Windows users, I'd imagine), driving a
requirement that the network provide (at least some) binary
packages in directly downloadable form.

OTOH, if compile time requirements are met (compiler, linker, SWIG,
whatever else the author uses), then distutils can easily produce
a binary package on the target machine, as long as the target's native
package manager is supported by distutils. And as long as all
dependency requirements are met.

So the catalog client will need to be able to acquire dependency
information from the network archive in order to ascertain whether
the module is buildable on the target system. If it is, it will
also need to acquire, invoke distutils to build and package, and
invoke the package manager to install all dependencies in order.

If it is not buildable, then it would need to seek out pre-built
binaries, download and install them. So, either the network archive 
interfaces with distutils such that a back-end can produce binaries,
OR archive maintainers manually produce binaries for whatever
platforms they choose to support, OR people without development
tools (Windows, Macs(?), disk-space tight OS's) are SOL.

I hope this clarifies distutils (potential) role a bit.

mwa




From doughellmann at home.com  Tue Feb 27 16:17:06 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Tue, 27 Feb 2001 10:17:06 -0500
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <Pine.A41.4.21.0102270924480.53710-100000@dakota.gate.net>
References: <Pine.A41.4.21.0102270924480.53710-100000@dakota.gate.net>
Message-ID: <01022710324002.20069@branagan>

On Tue, 27 Feb 2001, Mark W. Alexander wrote:
> I have no idea how this cc: list came to be. I'm adding catalog-sig
> and will drop individuals on subsequent posts....

I've done that, too.  It was getting a bit long.

> On Tue, 27 Feb 2001, Bruce Sass wrote:
> > Subject: [Distutils] Re: CPAN functionality for python - requirements
> >
> > > I sense a consensus that the "install" part should be handled by distutils.  Is
> > > that right?
> > 
> > As long as distutils does not presume to know how to do the
> > installation.
> > 
> > > Other requirements we might lay down up front:
> > >
> > > 1. Should run on all platforms where Python runs.
> > > 2. Must support mirrors on the server side.
> > > 3. Need to include documentation along with source for packages.
> > >
> > > The features related to dependencies and downloads could be handled by a
> > > platform-specific packaging system, but integrating with all of the different
> > > options on all of the different platforms where Python runs increases the
> > > complexity of the new tool.
> 
> It seems there's a misunderstanding of what distutils does. It provides
> a standard mechanism for producing installable python packages. IOW,

I think the misunderstanding was on my part, and between your description and
reviewing the distutils help I think I'm straightened out now.

With that said, I agree that the network needs to allow, optionally, for binary
distributions for platforms where building binary distributions is unusually
difficult (like having no compiler on Win32).  All packages should be available
in source form (assuming some OS license) by default.  The client should offer
the user the option of downloading any available package format, which can then
be installed using a separate tool ("./setup.py install" or rpm or whatever).

If as a user, I choose to download an RPM, that's up to me.  If no RPM is
available, the thing I do download should be able to *make* an RPM, so that's
where distutils comes in.

So, we don't quite get to a *standard* set of download and install
instructions because downloading different package types requires different
installation steps.  By trading that off we get package management using the
user's favorite package system.  That seems to be a reasonable trade.

> So the catalog client will need to be able to acquire dependency
> information from the network archive in order to ascertain whether
> the module is buildable on the target system. If it is, it will
> also need to acquire, invoke distutils to build and package, and
> invoke the package manager to install all dependencies in order.

I'm going to back off my request that this be an all-the-way tool.  Let's stick
to finding a package and all of its dependencies, and leave the installation to
the packages which are downloaded.  As a Python user, I may not have system
privileges sufficient to allow me to install a downloaded package.  As a system
administrator, I may not know how to use this tool to find packages.  So, the
Python user can find the package and provide it to the sysadmin for
installation.  Nice and clean.  

Optionally, there could be an argument which would invoke an installation
command for the package(s) being downloaded.

> If it is not buildable, then it would need to seek out pre-built
> binaries, download and install them. So, either the network archive 
> interfaces with distutils such that a back-end can produce binaries,

Can I build binaries for all platforms on whatever platform I use for the
network archive?

> OR archive maintainers manually produce binaries for whatever
> platforms they choose to support, 

OR automatically.  If the source distribution is always available, part of the
mirroring process could be to convert the source distributions into binary
distributions for the platform serviced by the mirror site.

> OR people without development tools (Windows, Macs(?), disk-space tight
> OS's) are SOL.

OR developers who do not have access to platforms where binary distributions
are prefered enlist other interested parties to assist them in preparing binary
packages.

OR third party volunteers post binary distributions to other sites on the
network

> I hope this clarifies distutils (potential) role a bit.

Significantly.

Thanks,
Doug



From mwa at gate.net  Tue Feb 27 16:55:03 2001
From: mwa at gate.net (Mark W. Alexander)
Date: Tue, 27 Feb 2001 10:55:03 -0500 (EST)
Subject: [Catalog-sig] Re: [Distutils] Re: CPAN functionality for python
 - requirements
In-Reply-To: <01022710324002.20069@branagan>
Message-ID: <Pine.A41.4.21.0102271045480.16050-100000@seminole.gate.net>

On Tue, 27 Feb 2001, Doug Hellmann wrote:

> Date: Tue, 27 Feb 2001 10:17:06 -0500
> From: Doug Hellmann <doughellmann at home.com>
> To: python-list at python.org, distutils-sig at python.org, catalog-sig at python.org
> Subject: [Catalog-sig] Re: [Distutils] Re: CPAN functionality for python
>     - requirements
> 
[snip]
> > It seems there's a misunderstanding of what distutils does. It provides
> > a standard mechanism for producing installable python packages. IOW,
> 
> I think the misunderstanding was on my part, and between your description and
> reviewing the distutils help I think I'm straightened out now.
> 
> With that said, I agree that the network needs to allow, optionally, for binary
> distributions for platforms where building binary distributions is unusually
> difficult (like having no compiler on Win32).  All packages should be available
> in source form (assuming some OS license) by default.  The client should offer
> the user the option of downloading any available package format, which can then
> be installed using a separate tool ("./setup.py install" or rpm or whatever).
> 
> If as a user, I choose to download an RPM, that's up to me.  If no RPM is
> available, the thing I do download should be able to *make* an RPM, so that's
> where distutils comes in.

Agreed.

[snip]

> I'm going to back off my request that this be an all-the-way tool.  Let's stick
> to finding a package and all of its dependencies, and leave the installation to
> the packages which are downloaded.  As a Python user, I may not have system
> privileges sufficient to allow me to install a downloaded package.  As a system
> administrator, I may not know how to use this tool to find packages.  So, the
> Python user can find the package and provide it to the sysadmin for
> installation.  Nice and clean.  
> 
> Optionally, there could be an argument which would invoke an installation
> command for the package(s) being downloaded.

Yes, after I sent this off, I started thinking this as well. It's
better as well for sites with multiple machines. Go get it once
and install everywhere.

> > If it is not buildable, then it would need to seek out pre-built
> > binaries, download and install them. So, either the network archive 
> > interfaces with distutils such that a back-end can produce binaries,
> 
> Can I build binaries for all platforms on whatever platform I use for the
> network archive?

That's an interesting question. What would it take to include
cross-compiling support in distutils?

> > OR archive maintainers manually produce binaries for whatever
> > platforms they choose to support, 
> 
> OR automatically.  If the source distribution is always available, part of the
> mirroring process could be to convert the source distributions into binary
> distributions for the platform serviced by the mirror site.

That was my original thought. I would hate to see, however, that some
platform gets "slighted" because there's no one to run a mirror for
them.

> > OR people without development tools (Windows, Macs(?), disk-space tight
> > OS's) are SOL.
> 
> OR developers who do not have access to platforms where binary distributions
> are prefered enlist other interested parties to assist them in preparing binary
> packages.
> 
> OR third party volunteers post binary distributions to other sites on the
> network

These last 2 are where I hoped the combination of distutils and a
network archive would reduce the number of volunteers needed. 
As long as distutils can produce a binary for a particular platform/
package manager combination, there's really no need for a dedicated
(official or UNofficial) "package mantainer". All that's needed
is a place where "python setup.py --bdist format=MyPackager" works.
How about an interarchive-API (or redirector?), so archives can
forward requests for binary packages to other sites known to support
that type of binary package?

Mark




From phrxy at csv.warwick.ac.uk  Tue Feb 27 19:03:56 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Tue, 27 Feb 2001 18:03:56 +0000 (GMT)
Subject: [Catalog-sig] Re: [Distutils] Re: CPAN functionality for python
 - requirements
In-Reply-To: <Pine.A41.4.21.0102271045480.16050-100000@seminole.gate.net>
Message-ID: <Pine.SOL.4.30.0102271758510.22762-100000@mimosa.csv.warwick.ac.uk>

On Tue, 27 Feb 2001, Mark W. Alexander wrote:
[...]
> As long as distutils can produce a binary for a particular platform/
> package manager combination, there's really no need for a dedicated
> (official or UNofficial) "package mantainer". All that's needed
> is a place where "python setup.py --bdist format=MyPackager" works.
> How about an interarchive-API (or redirector?), so archives can
> forward requests for binary packages to other sites known to support
> that type of binary package?
[...]

In which case:

1. Distutils should be able to pgp-sign binary (and source as well come to
think of it) packages.

2. CPyAN should be able to manage the signatures as appropriate.


John




From jafo at tummy.com  Wed Feb 28 05:15:25 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 21:15:25 -0700
Subject: [Catalog-sig] Re: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <Pine.A41.4.21.0102271045480.16050-100000@seminole.gate.net>; from mwa@gate.net on Tue, Feb 27, 2001 at 10:55:03AM -0500
References: <01022710324002.20069@branagan> <Pine.A41.4.21.0102271045480.16050-100000@seminole.gate.net>
Message-ID: <20010227211525.B24378@tummy.com>

On Tue, Feb 27, 2001 at 10:55:03AM -0500, Mark W. Alexander wrote:
>That was my original thought. I would hate to see, however, that some
>platform gets "slighted" because there's no one to run a mirror for
>them.

It seems to me that instead of having the "mirrors" do it, some work be done
on setting up "build nodes" that can take new source packages and turn them
into binaries...

Sean
-- 
 It often shows a fine command of a language to say nothing.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From jafo at tummy.com  Wed Feb 28 05:13:31 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 21:13:31 -0700
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <01022710324002.20069@branagan>; from doughellmann@home.com on Tue, Feb 27, 2001 at 10:17:06AM -0500
References: <Pine.A41.4.21.0102270924480.53710-100000@dakota.gate.net> <01022710324002.20069@branagan>
Message-ID: <20010227211331.A24378@tummy.com>

On Tue, Feb 27, 2001 at 10:17:06AM -0500, Doug Hellmann wrote:
>OR automatically.  If the source distribution is always available, part of the
>mirroring process could be to convert the source distributions into binary
>distributions for the platform serviced by the mirror site.

That seems like it would be a huge security issue.

Sean
-- 
 Passionate hatred can give meaning and purpose to an empty life.
                 -- Eric Hoffer
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From sanner at scripps.edu  Tue Feb 27 17:44:02 2001
From: sanner at scripps.edu (Michel Sanner)
Date: Tue, 27 Feb 2001 08:44:02 -0800
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: Doug Hellmann <doughellmann@home.com>
        "[Distutils] Re: CPAN functionality for python - requirements" (Feb 26, 10:45pm)
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> 
	<01022606501904.15046@branagan> 
	<20010226183606.F7531@tummy.com> 
	<01022623140707.15046@branagan>
Message-ID: <1010227084403.ZM33944@noah.scripps.edu>

Hi,

I followed this discussion (probably not as much as I should have).
We have setup a Python distribution site for the tools we have produced for
computational biology. It is quite primitive but let's us distribute to "ready
to go" packages as well as pre-compiled Python interpreters for some platforms
to "non-wizard" users.

One issue that we are strugling with is versions of packages.Is there a
"standard" way to deal with versions ? how do you plan to find out whether
someone has a compatible version of any given package ?

-Michel

On Feb 26, 10:45pm, Doug Hellmann wrote:
> Subject: [Distutils] Re: CPAN functionality for python - requirements
> On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> > On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
> > >What if Python kept up with its own packages and modules like XEmacs does?
 We
> > >could then use the same format on all platforms -- it might even be
something
> > >we roll ourself to make sure we can unpack it with *only* Python installed
on a
> > >platform.
> >
> > My understanding is that that's what distutils is meant to do, to an
extent.
> > The problem is that it's not entirely unreasonable to expect that an
> > "rpm -qa" produce a list of all the software that's installed on your
> > box.  If you don't use an RPM, you can't really make use of that, and
> > packages WILL get dropped after a re-install.
>
> The command "rpm -qa" only produces meaningful results on a system where RPM
is
> the default and primary package installation system.  There are more OSes
where
> that is *not* the case (and where RPM isn't installed at all) than where it
is.
>
> It still isn't clear to me why we would want to use more than one
distribution
> package format.  It appears that one argument is so that sysadmins can use
> those tools to track what is installed on the system.  But isn't that part of
> what this new thing (cyphon?) is supposed to do?  Do we want Python to
require
> an external package management system on each platform where this tool is
> supported, or do we want a Python-esq tool which just works the way it is
> supposed to, in the same way, everywhere Python is available?
>
> But I think we're arguing over a design before we've worked out requirements.
> The basic tasks of this new tool, as I see them, are:
>
> 1. Locate Python packages you want on the net.
> 2. Resolve dependencies between what you want, what you have, and what you
>    need.
> 3. Download the appropriate packages to give you what you want.
> 4. Install those packages.
>
> Do we agree these are the primary features?
>
> I sense a consensus that the "install" part should be handled by distutils.
 Is
> that right?
>
> Other requirements we might lay down up front:
>
> 1. Should run on all platforms where Python runs.
> 2. Must support mirrors on the server side.
> 3. Need to include documentation along with source for packages.
>
> The features related to dependencies and downloads could be handled by a
> platform-specific packaging system, but integrating with all of the different
> options on all of the different platforms where Python runs increases the
> complexity of the new tool.  (How do we handle RPMs on MacOS?  How do we
handle
> HQX files on Win32?  What about for Pippy, where files as such might not even
> be appropriate?  Does Jython have any special requirements?)  Do we deal with
> this complexity by deciding on a per-platform basis what format to use, or do
> we force the user to specify the format they want downloaded?
>
> If we're modeling what we're doing based on CPAN, maybe we should look at
> that design.  What format does CPAN use when downloading Perl packages?  What
> features does CPAN have that we want?  What does it not have that we want?
>
> Doug
>
> _______________________________________________
> Distutils-SIG maillist  -  Distutils-SIG at python.org
> http://mail.python.org/mailman/listinfo/distutils-sig
>-- End of excerpt from Doug Hellmann



-- 

-----------------------------------------------------------------------

>>>>>>>>>> AREA CODE CHANGE <<<<<<<<< we are now 858 !!!!!!!

Michel F. Sanner Ph.D.                   The Scripps Research Institute
Assistant Professor			Department of Molecular Biology
					  10550 North Torrey Pines Road
Tel. (858) 784-2341				     La Jolla, CA 92037
Fax. (858) 784-2860
sanner at scripps.edu                        http://www.scripps.edu/sanner
-----------------------------------------------------------------------




From jafo at tummy.com  Tue Feb 27 20:16:08 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 12:16:08 -0700
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <1010227084403.ZM33944@noah.scripps.edu>; from sanner@scripps.edu on Tue, Feb 27, 2001 at 08:44:02AM -0800
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022606501904.15046@branagan> <20010226183606.F7531@tummy.com> <01022623140707.15046@branagan> <doughellmann@home.com> <1010227084403.ZM33944@noah.scripps.edu>
Message-ID: <20010227121607.D1781@tummy.com>

On Tue, Feb 27, 2001 at 08:44:02AM -0800, Michel Sanner wrote:
>One issue that we are strugling with is versions of packages.Is there a
>"standard" way to deal with versions ? how do you plan to find out whether
>someone has a compatible version of any given package ?

That's something that I don't think we can do without human intervention.
If package X version 1.50 "fixes" something, the dependencies will have
to list "X >=1.20 <1.50".  Something like that...  In many cases it's
safe to say "X >=1.20", but there are certainly some places where that's
likely to have to be ammended on a case-by-case basis.

Sean
-- 
 Gone Postal Sort: Iterate over elements, any element that is out of order you
 blow away.  -- Evelyn, Kevin, and Sean, watching Monty Python and reading DDJ
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From mwa at gate.net  Tue Feb 27 22:52:01 2001
From: mwa at gate.net (Mark W. Alexander)
Date: Tue, 27 Feb 2001 16:52:01 -0500 (EST)
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <1010227084403.ZM33944@noah.scripps.edu>
Message-ID: <Pine.A41.4.21.0102271649460.21066-100000@seminole.gate.net>

I think this is a good reason for leveraging the native package
manager for distutil-produced-archive-managed packages. You should
be able to query your package manager for the installed version
and find newer ones in the archive.

Mark Alexander
mwa at gate.net


On Tue, 27 Feb 2001, Michel Sanner wrote:

> Date: Tue, 27 Feb 2001 08:44:02 -0800
> From: Michel Sanner <sanner at scripps.edu>
> To: Doug Hellmann <doughellmann at home.com>,
>      Sean Reifschneider <jafo at tummy.com>
> Cc: Bruce Sass <bsass at freenet.edmonton.ab.ca>,
>      Sean Reifschneider <jafo-pythonlist at tummy.com>,
>      Oleg Broytmann <phd at mail2.phd.pp.ru>,
>      Jeremy Hylton <jeremy at alum.mit.edu>,
>      Doug Hellmann <doughellmann at bigfoot.com>, python-list at python.org,
>      distutils-sig at python.org, efm at tummy.com
> Subject: Re: [Distutils] Re: CPAN functionality for python - requirements
> 
> Hi,
> 
> I followed this discussion (probably not as much as I should have).
> We have setup a Python distribution site for the tools we have produced for
> computational biology. It is quite primitive but let's us distribute to "ready
> to go" packages as well as pre-compiled Python interpreters for some platforms
> to "non-wizard" users.
> 
> One issue that we are strugling with is versions of packages.Is there a
> "standard" way to deal with versions ? how do you plan to find out whether
> someone has a compatible version of any given package ?
> 
> -Michel
> 




From doughellmann at bigfoot.com  Wed Feb 28 12:53:12 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Wed, 28 Feb 2001 11:53:12 GMT
Subject: CPAN functionality for python - requirements
References: <mailman.983303650.30338.python-list@python.org>
Message-ID: <IE5n6.77746$m5.4881842@news1.rdc1.ga.home.com>

In article <mailman.983303650.30338.python-list at python.org>, "Sean
Reifschneider" <jafo at tummy.com> wrote:

>>Right, so we need a single way of determining, maintaining, and
>>resolving dependencies.
> 
> One could say that the appropriate way to find out if a required module
> is available is to try importing it.

Some packages, especially in the context of Zope, require special
configuration of the path before they can be imported.  Besides, the
import statement might be for Package.Package.Module and the dependency
is really Package.Package -- how do we resolve that?

> I firmly believe that we have to allow not only system-wide installs,
> but also user installs (into their home directory, for example).

I like that.

Doug


From doughellmann at bigfoot.com  Wed Feb 28 12:53:16 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Wed, 28 Feb 2001 11:53:16 GMT
Subject: CPAN functionality for python - requirements
References: <mailman.983284693.1818.python-list@python.org>
Message-ID: <ME5n6.77748$m5.4881872@news1.rdc1.ga.home.com>

In article <mailman.983284693.1818.python-list at python.org>, "Bruce Sass"
<bsass at freenet.edmonton.ab.ca> wrote:

> On Tue, 27 Feb 2001, Doug Hellmann wrote:
>> On Tue, 27 Feb 2001, Bruce Sass wrote:
>> > On Mon, 26 Feb 2001, Doug Hellmann wrote:
>> > > On Mon, 26 Feb 2001, Sean Reifschneider wrote:
>> > > > On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
>> > <...>
>> > Dependencies are a big problem.  I've spoken with a local LUG member
>> > who was complaining that Debian does not know about the stuff he gets
>> > from CPAN, which makes mixing CPAN and debianized CPAN stuff a pain.
>> > Debian's solution is a user level program to debianize arbitrary CPAN
>> > modules... I want the same solution for Python modules, except I want
>> > it done transparently.
>>
>> I can see that as somewhat desirable, and it could be implemented, but
>> do you make the cyphon client download the right package format or
>> convert a standard distro into the right format?  Doesn't the user have
>> to specify a package type?
> 
> I prefer building from a generic package to a system specific package

This model would work. I didn't realize distutils would generate
system-specific package formats.  If we and agree on a standard format,
then it would be possible to populate the archive network with standard
formats and optional system-specific formats.  As a developer, I might
choose to provide several formats.  As a mirror operator, I might choose
to support one or many formats, and do some conversion for the end user.
As a user, I could choose to download in whatever format is available and
then convert to the format I want to use for installation.

> I would not use it if it means bypassing the native package management
> system, and would be willing to wait for "official" packages rather than
> use third-party packages.

What "official" packages do you mean here?

> Best of all scenarios would be if all the Python programmers used the
> same system I did and spit out packages in my favorite format, next best
> is having a package built by someone who knows my system just as well or
> better than I, then their proxy in the form of a scripted build,
> followed by someone I don't know building the package, a tarball or some
> other package format I can handle.
> 
> The first is a dream, the second could happen now and then if I was
> lucky, so I'll try for my third choice (and it sounds like distutils can
> handle it).

I think that's where we're headed.

>> > Downloading is independent of the native package management system.
>>
>> No, it isn't.  Even if *all* the tool does is download things I told it
>> to find and write them to the current directory, it has to know what
>> format I like or it may download (for example) debian packages for my
>> Mac -- and then I won't get useful results from the tool.
> 
> Build up the name of the download from the name of the package and
> system specific information, just like is done with the names of .deb
> and .rpm packages.

File type should not be dependent on file names.  We can add file type as
a field in the database.

>> > Complexity is only increased if the new tool thinks it should handle
>> > platform/OS/distro specific issues.
>>
>> It *must* handle those issues to be useful.
> 
> Isn't that is a lot of duplicated work.

Not if we use distutils for that part.

[About CPAN features...]

>> > Cons: Dependencies problems when mixed with modules fetched by other
>> > methods.
>>
>> Right, so we need a single way of determining, maintaining, and
>> resolving dependencies.
> 
> Or a way of enabling the native tools so they can do it.  I would not
> want to accept the job of keeping up with changing packaging standards
> for <how many?> different systems, and don't expect that anyone else
> would want to (or be able to keep it up).

But dependencies could be built into the network data structure.  Then it
is independent of the uploaded package format entirely.  AND, I would be
able to discover that the contents of the RPM I want depend on some other
package which exists only in source and deb formats.

Doug


From jafo at tummy.com  Tue Feb 27 02:26:01 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 18:26:01 -0700
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102260150060.28098-100000@bms>; from bsass@freenet.edmonton.ab.ca on Mon, Feb 26, 2001 at 02:56:51AM -0700
References: <20010226013903.B3221@tummy.com> <Pine.LNX.4.33.0102260150060.28098-100000@bms>
Message-ID: <20010226182601.D7531@tummy.com>

On Mon, Feb 26, 2001 at 02:56:51AM -0700, Bruce Sass wrote:
>Do you mean:

Ideally, it would be able to deal with both.  The ability to download
a distutils package and produce a .deb or a .rpm is more a function of
the client program than of the archive network...  It also largely depends
on distutils as well.

>As for the second... it seems like an awful waste of archive space to
>have the same code in many different formats, just because some want
>to put stuff in /usr/share and others in /usr/lib, etc.

I'm really not familiar with .deb packages.  Howeer, for RPMs, if the
author has included a .spec file in the tar file, you can simply run
"rpm -ta foo.tar.gz" and the results will (should) be a compiled binary
RPM.  That is, if the author of the package has thought ahead about
producing an RPM.

The question becomes: Is it reasonable to require authors of packages
for the archive network to do the same?  Convince me it is...

Sean
-- 
 "McGuyver stole all his tricks from Dr. Who."
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From bsass at freenet.edmonton.ab.ca  Tue Feb 27 09:22:34 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 01:22:34 -0700 (MST)
Subject: CPAN functionality for python
In-Reply-To: <20010226182601.D7531@tummy.com>
Message-ID: <Pine.LNX.4.33.0102270054250.28512-100000@bms>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> On Mon, Feb 26, 2001 at 02:56:51AM -0700, Bruce Sass wrote:
> >Do you mean:
>
> Ideally, it would be able to deal with both.  The ability to download
> a distutils package and produce a .deb or a .rpm is more a function of
> the client program than of the archive network...  It also largely depends
> on distutils as well.

So we are thinking along the same lines.

> >As for the second... it seems like an awful waste of archive space to
> >have the same code in many different formats, just because some want
> >to put stuff in /usr/share and others in /usr/lib, etc.
>
> I'm really not familiar with .deb packages.  Howeer, for RPMs, if the
> author has included a .spec file in the tar file, you can simply run
> "rpm -ta foo.tar.gz" and the results will (should) be a compiled binary
> RPM.  That is, if the author of the package has thought ahead about
> producing an RPM.

In a nutshell...
Debian's equivalent of a .spec is the "debian" subdir in the source
tree, containing administrivia and a "rules" makefile; doing
"debian/rules" compiles, "debian/rules binary" compiles and packages,
"debuild" compiles, packages and checks the package for errors and
conformance to Policy.  There can be a lot more to it, but nothing the
author needs to worry about, in fact the author does not need to know
anything about Debian.

There are tools that create the debian dir and pull information from
the source tree to customize the package, it is very easy to script
so it can debianize anything thrown at it - provided the source tree
is reasonably consistent.

> The question becomes: Is it reasonable to require authors of packages
> for the archive network to do the same?  Convince me it is...

No, it isn't.  The authors should not need to know anything about the
final format of their module, just that they need to provide specific
pieces of information if they want to be included in the network
(imo).

Hmmm, so RPM based systems don't have a way to automatically generate
a .spec file from a source tree... that would be a must.


- Bruce




From jafo at tummy.com  Tue Feb 27 09:38:32 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 01:38:32 -0700
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.33.0102270054250.28512-100000@bms>; from bsass@freenet.edmonton.ab.ca on Tue, Feb 27, 2001 at 01:22:34AM -0700
References: <20010226182601.D7531@tummy.com> <Pine.LNX.4.33.0102270054250.28512-100000@bms>
Message-ID: <20010227013832.U1781@tummy.com>

On Tue, Feb 27, 2001 at 01:22:34AM -0700, Bruce Sass wrote:
>Hmmm, so RPM based systems don't have a way to automatically generate
>a .spec file from a source tree... that would be a must.

RPMs include pieces of information such as:

   Description of package
   Summary of the package
   Source URL
   Copyright
   Type of package (network daemons, console tools, etc)
   What is needed to install it.
   What is needed to build it.
   Scripts used for adding and removing a package
      (probably less necessary for most modules packages, but is used
      for software for doing things like adding user IDs, inittab entries,
      and the like)
   What files are "documentation".
   Where to install files and what their permissions are.
   Which installed files are config files.

Some of it is obviously superfluous, but much of it is not, and little of
it can be automaticly pulled from an arbitrary tar file.

Now, at the minimum, an RPM file consists of the files that make it up
and the permissions, so you can fake a minimal RPM file if you need to.

Sean
-- 
 Brooks's Law of Prototypes: Plan to throw one away, you will anyhow.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From R.Liebscher at gmx.de  Tue Feb 27 10:02:46 2001
From: R.Liebscher at gmx.de (Rene Liebscher)
Date: Tue, 27 Feb 2001 10:02:46 +0100
Subject: [Distutils] Re: CPAN functionality for python
References: <14984.49127.853606.286789@w221.z064000254.bwi-md.dsl.cnc.net> <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru> <20010215170227.K23577@tummy.com> <20010226013903.B3221@tummy.com>
Message-ID: <3A9B6D36.20B1B3C2@gmx.de>

I don't know if I missed something, but if you want for 
all users install a package in their native format, why not so:

1. - download the binary in a format which can be unpacked 
     by python/distutils on every system where this binary works
     (Linux -> tar.gz?)
   - download some meta data

2. extract the binaries in a directory

3. feed the meta data and the directory in distutils
   bdist_{rpm|debian|...) and create a package for the
   users prefered system (RPM,...)

   (You probably have to extend the bdist_* commands, 
    but this way we also get people to write bdist_*
    commands for their platforms.)

4. use the finished package to install it 


Kind regards
Rene Liebscher



From jafo at tummy.com  Tue Feb 27 10:28:01 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 02:28:01 -0700
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: <3A9B6D36.20B1B3C2@gmx.de>; from R.Liebscher@gmx.de on Tue, Feb 27, 2001 at 10:02:46AM +0100
References: <14984.49127.853606.286789@w221.z064000254.bwi-md.dsl.cnc.net> <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru> <20010215170227.K23577@tummy.com> <20010226013903.B3221@tummy.com> <3A9B6D36.20B1B3C2@gmx.de>
Message-ID: <20010227022801.W1781@tummy.com>

On Tue, Feb 27, 2001 at 10:02:46AM +0100, Rene Liebscher wrote:
>3. feed the meta data and the directory in distutils
>   bdist_{rpm|debian|...) and create a package for the
>   users prefered system (RPM,...)

As I understand it, the bdist_* can create a binary RPM from a set of
pre-compiled binaries?  Can this be used in conjunction with the other
distutils tools to actually take the source file and turn it into
the binaries on a given platform?

If so, that's exactly what we need...

Once I have the archive network back-end done, I'm going to start looking at
the client.  I'm getting fairly close to a bare-bones server as it is
(without the admin tools -- adding a package requires some SQL).

I've only looked passingly at the distutils, but will obviously have to
get further into it shortly.

Thanks,
Sean
-- 
 Any sufficiently advanced technology is indistinguishable from a rigged demo.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From R.Liebscher at gmx.de  Tue Feb 27 10:44:51 2001
From: R.Liebscher at gmx.de (Rene Liebscher)
Date: Tue, 27 Feb 2001 10:44:51 +0100
Subject: [Distutils] Re: CPAN functionality for python
References: <14984.49127.853606.286789@w221.z064000254.bwi-md.dsl.cnc.net> <Pine.LNX.4.30.0102131941150.2534-100000@fep132.fep.ru> <20010215170227.K23577@tummy.com> <20010226013903.B3221@tummy.com> <3A9B6D36.20B1B3C2@gmx.de> <20010227022801.W1781@tummy.com>
Message-ID: <3A9B7713.334CD75D@gmx.de>

Sean Reifschneider wrote:
> 
> On Tue, Feb 27, 2001 at 10:02:46AM +0100, Rene Liebscher wrote:
> >3. feed the meta data and the directory in distutils
> >   bdist_{rpm|debian|...) and create a package for the
> >   users prefered system (RPM,...)
> 
> As I understand it, the bdist_* can create a binary RPM from a set of
> pre-compiled binaries?  Can this be used in conjunction with the other
> distutils tools to actually take the source file and turn it into
> the binaries on a given platform?

The bdist_* commands do an install in a temporary directory and create
from this a package (RPM,...).
I think it is not a big deal to provide an already existing directory
and
create from it a package.

-------------------------------------------------------
rob at cvs/distutils >python setup.py bdist_rpm --help
Global options:
  --verbose (-v)  run verbosely (default)
  --quiet (-q)    run quietly (turns verbosity off)
  --dry-run (-n)  don't actually do anything
  --help (-h)     show detailed help message

Options for 'bdist_rpm' command:
  --bdist-base         base directory for creating built distributions
  --rpm-base           base directory for creating RPMs (defaults to
"rpm"
                       under --bdist-base; must be specified for RPM 2)
  --dist-dir (-d)      directory to put final RPM files in (and .spec
files if
                       --spec-only)
  --python             path to Python interpreter to hard-code in the
.spec
                       file (default: "python")
  --fix-python         hard-code the exact path to the current Python
                       interpreter in the .spec file
  --spec-only          only regenerate spec file
  --source-only        only generate source RPM
  --binary-only        only generate binary RPM
  --use-bzip2          use bzip2 instead of gzip to create source
distribution
  --distribution-name  name of the (Linux) distribution to which this
RPM
                       applies (*not* the name of the module
distribution!)
  --group              package classification [default:
                       "Development/Libraries"]
  --release            RPM release number
  --serial             RPM serial number
  --vendor             RPM "vendor" (eg. "Joe Blow <joe at example.com>")
                       [default: maintainer or author from setup script]
  --packager           RPM packager (eg. "Jane Doe
                       <jane at example.net>")[default: vendor]
  --doc-files          list of documentation files (space or
comma-separated)
  --changelog          path to RPM changelog
  --icon               name of icon file
  --provides           capabilities provided by this package
  --requires           capabilities required by this package
  --conflicts          capabilities which conflict with this package
  --build-requires     capabilities required to build this package
  --obsoletes          capabilities made obsolete by this package
  --keep-temp (-k)     don't clean up RPM build directory
  --no-keep-temp       clean up RPM build directory [default]
  --use-rpm-opt-flags  compile with RPM_OPT_FLAGS when building from
source
                       RPM
  --no-rpm-opt-flags   do not pass any RPM CFLAGS to compiler
  --rpm3-mode          RPM 3 compatibility mode (default)
  --rpm2-mode          RPM 2 compatibility mode

usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
   or: setup.py --help [cmd1 cmd2 ...]
   or: setup.py --help-commands
   or: setup.py cmd --help
------------------------------------------------------------------------
May be it is enough to specify --bdist-base and --binary-only.
The other option could get values from some downloaded meta data.


> If so, that's exactly what we need...
> 
> Once I have the archive network back-end done, I'm going to start looking at
> the client.  I'm getting fairly close to a bare-bones server as it is
> (without the admin tools -- adding a package requires some SQL).
> 
> I've only looked passingly at the distutils, but will obviously have to
> get further into it shortly.
> 
> Thanks,
> Sean
> --
>  Any sufficiently advanced technology is indistinguishable from a rigged demo.
> Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
> tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From doughellmann at home.com  Wed Feb 14 01:26:56 2001
From: doughellmann at home.com (Doug Hellmann)
Date: Tue, 13 Feb 2001 19:26:56 -0500
Subject: CPAN functionality for python
In-Reply-To: <Pine.LNX.4.30.0102131928550.2534-100000@fep132.fep.ru>
References: <Pine.LNX.4.30.0102131928550.2534-100000@fep132.fep.ru>
Message-ID: <01021319314200.05138@branagan>

On Tue, 13 Feb 2001, Oleg Broytmann wrote:
> On Mon, 12 Feb 2001, Jeremy Hylton wrote:
> >   OB> This metadata can be put in static files. Thus, it can be easily
> >   OB> mirrored/reached by www/ftp/rsync/etc. In case of Zope, mirror
> >   OB> site must run Zope, so mirroring will be limited to probably 2-3
> >   OB> sites in the world.

How is the meta data indexed?  How is it searched?

It is possible to export data from a Zope server and import it into another
one, or even use it directly without having to import it.  This allows us to
actually *mirror* the content, and not have multiple Zope front ends to a
single ZEO-driven database on the backend.  I do this at my job and it works
fine.

> > I think this would be a fine project for Zope, in part because we're
> > going to be spinning up zope.python.org in the next few weeks and
> > could probably provide a home for it.  (Yeah, I work for Digital
> > Creations, too, but only because I think it's a great company.)
> 
>    Ok, so we've found one of these 2 or 3 mirrors.
> 
> >     - Each mirror runs a ZEO client program and its own Zope server.
> 
>    What will be these other 1 or 2 mirror sites that will run Zope?

Why do you assume no one will run Zope?

How many mirrors do we really need?  Not "We want as many as CPAN!" but how
many do we actually need to have in order to support the load we expect this
service to undertake and to give some fail-over capabilities.

>    I completeley agree that this task *can* be solved using Zope. We can
> develop data formats, protocols, etc. And I completely disagree that the
> task *should* be solved with Zope. It shouldn't.
>     Please count current mirrors of CPAN. The figure looks good, isn't it?
> This is because there are sites that provide FTP mirrors.

What is mirrored on all of these sites, the actual database or the service
which does the lookups?

Doug




From phd at phd.pp.ru  Wed Feb 14 10:25:28 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 14 Feb 2001 12:25:28 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <01021319314200.05138@branagan>
Message-ID: <Pine.LNX.4.30.0102141220520.8189-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Doug Hellmann wrote:
> >    Ok, so we've found one of these 2 or 3 mirrors.
> >
> > >     - Each mirror runs a ZEO client program and its own Zope server.
> >
> >    What will be these other 1 or 2 mirror sites that will run Zope?
>
> Why do you assume no one will run Zope?

   I didn't assume. I know for sure there are not much Zope installations,
and most FTP mirrors will not agree to run Zope for us.

> How many mirrors do we really need?  Not "We want as many as CPAN!" but how
> many do we actually need to have in order to support the load we expect this
> service to undertake and to give some fail-over capabilities.

   Initially we don't need much mirrors. But I personally hope sooner or
later the load will be high. I certainly want to develop Python community
and make it bigger and more eager for good modules and extensions! :)

> >     Please count current mirrors of CPAN. The figure looks good, isn't it?
> > This is because there are sites that provide FTP mirrors.
>
> What is mirrored on all of these sites, the actual database or the service
> which does the lookups?

   All things. Modules, docs, the perl distribution itself, etc.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From ssthapa at classes.cs.uchicago.edu  Tue Feb 13 20:14:29 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Tue, 13 Feb 2001 19:14:29 GMT
Subject: CPAN functionality for python
References: <mailman.982082114.10198.python-list@python.org>
Message-ID: <slrn98j2av.234.ssthapa@ntcs-ip45.uchicago.edu>

Oleg Broytmann <phd at phd.pp.ru> wrote:
>   I completeley agree that this task *can* be solved using Zope. We can
>develop data formats, protocols, etc. And I completely disagree that the
>task *should* be solved with Zope. It shouldn't.
>    Please count current mirrors of CPAN. The figure looks good, isn't it?
>This is because there are sites that provide FTP mirrors.
>   Who will agree to run Zope only to mirror CPyAN?

  It shouldn't be a problem to have both ftp and zope functionality at
the same time.  The way the code currently works is new toplevel commands
are implemented in separate modules.  The main program imports the module
and calls a function in the module which registers a top level command and
help for the command.  It wouldn't be any problem to have an ftpinstall and
zopeinstall module coexisting at the same time.  

-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From ssthapa at the-400.cs.uchicago.edu  Wed Feb 14 20:55:43 2001
From: ssthapa at the-400.cs.uchicago.edu (Suchandra S Thapa)
Date: Wed, 14 Feb 2001 19:55:43 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <slrn98lolv.40m.ssthapa@the-400.cs.uchicago.edu>

	The source code is now available on sourceforge under the project
name pythonsiphon.


From richard at iopen.co.nz  Wed Feb 14 22:40:14 2001
From: richard at iopen.co.nz (Richard Waid)
Date: Thu, 15 Feb 2001 10:40:14 +1300
Subject: CPAN functionality for python
References: <mailman.982142776.25775.python-list@python.org>
Message-ID: <96eu1l$1m3r$1@raewyn.paradise.net.nz>

In article <mailman.982142776.25775.python-list at python.org>, "Oleg
Broytmann" <phd at phd.pp.ru> wrote:

> On Tue, 13 Feb 2001, Doug Hellmann wrote:
>> >    Ok, so we've found one of these 2 or 3 mirrors.
>> >
>> > >     - Each mirror runs a ZEO client program and its own Zope
>> > >     server.
>> >
>> >    What will be these other 1 or 2 mirror sites that will run Zope?
>>
>> Why do you assume no one will run Zope?
> 
>    I didn't assume. I know for sure there are not much Zope
>    installations, and most FTP mirrors will not agree to run Zope for us.

Something that seems to be overlooked here - Zope might not be run by
large numbers of sites, but the sites that do run it are usually by
people fairly keen on Python. I'd suggest that while we might not get
huge numbers of mirrors initially, we'd get _enough_ (at least 1-3
per continent to begin with).

The advantage of Zope is that you could prototype the system quickly and
easily, as well as being an (almost) pure python solution, which just
seems 'right' somehow :)

Has anyone mentioned using XMLRPC to transfer the metadata and requests
around? It would be trivial to leverage that feature with Zope. The
actual transfers could be done out-of-band, using ordinary FTP or HTTP.

I'd suggest that there were two kinds of mirrors - meta data, and actual
data. The actual data could be stored on a regular FTP server, the
metadata could be accessed via XMLRPC from a Zope server. A little like
www.rpmfind.org in a way.

Richard Waid
Network/Software Engineer
http://iopen.co.nz


From andym at ActiveState.com  Thu Feb 15 00:44:03 2001
From: andym at ActiveState.com (Andy McKay)
Date: Wed, 14 Feb 2001 15:44:03 -0800
Subject: CPAN functionality for python
References: <mailman.982142776.25775.python-list@python.org> <96eu1l$1m3r$1@raewyn.paradise.net.nz>
Message-ID: <96f51s$gno$1@postum.activestate.com>

> I'd suggest that there were two kinds of mirrors - meta data, and actual
> data. The actual data could be stored on a regular FTP server, the
> metadata could be accessed via XMLRPC from a Zope server. A little like
> www.rpmfind.org in a way.

I wrote a product for Zope that provides a simple proof of concept called
the ZPM. It checks the product data on a Zope installation against the a
website (http://www.zopezen.org/products). Zope is serving back a simple xml
file containing the requested products information in XML. Zope could handle
the metadata quite easily, a regular FTP could do the hard work.

--
  Andy McKay.




From mbel44 at dial.pipex.net  Thu Feb 15 12:09:35 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Thu, 15 Feb 2001 11:09:35 +0000
Subject: CPAN functionality for python
References: <mailman.982142776.25775.python-list@python.org> <96eu1l$1m3r$1@raewyn.paradise.net.nz>
Message-ID: <drdn8tsl6c39p6r0s2tf2got2tuouv1v84@4ax.com>

"Richard Waid" <richard at iopen.co.nz> wrote:

>Has anyone mentioned using XMLRPC to transfer the metadata and requests
>around? It would be trivial to leverage that feature with Zope. The
>actual transfers could be done out-of-band, using ordinary FTP or HTTP.

XMLRPC method calls cant be cached by http proxy caches, which is a
serious disadvantage for metadata and content alike.

The requests that you propose passing around by XMLRPC are largely of
the 'has x changed since I last looked' variety, which is exactly what
http caches are designed to handle. It would be a shame if we couldnt
exploit them.

HTTP GET is your friend.



Toby Dickenson
tdickenson at geminidataloggers.com


From doughellmann at bigfoot.com  Tue Feb 27 04:42:17 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Tue, 27 Feb 2001 03:42:17 GMT
Subject: CPAN functionality for python
References: <mailman.983237829.25485.python-list@python.org>
Message-ID: <tmFm6.76393$m5.4733185@news1.rdc1.ga.home.com>

In article <mailman.983237829.25485.python-list at python.org>, "Sean
Reifschneider" <jafo at tummy.com> wrote:

> On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
>>What if Python kept up with its own packages and modules like XEmacs
>>does?  We could then use the same format on all platforms -- it might
>>even be something we roll ourself to make sure we can unpack it with
>>*only* Python installed on a platform.
> 
> My understanding is that that's what distutils is meant to do, to an
> extent. The problem is that it's not entirely unreasonable to expect
> that an
> "rpm -qa" produce a list of all the software that's installed on your
> box.  If you don't use an RPM, you can't really make use of that, and
> packages WILL get dropped after a re-install.

But "rpm -qa" only works on a system where RPM is the default and primary
package installation system.  There are more OSes where that is not the
case than where it is.

It isn't clear to me why we would want to use more than one distribution
package format.  It appears that the argument is so that sysadmins can
use those tools to track what is installed on the system.  But isn't that
part of what this new thing is supposed to do?  It finds newer versions
of what you have, resolves dependencies, downloads and installs the
packages.

Doug


From zamurai at gmx.net  Sun Feb 11 15:45:16 2001
From: zamurai at gmx.net (Zamurai)
Date: Sun, 11 Feb 2001 15:45:16 +0100
Subject: Does an modul exist, that plays mp3s?
Message-ID: <981902716.1338974340@news.t-online.de>



From phd at phd.pp.ru  Mon Feb 12 11:58:29 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 12 Feb 2001 13:58:29 +0300 (MSK)
Subject: Does an modul exist, that plays mp3s?
In-Reply-To: <981902716.1338974340@news.t-online.de>
Message-ID: <Pine.LNX.4.30.0102121358070.385-100000@fep132.fep.ru>

Snack Sound Toolkit (may be).

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From glenfant at equod.com.nospam  Sun Feb 11 18:46:27 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Sun, 11 Feb 2001 18:46:27 +0100
Subject: Where goes stdout ?
Message-ID: <966iqa$lu5$1@reader1.imaginet.fr>

Hi,

How to check in a script if stdout is a console (or xterm) or if it is
redirected to something else (file, pipe...) from command line.

Thanks in advance.

Gilles




From bsb at winnegan.de  Sun Feb 11 21:51:36 2001
From: bsb at winnegan.de (Siggy Brentrup)
Date: 11 Feb 2001 21:51:36 +0100
Subject: Where goes stdout ?
In-Reply-To: <966iqa$lu5$1@reader1.imaginet.fr>
References: <966iqa$lu5$1@reader1.imaginet.fr>
Message-ID: <87vgqh6izr.fsf@winnegan.de>

"Gillou" <glenfant at equod.com.nospam> writes:

> Hi,
> 
> How to check in a script if stdout is a console (or xterm) or if it is
> redirected to something else (file, pipe...) from command line.

import os, sys

if os.isatty(sys.stdout.fileno()):
    # output goes to a tty
    pass
else:
    # output is redirected
    pass

HTH
  Siggy




From fredrik at pythonware.com  Sun Feb 11 22:28:41 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 11 Feb 2001 21:28:41 GMT
Subject: Where goes stdout ?
References: <966iqa$lu5$1@reader1.imaginet.fr> <mailman.981924732.22939.python-list@python.org>
Message-ID: <duDh6.15658$AH6.2208571@newsc.telia.net>

Siggy Brentrup wrote:>
> > How to check in a script if stdout is a console (or xterm) or if it is
> > redirected to something else (file, pipe...) from command line.
>
> import os, sys
>
> if os.isatty(sys.stdout.fileno()):
>     # output goes to a tty

or:

    if sys.stdout.isatty():
        # output goes to a tty

Cheers /F




From ImranG at btinternet.com  Sun Feb 11 19:47:17 2001
From: ImranG at btinternet.com (Imran)
Date: Sun, 11 Feb 2001 18:47:17 +0000
Subject: smtplib problem
Message-ID: <966mnl$2dl$1@uranium.btinternet.com>

When using smtplib can you get the "header text" which mail servers
introduces themselves with ?

Imran
-- 
BTInternet FAQ V2.1 (Unofficial)
http://redrival.com/btifaq/

Petition for a Software Patent Free Europe 
http://petition.eurolinux.org/index_html?LANG=en


From fredrik at pythonware.com  Sun Feb 11 20:40:13 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 11 Feb 2001 19:40:13 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com>
Message-ID: <xUBh6.15619$AH6.2202087@newsc.telia.net>

Imran wrote:
> When using smtplib can you get the "header text" which mail servers
> introduces themselves with ?

>>> import smtplib
>>> server = smtplib.SMTP()
>>> server.connect("localhost")
(220, 'spam.egg ESMTP pymail 1.2.3/4.5.6; Sun, 11 Feb 2001 12:34:56 +0100')
>>> server.helo()
(250, 'spam.egg Hello IDENT:fredrik at spam.egg [1.2.3.4], pleased to meet you')

Cheers /F




From sheila at spamcop.net  Mon Feb 12 07:48:20 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 12 Feb 2001 06:48:20 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net>
Message-ID: <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com>

On Sun, 11 Feb 2001 19:40:13 GMT, "Fredrik Lundh" <fredrik at pythonware.com>
wrote in comp.lang.python in article
<xUBh6.15619$AH6.2202087 at newsc.telia.net>:

:Imran wrote:
:> When using smtplib can you get the "header text" which mail servers
:> introduces themselves with ?
:
:>>> import smtplib
:>>> server = smtplib.SMTP()
:>>> server.connect("localhost")
:(220, 'spam.egg ESMTP pymail 1.2.3/4.5.6; Sun, 11 Feb 2001 12:34:56 +0100')
:>>> server.helo()
:(250, 'spam.egg Hello IDENT:fredrik at spam.egg [1.2.3.4], pleased to meet you')
:
:Cheers /F
:

Nice. And how does one do SMTP authorization? (There must be a way. I see that
the NNTP module includes authorization commands, but the smtplib doesn't!)

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From matt at mondoinfo.com  Mon Feb 12 20:34:35 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Mon, 12 Feb 2001 19:34:35 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com>
Message-ID: <slrn98gemb.sn.matt@happy-hour.mondoinfo.com>

On Mon, 12 Feb 2001 06:48:20 GMT, Sheila King <sheila at spamcop.net>
wrote:

>Nice. And how does one do SMTP authorization? (There must be a way. I
>see that the NNTP module includes authorization commands, but the
>smtplib doesn't!)

Sheila,
Unfortunately SMTP authentication is a Big Mess. NNTP authentication
is generally just a matter of sending a username and a password. It's
described in a couple of hundred lines in RFC 2980. The information
necessary to implement SMTP authentication is scattered through a
stack of seemingly unrelated and difficult to follow RFCs (2104, 2195
2222, 2554, and 2595, I think). Figuring out the right thing to do is
no trivial task. So you wind up with a chicken-and-egg problem: Few
SMTP servers implement authentication because few clients want to use
it. Since few servers implement it, there's little reason for the
authors of mail clients to implement it.

I don't use it on my MTAs or MUAs. I just tell people to have ssh
forward port 25.

Regards,
Matt


From sheila at spamcop.net  Tue Feb 13 06:14:22 2001
From: sheila at spamcop.net (Sheila King)
Date: Tue, 13 Feb 2001 05:14:22 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com> <slrn98gemb.sn.matt@happy-hour.mondoinfo.com>
Message-ID: <6egh8t4vic38ufa4k8kbad5qr3301lhlo3@4ax.com>

On Mon, 12 Feb 2001 19:34:35 GMT, matt at mondoinfo.com (Matthew Dixon Cowles)
wrote in comp.lang.python in article
<slrn98gemb.sn.matt at happy-hour.mondoinfo.com>:

:On Mon, 12 Feb 2001 06:48:20 GMT, Sheila King <sheila at spamcop.net>
:wrote:
:
:>Nice. And how does one do SMTP authorization? (There must be a way. I
:>see that the NNTP module includes authorization commands, but the
:>smtplib doesn't!)
:
:Sheila,
:Unfortunately SMTP authentication is a Big Mess. NNTP authentication
:is generally just a matter of sending a username and a password. It's
:described in a couple of hundred lines in RFC 2980. The information
:necessary to implement SMTP authentication is scattered through a
:stack of seemingly unrelated and difficult to follow RFCs (2104, 2195
:2222, 2554, and 2595, I think). Figuring out the right thing to do is
:no trivial task. So you wind up with a chicken-and-egg problem: Few
:SMTP servers implement authentication because few clients want to use
:it. Since few servers implement it, there's little reason for the
:authors of mail clients to implement it.
:
:I don't use it on my MTAs or MUAs. I just tell people to have ssh
:forward port 25.

Thanks, Matt. Hmm. Well, quite a few SMTP's do authorization of some sort. I
guess some do POP before SMTP, which is probably not handled by the MTA at
all?

Most Windows MUA's seem to offer SMTP authorization. I think, with the
increasing spam situation, and the need to secure open mail servers, that more
and more are using authorization.

If I were writing mail software, for general purpose (not just for myself), I
would probably need to include an authorization option.

I suppose, from what you've written, that if I'm really interested in this,
I'd have to start following some of the mail newsgroups. I guess that is a
project for some other point in my life, since I've not time right now! (Hey,
maybe by then they'll get a SMTP authorization standard ironed out and in a
single document, just to make MY life easier?) I had heard that SMTP
authorization is just a proposed standard at this point.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/




From timr at probo.com  Wed Feb 14 07:12:50 2001
From: timr at probo.com (Tim Roberts)
Date: Tue, 13 Feb 2001 22:12:50 -0800
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com>
Message-ID: <k98k8tkp9k6undl9t71ogvlu0vse4pqa1k@4ax.com>

Sheila King <sheila at spamcop.net> wrote:
>
>Nice. And how does one do SMTP authorization? (There must be a way. I see that
>the NNTP module includes authorization commands, but the smtplib doesn't!)

There is no such thing.  The SMTP protocol does not include authorization.

Some providers have gone to a scheme whereby you have to check for POP mail
before you can use SMTP.  You have to send a password for POP, after which
they open a short window wherein you can send mail.  That doesn't take any
changes to smtplib.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From c.evans at clear.net.nz  Wed Feb 14 09:35:39 2001
From: c.evans at clear.net.nz (Carey Evans)
Date: 14 Feb 2001 21:35:39 +1300
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com>
	<xUBh6.15619$AH6.2202087@newsc.telia.net>
	<nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com>
	<k98k8tkp9k6undl9t71ogvlu0vse4pqa1k@4ax.com>
Message-ID: <87pugl1x2c.fsf@psyche.dnsalias.org>

Tim Roberts <timr at probo.com> writes:

> There is no such thing.  The SMTP protocol does not include authorization.

RFC2554 specifies an extension to SMTP, that adds authentication on
message submission to an SMTP server.  BCP46 (RFC3013) section 5.4
recommends its use.

smtplib.py doesn't implement SMTP AUTH, although you could do it
yourself using the docmd(), send() and getreply() methods.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

  "May not be representative of the experience of actual customers."


From sheila at spamcop.net  Wed Feb 14 15:44:04 2001
From: sheila at spamcop.net (Sheila King)
Date: Wed, 14 Feb 2001 14:44:04 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com> <k98k8tkp9k6undl9t71ogvlu0vse4pqa1k@4ax.com> <87pugl1x2c.fsf@psyche.dnsalias.org>
Message-ID: <bc6l8tod6qgtrbq8sha4qh30kst34isaeo@4ax.com>

On 14 Feb 2001 21:35:39 +1300, Carey Evans <c.evans at clear.net.nz> wrote in
comp.lang.python in article <87pugl1x2c.fsf at psyche.dnsalias.org>:

:Tim Roberts <timr at probo.com> writes:
:
:> There is no such thing.  The SMTP protocol does not include authorization.
:
:RFC2554 specifies an extension to SMTP, that adds authentication on
:message submission to an SMTP server.  BCP46 (RFC3013) section 5.4
:recommends its use.
:
:smtplib.py doesn't implement SMTP AUTH, although you could do it
:yourself using the docmd(), send() and getreply() methods.

Thank you! This is the type of tip I was looking for. Maybe in a few weeks,
when I have time, I'll mess around with it.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From sheila at spamcop.net  Wed Feb 14 15:42:50 2001
From: sheila at spamcop.net (Sheila King)
Date: Wed, 14 Feb 2001 14:42:50 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com> <k98k8tkp9k6undl9t71ogvlu0vse4pqa1k@4ax.com>
Message-ID: <276l8t4339e0i5nc1hrff42lon6v0f39r2@4ax.com>

On Tue, 13 Feb 2001 22:12:50 -0800, Tim Roberts <timr at probo.com> wrote in
comp.lang.python in article <k98k8tkp9k6undl9t71ogvlu0vse4pqa1k at 4ax.com>:

:Sheila King <sheila at spamcop.net> wrote:
:>
:>Nice. And how does one do SMTP authorization? (There must be a way. I see that
:>the NNTP module includes authorization commands, but the smtplib doesn't!)
:
:There is no such thing.  The SMTP protocol does not include authorization.
:
:Some providers have gone to a scheme whereby you have to check for POP mail
:before you can use SMTP.  You have to send a password for POP, after which
:they open a short window wherein you can send mail.  That doesn't take any
:changes to smtplib.

There is a proposed SMTP AUTHORIZATION protocol. I have personally used a
number of services that require it, and have mail reader software (several
different packages) that support it.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From ulf.engstrom at b2b-link.com  Fri Feb 16 12:59:03 2001
From: ulf.engstrom at b2b-link.com (=?iso-8859-1?Q?Ulf_Engstr=F8m?=)
Date: Fri, 16 Feb 2001 12:59:03 +0100
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com> <k98k8tkp9k6undl9t71ogvlu0vse4pqa1k@4ax.com>
Message-ID: <033701c0980f$dbadb660$9d8647c1@mobiliant.com>

Hello :)
As in RFC2554(http://www.landfield.com/rfcs/rfc2554.html)

(1) the name of the SMTP service extension is "Authentication"
(2) the EHLO keyword value associated with this extension is "AUTH"
(3) The AUTH EHLO keyword contains as a parameter a space separated
       list of the names of supported SASL mechanisms.
(4) a new SMTP verb "AUTH" is defined
(5) an optional parameter using the keyword "AUTH" is added to the
       MAIL FROM command, and extends the maximum line length of the
       MAIL FROM command by 500 characters.

This is not included in smtplib, but can be used with .ehlo(). One problem
arises; the MAILFROM command has to be accompanied with an AUTH as well on
most servers supporting RFC2554. I made a patch for this a long time ago,
but I've lost it since, don't know it it's possible to obtain it from the
patch-manager on sourceforge in some way.
The number of mailservers supporting this is not in majority, but it's
growing, especially in Europe. If it's interesting for more people I could
probably make a new patch to smtplib, or just create an authsmtplib. Or has
someone done this already?
Ulf Engstr?m

Sheila King <sheila at spamcop.net> wrote:
>
>Nice. And how does one do SMTP authorization? (There must be a way. I see
that
>the NNTP module includes authorization commands, but the smtplib doesn't!)

Tim Roberts:
> There is no such thing.  The SMTP protocol does not include authorization.
>
> Some providers have gone to a scheme whereby you have to check for POP
mail
> before you can use SMTP.  You have to send a password for POP, after which
> they open a short window wherein you can send mail.  That doesn't take any
> changes to smtplib.





From sheila at spamcop.net  Sat Feb 17 17:01:39 2001
From: sheila at spamcop.net (Sheila King)
Date: Sat, 17 Feb 2001 16:01:39 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com> <k98k8tkp9k6undl9t71ogvlu0vse4pqa1k@4ax.com> <mailman.982324940.17913.python-list@python.org>
Message-ID: <mt7t8tcsu43qdg70v235eq02jko79sekeh@4ax.com>

On Fri, 16 Feb 2001 12:59:03 +0100, Ulf Engstr?m <ulf.engstrom at b2b-link.com>
wrote in comp.lang.python in article
<mailman.982324940.17913.python-list at python.org>:

:Hello :)
:As in RFC2554(http://www.landfield.com/rfcs/rfc2554.html)
:
:(1) the name of the SMTP service extension is "Authentication"
:(2) the EHLO keyword value associated with this extension is "AUTH"
:(3) The AUTH EHLO keyword contains as a parameter a space separated
:       list of the names of supported SASL mechanisms.
:(4) a new SMTP verb "AUTH" is defined
:(5) an optional parameter using the keyword "AUTH" is added to the
:       MAIL FROM command, and extends the maximum line length of the
:       MAIL FROM command by 500 characters.

Ulf,

Thanks for providing more tips and suggestions on this topic.

:This is not included in smtplib, but can be used with .ehlo(). One problem
:arises; the MAILFROM command has to be accompanied with an AUTH as well on
:most servers supporting RFC2554. I made a patch for this a long time ago,
:but I've lost it since, don't know it it's possible to obtain it from the
:patch-manager on sourceforge in some way.
:The number of mailservers supporting this is not in majority, but it's
:growing, especially in Europe. If it's interesting for more people I could
:probably make a new patch to smtplib, or just create an authsmtplib. Or has
:someone done this already?

Well, clearly I'm interested. I wonder if others are? I'd suspect there is not
such a big demand, since I was even told in this thread, that there is no such
thing as SMTP authorization, and that hardly anyone uses it! Please don't go
to any trouble on only my account, as I don't know how soon I'd get around to
using it (many other projects on my slate at this time...)

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/


:Sheila King <sheila at spamcop.net> wrote:
:>
:>Nice. And how does one do SMTP authorization? (There must be a way. I see
:that
:>the NNTP module includes authorization commands, but the smtplib doesn't!)




From timr at probo.com  Mon Feb 12 07:24:56 2001
From: timr at probo.com (Tim Roberts)
Date: Sun, 11 Feb 2001 22:24:56 -0800
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com>
Message-ID: <2c0f8tkus5sui67f6jtss56l2s3bm3r6gq@4ax.com>

Imran <ImranG at btinternet.com> wrote:

>When using smtplib can you get the "header text" which mail servers
>introduces themselves with ?

Use the source, Luke.  You have the full source code for the smtplib.py
module at your disposal; the answer lies within:

C:\WINDOWS>python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import smtplib
>>> s = smtplib.SMTP()
>>> print s.connect('mail.probo.com')
(220, 'probo.com ESMTP Sendmail 8.9.3/8.9.3; Sun, 11 Feb 2001 22:21:44
-0800')
>>>

--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From dan at eevolved.com  Sun Feb 11 21:30:44 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 20:30:44 GMT
Subject: retry in try:...except:
Message-ID: <UDCh6.101352$Pm2.1931128@news20.bellglobal.com>

Does something like this exist

def getval(self, key)
        try:
                return self.__keys[key]
        except KeyError:
                self.__keys[key] = None
                retry #!!!!

thanks,
Dan
                


From tim.one at home.com  Sun Feb 11 21:57:39 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 11 Feb 2001 15:57:39 -0500
Subject: retry in try:...except:
In-Reply-To: <UDCh6.101352$Pm2.1931128@news20.bellglobal.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGELMIOAA.tim.one@home.com>

[Dan Parisien]
> Does something like this exist
>
> def getval(self, key)
>         try:
>                 return self.__keys[key]
>         except KeyError:
>                 self.__keys[key] = None
>                 retry #!!!!

No.  In this specific case, you could get the effect you're after by
replacing "retry" with "return self.getval(key)".  Then if you screw up,
you'll eventually get a nice "maximum recursion depth exceeded" error
instead of an infinite loop <0.5 wink>.  Slicker, replace the whole body
with:

    return self.__keys.setdefault(key, None)

or (more obscure to my eyes):

    return self.__keys.setdefault(key)

python-isn't-keen-for-magical-control-flow-ly y'rs  - tim




From dan at eevolved.com  Sun Feb 11 22:48:07 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 21:48:07 GMT
Subject: retry in try:...except:
References: <mailman.981925093.24324.python-list@python.org>
Message-ID: <rMDh6.101667$Pm2.1941818@news20.bellglobal.com>

Tim Peters wrote:

> [Dan Parisien]
> > Does something like this exist
> >
> > def getval(self, key)
> >         try:
> >                 return self.__keys[key]
> >         except KeyError:
> >                 self.__keys[key] = None
> >                 retry #!!!!

>     return self.__keys.setdefault(key, None)
ick :)
 
> or (more obscure to my eyes):
>     return self.__keys.setdefault(key)
yuck :) 

I understand though. 
I would be a big fan of things like this
#------
forever:
        data = socket.recv(...)
        do_something(data)
        if condition == 1:
                break

#------
do:
        something_fancy()
while(condition != false)

#-------
try:
        something()
except:
        fix_possible_error()
        retry

#-------
for i in indexes(list):
        do_something( list[i] )

#-------
obj = Widget()
        setSize(w, h)
        setPos(x, y)
        setBGColor(0)

Those things are lacking or are not documented (documented in a foot note 
to a sub page of a section of a chapter doesn't count ;) IMHO.

That would increase the ease of coding in python (which is already higher 
than any programming language I know).

I know these are things that keep getting brought up... Maybe there's a 
reason ;)

peace,
Dan


From fredrik at pythonware.com  Sun Feb 11 22:59:30 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 11 Feb 2001 21:59:30 GMT
Subject: retry in try:...except:
References: <mailman.981925093.24324.python-list@python.org> <rMDh6.101667$Pm2.1941818@news20.bellglobal.com>
Message-ID: <6XDh6.15667$AH6.2210881@newsc.telia.net>

Dan Parisien wrote:
> try:
>         something()
> except:
>         fix_possible_error()
>         retry

while 1:
    try:
        something()
    except:
        fix_possible_error()
    else:
        break

> for i in indexes(list):
>         do_something( list[i] )

for item in list:
    do_something(item)

(or "def indexes(list): return range(len(list))")

Cheers /F




From dan at eevolved.com  Sun Feb 11 23:09:20 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 22:09:20 GMT
Subject: retry in try:...except:
References: <mailman.981925093.24324.python-list@python.org> <rMDh6.101667$Pm2.1941818@news20.bellglobal.com> <6XDh6.15667$AH6.2210881@newsc.telia.net>
Message-ID: <k4Eh6.229328$JT5.8284353@news20.bellglobal.com>

Fredrik Lundh wrote:

> Dan Parisien wrote:

> while 1:
>     try:
>         something()
>     except:
>         fix_possible_error()
>     else:
>         break

that's what I ended up doing as a work-around. much less readable however

> for item in list:
>     do_something(item)
No. I want the indexes instead of using range(len(list)) over and over

> (or "def indexes(list): return range(len(list))")
I know ;)
but why code it over and over in each script you need it?

I am a big fan of code reuse. Every time I write something twice, I have a 
bad gut feeling because I know I'm doing it uselessly :(

Dan


From f8dy at my-deja.com  Mon Feb 12 00:03:12 2001
From: f8dy at my-deja.com (Mark Pilgrim)
Date: Sun, 11 Feb 2001 23:03:12 GMT
Subject: retry in try:...except:
References: <UDCh6.101352$Pm2.1931128@news20.bellglobal.com>
Message-ID: <9675nb$oli$1@nnrp1.deja.com>

In article <UDCh6.101352$Pm2.1931128 at news20.bellglobal.com>,
  dan at eevolved.com wrote:
> Does something like this exist
>
> def getval(self, key)
>         try:
>                 return self.__keys[key]
>         except KeyError:
>                 self.__keys[key] = None
>                 retry #!!!!
>
> thanks,
> Dan

def getval(self, key): return self.__keys.setdefault(key, None)

## For more information about the setdefault method, see
http://python.org/doc/current/lib/typesmapping.html

-M
--
You're smart; why haven't you learned Python yet?  http://diveintopython.org/


Sent via Deja.com
http://www.deja.com/


From phrxy at csv.warwick.ac.uk  Mon Feb 12 00:07:51 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Sun, 11 Feb 2001 23:07:51 +0000
Subject: retry in try:...except:
In-Reply-To: <UDCh6.101352$Pm2.1931128@news20.bellglobal.com>
References: <UDCh6.101352$Pm2.1931128@news20.bellglobal.com>
Message-ID: <Pine.SOL.4.30.0102112304370.16304-100000@mimosa.csv.warwick.ac.uk>

On Sun, 11 Feb 2001, Dan Parisien wrote:

> Does something like this exist
>
> def getval(self, key)
>         try:
>                 return self.__keys[key]
>         except KeyError:
>                 self.__keys[key] = None
>                 retry #!!!!
[...]

For this particular case,

>>> foo={"eggs": 1}
>>> print foo.get("eggs")
1
>>> print foo.get("spam")
None

But maybe you had another example in mind that this doesn't apply to?


John



From dan at eevolved.com  Mon Feb 12 08:06:29 2001
From: dan at eevolved.com (Dan Parisien)
Date: Mon, 12 Feb 2001 07:06:29 GMT
Subject: retry in try:...except:
References: <UDCh6.101352$Pm2.1931128@news20.bellglobal.com> <Pine.SOL.4.30.0102112304370.16304-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <VXLh6.104370$Pm2.1999546@news20.bellglobal.com>

John J. Lee wrote:

> On Sun, 11 Feb 2001, Dan Parisien wrote:
> 
> > Does something like this exist
> >
> > def getval(self, key)
> >         try:
> >                 return self.__keys[key]
> >         except KeyError:
> >                 self.__keys[key] = None
> >                 retry #!!!!
> [...]
> 
> For this particular case,
> 
> >>> foo={"eggs": 1}
> >>> print foo.get("eggs")
> 1
> >>> print foo.get("spam")
> None

Not this case in particular, just suggesting a retry keyword.

Dan



From slhath at home.com  Sun Feb 11 22:55:31 2001
From: slhath at home.com (Scott Hathaway)
Date: Sun, 11 Feb 2001 21:55:31 GMT
Subject: mxODBC question
Message-ID: <nTDh6.66287$bb.1038987@news1.rdc1.tx.home.com>

I am using mxODBC to access db's with no problem, but I have a question.  If
I use a sql call as follows:

select username from users;

Then, I make a call as follows:

data = c.fetchall()

Then, to access the first value of the first (and only) field, I have to
use:

theValue = data[0][0]

This is ugly.  Is there a way to do something like:

theValue = data[0]['firstname']

Or anything more easy to discern for code readability?

Thanks,
Scott




From emile at fenx.com  Mon Feb 12 00:12:35 2001
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 11 Feb 2001 15:12:35 -0800
Subject: mxODBC question
References: <nTDh6.66287$bb.1038987@news1.rdc1.tx.home.com>
Message-ID: <9676c6$jpmt8$1@ID-11957.news.dfncis.de>

Scott,

I've put together a class to handle this.  I'm only using it
at one site, so it probably won't scale well, but it allows
me to work with the data by using the database column names.
It's geared to handle only sql statements of the form
"select colA, colB, colC from tablename where conditions",
which in my case is about 90% of my usage.  It works by
parsing the sql select statement, mapping the retrieved cols
to class instance attributes with those names,
then returning a list of thoses instances.

--sample usage:

import SQLdata

all_open_ar = SQLdata.cache(('',),
    "select CUSTOMER_ID, INVOICE_ID, TYPE, INVOICE_DATE,
STATUS, \
    TOTAL_AMOUNT, PAID_AMOUNT, DISCOUNT_GIVEN,
LAST_PAID_DATE, ZERO_DATE \
    from RECEIVABLE \
    where INVOICE_DATE < %s",1)

open_ar = all_open_ar('09/30/2000')

for openInvoice in open_ar:
    bal = openInvoice.TOTAL_AMOUNT - openInvoice.PAID_AMOUNT
------

If you're interested, E-mail me, and I'll send it to you.
--

Emile van Sebille
emile at fenx.com
-------------------


"Scott Hathaway" <slhath at home.com> wrote in message
news:nTDh6.66287$bb.1038987 at news1.rdc1.tx.home.com...
> I am using mxODBC to access db's with no problem, but I
have a question.  If
> I use a sql call as follows:
>
> select username from users;
>
> Then, I make a call as follows:
>
> data = c.fetchall()
>
> Then, to access the first value of the first (and only)
field, I have to
> use:
>
> theValue = data[0][0]
>
> This is ugly.  Is there a way to do something like:
>
> theValue = data[0]['firstname']
>
> Or anything more easy to discern for code readability?
>
> Thanks,
> Scott
>
>




From glenfant at equod.com.nospam  Mon Feb 12 12:00:40 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Mon, 12 Feb 2001 12:00:40 +0100
Subject: mxODBC question
References: <nTDh6.66287$bb.1038987@news1.rdc1.tx.home.com>
Message-ID: <968fdg$1pd$1@reader1.imaginet.fr>

I use this small module with mxODBC. These functions (selectdict and
fetchdict) return a dictionnary of tuples.
The keys of the dictionnary are the names of the fields in the 'select'
statement. the tuples are the values of the columns.

Warning: like these function make a "curs.fetchall()", do not use select
statement like "select * from agianttable" or you'll run out of memory.

You can use these like below:

<sample1>
curs = aconnection.cursor()
curs.execute('select name city from people')
result = fetchdict(curs)
count = len(result['name'])
for i in (0, count):
    print result['name'][i], result['city'][i]
</sample1>

or like this if there is no reusable cursor:

<sample2>
result= selectdict(aconnection, 'select name city from people')
count = len(result['name'])
for i in (0, count):
    print result['name'][i], result['city'][i]
</sample2>

<script>
'''fetchdict.py
Maps the result of a "select ..." query to a dictionnary.
Each key of the dictionnary is a field name (depending on query)
Each value is a tuple of consecutive data of the field (order depending on
query)
Requires mxOdbc package.
'''
Ifield = 0

def _getcol(seq):
    global Ifield
    return seq[Ifield]


def fetchdict(curs):
    '''curs is the cursor which just executed a "select..." statement.'''
    global Ifield
    fdict={}
    fldnlist = map(lambda x: x[0], curs.description)
    fetchresult = curs.fetchall()
    Ifield = 0
    for fldname in fldnlist:
        fdict[fldname] = tuple(map(_getcol, fetchresult))
        Ifield = Ifield + 1
    return fdict

def selectdict(cnx, req):
    '''cnx is a connection object created with mxOdbc
    req is an SQL "select..." query'''
    curs = cnx.cursor()
    curs.execute(req)
    rtdict = fetchdict(curs)
    curs.close()
    return rtdict
</script>
"Scott Hathaway" <slhath at home.com> a ?crit dans le message news:
nTDh6.66287$bb.1038987 at news1.rdc1.tx.home.com...
> I am using mxODBC to access db's with no problem, but I have a question.
If
> I use a sql call as follows:
>
> select username from users;
>
> Then, I make a call as follows:
>
> data = c.fetchall()
>
> Then, to access the first value of the first (and only) field, I have to
> use:
>
> theValue = data[0][0]
>
> This is ugly.  Is there a way to do something like:
>
> theValue = data[0]['firstname']
>
> Or anything more easy to discern for code readability?
>
> Thanks,
> Scott
>
>



From sholden at holdenweb.com  Mon Feb 12 13:45:07 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 12 Feb 2001 07:45:07 -0500
Subject: mxODBC question
References: <nTDh6.66287$bb.1038987@news1.rdc1.tx.home.com>
Message-ID: <jZQh6.2381$o4.51421@e420r-atl1.usenetserver.com>

"Scott Hathaway" <slhath at home.com> wrote in message
news:nTDh6.66287$bb.1038987 at news1.rdc1.tx.home.com...
> I am using mxODBC to access db's with no problem, but I have a question.
If
> I use a sql call as follows:
>
> select username from users;
>
> Then, I make a call as follows:
>
> data = c.fetchall()
>
> Then, to access the first value of the first (and only) field, I have to
> use:
>
> theValue = data[0][0]
>
> This is ugly.  Is there a way to do something like:
>
> theValue = data[0]['firstname']
>
> Or anything more easy to discern for code readability?
>
> Thanks,
> Scott
>
The best tool I've come across for this is at

    http://www.lyra.org/greg/python/dtuple.py

It has worked in a number of such contexts for me, allowing access by
attribute name or index, with a well-thought-out set of methods..

regards
 Steve




From slhath at home.com  Wed Feb 14 00:15:05 2001
From: slhath at home.com (Scott Hathaway)
Date: Tue, 13 Feb 2001 23:15:05 GMT
Subject: mxODBC question
References: <nTDh6.66287$bb.1038987@news1.rdc1.tx.home.com>
Message-ID: <Zdji6.72238$bb.1086343@news1.rdc1.tx.home.com>

Thanks for all the feedback!

Scott


"Scott Hathaway" <slhath at home.com> wrote in message
news:nTDh6.66287$bb.1038987 at news1.rdc1.tx.home.com...
> I am using mxODBC to access db's with no problem, but I have a question.
If
> I use a sql call as follows:
>
> select username from users;
>
> Then, I make a call as follows:
>
> data = c.fetchall()
>
> Then, to access the first value of the first (and only) field, I have to
> use:
>
> theValue = data[0][0]
>
> This is ugly.  Is there a way to do something like:
>
> theValue = data[0]['firstname']
>
> Or anything more easy to discern for code readability?
>
> Thanks,
> Scott
>
>




From dan at eevolved.com  Mon Feb 12 00:12:23 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 23:12:23 GMT
Subject: why does this happen?
Message-ID: <r%Eh6.225470$f36.9097034@news20.bellglobal.com>

key = 0
dict = {0:''}
key = dict[key] = "__0"
print key
>>> __0
print dict
>>> {9: '', '__9': '__9'} 

It doesn't do it from right to left like in c? Obviously not. Does anyone 
know why?

thank you,
Dan




From dan at eevolved.com  Mon Feb 12 00:22:49 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 23:22:49 GMT
Subject: why does this happen? (messed up previous post)
References: <r%Eh6.225470$f36.9097034@news20.bellglobal.com>
Message-ID: <d9Fh6.229957$JT5.8294397@news20.bellglobal.com>

Dan Parisien wrote:

> key = 0
> dict = {0:''}
> key = dict[key] = "__0"
> print key
> >>> __0
> print dict
> >>> {9: '', '__9': '__9'}
> 
> It doesn't do it from right to left like in c? Obviously not. Does anyone
> know why?
> 
> thank you,
> Dan
> 
> 
haha. I wrote the example, _then_ copied and pasted from a terminal. serves 
me right.

Sorry for the confusion.

key = "key"
dict = {"key":None}
key = dict[key] = "newval"

print key
>>> newval
print dict
>>> {"key":None, "newval":"newval"}

I was expecting
print dict
>>>{"key":"newval"}

I'm using Python 2.0.1

thank you,
Dan



From scarblac at pino.selwerd.nl  Mon Feb 12 09:43:43 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 12 Feb 2001 08:43:43 GMT
Subject: why does this happen? (messed up previous post)
References: <r%Eh6.225470$f36.9097034@news20.bellglobal.com> <d9Fh6.229957$JT5.8294397@news20.bellglobal.com>
Message-ID: <slrn98f8el.rq.scarblac@pino.selwerd.nl>

Dan Parisien <dan at eevolved.com> wrote in comp.lang.python:
> key = "key"
> dict = {"key":None}
> key = dict[key] = "newval"
> 
> print key
> >>> newval
> print dict
> >>> {"key":None, "newval":"newval"}
> 
> I was expecting
> print dict
> >>>{"key":"newval"}

In C, = is an expression. a = b = c therefore means a = (b = c).

In Python, = is a statement, and a = b = c is simply a shorthand for
a = c; b = c. And that obviously happens left to right. This is also the
reason why in Python "a = (b = c)" is illegal.

I'd consider 'key = dict[key] = "newval"' to be deliberately confusing.
Once you're depending on the order of that sort of thing, your code is hard
to read. I had no idea what to expect when I read it, really.

Use the multiple = when you are setting several similar variables to the
same value, length = width = height = 1, but don't try to make the meaning
of one of them depend on the value of another...

-- 
Remco Gerlich


From gregj at pobox.com  Mon Feb 12 01:00:49 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 12 Feb 2001 00:00:49 GMT
Subject: why does this happen?
References: <r%Eh6.225470$f36.9097034@news20.bellglobal.com>
Message-ID: <96793e$rd6$1@nnrp1.deja.com>

In article <r%Eh6.225470$f36.9097034 at news20.bellglobal.com>,
dan at eevolved.com wrote:
> key = 0
> dict = {0:''}
> key = dict[key] = "__0"
> print key
> >>> __0
> print dict
> >>> {9: '', '__9': '__9'}

** I think you meant {0: '', '__0': '__0'}

>
> It doesn't do it from right to left like in c? Obviously not. Does
anyone
> know why?

In C assignment is an expression: the value of the expression a=1 is 1.
In Python, assignment is not an expression: print a=1 is an error.
Python allows assignment to several variables "simultaneously" (that's
what the Python Tutorial says) with the form a = b = 1. I don't see
anything in the Language Reference that says what order the assignments
are actually carried out, but from your experiment and my own it looks
like a = b = 1 is equivalent to:

a = 1
b = 1

In other words, the rightmost expression is assigned to each lvalue in
left-to-right order. But unless the order of assignment is explicitly
specified in the language spec you should not rely on it.

Perhaps a clearer example is:

>>> i = 0
>>> t = [0,0]
>>> i = t[i] = 1
>>> i
1
>>> t
[0,1]


--
Greg Jorgensen
Portland, Oregon, USA
gregj at pobox.com


Sent via Deja.com
http://www.deja.com/


From dan at eevolved.com  Mon Feb 12 01:14:24 2001
From: dan at eevolved.com (Dan Parisien)
Date: Mon, 12 Feb 2001 00:14:24 GMT
Subject: why does this happen?
References: <r%Eh6.225470$f36.9097034@news20.bellglobal.com> <96793e$rd6$1@nnrp1.deja.com>
Message-ID: <AVFh6.230276$JT5.8301482@news20.bellglobal.com>

> Greg Jorgensen
> Portland, Oregon, USA
> gregj at pobox.com

thank you very much :)

Dan



From cfelling at iae.nl  Mon Feb 12 02:22:05 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: Mon, 12 Feb 2001 01:22:05 GMT
Subject: why does this happen?
References: <r%Eh6.225470$f36.9097034@news20.bellglobal.com> <96793e$rd6$1@nnrp1.deja.com>
Message-ID: <967dus$iic$1@animus.fel.iae.nl>

> In C assignment is an expression: the value of the expression a=1 is 1.
> In Python, assignment is not an expression: print a=1 is an error.
> Python allows assignment to several variables "simultaneously" (that's
> what the Python Tutorial says) with the form a = b = 1. I don't see
> anything in the Language Reference that says what order the assignments
> are actually carried out, but from your experiment and my own it looks

>From the Python Language Reference 1.5.2:

          next up previous Python Reference Manual contents index

   Next: 6.4 The pass statement Up: 6. Simple statements Previous: 6.2
   Assert statements
     _________________________________________________________________

                           6.3 Assignment statements

   Assignment statements are used to (re)bind names to values and to
   modify attributes or items of mutable objects:

assignment_stmt: (target_list "=")+ expression_list
target_list:     target ("," target)* [","]
target:          identifier | "(" target_list ")" | "[" target_list "]"
               | attributeref | subscription | slicing

   (See section 5.3 for the syntax definitions for the last three
   symbols.)

   An assignment statement evaluates the expression list (remember that
   this can be a single expression or a comma-separated list, the latter
   yielding a tuple) and assigns the single resulting object to each of
   the target lists, from left to right.

So it is, from left to right:) 
-- 
groetjes, carel


From pretomal at yahoo.com  Mon Feb 12 00:41:56 2001
From: pretomal at yahoo.com (pretomal at yahoo.com)
Date: Sun, 11 Feb 2001 23:41:56 GMT
Subject: TclError: while using Tkinter()
Message-ID: <967802$qgq$1@nnrp1.deja.com>

Win98, ActivePython 2.0

I'm new to Python, and cannot get Tkinter to work.  Tk/Tcl8.32 is
installed, and I had to copy tcl83.dll and tk83.dll to my
\windows\system\ dir. to get Tkinter to load, but now I get the
following error while trying to call "Tkinter.Tk()" :

  File "c:\python20\lib\lib-tk\Tkinter.py", line 1482, in __init__
    self.tk = _tkinter.create(screenName, baseName, className)
TclError: invalid command name "tcl_findLibrary"

By searching comp.lang.tcl I'm led to believe that Tk.Init() is being
called before Tcl.Init(), but I'm clueless as how to solve that.
Any help is appreciated.

Pretomal


Sent via Deja.com
http://www.deja.com/


From rvprasad at cis.ksu.edu  Mon Feb 12 02:46:34 2001
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 11 Feb 2001 19:46:34 -0600
Subject: DOM creation
Message-ID: <m3bss8k70l.fsf@boss.dreamsoft.com>

Hi,

I have a question on how DOM for a XML document conforming to DOM 2 should be
constructed?

Now if there are no namespaces specified in the document then should attributes
be added to DOM using 
setAttributeNS('', Name, Value) 
or 
setAttribute(Name, Value)?

The problem I am facing is when reading in a XML document with no explicit
namespace specified in it through PyXML the attributes are added to the DOM
using setAttributeNS with an empty NameSpace.  So, I wanted to clarify if this
is a problem with PyXML or is this how other DOM Constructors work.

waiting for reply,

-- 
Venkatesh Prasad Ranganath


From alf at leo.logilab.fr  Mon Feb 12 15:50:53 2001
From: alf at leo.logilab.fr (Alexandre Fayolle)
Date: Mon, 12 Feb 2001 14:50:53 +0000 (UTC)
Subject: DOM creation
References: <m3bss8k70l.fsf@boss.dreamsoft.com>
Message-ID: <968t8d$23mg$1@norfair.nerim.net>

In comp.lang.python Venkatesh Prasad Ranganath <rvprasad at cis.ksu.edu> wrote:
> Now if there are no namespaces specified in the document then should attributes
> be added to DOM using 
> setAttributeNS('', Name, Value) 
> or 
> setAttribute(Name, Value)?

Use the first one, definitely. Event better, define a symbolic constant 
that you'll call, for example EMPTY_NS, and set it to '' for now. There 
has been  some discussion about what should be the prefix of empty namespace 
on xml-sig ('' or None), and it is likely that it will change from empty 
string to None soon (someone corrects me if I'm wrong, please.)

> The problem I am facing is when reading in a XML document with no explicit
> namespace specified in it through PyXML the attributes are added to the DOM
> using setAttributeNS with an empty NameSpace.  So, I wanted to clarify if this
> is a problem with PyXML or is this how other DOM Constructors work.

This is what the DOM L2 spec says. It also says that you should not mix 
the ns aware and non-ns aware APIs provided in DOM L2. So this means that 
you shoud use Document.createElementNS(EMPTY_NS,name) instead of 
Document.createElement(name) to create new element nodes.

I hope this helps.

Alexandre Fayolle
-- 
http://www.logilab.com 
Narval is the first software agent available as free software (GPL).
LOGILAB, Paris (France).


From uche at ogbuji.net  Wed Feb 14 15:12:30 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Wed, 14 Feb 2001 14:12:30 GMT
Subject: DOM creation
References: <m3bss8k70l.fsf@boss.dreamsoft.com>
Message-ID: <3A8A9254.C4F9FDC@ogbuji.net>

Venkatesh Prasad Ranganath wrote:

> I have a question on how DOM for a XML document conforming to DOM 2 should be
> constructed?
> 
> Now if there are no namespaces specified in the document then should attributes
> be added to DOM using
> setAttributeNS('', Name, Value)
> or
> setAttribute(Name, Value)?
> 
> The problem I am facing is when reading in a XML document with no explicit
> namespace specified in it through PyXML the attributes are added to the DOM
> using setAttributeNS with an empty NameSpace.  So, I wanted to clarify if this
> is a problem with PyXML or is this how other DOM Constructors work.

This is correct behavior.  Of course, if you use the
xml.dom.ext.reader.Sax reader, you get a tree with no namespace
specifiers at all, which is also correct.

If you plan to migrate to namespaces in future, or to mix namespace with
non-namespace behavior, I'd suggest sticking to the PyExpat and Sax2
readers and using the DOm Level 2 methods (with appended "NS").

> waiting for reply,

This reminds me.  I'm not sure I sent a reply to your last enquiry.

I had a few questions, such as which Reader you were trying to use (It
looked as if you didn't paste all of your example code in).

However, I'd suggest trying the latest 4Suite 0.10.2 beta that was
announced (since I noticed you're using XPath), and see if you still
have those problems.  If so, pleace copy follow-ups to
xml-sig at python.org, which I check more regularly than this newsgroup.

Thanks.

-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From rvprasad at cis.ksu.edu  Wed Feb 14 22:00:25 2001
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 14 Feb 2001 15:00:25 -0600
Subject: DOM creation
References: <m3bss8k70l.fsf@boss.dreamsoft.com> <3A8A9254.C4F9FDC@ogbuji.net>
Message-ID: <m3k86t9dzq.fsf@boss.dreamsoft.com>

>>>>> "Uche" == Uche Ogbuji <uche at ogbuji.net> writes:

    Uche> Venkatesh Prasad Ranganath wrote:
    >> I have a question on how DOM for a XML document conforming to DOM 2
    >> should be constructed?
    >> 
    >> Now if there are no namespaces specified in the document then should
    >> attributes be added to DOM using setAttributeNS('', Name, Value) or
    >> setAttribute(Name, Value)?
    >> 
    >> The problem I am facing is when reading in a XML document with no
    >> explicit namespace specified in it through PyXML the attributes are added
    >> to the DOM using setAttributeNS with an empty NameSpace.  So, I wanted to
    >> clarify if this is a problem with PyXML or is this how other DOM
    >> Constructors work.

    Uche> This is correct behavior.  Of course, if you use the
    Uche> xml.dom.ext.reader.Sax reader, you get a tree with no namespace
    Uche> specifiers at all, which is also correct.

    Uche> If you plan to migrate to namespaces in future, or to mix namespace
    Uche> with non-namespace behavior, I'd suggest sticking to the PyExpat and
    Uche> Sax2 readers and using the DOm Level 2 methods (with appended "NS").

If this is the case then should or shouldn't set/getAttribute() in DOM2
"intelligently" assume empty namespace ('')?  Also, is there any specification
on construction of DOM for XML documents?  Or does the DOM specs available at
W3C describe the construction process?  If so, can somebody tell me in which
section?

    >> waiting for reply,

    Uche> This reminds me.  I'm not sure I sent a reply to your last enquiry.

    Uche> I had a few questions, such as which Reader you were trying to use (It
    Uche> looked as if you didn't paste all of your example code in).

    Uche> However, I'd suggest trying the latest 4Suite 0.10.2 beta that was
    Uche> announced (since I noticed you're using XPath), and see if you still
    Uche> have those problems.  If so, pleace copy follow-ups to
    Uche> xml-sig at python.org, which I check more regularly than this newsgroup.

I guess it works fine on with 0.10.2b.

Thankx

-- 
Venkatesh Prasad Ranganath


From dsavitsk at e-coli.net  Mon Feb 12 03:04:14 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 11 Feb 2001 20:04:14 -0600
Subject: Python on FreeBSD
Message-ID: <3A87449E.47AACA64@e-coli.net>

greetings;

i am having some trouble wih python on FreeBSD.  I installed python 2.0
from the package collection.  Then when i installed idle from the ports
collection, it said that python 1.5.2 was a dependency and installed
that too.

That's not the problem (i don't think anyway).  The problem is that when
i tried to make install mod_python it couldn't find a checksum for
python15 (which as far as i know installed fine as idle works fine).  is
it safe to force the build w/o a checksum? or is it better to CVSup and
try to make install again? or will that do anything different?

os is FreeBSD 4.2.  i downloaded and burned the iso, so packages are
installed from cdrom, but i install ports via ftp.

any help, suggestions, places to look would be welcome as i am new to
the unix world.

doug


From chris at voodooland.net  Mon Feb 12 04:43:13 2001
From: chris at voodooland.net (Chris Watson)
Date: Sun, 11 Feb 2001 21:43:13 -0600 (CST)
Subject: Python on FreeBSD
In-Reply-To: <3A87449E.47AACA64@e-coli.net>
Message-ID: <Pine.BSF.4.21.0102112141470.23432-100000@open-systems.net>

> greetings;
> 
> 
> That's not the problem (i don't think anyway).  The problem is that when
> i tried to make install mod_python it couldn't find a checksum for
> python15 (which as far as i know installed fine as idle works fine).  is
> it safe to force the build w/o a checksum? or is it better to CVSup and
> try to make install again? or will that do anything different?

You need to update your ports tree with cvsup. The md5 checksum on
mod_python changed. And no it isnt going to hurt anything to over ride it
and install it with a mismatched checksum. Not a big deal at all :-)

--
=============================================================================
-Chris Watson         (316) 326-3862 | FreeBSD Consultant, FreeBSD Geek 
Work:              scanner at jurai.net | Open Systems Inc., Wellington, Kansas
Home:  scanner at deceptively.shady.org | http://open-systems.net
=============================================================================
WINDOWS: "Where do you want to go today?"
LINUX: "Where do you want to go tomorrow?"
BSD: "Are you guys coming or what?"
=============================================================================
GNU = Gnu's Non-portable & Unstable
irc.openprojects.net #FreeBSD -Join the revolution!
ICQ: 20016186




From gryphon01 at my-deja.com  Mon Feb 12 05:16:39 2001
From: gryphon01 at my-deja.com (gryphon01 at my-deja.com)
Date: Mon, 12 Feb 2001 04:16:39 GMT
Subject: test_pyexpat failed
Message-ID: <967o32$75r$1@nnrp1.deja.com>

I an trying to build Python 2.0 with expat 1.2. The compile doesn't complain
but test_pyexpat says: test test_pyexpat failed -- Writing:
"('xml=http://www.w3.org/XML/1998/namespace\\014external_entity', None,
'entity.file', None)", expected:
"('http://www.python.org/namespace=http://www.w3.org/XML/1998/namespace\\014e
xternal_entity'"

What is going on here?  I must be doing something
wrong.

Thanks

- Clint


Sent via Deja.com
http://www.deja.com/


From marc.vollmer at mahr.de  Mon Feb 12 07:45:24 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Mon, 12 Feb 2001 07:45:24 +0100
Subject: Newbie: Adress of Pointer from C
Message-ID: <9681bl$k4960$1@ID-59219.news.dfncis.de>

Hi,

I have an SocketServer (TCP/IP) as an C++ program and my SocketClient is a
Python-script. Now I  want to transfer the address of a pointer to Python,
but I don't think how I can read the address and can read-out the pointer
datas in the python script.

System: Python 2.0
OS: WinNT

Thanks and bye

Marc





From stephen_purcell at yahoo.com  Mon Feb 12 08:49:15 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 12 Feb 2001 08:49:15 +0100
Subject: Newbie: Adress of Pointer from C
In-Reply-To: <9681bl$k4960$1@ID-59219.news.dfncis.de>; from marc.vollmer@mahr.de on Mon, Feb 12, 2001 at 07:45:24AM +0100
References: <9681bl$k4960$1@ID-59219.news.dfncis.de>
Message-ID: <20010212084915.D24337@freedom.puma-ag.com>

Marc Vollmer wrote:
> 
> I have an SocketServer (TCP/IP) as an C++ program and my SocketClient is a
> Python-script. Now I  want to transfer the address of a pointer to Python,
> but I don't think how I can read the address and can read-out the pointer
> datas in the python script.

Eek! Don't expect Python to help you look at what the pointer is pointing
at. (Isn't it pointing at something in the server process anyway?). What you
can probably do is read the pointer value from the server across the socket,
but I can't think how that would help you later.

-Steve

-- 
Steve Purcell, Pythangelist
PythonTesting  = urlopen('http://pyunit.sourceforge.net/')
PythonServlets = urlopen('http://pyserv.sourceforge.net/')
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From marc.vollmer at mahr.de  Mon Feb 12 08:48:28 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Mon, 12 Feb 2001 08:48:28 +0100
Subject: Newbie: Windows DLL Access
Message-ID: <9684p5$jktge$1@ID-59219.news.dfncis.de>

I try to use a Visual C++ library (.lib or .dll) , but python give me this
error:

import medvc50f
ImportError: dynamic module does not define init function (initmedvc50f)

How I can use DLL od LIBS in python? Or where is a description?

Thanks
Marc





From stephen_purcell at yahoo.com  Mon Feb 12 10:06:54 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 12 Feb 2001 10:06:54 +0100
Subject: Newbie: Windows DLL Access
In-Reply-To: <9684p5$jktge$1@ID-59219.news.dfncis.de>; from marc.vollmer@mahr.de on Mon, Feb 12, 2001 at 08:48:28AM +0100
References: <9684p5$jktge$1@ID-59219.news.dfncis.de>
Message-ID: <20010212100654.B25640@freedom.puma-ag.com>

Marc Vollmer wrote:
> I try to use a Visual C++ library (.lib or .dll) , but python give me this
> error:
> 
> import medvc50f
> ImportError: dynamic module does not define init function (initmedvc50f)
> 
> How I can use DLL od LIBS in python? Or where is a description?

Check out 'calldll'; http://www.nightmare.com/software.html

-Steve

-- 
Steve Purcell, Pythangelist
PythonTesting  = urlopen('http://pyunit.sourceforge.net/')
PythonServlets = urlopen('http://pyserv.sourceforge.net/')
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From ransen_spam_me_not at nemo.it  Mon Feb 12 09:03:05 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Mon, 12 Feb 2001 08:03:05 GMT
Subject: Embedding Python on the Mac
Message-ID: <3a8797bb.2899561@news.newsguy.com>

I'm new to python and new to the mac...
...I've heard from various contacts that embedding
python in a Windows app is reasonably easy, but
has anyone experience of doing it for the Mac?


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From rtrocca at libero.it  Tue Feb 13 09:06:48 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Tue, 13 Feb 2001 08:06:48 GMT
Subject: Embedding Python on the Mac
References: <3a8797bb.2899561@news.newsguy.com>
Message-ID: <1eoqke6.gkpx5c1ydxrokN%rtrocca@libero.it>

Not so easy to be sincere. 
I mean: it is not so difficult if you are using C. It is uch more
difficult in C++. Anyway I didn't spend too much time on that. It is
better to ask the same question on the MacPython-SIG.
The MacPython distribution comes with some example (working) that show
how to do that.

Owen F. Ransen <ransen_spam_me_not at nemo.it> wrote:

> I'm new to python and new to the mac...
> ...I've heard from various contacts that embedding
> python in a Windows app is reasonably easy, but
> has anyone experience of doing it for the Mac?
> 
> 
> --
> Owen F. Ransen
> http://www.ransen.com/
> Home of Gliftic & Repligator Image Generators


From ransen_spam_me_not at nemo.it  Wed Feb 14 07:30:25 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Wed, 14 Feb 2001 06:30:25 GMT
Subject: Embedding Python on the Mac
References: <3a8797bb.2899561@news.newsguy.com> <1eoqke6.gkpx5c1ydxrokN%rtrocca@libero.it>
Message-ID: <3a8b6d8a.2151036@news.newsguy.com>

On Tue, 13 Feb 2001 08:06:48 GMT, rtrocca at libero.it (Riccardo Trocca)
wrote:

>Not so easy to be sincere. 
>I mean: it is not so difficult if you are using C. It is uch more
>difficult in C++. Anyway I didn't spend too much time on that. It is
>better to ask the same question on the MacPython-SIG.
>The MacPython distribution comes with some example (working) that show
>how to do that.

Thanks for the feedback - sei Italiano? Chiedo perche
abito vicino a Milano...


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From rtrocca at libero.it  Thu Feb 15 08:26:26 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Thu, 15 Feb 2001 07:26:26 GMT
Subject: Embedding Python on the Mac
References: <3a8797bb.2899561@news.newsguy.com> <1eoqke6.gkpx5c1ydxrokN%rtrocca@libero.it> <3a8b6d8a.2151036@news.newsguy.com>
Message-ID: <1eou9il.92hl7f16x7746N%rtrocca@libero.it>

Si si, sono Italiano, studio a genova. Se hai bisogno di qualche aiuto
con MacPython e filmati/immagini (vista la tua signature) fatti sentire.


Owen F. Ransen <ransen_spam_me_not at nemo.it> wrote:

> On Tue, 13 Feb 2001 08:06:48 GMT, rtrocca at libero.it (Riccardo Trocca)
> wrote:
> 
> >Not so easy to be sincere. 
> >I mean: it is not so difficult if you are using C. It is uch more
> >difficult in C++. Anyway I didn't spend too much time on that. It is
> >better to ask the same question on the MacPython-SIG.
> >The MacPython distribution comes with some example (working) that show
> >how to do that.
> 
> Thanks for the feedback - sei Italiano? Chiedo perche
> abito vicino a Milano...
> 
> 
> --
> Owen F. Ransen
> http://www.ransen.com/
> Home of Gliftic & Repligator Image Generators


From fredrik at pythonware.com  Mon Feb 12 10:22:00 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 12 Feb 2001 10:22:00 +0100
Subject: ANN: SRE 2.1a2 (separate distribution)
Message-ID: <054201c094d5$41b3a110$0900a8c0@SPIFF>

in case you don't want to upgrade Python right now:
 
    Secret Labs' Regular Expression Engine (SRE) is the new regular
    expression engine shipped with Python 1.6 and 2.0.  The latest
    SRE version (from Python 2.1a2) is now available separately:

        http://www.pythonware.com/products/sre

    Changes from the 2.0 include a couple of bug fixes, and improved
    portability: this version also works under Python 1.5.2.

    The distribution includes library source code, and prebuilt Windows
    binaries for Python 1.5.2 and 2.0.

enjoy,
the pythonware team

    "Secret Labs -- makers of fine pythonware since 1997."




From vbg-news at bitjungle.com  Mon Feb 12 11:31:37 2001
From: vbg-news at bitjungle.com (Vidar Gundersen)
Date: Mon, 12 Feb 2001 10:31:37 GMT
Subject: __file__ with pythonw
Message-ID: <wkwvawtct0.fsf@bitjungle.com>

how can i get a string with the path to
the current python script?
(should work both with .py and .pyw)


problem using __file__:
the statement below works as expected with python.exe,
but fails with pythonw.exe because:
-- variable __file__ does not exist (?)
-- sys.path[0] is empty
how come? is it a bug? system info: python 2.0, windows 95

try: filepath = os.path.split(__file__)[0]
except: filepath = sys.path[0]


-- 
Vidar Bronken Gundersen


From gavin at madzag.com  Mon Feb 12 12:07:49 2001
From: gavin at madzag.com (Gavin Tomlins)
Date: Mon, 12 Feb 2001 11:07:49 GMT
Subject: Permissions or mutable objects... Read only file contents template to write only?
Message-ID: <9uPh6.16771$o85.85546@news-server.bigpond.net.au>

Greetings

I'm currently faced with a problem regarding permissions. I'm trying to read in 
x# of template files to construct an output file.

eg.

templateHdr.txt
templateFooter.txt
outputResult.txt

I attempt to read in the files then  redirect the inputted template to the 
output file ergo,

cTemplateHdr = "templateHdr.txt"
cTemplateFooter = "templateFooter.txt"
cOutput = "outputResult.txt"

#open files for various operations
fileHdr = open(cTemplateHdr, "r")
fileFooter = open(cTemplateFooter, "r')
fileOutput = open(cOutput, "w")

cTempStr = ""
cTempStr = fileHdr.readlines()

fileOutput.write(cTempStr)

at this point I am presented with a type error indicating that I'm trying to 
write a 'read buffer'.

What is the process for rectifying this problem. Any assistance, reference 
would be greatly appreciated. I've searched the Python documentation but was 
unable to find appropriate documentation. If possible, could people cite the 
reference for my understanding. I can only put it down to lack of comprehension 
or poor choice of search criteria on my behalf.

Regards

Gavin

gavin at madzag.com


From ivnowa at hvision.nl  Mon Feb 12 13:56:43 2001
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 12 Feb 2001 12:56:43 GMT
Subject: Permissions or mutable objects... Read only file contents template to write only?
References: <9uPh6.16771$o85.85546@news-server.bigpond.net.au>
Message-ID: <3A87DBA4.710B@hvision.nl>

Gavin Tomlins wrote:

> #open files for various operations
> fileHdr = open(cTemplateHdr, "r")
> fileFooter = open(cTemplateFooter, "r')
> fileOutput = open(cOutput, "w")
> 
> cTempStr = ""
> cTempStr = fileHdr.readlines()
> 
> fileOutput.write(cTempStr)
> 
> at this point I am presented with a type error indicating that I'm trying to
> write a 'read buffer'.

fileHdr.readlines() returns a list. So, you are trying to write a list
of strings to fileOutput, while the .write method wants a string.

You can loop over the list, writing it line by line:

    for line in cTempStr:
        fileOutput.write(line)

or use the .writelines method:

    fileOutput.writelines(cTempStr)

Maybe you wanted to read only one line from the input file; in that case
you can use the .readline method.

> What is the process for rectifying this problem. Any assistance, reference
> would be greatly appreciated. I've searched the Python documentation but was
> unable to find appropriate documentation. If possible, could people cite the
> reference for my understanding. I can only put it down to lack of comprehension
> or poor choice of search criteria on my behalf.

See the Python library reference, 2.1.7.9.

HTH,


From walterm at cmat.edu.uy  Mon Feb 12 14:49:07 2001
From: walterm at cmat.edu.uy (Walter Moreira)
Date: Mon, 12 Feb 2001 10:49:07 -0300
Subject: root permissions
Message-ID: <20010212104907.A10918@cmat.edu.uy>

Hello Python people.

Hope this question not to be too off topic.
Inside a Python script I must read a file owned by root with permissions
600. But I want the script can be executed by a normal user. What can I do?

I am using sudo but this implies that the users must be listed in /etc/sudoers
and they get all the privileges. Is there an elegant solution?

Thanks.  -- Walter

-- 
--------------
Walter Moreira # Centro de Matematica # Universidad de la Republica
walterm at cmat.edu.uy     http://www.cmat.edu.uy/~walterm     Uruguay
        /OD\_        | "Agent Smith: Never send a human to do a
 O o . |_o_o_)       |  machine's job."      -- From ``The Matrix''



From cfelling at iae.nl  Tue Feb 13 00:13:26 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 13 Feb 2001 00:13:26 +0100
Subject: root permissions
References: <mailman.981985812.904.python-list@python.org>
Message-ID: <969qmm$8pd$1@animus.fel.iae.nl>

Walter Moreira <walterm at cmat.edu.uy> wrote:
> Hello Python people.

> Hope this question not to be too off topic.
> Inside a Python script I must read a file owned by root with permissions
> 600. But I want the script can be executed by a normal user. What can I do?

> I am using sudo but this implies that the users must be listed in /etc/sudoers
> and they get all the privileges. Is there an elegant solution?

There are basicly three ways to deal with this:

1) restrict sudo acces to your program only

2) wrap your program in a real executable and suid-it
   (example in the Python source: Misc/setuid-prog.c)

3) make a small suid-ed c program to access that root-owned file
   and nothing else, and call that program from within your Python script

1 and 2 share the same huge disadvantage of your program being run as
root!  Be very carefull in your programming, drop root privelige as
soon as possible, think twice and then refrain from it all to gether.
-- 
groetjes, carel


From erno-news at erno.iki.fi  Tue Feb 13 01:51:26 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Feb 2001 02:51:26 +0200
Subject: root permissions
References: <mailman.981985812.904.python-list@python.org>
Message-ID: <kur913xv5d.fsf@lasipalatsi.fi>

In article <mailman.981985812.904.python-list at python.org>, Walter
Moreira <walterm at cmat.edu.uy> writes:

| I am using sudo but this implies that the users must be listed in
| /etc/sudoers and they get all the privileges. Is there an elegant
| solution?

it is possible to configure sudo so that some users can only
execute some commands.

   -- erno


From un4e at rzstud1.rz.uni-karlsruhe.de  Mon Feb 12 15:35:40 2001
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 12 Feb 2001 14:35:40 GMT
Subject: PGP wrapper?
Message-ID: <968sbs$9of$1@news.rz.uni-karlsruhe.de>

Hi,

I'm looking for a PGP wrapper written in Python that supports key
generation. Unfortunately, the -kg command in PGP requires a lot of
interaction; I've tried to wrap it and despaired... Does anyone have such a
beast? Or is anyone aware of a library version of PGP, so that I might call
the key-generation process directly?

Thanks in advance,
		--Bj?rn


From matthias.horst at ruhr-uni-bochum.de  Mon Feb 12 15:50:12 2001
From: matthias.horst at ruhr-uni-bochum.de (matthias.horst at ruhr-uni-bochum.de)
Date: Mon, 12 Feb 2001 14:50:12 GMT
Subject: Python COM Object Return Values in Delphi
Message-ID: <968t70$304$1@nnrp1.deja.com>

God's blessing for anybody who can help me with the following problem:

I have written a Com server object in Python 2.0 similar to that (this
is not the problem)

class testClass:

    _public_methods_ = ['daddel']
    _reg_progid_ = "somthing"
    _reg_clsid_ = "{....}"

    def daddel(self):
        return ("hello",1)


if __name__=='__main__':
    import win32com.server.register
    win32com.server.register.UseCommandLine(testClass)

Does anybody know how I can access the return value of method daddel?
For a simple string or integer it's clear :

procedure callIt;
var obj:variant;
    dummy:string;
begin
   obj=createOleObject("putInHereTheName")
   dummy=obj.callHereTheMethod;
end;

But for a complex tupple...?!?!?

Thanks in advance.


Sent via Deja.com
http://www.deja.com/


From embed at geocities.com  Mon Feb 12 16:03:07 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 12 Feb 2001 10:03:07 -0500
Subject: Signals between threads
Message-ID: <UXSh6.2933$D3.13081@tor-nn1.netcom.ca>

I am limited by my embedded platform to a single process, with multiple
threads.

I have implemented a time.sleep() modification that makes it interruptable
(I use WaitForSingleObject( hInterruptSleepEvent,ms ) != WAIT_TIMEOUT) in
timemodule.c.

Now I'm considering some improvements to the thread-by-thread signalling
code in my personal python 1.5.2 variant:

import thread,time

def myfunc():
    print "starting"
    try:
        time.sleep(999999)
   except:
       print "time.sleep blew up. must have been sent an exception"
    print "stopping"

thrd = thread.start_new_thread( myfunc, () ) # currently thrd would be None.
time.sleep(1.0)
thrd.raise_exception(KeyboardInterrupt) # the wishfull-thinking bit.
time.sleep(1.0)


This is very much a practical effort to get real work done, but also in the
long term I'm intereseted in contributing threading improvements to the 2.xx
codebase.

In particular, the ability of one thread to send a signal (raise an
exception in another thread), and for the built-in thread module to return a
an object that can be used for this purpose.  (Currently
thread.start_new_thread returns None. I consider that to be Horrid
Behaviour, and the module threading.py to be a nice-API, but which should be
implemented in Python itself.)

I agree with posters who said that interruptable and "killable" threads are
a bad, or dangerous, idea.    I think the ability to inject an exception
into another thread could easily be abused, however I also think it needs to
be there, for when it's essential to getting a job done.

Is anyone else in a similar bind? What are you considering?

Warren Postma





From parz at home.com  Mon Feb 12 21:57:01 2001
From: parz at home.com (Parzival Herzog)
Date: Mon, 12 Feb 2001 20:57:01 GMT
Subject: Signals between threads
References: <UXSh6.2933$D3.13081@tor-nn1.netcom.ca>
Message-ID: <x6Yh6.17192$wI1.1584771@news1.rdc1.mb.home.com>

"Warren Postma" <embed at geocities.com> wrote in message
news:UXSh6.2933$D3.13081 at tor-nn1.netcom.ca...

> In particular, the ability of one thread to send a signal (raise an
> exception in another thread), and for the built-in thread module to return
a
> an object that can be used for this purpose.  (Currently
> thread.start_new_thread returns None. I consider that to be Horrid
> Behaviour, and the module threading.py to be a nice-API, but which should
be
> implemented in Python itself.)
>
> I agree with posters who said that interruptable and "killable" threads
are
> a bad, or dangerous, idea.    I think the ability to inject an exception
> into another thread could easily be abused, however I also think it needs
to
> be there, for when it's essential to getting a job done.

Absolutely: This is the single greatest defect in Python threads. You have
to have a way
to signal threads independent of whether they are blocked. In Modula-3, an
exception,
defined in the threading interface: "Thread.Alerted" is used to accomplish
this. This
makes Modula-3 threads practical to use. Python requires that threads poll
for
exceptional conditions, e.g "while self.keepRunning:" without providing a
guaranteed
way to ensure that the polling loop will actually run (i.e the thread could
be blocked
for abitrary amounts of time in i/O, or in consume arbitrary amounts of time
in computation
before returning to the polling loop. These requirements arise in real
applications that
need to be robust and reliable in an unreliable environment.

So yes, lets have a "thread.alerted" exception!

- Parzival




From Pieter_Claerhout at CreoScitex.com  Mon Feb 12 16:04:27 2001
From: Pieter_Claerhout at CreoScitex.com (Pieter Claerhout)
Date: Mon, 12 Feb 2001 16:04:27 +0100
Subject: Command prompts on a Windows station
Message-ID: <2B1262E83448D211AE4B00A0C9D61B030130191A@msgeuro1.177.115.149.IN-ADDR.ARPA>

Hello all,

is there a way in Python to catch that a command prompt was closed by
clicking on the close button in the title bar? I have a script that is
launching several other scripts, and I would like to kill those processes
again once the main script is closed again. I can catch this when he is
pressing CTRL+BREAK, but I can seem to catch is when the window is closed
using the close button? Any ideas? I remember that an email about this was
going around, but I can't seem to find it back...

Thanks,


Pieter



From timr at probo.com  Wed Feb 14 07:56:30 2001
From: timr at probo.com (Tim Roberts)
Date: Tue, 13 Feb 2001 22:56:30 -0800
Subject: Command prompts on a Windows station
References: <mailman.981990716.19562.python-list@python.org>
Message-ID: <tqak8togo545f4decs8hdtqjuv4jvo0kek@4ax.com>

Pieter Claerhout <Pieter_Claerhout at CreoScitex.com> wrote:
>
>is there a way in Python to catch that a command prompt was closed by
>clicking on the close button in the title bar?

Not as far as I know.  Clicking the 'close' button is a drastic
termination; it is a signal to the application that displays DOS windows
(WINOLDAP), which closes the VM without any further notice to the programs
running in the VM.  That's why, ordinarily, you get a "warning, is this
really what you want to do?" dialog when you close it that way.

If you were a clever sort, you might be able to do some magic with the
Win32 API.  You could use FindWindow to locate the window, subclass the
window with your own WndProc, and watch for WM_CLOSE messages.  However,
that's not very Python-like.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From bragib at my-deja.com  Mon Feb 12 16:59:25 2001
From: bragib at my-deja.com (bragib at my-deja.com)
Date: Mon, 12 Feb 2001 15:59:25 GMT
Subject: How to end process started in os.popen?
Message-ID: <96918n$6sn$1@nnrp1.deja.com>

I am running a csh script and getting the output through a pipe using
os.popen() on Unix.  I would like to be able to kill the running of the
csh script if a certain event occurs.

Just closing the pipe won't stop the csh script.

Bragi


Sent via Deja.com
http://www.deja.com/


From erno-news at erno.iki.fi  Mon Feb 12 17:51:03 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 12 Feb 2001 18:51:03 +0200
Subject: How to end process started in os.popen?
References: <96918n$6sn$1@nnrp1.deja.com>
Message-ID: <kuhf1zzvyg.fsf@lasipalatsi.fi>

In article <96918n$6sn$1 at nnrp1.deja.com>, bragib at my-deja.com writes:

| I am running a csh script and getting the output through a pipe using
| os.popen() on Unix.  I would like to be able to kill the running of the
| csh script if a certain event occurs.

| Just closing the pipe won't stop the csh script.

you could have the csh script write its pid to a file somewhere,
and use that. or you could roll your own popen with fork, pipe and exec
that returned the pid too.

  -- erno


From roymath at my-deja.com  Mon Feb 12 17:09:46 2001
From: roymath at my-deja.com (roymath at my-deja.com)
Date: Mon, 12 Feb 2001 16:09:46 GMT
Subject: Help: Non-capturing RE fails?
Message-ID: <9691s1$7f7$1@nnrp1.deja.com>

Hi folks,

I don't know if it is my understanding of non-capturing REs thats flawed, or
there is a bug in CPython 2.0. Anyhow, the following test indicates what I
think is the problem:

Please email me as well at roymath at yahoo.com, as I dont always have access
to dejanews.

Thanks,
Roy.

# ----------------------------------------------------------------------
import re

str = r' - <<ba>> @<<b>> <<c>> - '
desiredResult = r" - ba @<<b>> c - "

# The following fails. produces: " -ba @<<b>>c - " (Note. whitespace is eaten
up). # Use non-capturing RE. r = re.compile(r'(?:[^@])<<(.*?)>>') print
re.sub(r, r"\1", str)

# The following succeeds. produces: " - ba @<<b>> c - "
# Use a named group instead of non-capturing RE

def mysub(x):
    # print x.groups()
    return x.group('ws') + x.group('pat')

r = re.compile(r'(?P<ws>[^@])<<(?P<pat>.*?)>>')
print re.sub(r, mysub, str)

# ------------------------------------------------------------------------


Sent via Deja.com
http://www.deja.com/


From roymath at yahoo.com  Mon Feb 12 17:12:56 2001
From: roymath at yahoo.com (roymath at yahoo.com)
Date: Mon, 12 Feb 2001 16:12:56 GMT
Subject: Help: non-capturing group RE fails
Message-ID: <96921u$7j8$1@nnrp1.deja.com>

Hi folks,

I don't know if it is my understanding of non-capturing REs thats
flawed, or there
is a bug in CPython 2.0. Anyhow, the following test indicates what I
think is
the problem:

Please email me as well at roymath at yahoo.com, as I dont always have
access
to dejanews.

Thanks,
Roy.

# ----------------------------------------------------------------------
import re

str = r' - <<ba>> @<<b>> <<c>> - '
desiredResult = r" - ba @<<b>> c - "

# The following fails. produces: " -ba @<<b>>c - " (Note. whitespace is
eaten up).
# Use non-capturing RE.
r = re.compile(r'(?:[^@])<<(.*?)>>')
print re.sub(r, r"\1", str)

# The following succeeds. produces: " - ba @<<b>> c - "
# Use a named group instead of non-capturing RE

def mysub(x):
    # print x.groups()
    return x.group('ws') + x.group('pat')

r = re.compile(r'(?P<ws>[^@])<<(?P<pat>.*?)>>')
print re.sub(r, mysub, str)

#
------------------------------------------------------------------------


Sent via Deja.com
http://www.deja.com/


From robin.thomas at starmedia.net  Mon Feb 12 17:57:04 2001
From: robin.thomas at starmedia.net (Robin Thomas)
Date: Mon, 12 Feb 2001 11:57:04 -0500
Subject: Help: non-capturing group RE fails
In-Reply-To: <96921u$7j8$1@nnrp1.deja.com>
Message-ID: <4.3.1.2.20010212115249.00d2b100@exchange.starmedia.net>

At 04:12 PM 2/12/01 +0000, roymath at yahoo.com wrote:

># ----------------------------------------------------------------------
>import re
>
>str = r' - <<ba>> @<<b>> <<c>> - '
>desiredResult = r" - ba @<<b>> c - "
>
># The following fails. produces: " -ba @<<b>>c - " (Note. whitespace is
>eaten up).
># Use non-capturing RE.
>r = re.compile(r'(?:[^@])<<(.*?)>>')

(?:[^@]) is not a "group" that can be referenced. It is an "assertion", or 
"lookahead assertion", that is part of the match but does not get a group 
reference. That's why it is consumed along with the other ungrouped parts 
of the match, like << and >>.

r'([^@])<<(.*?)>>' is the quickest fix, along with

>print re.sub(r, r"\1", str)

print re.sub(r, r"\1\2", str)

completes the quick fix.


># The following succeeds. produces: " - ba @<<b>> c - "
># Use a named group instead of non-capturing RE
>
>def mysub(x):
>     # print x.groups()
>     return x.group('ws') + x.group('pat')
>
>r = re.compile(r'(?P<ws>[^@])<<(?P<pat>.*?)>>')
>print re.sub(r, mysub, str)

Here, (?P<ws>[^@]) really is a group definition, not merely a lookahead 
assertion. That's why it behaves differently. The fact that you've named 
the groups instead of using the implicit group numbering isn't material.


--
Robin Thomas
Director, Platform Engineering
StarMedia Network, Inc.
robin.thomas at starmedia.net




From robin.thomas at starmedia.net  Mon Feb 12 20:00:08 2001
From: robin.thomas at starmedia.net (Robin Thomas)
Date: Mon, 12 Feb 2001 14:00:08 -0500
Subject: Help: non-capturing group RE fails
In-Reply-To: <20010212173431.6556.qmail@web10004.mail.yahoo.com>
References: <4.3.1.2.20010212115249.00d2b100@exchange.starmedia.net>
Message-ID: <4.3.1.2.20010212124918.00d2f2c0@exchange.starmedia.net>

At 09:34 AM 2/12/01 -0800, Roy Mathew wrote:
>Robin,
>
>Thanks for your clear explanation; I had assumed that "non-capturing" meant
>"non-consuming" as well. I am still puzzled however, over the
>following behavior (slightly modified from before). ie: why does
>the "?=" which is a +ve lookahead assertion not produce the result
>we are looking for:

Damn, I'm sorry. I failed to notice that you want *nothing* touched in @<<b>>.

Anything not in a group will not be put in your replacement string, because 
there's no way to reference it. Things aren't "consumed", they're just 
considered part of the match, and the entire mathc substring is replaced 
with your replacement string.

"consumed" refers to anything that is not a lookahead or zero-width assertion.

?= does not consume any of the string. So

?=[^@]

says, "check the next character. As long as it's not '@', this re could 
still match. Continue matching at the same character I just checked!" So 
the next part of the re will start matching at the same character. That's 
why you get the weird behavior.

A negative lookbehind behaves differently. It will exclude the lookbehind's 
contents from the match, and do what you want:

# new code, works in 1.6 and up
import re
str = r' - <<ba>> @<<b>> <<c>> - '
r = re.compile(r'(?<!@)<<(.*?)>>')
print r.sub(r"\1", str)


--
Robin Thomas
Director, Platform Engineering
StarMedia Network, Inc.
robin.thomas at starmedia.net




From zope at thewebsons.com  Mon Feb 12 17:31:14 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Mon, 12 Feb 2001 08:31:14 -0800
Subject: How Best Get These Components?
Message-ID: <5.0.2.1.0.20010212082638.009f3320@thewebsons.com>

Hi;
I ran into some strange problems when I re-located my Apache box. Because 
of that, I (stupidly) ended up ripping out all things python and trying to 
re-install it. Well, python appears to work now, but I still appear to be 
missing a bunch of components. Then again, maybe I'm not: if python can't 
work with the components listed below, then perhaps I have a dead 
installation and a good one, too. (As an example, Zope works just fine.) 
What's the easiest way to test for if I really do need these tools? If I 
*do* need them, how should I go about installing them?
TIA,
BenO

the list...

missing    /usr/lib/python1.5/site-packages/_appletmodule.so
missing    /usr/lib/python1.5/site-packages/_cappletmodule.so
missing    /usr/lib/python1.5/site-packages/_gnomemodule.so
missing    /usr/lib/python1.5/site-packages/_gnomeuimodule.so
missing    /usr/lib/python1.5/site-packages/_gtkxmhtmlmodule.so
missing    /usr/lib/python1.5/site-packages/_zvtmodule.so
missing    /usr/lib/python1.5/site-packages/gettext.py
missing    /usr/lib/python1.5/site-packages/gettext.pyc
missing    /usr/lib/python1.5/site-packages/gettext.pyo
missing    /usr/lib/python1.5/site-packages/gnome/__init__.py
missing    /usr/lib/python1.5/site-packages/gnome/__init__.pyc
missing    /usr/lib/python1.5/site-packages/gnome/__init__.pyo
missing    /usr/lib/python1.5/site-packages/gnome/affine.py
missing    /usr/lib/python1.5/site-packages/gnome/affine.pyc
missing    /usr/lib/python1.5/site-packages/gnome/affine.pyo
missing    /usr/lib/python1.5/site-packages/gnome/applet.py
missing    /usr/lib/python1.5/site-packages/gnome/applet.pyc
missing    /usr/lib/python1.5/site-packages/gnome/applet.pyo
missing    /usr/lib/python1.5/site-packages/gnome/capplet.py
missing    /usr/lib/python1.5/site-packages/gnome/capplet.pyc
missing    /usr/lib/python1.5/site-packages/gnome/capplet.pyo
missing    /usr/lib/python1.5/site-packages/gnome/config.py
missing    /usr/lib/python1.5/site-packages/gnome/config.pyc
missing    /usr/lib/python1.5/site-packages/gnome/config.pyo
missing    /usr/lib/python1.5/site-packages/gnome/file.py
missing    /usr/lib/python1.5/site-packages/gnome/file.pyc
missing    /usr/lib/python1.5/site-packages/gnome/file.pyo
missing    /usr/lib/python1.5/site-packages/gnome/help.py
missing    /usr/lib/python1.5/site-packages/gnome/help.pyc
missing    /usr/lib/python1.5/site-packages/gnome/help.pyo
missing    /usr/lib/python1.5/site-packages/gnome/history.py
missing    /usr/lib/python1.5/site-packages/gnome/history.pyc
missing    /usr/lib/python1.5/site-packages/gnome/history.pyo
missing    /usr/lib/python1.5/site-packages/gnome/metadata.py
missing    /usr/lib/python1.5/site-packages/gnome/metadata.pyc
missing    /usr/lib/python1.5/site-packages/gnome/metadata.pyo
missing    /usr/lib/python1.5/site-packages/gnome/mime.py
missing    /usr/lib/python1.5/site-packages/gnome/mime.pyc
missing    /usr/lib/python1.5/site-packages/gnome/mime.pyo
missing    /usr/lib/python1.5/site-packages/gnome/score.py
missing    /usr/lib/python1.5/site-packages/gnome/score.pyc
missing    /usr/lib/python1.5/site-packages/gnome/score.pyo
missing    /usr/lib/python1.5/site-packages/gnome/triggers.py
missing    /usr/lib/python1.5/site-packages/gnome/triggers.pyc
missing    /usr/lib/python1.5/site-packages/gnome/triggers.pyo
missing    /usr/lib/python1.5/site-packages/gnome/ui.py
missing    /usr/lib/python1.5/site-packages/gnome/ui.pyc
missing    /usr/lib/python1.5/site-packages/gnome/ui.pyo
missing    /usr/lib/python1.5/site-packages/gnome/uiconsts.py
missing    /usr/lib/python1.5/site-packages/gnome/uiconsts.pyc
missing    /usr/lib/python1.5/site-packages/gnome/uiconsts.pyo
missing    /usr/lib/python1.5/site-packages/gnome/url.py
missing    /usr/lib/python1.5/site-packages/gnome/url.pyc
missing    /usr/lib/python1.5/site-packages/gnome/url.pyo
missing    /usr/lib/python1.5/site-packages/gnome/xmhtml.py
missing    /usr/lib/python1.5/site-packages/gnome/xmhtml.pyc
missing    /usr/lib/python1.5/site-packages/gnome/xmhtml.pyo
missing    /usr/lib/python1.5/site-packages/gnome/zvt.py
missing    /usr/lib/python1.5/site-packages/gnome/zvt.pyc
missing    /usr/lib/python1.5/site-packages/gnome/zvt.pyo
missing    /usr/lib/python1.5/site-packages/GDK.py
missing    /usr/lib/python1.5/site-packages/GDK.pyc
missing    /usr/lib/python1.5/site-packages/GDK.pyo
missing    /usr/lib/python1.5/site-packages/GTK.py
missing    /usr/lib/python1.5/site-packages/GTK.pyc
missing    /usr/lib/python1.5/site-packages/GTK.pyo
missing    /usr/lib/python1.5/site-packages/GdkImlib.py
missing    /usr/lib/python1.5/site-packages/GdkImlib.pyc
missing    /usr/lib/python1.5/site-packages/GdkImlib.pyo
missing    /usr/lib/python1.5/site-packages/GtkExtra.py
missing    /usr/lib/python1.5/site-packages/GtkExtra.pyc
missing    /usr/lib/python1.5/site-packages/GtkExtra.pyo
missing    /usr/lib/python1.5/site-packages/Gtkinter.py
missing    /usr/lib/python1.5/site-packages/Gtkinter.pyc
missing    /usr/lib/python1.5/site-packages/Gtkinter.pyo
missing    /usr/lib/python1.5/site-packages/_gdkimlibmodule.so
missing    /usr/lib/python1.5/site-packages/_gtkmodule.so
missing    /usr/lib/python1.5/site-packages/gtk.py
missing    /usr/lib/python1.5/site-packages/gtk.pyc
missing    /usr/lib/python1.5/site-packages/gtk.pyo
missing    /usr/lib/python1.5/site-packages/pyglade/__init__.py
missing    /usr/lib/python1.5/site-packages/pyglade/__init__.pyc
missing    /usr/lib/python1.5/site-packages/pyglade/__init__.pyo
missing    /usr/lib/python1.5/site-packages/pyglade/build.py
missing    /usr/lib/python1.5/site-packages/pyglade/build.pyc
missing    /usr/lib/python1.5/site-packages/pyglade/build.pyo
missing    /usr/lib/python1.5/site-packages/pyglade/style.py
missing    /usr/lib/python1.5/site-packages/pyglade/style.pyc
missing    /usr/lib/python1.5/site-packages/pyglade/style.pyo
missing    /usr/lib/python1.5/site-packages/pyglade/xmlparse.py
missing    /usr/lib/python1.5/site-packages/pyglade/xmlparse.pyc
missing    /usr/lib/python1.5/site-packages/pyglade/xmlparse.pyo
/usr/lib/python1.5/site-packages/rpmmodule.so
missing    /usr/lib/python1.5/lib-dynload/_tkinter.so
missing    /usr/lib/python1.5/lib-tk
missing    /usr/lib/python1.5/lib-tk/Canvas.py
missing    /usr/lib/python1.5/lib-tk/Canvas.pyc
missing    /usr/lib/python1.5/lib-tk/Canvas.pyo
missing    /usr/lib/python1.5/lib-tk/Dialog.py
missing    /usr/lib/python1.5/lib-tk/Dialog.pyc
missing    /usr/lib/python1.5/lib-tk/Dialog.pyo
missing    /usr/lib/python1.5/lib-tk/FileDialog.py
missing    /usr/lib/python1.5/lib-tk/FileDialog.pyc
missing    /usr/lib/python1.5/lib-tk/FileDialog.pyo
missing    /usr/lib/python1.5/lib-tk/FixTk.py
missing    /usr/lib/python1.5/lib-tk/FixTk.pyc
missing    /usr/lib/python1.5/lib-tk/FixTk.pyo
missing    /usr/lib/python1.5/lib-tk/ScrolledText.py
missing    /usr/lib/python1.5/lib-tk/ScrolledText.pyc
missing    /usr/lib/python1.5/lib-tk/ScrolledText.pyo
missing    /usr/lib/python1.5/lib-tk/SimpleDialog.py
missing    /usr/lib/python1.5/lib-tk/SimpleDialog.pyc
missing    /usr/lib/python1.5/lib-tk/SimpleDialog.pyo
missing    /usr/lib/python1.5/lib-tk/Tkconstants.py
missing    /usr/lib/python1.5/lib-tk/Tkconstants.pyc
missing    /usr/lib/python1.5/lib-tk/Tkconstants.pyo
missing    /usr/lib/python1.5/lib-tk/Tkdnd.py
missing    /usr/lib/python1.5/lib-tk/Tkdnd.pyc
missing    /usr/lib/python1.5/lib-tk/Tkdnd.pyo
missing    /usr/lib/python1.5/lib-tk/Tkinter.py
missing    /usr/lib/python1.5/lib-tk/Tkinter.pyc
missing    /usr/lib/python1.5/lib-tk/Tkinter.pyo
missing    /usr/lib/python1.5/lib-tk/tkColorChooser.py
missing    /usr/lib/python1.5/lib-tk/tkColorChooser.pyc
missing    /usr/lib/python1.5/lib-tk/tkColorChooser.pyo
missing    /usr/lib/python1.5/lib-tk/tkCommonDialog.py
missing    /usr/lib/python1.5/lib-tk/tkCommonDialog.pyc
missing    /usr/lib/python1.5/lib-tk/tkCommonDialog.pyo
missing    /usr/lib/python1.5/lib-tk/tkFileDialog.py
missing    /usr/lib/python1.5/lib-tk/tkFileDialog.pyc
missing    /usr/lib/python1.5/lib-tk/tkFileDialog.pyo
missing    /usr/lib/python1.5/lib-tk/tkFont.py
missing    /usr/lib/python1.5/lib-tk/tkFont.pyc
missing    /usr/lib/python1.5/lib-tk/tkFont.pyo
missing    /usr/lib/python1.5/lib-tk/tkMessageBox.py
missing    /usr/lib/python1.5/lib-tk/tkMessageBox.pyc
missing    /usr/lib/python1.5/lib-tk/tkMessageBox.pyo
missing    /usr/lib/python1.5/lib-tk/tkSimpleDialog.py
missing    /usr/lib/python1.5/lib-tk/tkSimpleDialog.pyc
missing    /usr/lib/python1.5/lib-tk/tkSimpleDialog.pyo




From dsh8290 at rit.edu  Mon Feb 12 21:00:16 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 12 Feb 2001 15:00:16 -0500
Subject: How Best Get These Components?
In-Reply-To: <5.0.2.1.0.20010212082638.009f3320@thewebsons.com>; from zope@thewebsons.com on Mon, Feb 12, 2001 at 08:31:14AM -0800
References: <5.0.2.1.0.20010212082638.009f3320@thewebsons.com>
Message-ID: <20010212150016.B5667@harmony.cs.rit.edu>

Whether or not you need them depends on what you want to do.


On Mon, Feb 12, 2001 at 08:31:14AM -0800, Ben Ocean wrote:
| Hi;
| I ran into some strange problems when I re-located my Apache box. Because 
| of that, I (stupidly) ended up ripping out all things python and trying to 
| re-install it. Well, python appears to work now, but I still appear to be 
| missing a bunch of components. Then again, maybe I'm not: if python can't 
| work with the components listed below, then perhaps I have a dead 
| installation and a good one, too. (As an example, Zope works just fine.) 
| What's the easiest way to test for if I really do need these tools? If I 
| *do* need them, how should I go about installing them?
| TIA,
| BenO
| 
| the list...
| 

These are the GTK/GNOME bindings.  There should be a link to James' 
site (the maintainer) from gtk.org and gnome.org.  You only need them
if you want to use programs that use gtk or gnome (for the user
interface).

| missing    /usr/lib/python1.5/site-packages/_appletmodule.so
| missing    /usr/lib/python1.5/site-packages/_cappletmodule.so
| missing    /usr/lib/python1.5/site-packages/_gnomemodule.so
| missing    /usr/lib/python1.5/site-packages/_gnomeuimodule.so
| missing    /usr/lib/python1.5/site-packages/_gtkxmhtmlmodule.so
| missing    /usr/lib/python1.5/site-packages/gnome/__init__.py
| missing    /usr/lib/python1.5/site-packages/gnome/__init__.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/__init__.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/affine.py
| missing    /usr/lib/python1.5/site-packages/gnome/affine.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/affine.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/applet.py
| missing    /usr/lib/python1.5/site-packages/gnome/applet.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/applet.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/capplet.py
| missing    /usr/lib/python1.5/site-packages/gnome/capplet.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/capplet.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/config.py
| missing    /usr/lib/python1.5/site-packages/gnome/config.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/config.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/file.py
| missing    /usr/lib/python1.5/site-packages/gnome/file.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/file.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/help.py
| missing    /usr/lib/python1.5/site-packages/gnome/help.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/help.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/history.py
| missing    /usr/lib/python1.5/site-packages/gnome/history.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/history.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/metadata.py
| missing    /usr/lib/python1.5/site-packages/gnome/metadata.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/metadata.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/mime.py
| missing    /usr/lib/python1.5/site-packages/gnome/mime.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/mime.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/score.py
| missing    /usr/lib/python1.5/site-packages/gnome/score.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/score.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/triggers.py
| missing    /usr/lib/python1.5/site-packages/gnome/triggers.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/triggers.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/ui.py
| missing    /usr/lib/python1.5/site-packages/gnome/ui.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/ui.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/uiconsts.py
| missing    /usr/lib/python1.5/site-packages/gnome/uiconsts.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/uiconsts.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/url.py
| missing    /usr/lib/python1.5/site-packages/gnome/url.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/url.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/xmhtml.py
| missing    /usr/lib/python1.5/site-packages/gnome/xmhtml.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/xmhtml.pyo
| missing    /usr/lib/python1.5/site-packages/gnome/zvt.py
| missing    /usr/lib/python1.5/site-packages/gnome/zvt.pyc
| missing    /usr/lib/python1.5/site-packages/gnome/zvt.pyo
| missing    /usr/lib/python1.5/site-packages/GDK.py
| missing    /usr/lib/python1.5/site-packages/GDK.pyc
| missing    /usr/lib/python1.5/site-packages/GDK.pyo
| missing    /usr/lib/python1.5/site-packages/GTK.py
| missing    /usr/lib/python1.5/site-packages/GTK.pyc
| missing    /usr/lib/python1.5/site-packages/GTK.pyo
| missing    /usr/lib/python1.5/site-packages/GdkImlib.py
| missing    /usr/lib/python1.5/site-packages/GdkImlib.pyc
| missing    /usr/lib/python1.5/site-packages/GdkImlib.pyo
| missing    /usr/lib/python1.5/site-packages/GtkExtra.py
| missing    /usr/lib/python1.5/site-packages/GtkExtra.pyc
| missing    /usr/lib/python1.5/site-packages/GtkExtra.pyo
| missing    /usr/lib/python1.5/site-packages/Gtkinter.py
| missing    /usr/lib/python1.5/site-packages/Gtkinter.pyc
| missing    /usr/lib/python1.5/site-packages/Gtkinter.pyo
| missing    /usr/lib/python1.5/site-packages/_gdkimlibmodule.so
| missing    /usr/lib/python1.5/site-packages/_gtkmodule.so
| missing    /usr/lib/python1.5/site-packages/gtk.py
| missing    /usr/lib/python1.5/site-packages/gtk.pyc
| missing    /usr/lib/python1.5/site-packages/gtk.pyo
| missing    /usr/lib/python1.5/site-packages/pyglade/__init__.py
| missing    /usr/lib/python1.5/site-packages/pyglade/__init__.pyc
| missing    /usr/lib/python1.5/site-packages/pyglade/__init__.pyo
| missing    /usr/lib/python1.5/site-packages/pyglade/build.py
| missing    /usr/lib/python1.5/site-packages/pyglade/build.pyc
| missing    /usr/lib/python1.5/site-packages/pyglade/build.pyo
| missing    /usr/lib/python1.5/site-packages/pyglade/style.py
| missing    /usr/lib/python1.5/site-packages/pyglade/style.pyc
| missing    /usr/lib/python1.5/site-packages/pyglade/style.pyo
| missing    /usr/lib/python1.5/site-packages/pyglade/xmlparse.py
| missing    /usr/lib/python1.5/site-packages/pyglade/xmlparse.pyc
| missing    /usr/lib/python1.5/site-packages/pyglade/xmlparse.pyo


Looks like this is the Tkinter bindings.  I don't know where to get
them.  It should have been included in the standard dist.  I don't
know what to do (since you built your own, right?)  These are only
necessary if you want to run programs using Tkinter for the UI.

| missing    /usr/lib/python1.5/lib-dynload/_tkinter.so
| missing    /usr/lib/python1.5/lib-tk
| missing    /usr/lib/python1.5/lib-tk/Canvas.py
| missing    /usr/lib/python1.5/lib-tk/Canvas.pyc
| missing    /usr/lib/python1.5/lib-tk/Canvas.pyo
| missing    /usr/lib/python1.5/lib-tk/Dialog.py
| missing    /usr/lib/python1.5/lib-tk/Dialog.pyc
| missing    /usr/lib/python1.5/lib-tk/Dialog.pyo
| missing    /usr/lib/python1.5/lib-tk/FileDialog.py
| missing    /usr/lib/python1.5/lib-tk/FileDialog.pyc
| missing    /usr/lib/python1.5/lib-tk/FileDialog.pyo
| missing    /usr/lib/python1.5/lib-tk/FixTk.py
| missing    /usr/lib/python1.5/lib-tk/FixTk.pyc
| missing    /usr/lib/python1.5/lib-tk/FixTk.pyo
| missing    /usr/lib/python1.5/lib-tk/ScrolledText.py
| missing    /usr/lib/python1.5/lib-tk/ScrolledText.pyc
| missing    /usr/lib/python1.5/lib-tk/ScrolledText.pyo
| missing    /usr/lib/python1.5/lib-tk/SimpleDialog.py
| missing    /usr/lib/python1.5/lib-tk/SimpleDialog.pyc
| missing    /usr/lib/python1.5/lib-tk/SimpleDialog.pyo
| missing    /usr/lib/python1.5/lib-tk/Tkconstants.py
| missing    /usr/lib/python1.5/lib-tk/Tkconstants.pyc
| missing    /usr/lib/python1.5/lib-tk/Tkconstants.pyo
| missing    /usr/lib/python1.5/lib-tk/Tkdnd.py
| missing    /usr/lib/python1.5/lib-tk/Tkdnd.pyc
| missing    /usr/lib/python1.5/lib-tk/Tkdnd.pyo
| missing    /usr/lib/python1.5/lib-tk/Tkinter.py
| missing    /usr/lib/python1.5/lib-tk/Tkinter.pyc
| missing    /usr/lib/python1.5/lib-tk/Tkinter.pyo
| missing    /usr/lib/python1.5/lib-tk/tkColorChooser.py
| missing    /usr/lib/python1.5/lib-tk/tkColorChooser.pyc
| missing    /usr/lib/python1.5/lib-tk/tkColorChooser.pyo
| missing    /usr/lib/python1.5/lib-tk/tkCommonDialog.py
| missing    /usr/lib/python1.5/lib-tk/tkCommonDialog.pyc
| missing    /usr/lib/python1.5/lib-tk/tkCommonDialog.pyo
| missing    /usr/lib/python1.5/lib-tk/tkFileDialog.py
| missing    /usr/lib/python1.5/lib-tk/tkFileDialog.pyc
| missing    /usr/lib/python1.5/lib-tk/tkFileDialog.pyo
| missing    /usr/lib/python1.5/lib-tk/tkFont.py
| missing    /usr/lib/python1.5/lib-tk/tkFont.pyc
| missing    /usr/lib/python1.5/lib-tk/tkFont.pyo
| missing    /usr/lib/python1.5/lib-tk/tkMessageBox.py
| missing    /usr/lib/python1.5/lib-tk/tkMessageBox.pyc
| missing    /usr/lib/python1.5/lib-tk/tkMessageBox.pyo
| missing    /usr/lib/python1.5/lib-tk/tkSimpleDialog.py
| missing    /usr/lib/python1.5/lib-tk/tkSimpleDialog.pyc
| missing    /usr/lib/python1.5/lib-tk/tkSimpleDialog.pyo

Not sure where these come from, but gettext is the GNU library for
internationalizing programs.  These are only necessary if you want to
run programs that use gettext for internationalization.

| missing    /usr/lib/python1.5/site-packages/_zvtmodule.so
| missing    /usr/lib/python1.5/site-packages/gettext.py
| missing    /usr/lib/python1.5/site-packages/gettext.pyc
| missing    /usr/lib/python1.5/site-packages/gettext.pyo


If you don't want to run programs that need these modules, then you
don't need to install them.  If you do (say, netcfg or printtool --
both need Tkinter) then they will crash with a Import Failed, No Such
Module type of message.

You have a RH system right?  You should just be able to install the
rpms.  Of course, since you compiled python from source, rpm will
complain that you are missing python.  Just use --nodeps when you
install.  Also realize that you will "forever" have such issues since
part of your system is rpm-based and part is hand installed from
tarballs.

HTH,
-D




From gvwilson at interlog.com  Mon Feb 12 17:53:59 2001
From: gvwilson at interlog.com (Greg Wilson)
Date: 12 Feb 2001 11:53:59 -0500
Subject: Python Conference BOF Mailing lists
Message-ID: <9694f7$q5e@shell1.interlog.com>

Two mailing lists have been set up to discuss admin issues
for the BOFs on iterators and sets:

    Iterator-BOF at python9.org   (March 6)
    Set-BOF at python9.org        (March 7)

If you are planning to attend, please send mail to the list,
so that we can plan room allocation, make sure we get served
quickly for lunch so that we can get discussion started early,
etc.

Greg


From paul.moore at uk.origin-it.com  Mon Feb 12 18:03:55 2001
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Mon, 12 Feb 2001 18:03:55 +0100
Subject: How do I use a pyc file created with an older version of Python?
Message-ID: <FheIOmKxaMKoDmEgrL3aSCRNu7QW@4ax.com>

Hi,
This is probably a silly question, but I have an old pyc file
generated under Python 1.5.? (I think), and I now use Python 2.0. I
don't have the source easily accessible, unortunately, and I don't
really want to install Python 1.5 as well, just for this one bit of
code.

Is it possible to convert a pyc from an older version of Python to a
newer one, without the source? It seems like it should be possible, as
a pyc is (I believe) just a magic number plus a marshalled code
object. Assuming the code itself is compatible, could I just "fix" the
magic number? Or would I need to "upgrade" the marshalled code
somehow? Or is this not possible, and I just have to get the source
back...?

Thanks,
Paul



From mwh21 at cam.ac.uk  Mon Feb 12 18:27:39 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 12 Feb 2001 17:27:39 +0000
Subject: How do I use a pyc file created with an older version of Python?
References: <FheIOmKxaMKoDmEgrL3aSCRNu7QW@4ax.com>
Message-ID: <m3g0hjg6b8.fsf@atrus.jesus.cam.ac.uk>

Paul Moore <paul.moore at uk.origin-it.com> writes:

> Is it possible to convert a pyc from an older version of Python to a
> newer one, without the source? It seems like it should be possible,
> as a pyc is (I believe) just a magic number plus a marshalled code
> object. Assuming the code itself is compatible, could I just "fix"
> the magic number?

The reason the magic number is different is because the code is not
compatible.

> Or would I need to "upgrade" the marshalled code somehow? Or is this
> not possible, and I just have to get the source back...?

Getting the source back is your only real hope.  I suppose it would be
possible to write a 1.5.2 -> 2.0 bytecode translator (the differences
aren't all that great), but it doesn't seem worth the effort (to me,
at least).

There's also decompyle (google for it) that may, or may not, be able
to translate your .pyc into a .py.

Cheers,
M.

-- 
  You owe The Oracle a TV with an 'intelligence' control - I've tried
  'brightness' but that didn't work.
                                      -- Internet Oracularity #1192-01


From stephen_purcell at yahoo.com  Mon Feb 12 18:38:08 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 12 Feb 2001 18:38:08 +0100
Subject: How do I use a pyc file created with an older version of Python?
In-Reply-To: <FheIOmKxaMKoDmEgrL3aSCRNu7QW@4ax.com>; from paul.moore@uk.origin-it.com on Mon, Feb 12, 2001 at 06:03:55PM +0100
References: <FheIOmKxaMKoDmEgrL3aSCRNu7QW@4ax.com>
Message-ID: <20010212183808.A1750@freedom.puma-ag.com>

Paul Moore wrote:
> Is it possible to convert a pyc from an older version of Python to a
> newer one, without the source? It seems like it should be possible, as
> a pyc is (I believe) just a magic number plus a marshalled code
> object.

Bad luck. The marshalled code objects aren't compatible from 1.5.2 to 2.0.

-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From thomas at xs4all.net  Wed Feb 14 16:35:59 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 14 Feb 2001 16:35:59 +0100
Subject: How do I use a pyc file created with an older version of Python?
In-Reply-To: <FheIOmKxaMKoDmEgrL3aSCRNu7QW@4ax.com>; from paul.moore@uk.origin-it.com on Mon, Feb 12, 2001 at 06:03:55PM +0100
References: <FheIOmKxaMKoDmEgrL3aSCRNu7QW@4ax.com>
Message-ID: <20010214163559.I4924@xs4all.nl>

On Mon, Feb 12, 2001 at 06:03:55PM +0100, Paul Moore wrote:

> This is probably a silly question, but I have an old pyc file
> generated under Python 1.5.? (I think), and I now use Python 2.0. I
> don't have the source easily accessible, unortunately, and I don't
> really want to install Python 1.5 as well, just for this one bit of
> code.

You'll have to do one of the latter two, unfortunately.

> Is it possible to convert a pyc from an older version of Python to a
> newer one, without the source? It seems like it should be possible, as
> a pyc is (I believe) just a magic number plus a marshalled code
> object. Assuming the code itself is compatible, could I just "fix" the
> magic number? Or would I need to "upgrade" the marshalled code
> somehow? Or is this not possible, and I just have to get the source
> back...?

A .pyc file is indeed 'just' a magic number and a marshalled code object.
However, the magic number is there for a reason: the actual bytecode
contained in the marshalled object *changed meaning*. For instance, in the
1.5.2 -> 2.0 change, the 'IMPORT_FROM' bytecode was split into 'IMPORT_FROM'
and 'IMPORT_STAR'. 'IMPORT_FROM' no longer handles the special case of the
"*" name, and if you try to run Python 1.5 (or older) bytecode that uses
'IMPORT_FROM' with a name of "*", you'll probably get a system error. Or
worse, it doesn't raise an error, but it doesn't import any of the desired
names either :)

There are more problems, though (before you think you're safe because the
code doesn't do 'from foo import *' :) between 1.5 and 2.0, the
UNPACK_TUPLE, UNPACK_LIST, UNPACK_ARG and UNPACK_VARARG opcodes were merged
into a single UNPACK_SEQUENCE opcode. UNPACK_TUPLE will still work, since it
has the same number and functionality as the 'new' UNPACK_SEQUENCE opcode,
but the other ones will raise a system error. Except for UNPACK_VARARG,
which is now DUP_TOPX, and would cause a lot of confusion :-) Several other
opcodes were removed as well.

So, bottom line, you're out of luck.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From scarpenter at innerx.net  Mon Feb 12 18:26:45 2001
From: scarpenter at innerx.net (scarpenter at innerx.net)
Date: Mon, 12 Feb 2001 12:26:45 -0500
Subject: New Online Auction Site for Brand Name Tools, Hardware, Home Improvement Products
Message-ID: <E14SMkP-0005b5-00@mail.python.org>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010212/a9a7129f/attachment.htm>

From tjg at exceptionalminds.com  Mon Feb 12 19:03:37 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Mon, 12 Feb 2001 10:03:37 -0800
Subject: I don't think I understand popen2() behaviour
Message-ID: <20010212100337.Y17173@trufflehunter.avalongroup.net>

Hi,

I need some enlightenment. For the first time ever, I'm playing with pipes
in python and I must be missing something. I'll paste in a couple of
examples to see if I can explain myself.

>>> import os
>>> x = os.popen('tail -f /var/log/messages')
>>> tail: /var/log/messages: Permission denied
tail: no files remaining

>>> 

OK, so I'd expect that behaviour, as I'm logged in as me, not root.

Now, my understanding of popen2() is that it gives me a file handle to both
stdin and stderr, so since I don't want those error messages popping up on
the screen I tried the following

>>> import popen2
>>> x,y = popen2.popen2('tail -f /var/log/messages')
>>> tail: /var/log/messages: Permission denied
tail: no files remaining

>>>

Now that was unexpected, as I would have thought that the error messages
should have been hidden away in my file object y, and that I wouldn't see
them until I did a y.readlines() or something. However, when I try to access
y, I get the following:

>>> y.readlines()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
IOError: [Errno 9] Bad file descriptor
>>> 

So, obviously I don't understand something, but I'm not even sure what it is
I don't understand yet!

Thanks for your assistance.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  26 days 22:08 hours ago<<



From bsb at winnegan.de  Mon Feb 12 19:48:35 2001
From: bsb at winnegan.de (Siggy Brentrup)
Date: 12 Feb 2001 19:48:35 +0100
Subject: I don't think I understand popen2() behaviour
In-Reply-To: <20010212100337.Y17173@trufflehunter.avalongroup.net>
References: <20010212100337.Y17173@trufflehunter.avalongroup.net>
Message-ID: <87wvavd9fg.fsf@winnegan.de>

Timothy Grant <tjg at exceptionalminds.com> writes:

> Hi,

> I need some enlightenment. For the first time ever, I'm playing with
> pipes in python and I must be missing something. I'll paste in a
> couple of examples to see if I can explain myself.

> >>> import os
> >>> x = os.popen('tail -f /var/log/messages')
> >>> tail: /var/log/messages: Permission denied
> tail: no files remaining
> 
> >>> 

> OK, so I'd expect that behaviour, as I'm logged in as me, not root.

> Now, my understanding of popen2() is that it gives me a file handle
> to both stdin and stderr, so since I don't want those error messages
> popping up on the screen I tried the following

> >>> import popen2
> >>> x,y = popen2.popen2('tail -f /var/log/messages')
> >>> tail: /var/log/messages: Permission denied
> tail: no files remaining
> 
> >>>

> Now that was unexpected, as I would have thought that the error
> messages should have been hidden away in my file object y, and that
> I wouldn't see them until I did a y.readlines() or something.

popen2 connects the command's stdin and stdout, but not to stderr.

> However, when I try to access y, I get the following:

> >>> y.readlines()
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> IOError: [Errno 9] Bad file descriptor
> >>> 

>>> y
<open file '(fdopen)', mode 'w' at 807c4b0>

shows that you are reading from a file open for output, hence the
IOError.
 
> So, obviously I don't understand something, but I'm not even sure
> what it is I don't understand yet!

Try the following:

>>> cld = popen2.Popen3('tail -f /var/log/messages', 1)

For details cf.
http://www.python.org/doc/current/lib/popen3-objects.html

HIH
  Siggy




From tjg at exceptionalminds.com  Mon Feb 12 20:09:48 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Mon, 12 Feb 2001 11:09:48 -0800
Subject: I don't think I understand popen2() behaviour
In-Reply-To: <87wvavd9fg.fsf@winnegan.de>; from bsb@winnegan.de on Mon, Feb 12, 2001 at 07:48:35PM +0100
References: <20010212100337.Y17173@trufflehunter.avalongroup.net> <87wvavd9fg.fsf@winnegan.de>
Message-ID: <20010212110947.Z17173@trufflehunter.avalongroup.net>

On Mon, Feb 12, 2001 at 07:48:35PM +0100, Siggy Brentrup wrote:
> > I wouldn't see them until I did a y.readlines() or something.
> 
> popen2 connects the command's stdin and stdout, but not to stderr.
> 
>  
> > So, obviously I don't understand something, but I'm not even sure
> > what it is I don't understand yet!
> 
> Try the following:
> 
> >>> cld = popen2.Popen3('tail -f /var/log/messages', 1)
> 
> For details cf.
> http://www.python.org/doc/current/lib/popen3-objects.html
> 
Hmm, another case of my brain interpreting what I read in light of what my
brain wanted me to read.

Thanks for the pointer. I'll try and pay closer attention next time.


-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  26 days 23:20 hours ago<<



From stephen_purcell at yahoo.com  Mon Feb 12 19:54:15 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 12 Feb 2001 19:54:15 +0100
Subject: I don't think I understand popen2() behaviour
In-Reply-To: <20010212100337.Y17173@trufflehunter.avalongroup.net>; from tjg@exceptionalminds.com on Mon, Feb 12, 2001 at 10:03:37AM -0800
References: <20010212100337.Y17173@trufflehunter.avalongroup.net>
Message-ID: <20010212195415.A3546@freedom.puma-ag.com>

Timothy Grant wrote:
> >>> import popen2
> >>> x,y = popen2.popen2('tail -f /var/log/messages')
> >>> tail: /var/log/messages: Permission denied
> tail: no files remaining
> 
> >>>
> 
> Now that was unexpected, as I would have thought that the error messages
> should have been hidden away in my file object y,

popen2 gives you (stdin, stdout) for the child process. You probably want
'popen3' instead if you want to grab stderr.

-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From stephen_purcell at yahoo.com  Mon Feb 12 19:55:48 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 12 Feb 2001 19:55:48 +0100
Subject: I don't think I understand popen2() behaviour
In-Reply-To: <20010212195415.A3546@freedom.puma-ag.com>; from stephen_purcell@yahoo.com on Mon, Feb 12, 2001 at 07:54:15PM +0100
References: <20010212100337.Y17173@trufflehunter.avalongroup.net> <20010212195415.A3546@freedom.puma-ag.com>
Message-ID: <20010212195548.B3546@freedom.puma-ag.com>

Steve Purcell wrote:
> 
> popen2 gives you (stdin, stdout) for the child process.

Sorry, that should read (stdout, stdin).

Cribbed from http://www.python.org/doc/lib/module-popen2.html

-Steve



From stevena at neosynapse.net  Mon Feb 12 19:53:02 2001
From: stevena at neosynapse.net (Steven D. Arnold)
Date: Mon, 12 Feb 2001 13:53:02 -0500
Subject: ANN: Pure-Python Encryption Tools
Message-ID: <5.0.2.1.2.20010212133300.03a64ae0@phear.dementian.com>

Neosynapse Consulting Group announces the release of pure-Python encryption 
tools.  These modules may be used as building blocks for programs that need 
cryptographic services.  The modules are released as open-source software 
under the Neosynapse Artistic License.

So far, we have an implementation of IDEA and a module for handling modular 
arithmetic.  Coming soon, look for modules to handle problems relating to 
prime numbers, a Blowfish module and an RSA public-key cryptography module.

Interested people are welcomed to participate in the development of these 
open-source modules.  The software may be downloaded at:

         http://www.neosynapse.net/free_encrypt.htm


--
Steven D. Arnold                   President, Neosynapse Consulting Group
stevena at neosynapse.net                          http://www.neosynapse.net




From glennj at pcard31e.ca.nortel.com  Tue Feb 13 21:36:14 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 13 Feb 2001 20:36:14 GMT
Subject: ANN: Pure-Python Encryption Tools
References: <mailman.982016894.21075.python-list@python.org>
Message-ID: <96c5ru$s1f$1@bcarh8ab.ca.nortel.com>

Steven D. Arnold wrote:
>Neosynapse Consulting Group announces the release of pure-Python encryption 
>tools.  These modules may be used as building blocks for programs that need 
>cryptographic services.  The modules are released as open-source software 
>under the Neosynapse Artistic License.
>
>So far, we have an implementation of IDEA and a module for handling modular 
>arithmetic.  Coming soon, look for modules to handle problems relating to 
>prime numbers, a Blowfish module and an RSA public-key cryptography module.
>
>Interested people are welcomed to participate in the development of these 
>open-source modules.  The software may be downloaded at:
>
>         http://www.neosynapse.net/free_encrypt.htm

Is this url active?
    File Not Found
    requested URL /free_encrypt.htm was not found on this server.

-- 
Glenn 


From emile at fenx.com  Tue Feb 13 22:43:14 2001
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 13 Feb 2001 13:43:14 -0800
Subject: ANN: Pure-Python Encryption Tools
References: <mailman.982016894.21075.python-list@python.org> <96c5ru$s1f$1@bcarh8ab.ca.nortel.com>
Message-ID: <96c9s8$kpchj$1@ID-11957.news.dfncis.de>

<snip>
> >Interested people are welcomed to participate in the
development of these
> >open-source modules.  The software may be downloaded at:
> >
> >         http://www.neosynapse.net/free_encrypt.htm
>
> Is this url active?
>     File Not Found
>     requested URL /free_encrypt.htm was not found on this
server.
>
> --
> Glenn

I got right through.

--

Emile van Sebille
emile at fenx.com
-------------------





From fdrake at acm.org  Mon Feb 12 19:53:08 2001
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Mon, 12 Feb 2001 13:53:08 -0500 (EST)
Subject: Fwd: test_pyexpat failed (gryphon01@my-deja.com)
In-Reply-To: <14983.63939.371712.913299@w221.z064000254.bwi-md.dsl.cnc.net>
References: <14983.63939.371712.913299@w221.z064000254.bwi-md.dsl.cnc.net>
Message-ID: <14984.12564.945348.622604@cj42289-a.reston1.va.home.com>

[Jeremy Hylton forwarded this message to me.]
 > From: gryphon01 at my-deja.com
 > To: python-list at python.org
 > Subject: test_pyexpat failed
 > Date: Mon, 12 Feb 2001 04:16:39 GMT
 > 
 > I an trying to build Python 2.0 with expat 1.2. The compile doesn't complain
 > but test_pyexpat says: test test_pyexpat failed -- Writing:
 > "('xml=http://www.w3.org/XML/1998/namespace\\014external_entity', None,
 > 'entity.file', None)", expected:
 > "('http://www.python.org/namespace=http://www.w3.org/XML/1998/namespace\\014e
 > xternal_entity'"

  You're not doing anything wrong; the test suite exposes some
internal data that expat passes through Python to allow the
.ExternalEntityParserCreate() method to pick up context information.
The format of the context information can and has changed over the
life of Expat, but should not be exposed to begin with.  The current
version of the test (in the alphas for 2.1) does not expose this
implementation detail.  You should be able to pick up the files
Lib/test/test_pyexpat.py and Lib/test/output/test_pyexpat from the
Python CVS and use them instead.
  I'll add an entry to the Python 2.0 FAQ.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Digital Creations




From timo at lightdog.com  Wed Feb 14 14:49:23 2001
From: timo at lightdog.com (timo at lightdog.com)
Date: Wed, 14 Feb 2001 08:49:23 -0500
Subject: International/Localized Python GUI apps
In-Reply-To: <14984.12564.945348.622604@cj42289-a.reston1.va.home.com>
References: <14983.63939.371712.913299@w221.z064000254.bwi-md.dsl.cnc.net>
Message-ID: <3A8A4693.16262.33F62C@localhost>

I (desparately) want to be using python for our GUI apps here, but 
we're currently using M$ Visual C++. I'm not very happy with it, 
MFC in particular. It's a horrible set of wrappers on an unhappy api.

Still, we have some opportunity. I need to be able to do UIs that are 
easily internationalizable, and I need to do a lot of work in grids. I've 
looked at WxPython, but it doesn't seem to do wide characters 
and/or unicode. I really need multilingual support, some of those 
languages being asian ones. 

Is there a way for me to propose python and tkinter or something 
for this work? I really need grid controls, and i really need foreign 
language support. If I have these, and can make an argument for it 
being easier than MFC/VC++, then I have a really good chance of 
getting it pushed through here. 

Any help? 

Tim



From boud at valdyas.org  Wed Feb 14 21:33:46 2001
From: boud at valdyas.org (Boudewijn Rempt)
Date: Wed, 14 Feb 2001 21:33:46 +0100
Subject: International/Localized Python GUI apps
References: <14983.63939.371712.913299@w221.z064000254.bwi-md.dsl.cnc.net> <mailman.982158446.26260.python-list@python.org>
Message-ID: <3a8aebcb$0$36171$1a0eedbc@dreader2.news.xs4all.nl>

timo at lightdog.com wrote:

> I (desparately) want to be using python for our GUI apps here, but
> we're currently using M$ Visual C++. I'm not very happy with it,
> MFC in particular. It's a horrible set of wrappers on an unhappy api.
> 
> Still, we have some opportunity. I need to be able to do UIs that are
> easily internationalizable, and I need to do a lot of work in grids. I've
> looked at WxPython, but it doesn't seem to do wide characters
> and/or unicode. I really need multilingual support, some of those
> languages being asian ones.
> 
> Is there a way for me to propose python and tkinter or something
> for this work? I really need grid controls, and i really need foreign
> language support. If I have these, and can make an argument for it
> being easier than MFC/VC++, then I have a really good chance of
> getting it pushed through here.
> 

If you really need Asian character set support, and _good_ support, you 
should try PyQt. I've just released an application that uses it - you can 
see a screenshot at http://www.valdyas.org/linguistics/screenshots.html. If 
you need good BiDi, you might have to wait for version 3.0 of Qt. PyQt also 
offers a strong Table control, grid layouts and a good canvas. And, of 
course, it's very, very easy to work with.

It's not free for Windows, but that should make it _easier_ to get it into 
a a shop using Visual C++ ;-).

Boudewijn Rempt



From loewis at informatik.hu-berlin.de  Tue Feb 20 10:58:12 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 20 Feb 2001 10:58:12 +0100
Subject: International/Localized Python GUI apps
References: <14983.63939.371712.913299@w221.z064000254.bwi-md.dsl.cnc.net> <mailman.982158446.26260.python-list@python.org>
Message-ID: <p6q7l2ln0az.fsf@informatik.hu-berlin.de>

timo at lightdog.com writes:

> Still, we have some opportunity. I need to be able to do UIs that are 
> easily internationalizable, and I need to do a lot of work in grids. I've 
> looked at WxPython, but it doesn't seem to do wide characters 
> and/or unicode. I really need multilingual support, some of those 
> languages being asian ones. 

For the internationalization, I recommend to use the gettext module of
Python 2.0.

Regards,
Martin



From corour01 at motorola.com  Mon Feb 12 20:11:31 2001
From: corour01 at motorola.com (O'Rourke Clodagh-corour01)
Date: Mon, 12 Feb 2001 19:11:31 -0000
Subject: Embedding UNIX Commands
Message-ID: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com>

Hi All,
I'm wondering could anyone advise me on what is the best way to embed unix commands such as setenv and whoami in a Python/Tkinter application.  Would it be to put them into a C module or can it be done directly in python?

Thanks,
Clodagh O' Rourke.



From dsh8290 at rit.edu  Mon Feb 12 21:07:49 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 12 Feb 2001 15:07:49 -0500
Subject: Embedding UNIX Commands
In-Reply-To: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com>; from corour01@motorola.com on Mon, Feb 12, 2001 at 07:11:31PM +0000
References: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com>
Message-ID: <20010212150748.C5667@harmony.cs.rit.edu>

You can use the system function :

import os
os.system( "whoami" )

I don't think you can use setenv -- that is a csh thing, and system()
won't even start up a shell (AFAIK).  Even if it did, it wouldn't do
you any good since that shell would terminate and its environment
would be meaningless.  (Also, on Linux systems bash is the default
shell)

It would probably be better if you described what you want to
accomplish with these commands and then people can come up with a
different way that wouldn't rely on the underlying system (and PATH).

If you simply want to change the environment of your python process
you can :

import sys
sys.env[ "VAR" ] = "value"


(also possibly the sys.setenv() function, try  dir( sys ) , and  print
sys.setenv.__doc__ )

HTH,
-D

On Mon, Feb 12, 2001 at 07:11:31PM +0000, O'Rourke Clodagh-corour01 wrote:
| Hi All,
| I'm wondering could anyone advise me on what is the best way to embed unix commands such as setenv and whoami in a Python/Tkinter application.  Would it be to put them into a C module or can it be done directly in python?
| 
| Thanks,
| Clodagh O' Rourke.
| 



From kalle at gnupung.net  Tue Feb 13 01:42:58 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Tue, 13 Feb 2001 01:42:58 +0100
Subject: Embedding UNIX Commands
In-Reply-To: <20010212150748.C5667@harmony.cs.rit.edu>; from dsh8290@rit.edu on Mon, Feb 12, 2001 at 03:07:49PM -0500
References: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com> <20010212150748.C5667@harmony.cs.rit.edu>
Message-ID: <20010213014258.A17495@apone.network.loc>

Sez D-Man:
> If you simply want to change the environment of your python process
> you can :
> 
> import sys
> sys.env[ "VAR" ] = "value"
> 
> 
> (also possibly the sys.setenv() function, try  dir( sys ) , and  print
> sys.setenv.__doc__ )

That should be
 import os
 os.environ["VAR"] = "value"
 os.putenv()
 dir(os)
and
 print os.putenv.__doc__
I believe.  At least with Python 2.x.

Check http://www.python.org/doc/current/lib/os-procinfo.html for more info.

Otherwise, I agree with the previous speaker...

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010213/412acf01/attachment.pgp>

From erno-news at erno.iki.fi  Tue Feb 13 01:33:45 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Feb 2001 02:33:45 +0200
Subject: Embedding UNIX Commands
References: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com> <mailman.982008497.29571.python-list@python.org>
Message-ID: <kuwvavxvyu.fsf@lasipalatsi.fi>

In article <mailman.982008497.29571.python-list at python.org>, D-Man
<dsh8290 at rit.edu> writes:

| I don't think you can use setenv -- that is a csh thing, and system()
| won't even start up a shell (AFAIK).  Even if it did, it wouldn't do
| you any good since that shell would terminate and its environment
| would be meaningless.  (Also, on Linux systems bash is the default
| shell)

actually, system() (and popen()) pass the command line to /bin/sh
under unix.

| It would probably be better if you described what you want to
| accomplish with these commands and then people can come up with a
| different way that wouldn't rely on the underlying system (and PATH).

| If you simply want to change the environment of your python process
| you can :

| import sys
| sys.env[ "VAR" ] = "value"


| (also possibly the sys.setenv() function, try  dir( sys ) , and  print
| sys.setenv.__doc__ )

those should be os.environ and os.putenv instead of sys.env and sys.setenv.

  -- erno


From bill at libc.org  Mon Feb 12 21:00:55 2001
From: bill at libc.org (Bill Anderson)
Date: Tue, 13 Feb 2001 03:00:55 +0700
Subject: Embedding UNIX Commands
References: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com> <mailman.982008497.29571.python-list@python.org> <kuwvavxvyu.fsf@lasipalatsi.fi>
Message-ID: <rB7i6.1898$AX.389928@news03.micron.net>

In article <kuwvavxvyu.fsf at lasipalatsi.fi>, "Erno Kuusela"
<erno-news at erno.iki.fi> wrote:

> In article <mailman.982008497.29571.python-list at python.org>, D-Man
> <dsh8290 at rit.edu> writes:
> 
> | I don't think you can use setenv -- that is a csh thing, and system()
> | won't even start up a shell (AFAIK).  Even if it did, it wouldn't do
> | you any good since that shell would terminate and its environment
> | would be meaningless.  (Also, on Linux systems bash is the default
> | shell)
> 
> actually, system() (and popen()) pass the command line to /bin/sh under
> unix.

Which _is_ bash (on most Linux Dists):
ucntcme at locutus in /home/ucntcme 
$ ll /bin/sh lrwxrwxrwx    1 root     root            4 Jan 13 19:58 /bin/sh -> bash*


From chris at voodooland.net  Wed Feb 14 01:28:17 2001
From: chris at voodooland.net (Chris Watson)
Date: Tue, 13 Feb 2001 18:28:17 -0600 (CST)
Subject: Embedding UNIX Commands
In-Reply-To: <rB7i6.1898$AX.389928@news03.micron.net>
Message-ID: <Pine.BSF.4.21.0102131826510.41335-100000@open-systems.net>

> Which _is_ bash (on most Linux Dists):
> ucntcme at locutus in /home/ucntcme 
> $ ll /bin/sh lrwxrwxrwx    1 root     root            4 Jan 13 19:58 /bin/sh -> bash*

	Non portable. Non standard. Big surprise. Please dont inject code
out their that uses bash when you really mean to use sh. It wont work on
other unix without bash.

--
=============================================================================
-Chris Watson         (316) 326-3862 | FreeBSD Consultant, FreeBSD Geek 
Work:              scanner at jurai.net | Open Systems Inc., Wellington, Kansas
Home:  scanner at deceptively.shady.org | http://open-systems.net
=============================================================================
WINDOWS: "Where do you want to go today?"
LINUX: "Where do you want to go tomorrow?"
BSD: "Are you guys coming or what?"
=============================================================================
GNU = Gnu's Non-portable & Unstable
irc.openprojects.net #FreeBSD -Join the revolution!
ICQ: 20016186




From csrabak at uol.com.br  Tue Feb 13 01:56:07 2001
From: csrabak at uol.com.br (Cesar Rabak)
Date: Mon, 12 Feb 2001 21:56:07 -0300
Subject: Embedding UNIX Commands
References: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com> <mailman.982008497.29571.python-list@python.org>
Message-ID: <3A888627.4EB23450@uol.com.br>

D-Man wrote:
> 
> You can use the system function :
> 
> import os
> os.system( "whoami" )
> 
> I don't think you can use setenv -- that is a csh thing, and system()
> won't even start up a shell (AFAIK).  Even if it did, it wouldn't do
> you any good since that shell would terminate and its environment
> would be meaningless.  (Also, on Linux systems bash is the default
> shell)

setenv is also a non standard (albeit very commonplace) function which
is available in several C libraries.

Perhaps Python can call it if it was compiled with a library that has
it?


From bpederson at geocities.com  Mon Feb 12 21:17:27 2001
From: bpederson at geocities.com (Barry Pederson)
Date: Mon, 12 Feb 2001 14:17:27 -0600
Subject: QuakeWorld Python 1.0
Message-ID: <3A8844D7.1CE2CC58@geocities.com>

QWPython (http://qwpython.sourceforge.net) is a Python-powered
QuakeWorld dedicated server. The core engine has been wrapped up as a
Python module and altered to call back to Python to execute Quake game
logic. A QuakeC -> Python translator is also included, along with
pre-translated versions of Deathmatch and Capture the Flag.

Version 1.0 includes sourcecode which builds and runs on both Windows
and Unix platforms, and Win32 binaries.

        Barry


From chris.gonnerman at usa.net  Tue Feb 13 03:18:03 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Mon, 12 Feb 2001 20:18:03 -0600
Subject: QuakeWorld Python 1.0
References: <3A8844D7.1CE2CC58@geocities.com>
Message-ID: <011e01c09563$837ecc80$a100000a@local>

----- Original Message -----
From: "Barry Pederson" <bpederson at geocities.com>
Subject: QuakeWorld Python 1.0
> Version 1.0 includes sourcecode which builds and runs on both Windows
> and Unix platforms, and Win32 binaries.

Where ARE the Win32 binaries?  I downloaded the Win32 package but got no
binaries
(no .exe or .dll files).






From bpederson at geocities.com  Tue Feb 13 04:50:43 2001
From: bpederson at geocities.com (Barry Pederson)
Date: Mon, 12 Feb 2001 21:50:43 -0600
Subject: QuakeWorld Python 1.0
References: <3A8844D7.1CE2CC58@geocities.com> <mailman.982030514.9176.python-list@python.org>
Message-ID: <3A88AF13.7B0B5275@geocities.com>


Chris Gonnerman wrote:
> 
> ----- Original Message -----
> From: "Barry Pederson" <bpederson at geocities.com>
> Subject: QuakeWorld Python 1.0
> > Version 1.0 includes sourcecode which builds and runs on both Windows
> > and Unix platforms, and Win32 binaries.
> 
> Where ARE the Win32 binaries?  I downloaded the Win32 package but got no
> binaries
> (no .exe or .dll files).

Under Windows, Python uses .pyd instead of .dll for native-code
extensions.  

In the QWPython Win32 distribution, qwpython\qwsv.pyd is what you're
looking for.  To actually run a game, you run the ctf.py or
deathmatch.py scripts, which take care of loading the native-code
module, along with the pythonized game code (which should be covered in
the README.html file)

	Barry


From jepler at inetnebr.com  Tue Feb 13 14:50:43 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Tue, 13 Feb 2001 13:50:43 GMT
Subject: QuakeWorld Python 1.0
References: <3A8844D7.1CE2CC58@geocities.com>
Message-ID: <slrn98ieti.1i6.jepler@potty.housenet>

On Mon, 12 Feb 2001 14:17:27 -0600, Barry Pederson
 <bpederson at geocities.com> wrote:
>QWPython (http://qwpython.sourceforge.net) is a Python-powered
>QuakeWorld dedicated server.

This is very cool.  I've been wishing that something like this would happen
for quite awhile.  (In fact, I tried my hand at a pure-Python Quake server
once, long before the source was available---Like most of my big projects
seem to do, it ultimately went nowhere)

Jeff


From shredwheat at mediaone.net  Wed Feb 14 07:11:50 2001
From: shredwheat at mediaone.net (Pete Shinners)
Date: Wed, 14 Feb 2001 06:11:50 GMT
Subject: QuakeWorld Python 1.0
References: <3A8844D7.1CE2CC58@geocities.com> <slrn98ieti.1i6.jepler@potty.housenet>
Message-ID: <Gkpi6.10116$_D.574713@typhoon.we.rr.com>

> This is very cool.  I've been wishing that something like this would happen
> for quite awhile.  (In fact, I tried my hand at a pure-Python Quake server
> once, long before the source was available---Like most of my big projects
> seem to do, it ultimately went nowhere)

this is neat, i spent some time looking at it this afternoon.
the origin worlds online game in development is using a very
similar approach. the server is using python for all the
game logic





From hamish_lawson at yahoo.co.uk  Mon Feb 12 21:18:00 2001
From: hamish_lawson at yahoo.co.uk (=?iso-8859-1?q?Hamish=20Lawson?=)
Date: Mon, 12 Feb 2001 20:18:00 +0000 (GMT)
Subject: Help: non-capturing group RE fails
Message-ID: <20010212201800.11411.qmail@web215.mail.yahoo.com>

Robin Thomas wrote:

> (?:[^@]) is not a "group" that can be referenced. It is an
> "assertion", or "lookahead assertion", that is part of the match but
> does not get a group 

I don't mean to pick holes in Robin's explanation, but I believe the
terminology used above isn't quite correct regarding assertions. I
point this out just so that Roy doesn't get confused when reading
reference material on REs - the concepts can be tricky enough!

Robin is correct in saying that (?:) can't be referenced since it's not
memorised; it's what's called a non-capturing group. However it does
consume the string, in that it causes the RE engine's 'pointer' to be
advanced through the string, and hence by definition isn't an
assertion. An assertion specifies a condition that must be fulfilled,
but doesn't causing the string to be consumed (i.e. doesn't advance the
pointer). (But to be fair to Robin, he makes the difference between
consuming and capturing clearer in another post.) The term 'capturing'
relates to whether the content is memorised for later reference
(whether by index or by name); 'consuming' to whether it causes the
pointer to be advanced.

Consider the code below.

r = re.compile("a(?:bd)+(?=c)(..)")
s = "HKabdbdGKMDabdbdbdcWFGT"
m = r.search(s)
print m.group(1)

This prints out 'cW'.

I want the '+' to apply to the pair 'bd' rather than just 'd', but
since I don't care about referring to the 'bd' pairs that are matched,
I use a non-capturing group '(?:bd)' rather than a capturing group
(bd); this means that the later '(..)' group in the pattern will be the
first referrable group.

The '(?=c)' in the pattern is a look-ahead assertion that specifies
that the 'bd' pairs must be followed by 'c'. But because it is an
assertion, it doesn't cause the string to be consumed. Hence the 'c' is
still available to be consumed by the '(..)' expression.

Hamish Lawson



____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From andreas.dietrich at news.online.de  Mon Feb 12 21:27:10 2001
From: andreas.dietrich at news.online.de (Andreas Dietrich)
Date: Mon, 12 Feb 2001 21:27:10 +0100
Subject: How do I modify my callers local variables?
Message-ID: <uug969.9s8.ln@news.onlinehome.de>

Hallo, Pythoneers!

I tried to simplify keyword parsing and wrote the following routine:

def parse_keywords(defaults,kw):
    import sys
    try:
        raise None
    except:
        frame = sys.exc_info()[2].tb_frame.f_back
        locals_dict=frame.f_locals
    for i in defaults.keys():
        try:
            locals_dict[i]=kw[i]
        except KeyError:
            try:
                locals_dict[i]=defaults[i]
            except KeyError:
	        #We should not be here
                print 'Moo. Moo'
    return

def test(**kw):
    parse_keywords({'a':1,'b':2,'c':3},kw)
    #didn't work
    print a,b,c

test()

Unfortunately it doesn't work.
Is there a way to do it? 
I'm tired of the repetitious code involved in keyword parsing.

I'd really appreciate any insights into this matter.

Thank you. Andreas

-- 
"We've heard that a million monkeys at a million keyboards could produce the
complete works of Shakespeare. Now, thanks to the Internet, we know this is not
true."  -- Robert Wilensky



From jeremy at alum.mit.edu  Mon Feb 12 23:05:41 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Mon, 12 Feb 2001 17:05:41 -0500 (EST)
Subject: How do I modify my callers local variables?
In-Reply-To: <uug969.9s8.ln@news.onlinehome.de>
References: <uug969.9s8.ln@news.onlinehome.de>
Message-ID: <14984.24117.294067.705291@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "AD" == Andreas Dietrich <andreas.dietrich at news.online.de> writes:

  AD> def parse_keywords(defaults,kw):
  AD>     import sys try:
  AD>         raise None
  AD>     except:
  AD>         frame = sys.exc_info()[2].tb_frame.f_back
  AD>         locals_dict=frame.f_locals
  AD>     for i in defaults.keys():
  AD>         try:
  AD>             locals_dict[i]=kw[i]
  AD>         except KeyError:
  AD>             try:
  AD>                 locals_dict[i]=defaults[i]
  AD>             except KeyError:
  AD> 	        #We should not be here
  AD>                 print 'Moo. Moo'
  AD>     return

  AD> def test(**kw):
  AD>     parse_keywords({'a':1,'b':2,'c':3},kw)
  AD>     #didn't work
  AD>     print a,b,c

  AD> test()

  AD> Unfortunately it doesn't work.  Is there a way to do it?  I'm
  AD> tired of the repetitious code involved in keyword parsing.

There is no way to do it.  If there were, it would be tempting to
disable it <0.5 wink>.

  AD> I'd really appreciate any insights into this matter.

When test() is compiled, the compiler determines that a, b, and c are
all global variables.  The body of test() does not contains
assignments or other name binding operations.

If you want to have default arguments for local variables, I suggest
using ... default arguments!

def test(a=1, b=2, c=3):
    print a, b, c

Jeremy



From andreas.dietrich at news.online.de  Mon Feb 12 23:48:08 2001
From: andreas.dietrich at news.online.de (Andreas Dietrich)
Date: Mon, 12 Feb 2001 23:48:08 +0100
Subject: How do I modify my callers local variables?
References: <uug969.9s8.ln@news.onlinehome.de>
Message-ID: <87p969.t89.ln@news.onlinehome.de>

On Mon, 12 Feb 2001 21:27:10 +0100, I wrote:
>def test(**kw):
>    parse_keywords({'a':1,'b':2,'c':3},kw)
>    #didn't work
>    print a,b,c
>
>test()

I?hate to post a followup to my own posting, but I forgot to point out the
_real_ problem. It doesn't work even if I replace test() with

def test(**kw):
    a=b=c=0
    parse_keywords({'a':1,'b':2,'c':3},kw)
    #didn't work
    print a,b,c

I'm aware of the fact that python considers variables as global that are
not assigned to (Could I modify _that_ from parse_keywords ? If not, this
not a showstopper. )

Andreas

-- 
"We've heard that a million monkeys at a million keyboards could produce the
complete works of Shakespeare. Now, thanks to the Internet, we know this is not
true."  -- Robert Wilensky



From tjreedy at udel.edu  Tue Feb 13 03:57:50 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 12 Feb 2001 21:57:50 -0500
Subject: How do I modify my callers local variables?
References: <uug969.9s8.ln@news.onlinehome.de>
Message-ID: <96a7jt$6no$1@news.udel.edu>

> def parse_keywords(defaults,kw):
>     import sys
>     try:
>         raise None
>     except:
>         frame = sys.exc_info()[2].tb_frame.f_back
>         locals_dict=frame.f_locals

This only gives you a PyDict copy of the function locals, which happen to
be internally implemented as an indexed C array.

>     for i in defaults.keys():
>         try:
>             locals_dict[i]=kw[i]
>         except KeyError:
>             try:
>                 locals_dict[i]=defaults[i]

Modifying your PyDict copy has no effect on the internal C array that the
function actually uses.  Thus this assignment in useless in terms of what
you hope to accomplish.

> ...
> I'd really appreciate any insights into this matter.

I hope this helps.

Terry J. Reedy




From sandj.williams at gte.net  Mon Feb 12 21:42:16 2001
From: sandj.williams at gte.net (Steve Williams)
Date: Mon, 12 Feb 2001 20:42:16 GMT
Subject: Google buys Deja
Message-ID: <3A884C33.B253C411@gte.net>

http://www.theregister.co.uk/content/6/16836.html



From bs1535 at sbc.com  Mon Feb 12 22:17:05 2001
From: bs1535 at sbc.com (SNYDER, BARRON F. (AIT))
Date: Mon, 12 Feb 2001 16:17:05 -0500
Subject: Can a Tkinter Scrollbar scroll 2 widgets at the same time?
Message-ID: <35BD410BA148D411A7ED00508BCFFBDA05CB2824@msgil65170u05.nbk2305.il.ameritech.com>

Since Tkinter doesn't have a table widget, I'm using a ScrolledListBox and a
ScrolledFrame (to hold the column headings as Buttons). I would like the
scrollbar from the ScrolledListBox to scroll both ScrolledListBox and
ScrolledFrame at the same time. Is this possible?

If not, can a separate Scrollbar scroll more than one widget (can I scroll,
say, 2 listboxes with one scrollbar?)?

Thanks,
Barron



From sholden at holdenweb.com  Mon Feb 12 22:49:46 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 12 Feb 2001 16:49:46 -0500
Subject: Can a Tkinter Scrollbar scroll 2 widgets at the same time?
References: <mailman.982012694.3832.python-list@python.org>
Message-ID: <JXYh6.9517$n4.75608@e420r-atl2.usenetserver.com>

"SNYDER, BARRON F. (AIT)" <bs1535 at sbc.com> wrote in message
news:mailman.982012694.3832.python-list at python.org...
>
> Since Tkinter doesn't have a table widget, I'm using a ScrolledListBox and
a
> ScrolledFrame (to hold the column headings as Buttons). I would like the
> scrollbar from the ScrolledListBox to scroll both ScrolledListBox and
> ScrolledFrame at the same time. Is this possible?
>
> If not, can a separate Scrollbar scroll more than one widget (can I
scroll,
> say, 2 listboxes with one scrollbar?)?
>
> Thanks,
> Barron
>

Traditionally you do this by registering the set() method of the scroll bar
as the yscrollcommand of each of the listboxes, and registering a callback
for the scroll bar which adjusts the positions of each listbox.

Here's an extract from the __init__() method of a mail searchwindow, which
shows you how this works in practice:

        self.msgScrollBar.config(command=self.scrollMsg)
        self.msgDateList = Listbox(self.msgFrame, borderwidth=1, bg='white',
                        yscrollcommand=self.msgScrollBar.set,
exportselection=0)
        self.msgDateList.bind("<Double-Button-1>", self.showMessage)
        self.msgSenderList = Listbox(self.msgFrame, borderwidth=1,
bg='white',
                        yscrollcommand=self.msgScrollBar.set,
exportselection=0)
        self.msgSubjectList = Listbox(self.msgFrame, borderwidth=1,
bg='white',
                        yscrollcommand=self.msgScrollBar.set,
exportselection=0)

And here's the code of the scrollMsg callback from msgScrollBar events:

    def scrollMsg(self, *args):
        apply(self.msgDateList.yview, args)
        apply(self.msgSenderList.yview, args)
        apply(self.msgSubjectList.yview, args)

This does have limitations, though, since any changes to the scrolling lists
made by keyboard aren't reflected in the others.

regards
 Steve

PS: Can anyone explain how I might implement a columnar scrolling list,
allowing different text elements to align vertically inside each row of a
scrolling list -- without using a fixed-pitch font? This would solve an
interface problem for me.




From ken at lt.com  Mon Feb 12 22:50:35 2001
From: ken at lt.com (Kenneth Loafman)
Date: Mon, 12 Feb 2001 15:50:35 -0600
Subject: How do I change a process name or even a thread name in python!
References: <3a884bc2@si-nic.hrz.uni-siegen.de>
Message-ID: <3A885AAB.863BC0B8@lt.com>

C manages because the executable is the same as the program.  A script
is different.  The interpreter name (python, sh, bash, etc.) will appear
as the name if you run it like '$ python foo.py'.  Same applies to '$ sh
script'.

Try putting a '#! /usr/bin/python' as the first line, then running it as
'$ ./foo.py'.  Under Linux/UNIX, that picks up the right name and
arguments for ps aux.

If you have to run it as 'python foo.py ...', I have no idea.

...Ken

Fionn Behrens wrote:
> 
> I am sick of seeing 10x"python" in top, without being able to say which is
> which.
> 
> I mean the name of the process as it appears in "ps" or "top". Obviously,
> changing argv[0] like in C does not do the trick (which would not be of
> much use for distinguishable threads anyway)
> 
> Any suggestions?
> 
> Regards,
>         Fionn
> --
> http://mail.python.org/mailman/listinfo/python-list



From kalle at gnupung.net  Tue Feb 13 00:02:20 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Tue, 13 Feb 2001 00:02:20 +0100
Subject: How do I change a process name or even a thread name in python!
In-Reply-To: <3A885AAB.863BC0B8@lt.com>; from ken@lt.com on Mon, Feb 12, 2001 at 03:50:35PM -0600
References: <3a884bc2@si-nic.hrz.uni-siegen.de> <3A885AAB.863BC0B8@lt.com>
Message-ID: <20010213000220.A13868@apone.network.loc>

[reformatted included text]
Sez Kenneth Loafman:
> Fionn Behrens wrote:
> > 
> > I am sick of seeing 10x"python" in top, without being able to say which is
> > which.
> 
> Try putting a '#! /usr/bin/python' as the first line, then running it as
> '$ ./foo.py'.  Under Linux/UNIX, that picks up the right name and
> arguments for ps aux.

As long as the magic line isn't
#!/usr/bin/env python

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010213/64d8fc80/attachment.pgp>

From donn at u.washington.edu  Tue Feb 13 01:37:39 2001
From: donn at u.washington.edu (Donn Cave)
Date: 13 Feb 2001 00:37:39 GMT
Subject: How do I change a process name or even a thread name in python!
References: <3a884bc2@si-nic.hrz.uni-siegen.de> <mailman.982014674.11647.python-list@python.org>
Message-ID: <969vkj$sak$1@nntp6.u.washington.edu>

Quoth Kenneth Loafman <ken at lt.com>:
| C manages because the executable is the same as the program.  A script
| is different.  The interpreter name (python, sh, bash, etc.) will appear
| as the name if you run it like '$ python foo.py'.  Same applies to '$ sh
| script'.
|
| Try putting a '#! /usr/bin/python' as the first line, then running it as
| '$ ./foo.py'.  Under Linux/UNIX, that picks up the right name and
| arguments for ps aux.

It does?  I don't think so!

$ ./iam &
[1] 28372
$ ps wwaux | fgrep iam
donn     28372  0.8  0.4  2768 1140 pts/1    S    16:31   0:00 /usr/local/bin/python ./iam

	Donn Cave, donn at u.washington.edu


From erno-news at erno.iki.fi  Tue Feb 13 04:49:10 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 13 Feb 2001 05:49:10 +0200
Subject: How do I change a process name or even a thread name in python!
References: <3a884bc2@si-nic.hrz.uni-siegen.de> <mailman.982014674.11647.python-list@python.org> <969vkj$sak$1@nntp6.u.washington.edu>
Message-ID: <kun1brxmx5.fsf@lasipalatsi.fi>

In article <969vkj$sak$1 at nntp6.u.washington.edu>, Donn Cave
<donn at u.washington.edu> writes:

| Quoth Kenneth Loafman <ken at lt.com>:
|| Try putting a '#! /usr/bin/python' as the first line, then running it as
|| '$ ./foo.py'.  Under Linux/UNIX, that picks up the right name and
|| arguments for ps aux.

| It does?  I don't think so!

| $ ./iam &
| [1] 28372
| $ ps wwaux | fgrep iam
| donn     28372  0.8  0.4  2768 1140 pts/1    S    16:31   0:00 /usr/local/bin/python ./iam

clearly, python needs setproctitle() in the posix module.

  -- erno


From phd at phd.pp.ru  Tue Feb 13 10:32:50 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 13 Feb 2001 12:32:50 +0300 (MSK)
Subject: How do I change a process name or even a thread name in python!
In-Reply-To: <kun1brxmx5.fsf@lasipalatsi.fi>
Message-ID: <Pine.LNX.4.30.0102131232060.2544-100000@fep132.fep.ru>

On 13 Feb 2001, Erno Kuusela wrote:
> clearly, python needs setproctitle() in the posix module.

   For a long long while I longing to write one, but always short on time!

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From sdm7g at virginia.edu  Wed Feb 14 01:16:53 2001
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 13 Feb 2001 19:16:53 -0500 (EST)
Subject: setproctitle [was: How do I change a process name or even a thread
 name in python!]
In-Reply-To: <Pine.LNX.4.30.0102131232060.2544-100000@fep132.fep.ru>
Message-ID: <Pine.NXT.4.21.0102131906450.268-100000@localhost.virginia.edu>


On Tue, 13 Feb 2001, Oleg Broytmann wrote:

> On 13 Feb 2001, Erno Kuusela wrote:
> > clearly, python needs setproctitle() in the posix module.
> 
>    For a long long while I longing to write one, but always short on time!
> 

Because of the posix_1str wrapper in posixmodule, it ought to be pretty
trivial ... something like:

- #ifdef HAVE_SETPROCTITLE
- 
- static char posix_setproctitle__doc__[] = 
- "setproctitle(path) -> None\n\
- Set the process title."; 
- 
- static PyObject *
- posix_setproctitle(PyObject *self, PyObject *args)
- {
-       return posix_1str(args, "s:setproctitle", setproctitle);
- }
- 
- #endif HAVE_SETPROCTITLE


I was about to try this, but I discovered that althought there's
a man page to setproctitle(3) in macosx/darwin, they inadvertently
left that function out of the public beta. ( Well -- more precisely,
I'm told the code is there, but it's not in the library. )  

The actual args to (C) setproctitle is a format string and optional
args, but it makes more sense to do any string formatting in Python,
and just pass that string. 

A nicer implementation would allow no arg to reset to the default
proctitle.

-- Steve Majewski





From donn at u.washington.edu  Wed Feb 14 02:00:10 2001
From: donn at u.washington.edu (Donn Cave)
Date: 14 Feb 2001 01:00:10 GMT
Subject: setproctitle [was: How do I change a process name or even a thread  name in python!]
References: <mailman.982109894.26467.python-list@python.org>
Message-ID: <96claq$gro$1@nntp6.u.washington.edu>

Quoth "Steven D. Majewski" <sdm7g at virginia.edu>:
| On Tue, 13 Feb 2001, Oleg Broytmann wrote:
|> On 13 Feb 2001, Erno Kuusela wrote:
|>> clearly, python needs setproctitle() in the posix module.
|> 
|>    For a long long while I longing to write one, but always short on time!

| Because of the posix_1str wrapper in posixmodule, it ought to be pretty
| trivial ... something like:
|
| - #ifdef HAVE_SETPROCTITLE
| - 
| - static char posix_setproctitle__doc__[] = 
| - "setproctitle(path) -> None\n\
| - Set the process title."; 
| - 
| - static PyObject *
| - posix_setproctitle(PyObject *self, PyObject *args)
| - {
| -       return posix_1str(args, "s:setproctitle", setproctitle);
| - }
| - 
| - #endif HAVE_SETPROCTITLE
|
|
| I was about to try this, but I discovered that althought there's
| a man page to setproctitle(3) in macosx/darwin, they inadvertently
| left that function out of the public beta. ( Well -- more precisely,
| I'm told the code is there, but it's not in the library. )  
|
| The actual args to (C) setproctitle is a format string and optional
| args, but it makes more sense to do any string formatting in Python,
| and just pass that string. 
|
| A nicer implementation would allow no arg to reset to the default
| proctitle.

It looks to me like the only platforms that have setproctitle()
at this point are NetBSD, FreeBSD, OpenBSD and BSDI.  Most of the
rest of the UNIX world does have some way to overwrite the command
line as seen from "ps", but not all very pretty.

	Donn Cave, donn at u.washington.edu


From sdm7g at virginia.edu  Wed Feb 14 04:15:42 2001
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 13 Feb 2001 22:15:42 -0500 (EST)
Subject: setproctitle [was: How do I change a process name or even a thread name in python!]
In-Reply-To: <96claq$gro$1@nntp6.u.washington.edu>
Message-ID: <Pine.A32.3.90.1010213221029.14204A-100000@elvis.med.Virginia.EDU>

On 14 Feb 2001, Donn Cave wrote:

> 
> It looks to me like the only platforms that have setproctitle()
> at this point are NetBSD, FreeBSD, OpenBSD and BSDI.  Most of the
> rest of the UNIX world does have some way to overwrite the command
> line as seen from "ps", but not all very pretty.
> 

Maybe there's a evil curse on the function, Donn. 

The man page for it on macosx seems to die in midsentence,
as if the author was struck dead for trying to tell us it's
bizarre history.  

SEE ALSO
     ps(1),  w(1),  printf(3)

HISTORY
     The setproctitle() function first appeared in

BSD Experimental                April 13, 1994                               1



In fact, not dissimilar from how the Python Underground has 
been known to



From jepler at inetnebr.com  Wed Feb 14 15:14:30 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 14 Feb 2001 14:14:30 GMT
Subject: setproctitle [was: How do I change a process name or even a thread name in python!]
References: <mailman.982109894.26467.python-list@python.org>
Message-ID: <slrn98l4m5.2s8.jepler@potty.housenet>

On Tue, 13 Feb 2001 19:16:53 -0500 (EST), Steven D. Majewski
 <sdm7g at virginia.edu> wrote:
>The actual args to (C) setproctitle is a format string and optional
>args, but it makes more sense to do any string formatting in Python,
>and just pass that string. 

Actually, this is unsafe, and has been the cause of a number of bugs 
recently exposed on Bugtraq.

Imagine that you are writing an ftp server which uses 'setproctitle' to
change its process name to give information including the e-mail address an
anonymous user gave when he logged in.  The user might pass a string like
"%x%x%x%x", which will be treated as printf-style specifiers by setproctitle.

So, rather than
	setproctitle(new_title);
you must
	setproctitle("%s", new_title);
to be safe.

See bugtraq or search for "setproctitle format vulnerability" for more
details on this potential security risk.  One URL:
	http://www.hert.org/papers/format.html

Unfortunately, that means the convenience function in the posix module
can't be used.

Jeff
jepler at inetnebr.com


From phd at phd.pp.ru  Wed Feb 14 10:20:47 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 14 Feb 2001 12:20:47 +0300 (MSK)
Subject: setproctitle [was: How do I change a process name or even a
 thread name in python!]
In-Reply-To: <Pine.NXT.4.21.0102131906450.268-100000@localhost.virginia.edu>
Message-ID: <Pine.LNX.4.30.0102141216550.8189-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Steven D. Majewski wrote:
> -       return posix_1str(args, "s:setproctitle", setproctitle);

   setproctitle is available in libc only in *BSD. Other unicies require
different tricks. Have you ever looked at sendmail source code? You'll find
there many surprises. The problem is their license.
   May be I'll steal the code from Postgres (which has similar, but
different implementation, covered under BSD license).

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From chris.gonnerman at usa.net  Wed Feb 14 14:49:49 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Wed, 14 Feb 2001 07:49:49 -0600
Subject: setproctitle [was: How do I change a process name or even athread name in python!]
References: <Pine.LNX.4.30.0102141216550.8189-100000@fep132.fep.ru>
Message-ID: <000201c096d7$a8d76f80$a100000a@local>

----- Original Message -----
From: "Oleg Broytmann" <phd at phd.pp.ru>
Subject: Re: setproctitle [was: How do I change a process name or even
athread name in python!]
>    setproctitle is available in libc only in *BSD. Other unicies require
> different tricks. Have you ever looked at sendmail source code? You'll
find
> there many surprises. The problem is their license.
>    May be I'll steal the code from Postgres (which has similar, but
> different implementation, covered under BSD license).

I looked at several implementations of setproctitle... they all seem to need
to be initialized at
program startup with the argc, argv and envp values from the call to main().
I'm not sure how
to do this in a module for Python without changing the main interpreter (but
I haven't looked
at the interpreter source yet...)





From ken at lt.com  Tue Feb 13 14:48:58 2001
From: ken at lt.com (Kenneth Loafman)
Date: Tue, 13 Feb 2001 07:48:58 -0600
Subject: How do I change a process name or even a thread name in python!
References: <3a884bc2@si-nic.hrz.uni-siegen.de> <mailman.982014674.11647.python-list@python.org> <969vkj$sak$1@nntp6.u.washington.edu>
Message-ID: <3A893B4A.9ECCB53C@lt.com>

Donn Cave wrote:
> 
> Quoth Kenneth Loafman <ken at lt.com>:
> | C manages because the executable is the same as the program.  A script
> | is different.  The interpreter name (python, sh, bash, etc.) will appear
> | as the name if you run it like '$ python foo.py'.  Same applies to '$ sh
> | script'.
> |
> | Try putting a '#! /usr/bin/python' as the first line, then running it as
> | '$ ./foo.py'.  Under Linux/UNIX, that picks up the right name and
> | arguments for ps aux.
> 
> It does?  I don't think so!
> 
> $ ./iam &
> [1] 28372
> $ ps wwaux | fgrep iam
> donn     28372  0.8  0.4  2768 1140 pts/1    S    16:31   0:00 /usr/local/bin/python ./iam

You're right.  My top shows it as foo.py, but ps does not.  I may have
been confusing the behavior with Solaris.  Its been a while since I
touched a Sun box though.

...Ken



From danw at rtp.ericsson.se  Mon Feb 12 23:10:44 2001
From: danw at rtp.ericsson.se (Daniel Wickstrom)
Date: 12 Feb 2001 17:10:44 -0500
Subject: Calling static methods in Jython
Message-ID: <5cn1brbli3.fsf@rtp.ericsson.se>

I've already posted this on the jython-users list, but I haven't
received a response.

I want to use Jython with some existing java classes that contain
static methods.  How does one call java static methods from Jython?



-Dan


From rgruet at intraware.com  Mon Feb 12 23:25:59 2001
From: rgruet at intraware.com (Richard Gruet)
Date: Mon, 12 Feb 2001 22:25:59 GMT
Subject: Calling static methods in Jython
References: <5cn1brbli3.fsf@rtp.ericsson.se>
Message-ID: <3A8862F6.6CC23EE7@intraware.com>

If the Java class is C and the static method M, simply do from Python:
C.M(args...)

Daniel Wickstrom wrote:

> I've already posted this on the jython-users list, but I haven't
> received a response.
>
> I want to use Jython with some existing java classes that contain
> static methods.  How does one call java static methods from Jython?
>
> -Dan
-------------- next part --------------
A non-text attachment was scrubbed...
Name: rgruet.vcf
Type: text/x-vcard
Size: 370 bytes
Desc: Card for Richard Gruet
URL: <http://mail.python.org/pipermail/python-list/attachments/20010212/e70fec25/attachment.vcf>

From stephen_purcell at yahoo.com  Tue Feb 13 09:54:12 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 13 Feb 2001 09:54:12 +0100
Subject: Calling static methods in Jython
In-Reply-To: <5cn1brbli3.fsf@rtp.ericsson.se>; from danw@rtp.ericsson.se on Mon, Feb 12, 2001 at 05:10:44PM -0500
References: <5cn1brbli3.fsf@rtp.ericsson.se>
Message-ID: <20010213095412.E5558@freedom.puma-ag.com>

Daniel Wickstrom wrote:
> 
> I've already posted this on the jython-users list, but I haven't
> received a response.
> 
> I want to use Jython with some existing java classes that contain
> static methods.  How does one call java static methods from Jython?

Just as you would do it in Java:

Jython 2.0 on java1.2.2 (JIT: javacomp)
Type "copyright", "credits" or "license" for more information.
>>> import java.text.DateFormat
>>> df = java.text.DateFormat.getDateInstance(java.text.DateFormat.MEDIUM)
>>> df
java.text.SimpleDateFormat at ce9bf0a5


-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From rgruet at intraware.com  Mon Feb 12 23:13:43 2001
From: rgruet at intraware.com (Richard Gruet)
Date: Mon, 12 Feb 2001 22:13:43 GMT
Subject: sluggish startup with 2.0
Message-ID: <3A886015.53BF722E@intraware.com>

Hey Pythoners,

Apologies if that question has already been raised.
Startup/launch time seems far longer in Python 2.0 than it was in 1.5.2.

Is it normal or is it specific to my configuration ? I use ActivePython
build 202 on Windows NT 4 SP6.

Thanks

-------------- next part --------------
A non-text attachment was scrubbed...
Name: rgruet.vcf
Type: text/x-vcard
Size: 370 bytes
Desc: Card for Richard Gruet
URL: <http://mail.python.org/pipermail/python-list/attachments/20010212/85532fa4/attachment.vcf>

From rgruet at intraware.com  Tue Feb 13 18:07:40 2001
From: rgruet at intraware.com (Richard Gruet)
Date: Tue, 13 Feb 2001 17:07:40 GMT
Subject: sluggish startup with 2.0
References: <3A886015.53BF722E@intraware.com>
Message-ID: <3A8969DA.2FF1D856@intraware.com>

I answer to my own question. In fact, Python 2.0 sluggishness on startup was
due to some misconfiguration in the Windows shell associations on my PC. Now
it's fixed.

Richard Gruet wrote:

> Hey Pythoners,
>
> Apologies if that question has already been raised.
> Startup/launch time seems far longer in Python 2.0 than it was in 1.5.2.
>
> Is it normal or is it specific to my configuration ? I use ActivePython
> build 202 on Windows NT 4 SP6.
>
> Thanks
-------------- next part --------------
A non-text attachment was scrubbed...
Name: rgruet.vcf
Type: text/x-vcard
Size: 370 bytes
Desc: Card for Richard Gruet
URL: <http://mail.python.org/pipermail/python-list/attachments/20010213/523a4af6/attachment.vcf>

From nickb at earth.ox.ac.uk  Mon Feb 12 23:56:24 2001
From: nickb at earth.ox.ac.uk (Nick Belshaw)
Date: Mon, 12 Feb 2001 22:56:24 +0000
Subject: help: silent failure on socket termination?
Message-ID: <3A886A17.B04EE140@earth.ox.ac.uk>

I was wondering if this is a problem which can be fixed?

It is very upleasant and I have not found any references to similar.

I am running Python1.5.2 on Redhat 6.2

Using  the ayncore/asynchat modules to support customized communications
between
client and server  code-modules on the same or different machines I am
suffering an occasional silent failure of my server module when
terminating the Client during
conditions when the client-server link is busy.

9 times out of 10 there is no problem and the socket connection
terminates correctly
but the other time the server code simply exits without any error
message!!!????
No core dump, error messages or anything anywhere.....the program simply
exits....
I didn't know such a thing could happen !!!!!!!

I am presuming the failure is inside the socket module since the
asyncore.py routines seems
to log/flag all eventualities and so can not be seeing the problem.

Posting code here would really be impossible - too much is going on  :-(

My question is therefore, are there any known problems with socket.py or
am I looking at
a problem specific to my configuration such as a marginal Network card
or similar?

Unfortunately the problem only appeared on 'upgrading' from a Redhat6.0
configuration
which was running very stably for several years.....so hardware seems
unlikely.

any suggestions would be greatly appreciated.

Nick/Oxford




From amyk at foretec.com  Tue Feb 13 00:43:10 2001
From: amyk at foretec.com (Amy Katherine)
Date: Mon, 12 Feb 2001 18:43:10 -0500
Subject: ***Python 9 New Updates and Activities***
Message-ID: <4.2.2.20010212184223.00b5aed0@colossus.foretec.com>

Greetings!

The Python 9 conference agenda is now available on-line at 
http://www1.python9.org/p9-agenda.html.

Python 9 will offer two lunchtime "Birds of a Feather" (BOF) meetings, one 
on Tuesday, March 6 and one on Wednesday, March 7.  The topic for the 
Tuesday meeting is "Adding Iterators to Python", and the topic for the 
Wednesday meeting is "Adding Sets to Python".  To sign up for one or both 
of these BOFs, please send an email note to:  Iterator-BOF at python9.org 
and/or Set-BOF at python9.org.  On each day, tables will be reserved for BOF 
participants at lunchtime, and lunch will be served promptly at 12 
Noon.  Following lunch, BOF participants will adjourn to a meeting room for 
the remainder of the lunch break.  The location of the meeting room will be 
posted on the bulletin board in the registration area.

On-line registration for Python 9 closes on Sunday, February 25, 2001.  To 
register, please go to http://www1.python9.org/p9-reginfo.html.

Rooms at the Hilton Long Beach are no longer available at the special group 
rate of $134 per night.  However, rooms for Python 9 attendees are now 
available at the Holiday Inn Downtown Long Beach and the Westin Long Beach 
hotels at favorable rates. For more information, please go to 
http://www1.python9.org/p9-hotelinfo.html.

ActiveState (www.ActiveState.com) is pleased to launch its first annual 
Programmer's Choice award for the Python programmer who's actively 
contributing to the community.  Nominate your favorite programmer by 
February 20, 2001 by going to  http://www.ActiveState.com/Awards.  The 
winner will be announced at the Python 9 conference.
__________________________________________________________

Sponsors:
Platinum Sponsor: ActiveState (www.ActiveState.com)
Gold Sponsor: Digital Creations (www.digicool.com)
Silver Sponsor: O'Reilly & Associates (www.oreilly.com)

Exhibitors:
ActiveState (www.ActiveState.com)
Digital Creations (www.digicool.com)
Archaeopteryx (www.archaeopteryx.com)
Perforce Software (www.perforce.com)
Irvine Sci-Tech Books (www.scitechbooks.com) 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010212/a2677a60/attachment.htm>

From qiang.li at aya.yale.edu  Tue Feb 13 00:43:15 2001
From: qiang.li at aya.yale.edu (qiang)
Date: Mon, 12 Feb 2001 15:43:15 -0800
Subject: problem with import cephes under windowsNT
Message-ID: <982021397.659463@emperor.labs.agilent.com>

Hi,

I am a newbie. I tried to import cephes module under windowsNT but kept
getting error messages saying
"ImportError: DLL load failed: The specified module could not be found."
I have tried to place cephes.pyd in various paths including
c:\python20\dlls and c:\python20 with no success.
Perhaps someone can point out the problem?

Thanks a lot!






From MarkH at ActiveState.com  Tue Feb 13 08:34:11 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 13 Feb 2001 07:34:11 GMT
Subject: problem with import cephes under windowsNT
References: <982021397.659463@emperor.labs.agilent.com>
Message-ID: <3A88E292.9020103@ActiveState.com>

qiang wrote:

> I am a newbie. I tried to import cephes module under windowsNT but kept
> getting error messages saying
> "ImportError: DLL load failed: The specified module could not be found."
> I have tried to place cephes.pyd in various paths including
> c:\python20\dlls and c:\python20 with no success.
> Perhaps someone can point out the problem?


This generally means that cephes.pyd is dependent on _another_ DLL, and that DLL can't be found.

If you have msvc, run "dumpbin /imports cephes.pyd | grep -i dll" and you will get a reasonable list...  There are also freeware programs around that can tell you dependencies in a much nicer way, but I don't use one.

Mark.



From luckow at netvision.com.br  Tue Feb 13 01:34:28 2001
From: luckow at netvision.com.br (Luckow)
Date: Mon, 12 Feb 2001 21:34:28 -0300
Subject: Can I create HTML mails in Python,?? How??
Message-ID: <003d01c09554$f23190e0$675ed7c8@decio>

At the web site that I work I only get to send mails with the follow command.

os.popen("echo 'Message of the mail' | su myself -c mail 'Subject of mail' pedro at provedor.com.br")

However I find an commmand that I can create email in HTML.

 f = os.popen ( "/usr/sbin/sendmail", "w" )

 f.write ( """\
 From: pedro at terra.com.br
 To: joao at uol.com.br
 Subject: some subject
 
 body os de messagem
 """ )

this comand allow me to create since an simple text mail until an HTML email, but this command not run in my FreeBSD (with SendMail). The browser show an I/O Error.

Or some body have some source-code with this that I want.

SOME BODY CAN HELP-ME please.

tanks for all.

D?cio H. Luckow
(sorry my bad english)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010212/8a909f03/attachment.htm>

From sandipan at vsnl.com  Tue Feb 13 03:45:02 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Tue, 13 Feb 2001 08:15:02 +0530
Subject: Can I create HTML mails in Python,?? How??
References: <003d01c09554$f23190e0$675ed7c8@decio>
Message-ID: <00d601c09566$f7ad33a0$020000c3@node02>

You can use smtplib or os.popen (like you have written) for sending the mail, but need to use MimeWriter or such module to build a MIME email. And put the HTML document in the MIME body appropriately.

Regards,

Sandipan
  ----- Original Message ----- 
  From: Luckow 
  To: python-list at python.org 
  Sent: Tuesday, February 13, 2001 6:04 AM
  Subject: Can I create HTML mails in Python,?? How??


  At the web site that I work I only get to send mails with the follow command.

  os.popen("echo 'Message of the mail' | su myself -c mail 'Subject of mail' pedro at provedor.com.br")

  However I find an commmand that I can create email in HTML.

   f = os.popen ( "/usr/sbin/sendmail", "w" )

   f.write ( """\
   From: pedro at terra.com.br
   To: joao at uol.com.br
   Subject: some subject

   body os de messagem
   """ )

  this comand allow me to create since an simple text mail until an HTML email, but this command not run in my FreeBSD (with SendMail). The browser show an I/O Error.

  Or some body have some source-code with this that I want.

  SOME BODY CAN HELP-ME please.

  tanks for all.

  D?cio H. Luckow
  (sorry my bad english)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010213/2cc32250/attachment.htm>

From tjg at exceptionalminds.com  Tue Feb 13 05:57:46 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Mon, 12 Feb 2001 20:57:46 -0800
Subject: killing a task started with popen()
Message-ID: <20010212205746.A19651@trufflehunter.avalongroup.net>

Hi again.

I'm using the following code

x = os.popen('tail -fn1 /var/log/messages')

works beautifully.

however, x.close() doesn't work. I understand why, but I don't know
what to do about it. So, what method would be best to terminate the process
I started?

I guess I should just let it run until the parent exits, but that seems kind
of klunky.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  27 days  9:02 hours ago<<



From wcanopy at covad.net  Tue Feb 13 07:51:34 2001
From: wcanopy at covad.net (Wiwih Gunadi)
Date: Tue, 13 Feb 2001 00:51:34 -0600
Subject: Setting boolean COM properties
Message-ID: <96ak75$ru$1@newstest.laserlink.net>

Hi,

Does anybody know how to set a boolean COM object property?
In his book, Mark indicated two objects PyTrue and PyFalse but he didn't
tell how to get these objects.

Searching the internet yields dissapointing results leading me to believe
that these objects are obsolete or nobody ever use'em.

Pls help,
-w--






From MarkH at ActiveState.com  Tue Feb 13 08:37:11 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 13 Feb 2001 07:37:11 GMT
Subject: Setting boolean COM properties
References: <96ak75$ru$1@newstest.laserlink.net>
Message-ID: <3A88E346.2080106@ActiveState.com>

Wiwih Gunadi wrote:


> Does anybody know how to set a boolean COM object property?
> In his book, Mark indicated two objects PyTrue and PyFalse but he didn't
> tell how to get these objects.
> 
> Searching the internet yields dissapointing results leading me to believe
> that these objects are obsolete or nobody ever use'em.

You can just use integers 1 and 0, and they should work fine.

PyTrue and PyFalse are hidden objects.  You can get at them via:

true = 1==1
false = 1==0

Due to a quirk in Python, these variables will then look like ints, but 
are really the PyTrue and PyFalse objects, as shown here:

 >>> true = 1==1
 >>> true
1
 >>> true == 1
1
 >>> true is 1
0
 >>>

HTH,

Mark.



From max at alcyone.com  Tue Feb 13 08:36:17 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 12 Feb 2001 23:36:17 -0800
Subject: Setting boolean COM properties
References: <96ak75$ru$1@newstest.laserlink.net> <3A88E346.2080106@ActiveState.com>
Message-ID: <3A88E3F1.587F70EF@alcyone.com>

Mark Hammond wrote:

> Due to a quirk in Python, these variables will then look like ints,
> but
> are really the PyTrue and PyFalse objects, as shown here:

Which brings up a fairly obvious question:  Especially given that there
are hidden true and false objects in the Python core, how come there
isn't a distinct Boolean type in Python, with values true and false? 
There's nothing on this in the FAQ.

This is one of the few things I really wish was in Python, for
cleanliness sake.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There are no dull subjects.  There are only dull writers.
\__/ H.L. Mencken
    Alcyone Systems / http://www.alcyone.com/
 Alcyone Systems, San Jose, California.


From wcanopy at covad.net  Wed Feb 14 01:14:46 2001
From: wcanopy at covad.net (Wiwih Gunadi)
Date: Tue, 13 Feb 2001 18:14:46 -0600
Subject: Setting boolean COM properties
References: <96ak75$ru$1@newstest.laserlink.net> <3A88E346.2080106@ActiveState.com>
Message-ID: <96chau$54l$1@newstest.laserlink.net>

I see, in that case, the problem lies with the COM object property,
specifically:

Word.Application.ActiveDocument.Footnotes.Item(1).Range.TextRetrievalMode.In
cludeFieldCodes

In the msdn documentation, this property is read-write, but no matter what I
assign to it, it won't change value.

Anybody has different experience with this?

-Will--

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A88E346.2080106 at ActiveState.com...> Wiwih Gunadi wrote:
>
>
> > Does anybody know how to set a boolean COM object property?
> > In his book, Mark indicated two objects PyTrue and PyFalse but he didn't
> > tell how to get these objects.
> >
> > Searching the internet yields dissapointing results leading me to
believe
> > that these objects are obsolete or nobody ever use'em.
>
> You can just use integers 1 and 0, and they should work fine.
>
> PyTrue and PyFalse are hidden objects.  You can get at them via:
>
> true = 1==1
> false = 1==0
>
> Due to a quirk in Python, these variables will then look like ints, but
> are really the PyTrue and PyFalse objects, as shown here:
>
>  >>> true = 1==1
>  >>> true
> 1
>  >>> true == 1
> 1
>  >>> true is 1
> 0
>  >>>
>
> HTH,
>
> Mark.
>




From dhjung7 at orgio.net  Tue Feb 13 08:52:54 2001
From: dhjung7 at orgio.net (iact21)
Date: Tue, 13 Feb 2001 16:52:54 +0900
Subject: How can I use "tkFileDialog.askopenfilename()"
Message-ID: <96ap4f$5qp$1@b5nntp2.channeli.net>

>> import tkSimpleDialg, Tkinter
>> root = Tk()
>> fn = tkFileDialog,askopenfilename()

error...
Traceback (innermost last):
    File "<stdin>", line 1, in ?
NameError: tkFileDialog

Please..help..me...!!!





From b012414 at mail.dvz.fh-koeln.de  Tue Feb 13 12:40:28 2001
From: b012414 at mail.dvz.fh-koeln.de (Stephan Effelsberg)
Date: Tue, 13 Feb 2001 12:40:28 +0100
Subject: How can I use "tkFileDialog.askopenfilename()"
References: <96ap4f$5qp$1@b5nntp2.channeli.net>
Message-ID: <3A891D2B.D1E3717B@mail.dvz.fh-koeln.de>

iact21 schrieb:

> >> import tkSimpleDialg, Tkinter

It seems that you want tkFileDialog, not tkSimpleDialog.

> >> root = Tk()
> >> fn = tkFileDialog,askopenfilename()

Also take care of typos, dot instead of comma.

> error...
> Traceback (innermost last):
>     File "<stdin>", line 1, in ?
> NameError: tkFileDialog

See your import line ...

> Please..help..me...!!!

import tkFileDialog
# no own root window needed
fn = tkFileDialog.askopenfilename()




From Michael.Husmann at teleatlas.com  Tue Feb 13 10:17:20 2001
From: Michael.Husmann at teleatlas.com (Michael Husmann)
Date: Tue, 13 Feb 2001 10:17:20 +0100
Subject: How to pickle a C struct?
Message-ID: <3A88FBA0.939243C5@teleatlas.com>

I have SWIG(ed) a C library with lots of structs and wanted
to use the shelve module to store some data on harddisk.
Those data contains a simple C struct.
I have read that I have to add the two functions
__getstate__() and __setstate__() in addmethods for storing
and retrieving. 

Is there someone who has done so to give me an example.

Michael


From djc at object-craft.com.au  Wed Feb 14 02:05:28 2001
From: djc at object-craft.com.au (Dave Cole)
Date: 14 Feb 2001 12:05:28 +1100
Subject: How to pickle a C struct?
References: <3A88FBA0.939243C5@teleatlas.com>
Message-ID: <m3n1bqnkfb.fsf@vole.object-craft.com.au>

>>>>> "Michael" == Michael Husmann <Michael.Husmann at teleatlas.com> writes:

Michael> I have SWIG(ed) a C library with lots of structs and wanted
Michael> to use the shelve module to store some data on harddisk.
Michael> Those data contains a simple C struct.  I have read that I
Michael> have to add the two functions __getstate__() and
Michael> __setstate__() in addmethods for storing and retrieving.

Michael> Is there someone who has done so to give me an example.

I thought you had to do that too.  You need to use the copy_reg
module.  Here is some code from my Sybase module which allows you to
pickle the numeric type I implemented.

There is probably more code here than you need, but I would prefer to
err on the side of caution.

- Dave

- - from numeric.c - - - - - - - - - - - - - - - - - - - - - - - - - - - -

/* Implement the Sybase.numeric() method
 */
PyObject *NumericType_new(PyObject *module, PyObject *args)
{
    NumericObj *self;
    int precision, scale;
    PyObject *obj;

    self = PyObject_NEW(NumericObj, &NumericType);
    if (self == NULL)
        return NULL;

    precision = -1;
    scale = -1;
    if (PyArg_ParseTuple(args, "O|ii", &obj, &precision, &scale)) {
        NumericObj *num = NULL;

        if (PyInt_Check(obj))
            num = Numeric_FromInt(obj, precision, scale);
        else if (PyLong_Check(obj))
            num = Numeric_FromLong(obj, precision, scale);
        else if (PyFloat_Check(obj))
            num = Numeric_FromFloat(obj, precision, scale);
        else if (PyString_Check(obj))
            num = Numeric_FromString(obj, precision, scale);
        else if (Numeric_Check(obj))
            num = Numeric_FromNumeric(obj, precision, scale);
        else
            raise_exception_string(DataError, "could not convert to Numeric");
        if (num)
            return (PyObject*)num;
    }

    Py_DECREF(self);
    return NULL;
}

/* Used in unpickler
 */
static PyObject *numeric_constructor = NULL;

/* Register the numeric type with the copy_reg module.  This allows
 * Python to (un)pickle numeric objects.  The equivalent Python code
 * is this:
 *
 * def pickle_numeric(n):
 *     return numeric, (str(n), n.precision, n.scale)
 * 
 * copy_reg.pickle(type(numeric(1)), pickle_numeric, numeric)
 */
/* Numeric pickling function
 */
PyObject *pickle_numeric(PyObject *module, PyObject *args)
{
    NumericObj *obj = NULL;
    PyObject *values = NULL,
        *tuple = NULL;
    char text[NUMERIC_LEN];

    if (!PyArg_ParseTuple(args, "O!", &NumericType, &obj))
        goto error;
    numeric_as_string(obj, text);
    if ((values = Py_BuildValue("(sii)", text,
                                obj->num.precision, obj->num.scale)) == NULL)
        goto error;
    tuple = Py_BuildValue("(OO)", numeric_constructor, values);

error:
    Py_XDECREF(values);
    return tuple;
}

/* Register Numeric type pickler
 */
void copy_reg_numeric(PyObject *dict)
{
    PyObject *module = NULL,
        *pickle_func = NULL,
        *pickler = NULL,
        *obj = NULL;

    module = PyImport_ImportModule("copy_reg");
    if (module == NULL)
        goto error;
    if ((pickle_func = PyObject_GetAttrString(module, "pickle")) == NULL)
        goto error;
    if ((numeric_constructor = PyDict_GetItemString(dict, "numeric")) == NULL)
        goto error;
    if ((pickler = PyDict_GetItemString(dict, "pickle_numeric")) == NULL)
        goto error;
    Py_XINCREF(numeric_constructor);
    obj = PyObject_CallFunction(pickle_func, "OOO",
                                &NumericType, pickler, numeric_constructor);

error:
    Py_XDECREF(obj);
    Py_XDECREF(pickler);
    Py_XDECREF(pickle_func);
    Py_XDECREF(module);
}

- - from Sybase.c - - - - - - - - - - - - - - - - - - - - - - - - - - - -

static struct PyMethodDef Sybase_methods[] = {
    { "connect", (PyCFunction)ConnectType_new, METH_VARARGS | METH_KEYWORDS },
    { "numeric", (PyCFunction)NumericType_new, METH_VARARGS },
    { "pickle_numeric", (PyCFunction)pickle_numeric, METH_VARARGS },
    { NULL, NULL }
};

void initSybase(void)
{
    PyObject *module;
    PyObject *dict;

    module = Py_InitModule(module_name, Sybase_methods);
    dict = PyModule_GetDict(module);
        :
        :
    copy_reg_numeric(dict);
}


-- 
http://www.object-craft.com.au


From niessink at serc.nl  Tue Feb 13 11:17:54 2001
From: niessink at serc.nl (Frank Niessink)
Date: 13 Feb 2001 10:17:54 GMT
Subject: How to import from filename with dots in it?
Message-ID: <96b1ki$j26$1@newshost.accu.uu.nl>

Hi all,

I want to import from a file called 'foo-0.1.py'. Is this at all possible?
Maybe using __import__?

Thanks, Frank

-- 
The Hitch Hiker's Guide to the Galaxy has this to say on the subject of 
flying. There is an art, it says, or rather a knack to flying. The knack lies
in learning how to throw yourself at the ground and miss. Pick a nice day, it
suggests, and try it.
	-- Douglas Adams, 'Life, the Universe, and Everything'


From niessink at serc.nl  Tue Feb 13 12:30:07 2001
From: niessink at serc.nl (Frank Niessink)
Date: 13 Feb 2001 11:30:07 GMT
Subject: How to import from filename with dots in it?
References: <96b1ki$j26$1@newshost.accu.uu.nl>
Message-ID: <96b5rv$his$1@newshost.accu.uu.nl>

Frank Niessink <niessink at serc.nl> wrote:
> 
> I want to import from a file called 'foo-0.1.py'. Is this at all possible?
> Maybe using __import__?

After some experimenting I found a solution:

>>> import imp
>>> location = imp.find_module('foo-0.1')
>>> foo_0_1 = imp.load_module('dummy', *location)

The first argument to load_module does not seem to matter?

Cheers, Frank
-- 
"An SEP," he said, "is something that we can't see, or don't see, or our brain
doesn't let us see, because we think that it's somebody else's problem. That's
what SEP means. Somebody Else's Problem. The brain just edits it out, it's
like a blind spot."
	-- Douglas Adams, 'Life, the Universe, and Everything'


From jfontain at winealley.com  Tue Feb 13 12:51:50 2001
From: jfontain at winealley.com (Jean-Luc Fontaine)
Date: Tue, 13 Feb 2001 12:51:50 +0100
Subject: how to get the output of embedded python?
Message-ID: <96b6o6$a6e$1@s1.read.news.oleane.net>

I am trying to write a tclpython extension so that modules for my system 
monitoring application (moodss at http://jfontain.free.fr/moodss11.htm) can 
also be written in python, on top of tcl and perl.

It has been quite simple to write the tclperl extension, since perl allows 
retrieving the result in string form of any script, using the internal eval 
facilities.

With python, according to my research on the subject (which may be wrong 
since I am new in python), there does not seem to be any simple way to 
gather the output of a script passed for example to PyRun_SimpleString. I 
also looked at the huge piece of code that handles the interactive part of 
python so that typing 1+1 at the prompt prints 2.

Did I miss something in the C API and is there an implementation of say:

char *PyEval(char *script)

somewhere?

Additionally, I would also need the above functionality on a per 
interpreter basis (I need to be able to use several interpreters 
concurrently).

My apologies if the questions above seem trivial.

-- 
Jean-Luc Fontaine mailto:jfontain at winealley.com http://www.winealley.com



From db3l at fitlinxx.com  Tue Feb 13 19:28:33 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 13 Feb 2001 13:28:33 -0500
Subject: how to get the output of embedded python?
References: <96b6o6$a6e$1@s1.read.news.oleane.net>
Message-ID: <un1bqcu9a.fsf@ctwd0143.fitlinxx.com>

Jean-Luc Fontaine <jfontain at winealley.com> writes:

> With python, according to my research on the subject (which may be wrong 
> since I am new in python), there does not seem to be any simple way to 
> gather the output of a script passed for example to PyRun_SimpleString. I 
> also looked at the huge piece of code that handles the interactive part of 
> python so that typing 1+1 at the prompt prints 2.

You may need to be a little more clear on what you mean by "output" -
is it anything the script generates, or are you looking for
information that is otherwise sent to sys.stdout?  The former may be
impossible, given that scripts may write files or do various
operations.  The latter isn't too bad, since you can just redirect
sys.stdout before executing the code, unless you need a highly secure
or guaranteed environment.

I took a fairly low tech approach in one of my embedded applications.
First, I just implemented some functions that expected to be called
with any data being generated by the script.  In my case the
application was a simulator and script output was just placed into a
main scrolling log window (highlighted if it was to stderr).  Then,
after instantiating an interpreter (and you could do this for multiple
interpreters although you'd need to separate their output in your
embedded processing function), I first forced the execution of an
initialization script that rebound sys.stdout/stderr to my code, e.g.:

    import sys

    class StdoutCatcher:
	def write(self, stuff):
	    ecna.stdout(stuff)

    class StderrCatcher:
	def write(self, stuff):
	    ecna.stderr(stuff)

    sys.stdout = StdoutCatcher()
    sys.stderr = StderrCatcher()

(in the above, the "ecna." functions are those already registered by
the embedding application)

And after that, any scripts that I executed within that same
interpreter would send output normally destined for sys.stdout/stderr
to my internal function which could do whatever it wanted with it.  In
your case, you could just accumulate it until some other part of your
application requested it.

Note that this doesn't preclude the script being run from trying to
reset sys.stdout, and it does involve an extra "init" script (hidden
though it was from the main script).  You could be a bit more
self-contained by actually implementing the catcher classes directly
in your embedding application, but for simply embedding typical
scripts that you want to trap the output of, this should work well.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From jfontain at winealley.com  Wed Feb 14 11:06:53 2001
From: jfontain at winealley.com (Jean-Luc Fontaine)
Date: Wed, 14 Feb 2001 11:06:53 +0100
Subject: how to get the output of embedded python?
References: <96b6o6$a6e$1@s1.read.news.oleane.net> <un1bqcu9a.fsf@ctwd0143.fitlinxx.com>
Message-ID: <96dkvb$eb8$1@s1.read.news.oleane.net>

David Bolen wrote:

> Jean-Luc Fontaine <jfontain at winealley.com> writes:
> 
> > With python, according to my research on the subject (which may be wrong
> > since I am new in python), there does not seem to be any simple way to
> > gather the output of a script passed for example to PyRun_SimpleString.
> > I also looked at the huge piece of code that handles the interactive
> > part of python so that typing 1+1 at the prompt prints 2.
> 
> You may need to be a little more clear on what you mean by "output" -
> is it anything the script generates,

Yes. That is whatever output you would see when in interactive mode, if 
that makes any sense...

> or are you looking for
> information that is otherwise sent to sys.stdout?  The former may be
> impossible, given that scripts may write files or do various
> operations.

That does not seem to matter in other scripting languages, such as Tcl or 
Perl: calling an internal eval C function with a script as argument returns 
the result of the script (not what comes out on stdout or stderr), which 
may be empty.

For example, if I define a function foo that returns a string, invoking 
eval("foo()") in C would return that string. Is not that the behavior of 
python in interactive mode, for example?

-- 
Jean-Luc Fontaine mailto:jfontain at winealley.com http://www.winealley.com



From db3l at fitlinxx.com  Thu Feb 15 02:16:04 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 14 Feb 2001 20:16:04 -0500
Subject: how to get the output of embedded python?
References: <96b6o6$a6e$1@s1.read.news.oleane.net> <un1bqcu9a.fsf@ctwd0143.fitlinxx.com> <96dkvb$eb8$1@s1.read.news.oleane.net>
Message-ID: <uzofod9uz.fsf@ctwd0143.fitlinxx.com>

Jean-Luc Fontaine <jfontain at winealley.com> writes:

> Yes. That is whatever output you would see when in interactive mode, if 
> that makes any sense...

Oh, I think I understand better.  You don't actually want to trap any
output generated during the execution of the code - you just want to
deal with the result of the expression you may be evaluating.

> That does not seem to matter in other scripting languages, such as Tcl or 
> Perl: calling an internal eval C function with a script as argument returns 
> the result of the script (not what comes out on stdout or stderr), which 
> may be empty.
> 
> For example, if I define a function foo that returns a string, invoking 
> eval("foo()") in C would return that string. Is not that the behavior of 
> python in interactive mode, for example?

On an expression by expression basis, yes, each expression has a
result object (but not statements such as "print" for example).

You should be able to handle the same thing from your controlling code
by dipping a little lower than the "Simple" high level functions.  By
using either PyRun_File or PyRun_String you can evaluate Python code,
and the result of the function is the resulting Python object (a
PyObject *).  Use PyEval_GetGlobals and PyEval_GetLocals for the
dictionaries to supply to the execution.

Once you have the resulting object pointer you can do anything you
want with it (using any of the Python functions available for
extending/embedding applications), including converting it into a
string representation if you like.  Note that the result does not
itself have to be a string object, since a Python expression can
result in any Python object.  Check out the abstract and concrete
objects layer in the C/API reference or the Python source for all
sorts of functions.

But if you know you want a string representation, you should be able
to use something like PyObject_Str to return a new Python object with
a string representation of the result object.  Then, if you just want
a char*, use PyString_AsString.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From jfontain at winealley.com  Thu Feb 15 14:56:45 2001
From: jfontain at winealley.com (Jean-Luc Fontaine)
Date: Thu, 15 Feb 2001 14:56:45 +0100
Subject: how to get the output of embedded python?
References: <96b6o6$a6e$1@s1.read.news.oleane.net> <un1bqcu9a.fsf@ctwd0143.fitlinxx.com> <96dkvb$eb8$1@s1.read.news.oleane.net> <uzofod9uz.fsf@ctwd0143.fitlinxx.com>
Message-ID: <96gmq7$p9r$1@s1.read.news.oleane.net>

David Bolen wrote:

> Jean-Luc Fontaine <jfontain at winealley.com> writes:
> 
> > Yes. That is whatever output you would see when in interactive mode, if
> > that makes any sense...
> 
> Oh, I think I understand better.  You don't actually want to trap any
> output generated during the execution of the code - you just want to
> deal with the result of the expression you may be evaluating.
> 
> > That does not seem to matter in other scripting languages, such as Tcl
> > or Perl: calling an internal eval C function with a script as argument
> > returns the result of the script (not what comes out on stdout or
> > stderr), which may be empty.
> > 
> > For example, if I define a function foo that returns a string, invoking
> > eval("foo()") in C would return that string. Is not that the behavior of
> > python in interactive mode, for example?
> 
> On an expression by expression basis, yes, each expression has a
> result object (but not statements such as "print" for example).
> 
> You should be able to handle the same thing from your controlling code
> by dipping a little lower than the "Simple" high level functions.  By
> using either PyRun_File or PyRun_String you can evaluate Python code,
> and the result of the function is the resulting Python object (a
> PyObject *).  Use PyEval_GetGlobals and PyEval_GetLocals for the
> dictionaries to supply to the execution.
> 
> Once you have the resulting object pointer you can do anything you
> want with it (using any of the Python functions available for
> extending/embedding applications), including converting it into a
> string representation if you like.  Note that the result does not
> itself have to be a string object, since a Python expression can
> result in any Python object.  Check out the abstract and concrete
> objects layer in the C/API reference or the Python source for all
> sorts of functions.
> 
> But if you know you want a string representation, you should be able
> to use something like PyObject_Str to return a new Python object with
> a string representation of the result object.  Then, if you just want
> a char*, use PyString_AsString.
> 
> --

Thank you so very much for all this information: that is what I needed to 
get started.

-- 
Jean-Luc Fontaine mailto:jfontain at winealley.com http://www.winealley.com



From jh at web.de  Thu Feb 15 23:00:29 2001
From: jh at web.de (Jürgen Hermann)
Date: Thu, 15 Feb 2001 23:00:29 +0100
Subject: how to get the output of embedded python?
References: <96b6o6$a6e$1@s1.read.news.oleane.net>
Message-ID: <96hjhn$hni$05$1@news.t-online.com>

"Jean-Luc Fontaine" <jfontain at winealley.com> schrieb im Newsbeitrag
news:96b6o6$a6e$1 at s1.read.news.oleane.net...
> Did I miss something in the C API and is there an implementation of say:
>
> char *PyEval(char *script)
>
> somewhere?

BTW, the easiest way to handle it is to create a cStringIO object,
PySys_SetItem() that as your stdout and then start your script. After it did
run, you can getvalue() the output.

This is much easier than writing your own file-like C extension.




From boswell at caltech.edu  Tue Feb 13 13:24:30 2001
From: boswell at caltech.edu (Dustin Boswell)
Date: Tue, 13 Feb 2001 04:24:30 -0800
Subject: https support in Python?
Message-ID: <3A89277E.68C09AB1@caltech.edu>

I noticed that although python 1.6 and up are claimed to support code
like:
urllib.urlopen("https://secure-server.com")
the Modules/Setup file (in the source distribution for Python) has the
SSL
part commented out since it "breaks" the Python build.  It sure does.
;-)

I've tried downloading and installing various openssl, ssleay, and other
packages
but it always complains one way or another during build.  And it seems
that
there are a lot of hacks, but no official Python releases.
My question is, what is the "standard" way (what packages, etc...) to
get my
python interpreter built so I can do urlopen("https://")?

Any links to good packages (third party ok) would be great help.

P.S.
I'm running Redhat 6.2

Thanks in advance.



From olafb+usenet at pvv.org  Tue Feb 13 14:11:18 2001
From: olafb+usenet at pvv.org (Olaf Trygve Berglihn)
Date: 13 Feb 2001 14:11:18 +0100
Subject: https support in Python?
References: <3A89277E.68C09AB1@caltech.edu>
Message-ID: <dn1ae7qhgnd.fsf@lizard.pvv.ntnu.no>

Dustin Boswell <boswell at caltech.edu> writes:

> I noticed that although python 1.6 and up are claimed to support code
> like:
> urllib.urlopen("https://secure-server.com")
> the Modules/Setup file (in the source distribution for Python) has the
> SSL
> part commented out since it "breaks" the Python build.  It sure does.
> ;-)
> 
> I've tried downloading and installing various openssl, ssleay, and other
> packages
> but it always complains one way or another during build.  And it seems
> that
> there are a lot of hacks, but no official Python releases.
> My question is, what is the "standard" way (what packages, etc...) to
> get my
> python interpreter built so I can do urlopen("https://")?
> 
> Any links to good packages (third party ok) would be great help.

You can try my (unofficial) build of python 2.0 found at:
<URI: http://www.pvv.org/~olafb/software/python/ >

Try rebuilding the srpm on your redhat. The binary rpms were build on
a Trustix Secure Linux v1.2 and link against openssl-0.9.6. I guess
rh-6.2 uses openssl-0.9.5a. It should work with the older openssl.
The Tkinter module is left out in this build.

Olaf
-- 
Olaf Trygve Berglihn <olafb+usenet at pvv.org>


From gustafl at algonet.se  Tue Feb 13 13:35:31 2001
From: gustafl at algonet.se (Gustaf Liljegren)
Date: Tue, 13 Feb 2001 13:35:31 +0100
Subject: Delete a certain line from a file
Message-ID: <96b9n0$amd$1@cubacola.tninet.se>

Excuse another newbie question. My script is reading a textfile with a list
of words (separated with newlines), and now I need something to delete a
certain record when I encounter it in a loop. Something like:

for line in f.readlines():
  if line == "somethingbad":
    # delete the whole line with newline
    break

I have successfully done this by adding each line to a list and then used
list.remove("somethingbad"), but there's another way without having to put
the whole file in a list, isn't it?

Regards,

Gustaf Liljegren




From stephen_purcell at yahoo.com  Tue Feb 13 15:39:06 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 13 Feb 2001 15:39:06 +0100
Subject: Delete a certain line from a file
In-Reply-To: <96b9n0$amd$1@cubacola.tninet.se>; from gustafl@algonet.se on Tue, Feb 13, 2001 at 01:35:31PM +0100
References: <96b9n0$amd$1@cubacola.tninet.se>
Message-ID: <20010213153906.A759@freedom.puma-ag.com>

Gustaf Liljegren wrote:
> Excuse another newbie question. My script is reading a textfile with a list
> of words (separated with newlines), and now I need something to delete a
> certain record when I encounter it in a loop. Something like:
> 
> for line in f.readlines():
>   if line == "somethingbad":
>     # delete the whole line with newline
>     break

Note that 'readlines' always reads the whole file into memory anyway...

> 
> I have successfully done this by adding each line to a list and then used
> list.remove("somethingbad"), but there's another way without having to put
> the whole file in a list, isn't it?

Yes, filter the file into another file then rename it:

  infile = open(inputfilename,'r')
  outputfile = open(outputfilename, 'w')

  while 1:
    line = infile.readline()
    if not line: break
    if line != "somethingbad":
      outfile.write(line)

  infile.close()
  outfile.close()
  os.rename(outputfilename, inputfilename)

-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From cfelling at iae.nl  Tue Feb 13 15:41:35 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 13 Feb 2001 15:41:35 +0100
Subject: Delete a certain line from a file
References: <96b9n0$amd$1@cubacola.tninet.se>
Message-ID: <96bh2v$26t$1@animus.fel.iae.nl>

Gustaf Liljegren <gustafl at algonet.se> wrote:
> Excuse another newbie question. My script is reading a textfile with a list
> of words (separated with newlines), and now I need something to delete a
> certain record when I encounter it in a loop. Something like:

> for line in f.readlines():
>   if line == "somethingbad":
>     # delete the whole line with newline
>     break

use the marvelous fileinput module like in:

    import fileinput

    for line in fileinput.input('yourfile', inplace=1):
        if line != "somethingbad":
            print line,

The keyword argument "inplace=1", tells fileinput.input to handle this
file special: make a copy, read from that copy and redirect stdout to
a new version of that file, effectively allowing you to change that
file `in place'.  Notice the trailing "," in the print line, this tells
print not to add (yet an other) end-of-line character.


> I have successfully done this by adding each line to a list and then used
> list.remove("somethingbad"), but there's another way without having to put
> the whole file in a list, isn't it?

see above:),
your approach seems correct though, though could be spedup a little:

    data = filter(lambda x: x != "somethingbad", f.readlines())

or:

    data = []
    for line in f.readlines()
        if line != "somethingbad"
            data.append(line)
-- 
groetjes, carel


From csrabak at uol.com.br  Tue Feb 13 15:56:45 2001
From: csrabak at uol.com.br (Cesar Rabak)
Date: Tue, 13 Feb 2001 11:56:45 -0300
Subject: Delete a certain line from a file
References: <96b9n0$amd$1@cubacola.tninet.se>
Message-ID: <3A894B2D.8D48EAFF@uol.com.br>

Gustaf Liljegren wrote:
> 
> Excuse another newbie question. My script is reading a textfile with a list
> of words (separated with newlines), and now I need something to delete a
> certain record when I encounter it in a loop. Something like:
> 
> for line in f.readlines():
>   if line == "somethingbad":
>     # delete the whole line with newline
>     break
> 
> I have successfully done this by adding each line to a list and then used
> list.remove("somethingbad"), but there's another way without having to put
> the whole file in a list, isn't it?

Unfortunately no ;-)

If you feel you'll need this kind of functionality more often than not,
you should consider changing from a textfile to a database.

HTH

Cesar


From mike.mellor at tbe.com  Tue Feb 13 13:49:32 2001
From: mike.mellor at tbe.com (mike.mellor at tbe.com)
Date: Tue, 13 Feb 2001 12:49:32 -0000
Subject: Passing Values
Message-ID: <96bags+tjse@eGroups.com>

I am working on an RPN calculator program, using a four level stack.  
I initialize it as:

stack = [0.0,0.0,0.0,0.0]

I am noticing problems with the following two functions (xStack and 
yStack are Entry widgets).  clearer is supposed to clear the values 
of the stack, and RollDown is supposed to "roll the stack" down.

def clearer(event):
    yStack.delete(0,END)
    xStack.delete(0,END)
    stack = [0.0, 0.0, 0.0, 0.0]
    print 'Clear ' + str(stack)
    for i in range(4):
       print 'Stack ' + `i` + ' ' + str(stack[i])

def RollDown(event):
    stack[0] = stack[1]
    stack[1] = stack[2]
    stack[2] = stack[3]
    try:
        stack[3] = float(xStack.get())
    except ValueError:
        stack[3] = 0.0
    xStack.delete(0,END)
    xStack.insert(0,stack[0])
    yStack.delete(0,END)
    yStack.insert(0,str(stack[1]))
    print 'Roll down ' + str(stack)
    for i in range(4):
       print 'Stack ' + `i` + ' ' + str(stack[i])

The console reports that the stack is cleared, but when I roll the 
stack down, old values pop in.  Do I need to use a deepcopy to pass 
the values from one stack location to the next (stack[0] = deepcopy
(stack[1])) or is there a better way to do this?

Thanks.

Mike





From scarblac at pino.selwerd.nl  Wed Feb 14 00:06:50 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 13 Feb 2001 23:06:50 GMT
Subject: Passing Values
References: <mailman.982068614.19128.python-list@python.org>
Message-ID: <slrn98jfcs.435.scarblac@pino.selwerd.nl>

mike.mellor at tbe.com <mike.mellor at tbe.com> wrote in comp.lang.python:
> I am working on an RPN calculator program, using a four level stack.  
> I initialize it as:
> 
> stack = [0.0,0.0,0.0,0.0]
> 
> I am noticing problems with the following two functions (xStack and 
> yStack are Entry widgets).  clearer is supposed to clear the values 
> of the stack, and RollDown is supposed to "roll the stack" down.
> 
> def clearer(event):
>     yStack.delete(0,END)
>     xStack.delete(0,END)
>     stack = [0.0, 0.0, 0.0, 0.0]
>     print 'Clear ' + str(stack)
>     for i in range(4):
>        print 'Stack ' + `i` + ' ' + str(stack[i])

The stack you use here is a local variable. It has no effect on the global
variable also named stack.

Solution: put 'global stack' as the first line of the function.

> def RollDown(event):
>     stack[0] = stack[1]
>     stack[1] = stack[2]
>     stack[2] = stack[3]
>     try:
>         stack[3] = float(xStack.get())
>     except ValueError:
>         stack[3] = 0.0
>     xStack.delete(0,END)
>     xStack.insert(0,stack[0])
>     yStack.delete(0,END)
>     yStack.insert(0,str(stack[1]))
>     print 'Roll down ' + str(stack)
>     for i in range(4):
>        print 'Stack ' + `i` + ' ' + str(stack[i])

Here, you never directly assign to stack, so Python assumes you meant the
global. This function uses the global stack.

> The console reports that the stack is cleared, but when I roll the 
> stack down, old values pop in.  Do I need to use a deepcopy to pass 
> the values from one stack location to the next (stack[0] = deepcopy
> (stack[1])) or is there a better way to do this?

Instead of stack[0]=stack[1]; stack[1]=stack[2]; stack[2]=etc, you
simply want to do 'del stack[0]'. Then add the new element with
stack.append(0.0).

-- 
Remco Gerlich


From danielk at aracnet.com  Tue Feb 13 16:32:23 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Tue, 13 Feb 2001 07:32:23 -0800
Subject: spawnv question
Message-ID: <bcki8tc9bmoophugldu9k4mfju3emm4jc2@4ax.com>

Here is a snippet of code I'm using with the 'spawnv' command:

import os
args = ('http://www.python.org', 'http://www.python.org') # This works
os.spawnv(os.P_NOWAIT, 'c:\progra~1\intern~1\iexplore', args)
print "done"

The Python Reference for the 'spawnv' command says:

 "Execute the program path in a new process, passing the arguments specified in
args as command-line parameters. args may be a list or a tuple."

My question is, why do I have to specify the url argument twice in the 'args'
variable. It doesn't seem to work properly if I specify it as a 1-element
tuple, ie...

args = ('http://www.python.org',) # This doesn't work

Thanks,
Daniel Klein



From fredrik at pythonware.com  Tue Feb 13 16:48:32 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 13 Feb 2001 15:48:32 GMT
Subject: spawnv question
References: <bcki8tc9bmoophugldu9k4mfju3emm4jc2@4ax.com>
Message-ID: <kHci6.12999$Qb7.2088543@newsb.telia.net>

Daniel Klein wrote:
> Here is a snippet of code I'm using with the 'spawnv' command:
>
> import os
> args = ('http://www.python.org', 'http://www.python.org') # This works
> os.spawnv(os.P_NOWAIT, 'c:\progra~1\intern~1\iexplore', args)
> print "done"
>
> The Python Reference for the 'spawnv' command says:
>
>  "Execute the program path in a new process, passing the arguments specified in
> args as command-line parameters. args may be a list or a tuple."
>
> My question is, why do I have to specify the url argument twice in the 'args'
> variable.

because internet explorer, like any other program, expects to
find its first argument in argv[1], not argv[0].

by convention (and sometimes necessity), the first argument
should be the name of the program's executable.  try this:

    os.spawnv(mode, program, (program,) + tuple(args))

under unix, (program,) should be (os.path.basename(program),)
instead.

Cheers /F




From dbrueck at edgix.com  Tue Feb 13 16:50:12 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 13 Feb 2001 08:50:12 -0700
Subject: spawnv question
In-Reply-To: <bcki8tc9bmoophugldu9k4mfju3emm4jc2@4ax.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHCEDJPJAA.dbrueck@edgix.com>

argv[0] is reserved for the program name.

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Daniel Klein
> Sent: Tuesday, February 13, 2001 8:32 AM
> To: python-list at python.org
> Subject: spawnv question
> 
> 
> Here is a snippet of code I'm using with the 'spawnv' command:
> 
> import os
> args = ('http://www.python.org', 'http://www.python.org') # This works
> os.spawnv(os.P_NOWAIT, 'c:\progra~1\intern~1\iexplore', args)
> print "done"
> 
> The Python Reference for the 'spawnv' command says:
> 
>  "Execute the program path in a new process, passing the 
> arguments specified in
> args as command-line parameters. args may be a list or a tuple."
> 
> My question is, why do I have to specify the url argument twice 
> in the 'args'
> variable. It doesn't seem to work properly if I specify it as a 1-element
> tuple, ie...
> 
> args = ('http://www.python.org',) # This doesn't work
> 
> Thanks,
> Daniel Klein
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list



From cerutti at together.net  Tue Feb 13 17:24:44 2001
From: cerutti at together.net (Neil Cerutti)
Date: 13 Feb 2001 16:24:44 GMT
Subject: spawnv question
References: <mailman.982079535.31685.python-list@python.org>
Message-ID: <slrn98iocg.16.cerutti@fiad06.norwich.edu>

Dave Brueck posted:
>argv[0] is reserved for the program name.

Python gives you the ability to call programs while supplying an
erroneous program names to them. This is very useful, and it
makes perfect sense if you know various C libraries. After all,
spawning external programs should only be attempted by
experienced programmers.

-- 
Neil Cerutti <cerutti at together.net>


From erno-news at erno.iki.fi  Wed Feb 14 09:16:12 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 14 Feb 2001 10:16:12 +0200
Subject: spawnv question
References: <mailman.982079535.31685.python-list@python.org> <slrn98iocg.16.cerutti@fiad06.norwich.edu>
Message-ID: <kun1bpwugj.fsf@lasipalatsi.fi>

In article <slrn98iocg.16.cerutti at fiad06.norwich.edu>,
cerutti at together.net (Neil Cerutti) writes:

| Dave Brueck posted:
|| argv[0] is reserved for the program name.

| Python gives you the ability to call programs while supplying an
| erroneous program names to them. This is very useful, and it
| makes perfect sense if you know various C libraries. After all,
| spawning external programs should only be attempted by
| experienced programmers.

there is no way to tell what is an erroneous program name
to you (short of popping up a dialog box and asking) :)

  -- erno


From jurgen.defurne at philips.com  Wed Feb 14 08:26:52 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Wed, 14 Feb 2001 08:26:52 +0100
Subject: spawnv question
Message-ID: <0056900015894399000002L092*@MHS>

To know why, you should consult the MS documentation about the spawn calls.
Their standard is to use as first argument in the argumentlist the name of the program
that you called.

The Python documentation is very unclear about the spawn calls. What is not
mentioned also is that both spawnv(e) and spawnl(e) are available in Python 2.0.

Jurgen




danielk at aracnet.com@SMTP at python.org on 13/02/2001 16:53:38
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	spawnv question
Classification:	

Here is a snippet of code I'm using with the 'spawnv' command:

import os
args = ('http://www.python.org', 'http://www.python.org') # This works
os.spawnv(os.P_NOWAIT, 'c:\progra~1\intern~1\iexplore', args)
print "done"

The Python Reference for the 'spawnv' command says:

 "Execute the program path in a new process, passing the arguments specified in
args as command-line parameters. args may be a list or a tuple."

My question is, why do I have to specify the url argument twice in the 'args'
variable. It doesn't seem to work properly if I specify it as a 1-element
tuple, ie...

args = ('http://www.python.org',) # This doesn't work

Thanks,
Daniel Klein

--
http://mail.python.org/mailman/listinfo/python-list





From hamish_lawson at yahoo.co.uk  Tue Feb 13 16:44:08 2001
From: hamish_lawson at yahoo.co.uk (=?iso-8859-1?q?Hamish=20Lawson?=)
Date: Tue, 13 Feb 2001 15:44:08 +0000 (GMT)
Subject: MySQL + SQL Statements + Quote escaping
Message-ID: <20010213154408.14748.qmail@web214.mail.yahoo.com>

Christian Theune wrote:

    i just tuned into python and it's really cool. I like it.
    Before I came to python i used to write scripts in php.
    There was a function called "addslashes" it escaped quotes and
    slashes to make strings sql-safe.

Database modules that conform to the DB-API specfication, such as
MySQLdb, provide a placeholder mechanism that will take care of the
quoting automatically. For example:

    cursor.execute(
        "select * from customers where surname = %s and age < %s", 
        ("O'Hara", 40)
    )

The %s placeholders are replaced by correctly quoted values.

There are two advantages to this approach.

Firstly, database engines differ in how they escape quotes. Many, such
as MySQLdb, prefix them with a backslash, but other DB engines escape
quotes by doubling them (thus 'O''Reilly'). Using the placeholder
mechanism lets the particular database module deal with that for you.

The second advantage is that the execute() method can optimise its
performance when the same operation is repeatedly executed but with
different bound values for the placeholders. But if you build the query
string yourself on each iteration, then the execute() method will have
to parse it each time for correctness, thus losing the opportunity for
optimisation.

Hamish Lawson


____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From ct at gocept.com  Tue Feb 13 23:01:17 2001
From: ct at gocept.com (Christian Theune)
Date: Tue, 13 Feb 2001 23:01:17 +0100
Subject: MySQL + SQL Statements + Quote escaping
References: <mailman.982079115.29877.python-list@python.org>
Message-ID: <3A89AEAD.70003@gocept.com>

thanks to all of you, who answered.

I think the most reliable and effictive solution was brought up from 
Hamish Lawson.


have a nice day


Christian Theune



From ben.held at staarinc.com  Tue Feb 13 16:45:51 2001
From: ben.held at staarinc.com (Ben Held)
Date: Tue, 13 Feb 2001 09:45:51 -0600
Subject: returning local character string from python extension dll
Message-ID: <3a8956ec$0$46002$272ea4a1@news.execpc.com>

I am building a python extension dll using C++ and SWIG.  I need to return a
character string from a class member function.  The problem is the that the
string is not a member of the class, but it is a local string:

char* CFoo::GetName() {
    char pc[] = "blah";
.........
    return pc;
}

I want to be able to call "print pFoo.GetName()" in python and have "blah"
printed.  If I define my function just as above, I get garbage printed since
the data is local and destroyed when the function exits.  I know I can
return a PyObject* using PyString_FromString, but then print on that object
doesn't show the string like I need.

Ideas?

Ben
ben.held at staarinc.com




From vonehr at ira.uka.de  Tue Feb 13 17:55:56 2001
From: vonehr at ira.uka.de (Markus von Ehr)
Date: Tue, 13 Feb 2001 17:55:56 +0100
Subject: tkinter:
Message-ID: <3A89671C.58AC2833@ira.uka.de>

Hi,

I have a class that creates a toplevel window and want to destroy the
toplevel
from a thread within this class.
I get an error in python.exe, does anybody know why and how I can
destroy the toplevel in this thread?

class MyWindow:
    def __init__(self, root, duration):

        self.duration = int(duration * 10 + 0.5)
        self.toplevel = Toplevel(root)
        thread.start_new_thread(self.my_thread, (10,))

    def my_thread(self, start):
        time.sleep(self.duration)
        self.toplevel.grab_release()
        self.root.focus_set()
        self.root.focus_set()
        self.toplevel.destroy()

Thanks for any answer,

Markus




From u504516601 at spawnkill.ip-mobilphone.net  Tue Feb 13 17:58:38 2001
From: u504516601 at spawnkill.ip-mobilphone.net (u504516601 at spawnkill.ip-mobilphone.net)
Date: Tue, 13 Feb 2001 16:58:38 GMT
Subject: pattern for processing HTML form with multiple rows?
Message-ID: <l.982083519.1065521240@fb0.ipathhq.com>

Architecture: Python, ASP, MsSQL

Want to have a form for inserting records into a database. 
Would like to offer a grid so that user can enter multiple 
records at one time. Assume data validation provided through 
pulldowns for key fields.

What's the "best" approach for looping over the fields 
and performing the inserts?

 



-- 
Sent by fluxent  from yahoo piece  of com
This is a spam protected message. Please answer with reference header.
Posted via http://www.usenet-replayer.com/cgi/content/new


From gregj at pobox.com  Sun Feb 18 03:15:56 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 18 Feb 2001 02:15:56 GMT
Subject: pattern for processing HTML form with multiple rows?
References: <l.982083519.1065521240@fb0.ipathhq.com>
Message-ID: <3a8f2fd9.161195146@news.potlnd1.or.home.com>

On Tue, 13 Feb 2001 16:58:38 GMT,
u504516601 at spawnkill.ip-mobilphone.net wrote:

>Architecture: Python, ASP, MsSQL
>
>Want to have a form for inserting records into a database. 
>Would like to offer a grid so that user can enter multiple 
>records at one time. Assume data validation provided through 
>pulldowns for key fields.
>
>What's the "best" approach for looping over the fields 
>and performing the inserts?

You can set up your HTML form "grid" a couple of ways. One way is to
give the fields in each grid row different names. Another way is to
reuse the form field names.

Example 1:

<form method="POST" action="cgi.py">
<input type="hidden" name="numrows" value="10">
<table>
<tr>
    <td><input type="text" name="r1f1"></td>
    <td><input type="text" name="r1f2"></td>
</tr>
<tr>
    <td><input type="text" name="r2f1"></td>
    <td><input type="text" name="r2f2"></td>
</tr>
...
<tr>
    <td><input type="text" name="r10f1"></td>
    <td><input type="text" name="r10f2"></td>
</tr>
</table>
</form>


In your Python CGI program:

import cgi
form = cgi.FieldStorage()
numrows = form["numrows"]     # passed from form
for r in range(1,numrows+1):    # for each row...
    f1 = form["r" + str(r) + "f1"].value    # get field f1
    f2 = form["r" + str(r) + "f2"].value    # get field f2
    # do something with fields f1 & f2
    # note that the fields may be empty


Example 2:

<form method="POST" action="cgi.py">
<table>
<tr>
    <td><input type="text" name="f1"></td>
    <td><input type="text" name="f2"></td>
</tr>
<tr>
    <td><input type="text" name="f1"></td>
    <td><input type="text" name="f2"></td>
</tr>
...
<tr>
    <td><input type="text" name="f1"></td>
    <td><input type="text" name="f2"></td>
</tr>
</table>
</form>

Note that the fields have the same name in each row. The Python code
is:

import cgi
form = cgi.FieldStorage()
rows = len(form["f1"])    # f1, f2 are lists, one entry per row
for i in range(rows):
    f1 = form["f1"][i].value
    f2 = form["f2"][i].value
    # do something with fields f1 & f2
    # note that the fields may be empty


I can't say which method may work best for you. You can only go so far
simulating a grid-type data entry widget in HTML. The main point to
remember is that if an HTML form has more than one field with the same
name, the values are sent to the receiving CGI program in a
comma-separated list.


Greg Jorgensen
Deschooling Society
Portland, Oregon USA


From chrisa at ASPATECH.COM.BR  Tue Feb 13 19:29:00 2001
From: chrisa at ASPATECH.COM.BR (Chris Richard Adams)
Date: Tue, 13 Feb 2001 15:29:00 -0300
Subject: while loop with f.readline()
Message-ID: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR>

I'm trying to create a loop that reads through a simple file with
strings on each line.

f=open('/tmp/xferlog', 'r')
while f.readline():
 line = f.readline()
 print line

Problem 1: Although I know there are five lines in the file - it only
prints the 3rd and 4th line.  Is my usage above correct for this - I'd
expect to see all lines.

Problem 2: How do define the range of the while loop: for example if I
place f.close() in the list of items wto go through in the loop:


f=open('/tmp/xferlog', 'r')
while f.readline():
 line = f.readline()
 print line
 f.close()

How do i code the loop to know that it should be executed after the
loop...not after each iteration???

Thanks!



From dsh8290 at rit.edu  Tue Feb 13 18:39:08 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 13 Feb 2001 12:39:08 -0500
Subject: while loop with f.readline()
In-Reply-To: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR>; from chrisa@aspatech.com.br on Tue, Feb 13, 2001 at 03:29:00PM -0300
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR>
Message-ID: <20010213123908.B11342@harmony.cs.rit.edu>

On Tue, Feb 13, 2001 at 03:29:00PM -0300, Chris Richard Adams wrote:
| I'm trying to create a loop that reads through a simple file with
| strings on each line.
| 
| f=open('/tmp/xferlog', 'r')
| while f.readline():

This reads line 1 , 3, 5

|  line = f.readline()

this reads lines 2 , 4 

|  print line

this prints the lines stored in 'line', lines 2, 4

| 
| Problem 1: Although I know there are five lines in the file - it only
| prints the 3rd and 4th line.  Is my usage above correct for this - I'd
| expect to see all lines.
| 
| Problem 2: How do define the range of the while loop: for example if I
| place f.close() in the list of items wto go through in the loop:
| 
| 
| f=open('/tmp/xferlog', 'r')
| while f.readline():
|  line = f.readline()
|  print line
|  f.close()
| 
| How do i code the loop to know that it should be executed after the
| loop...not after each iteration???
| 
| Thanks!
| 

f = open( '/tmp/xferlog' , 'r' )
for line in f.readlines() :
    print line
f.close() # not indented, this happens after the loop

If you are using Python 2.1 or greater you can use the f.xreadlines()
function which won't read the entire file at once.

Alternatively :

f = open( '/tmp/xferlog' , 'r' )
while 1 :
    line = f.readline()
    if not line : break # stop the loop if no line was read
    print line
f.close() # after the loop, not in it


or

f = open( '/tmp/xferlog' , 'r' )
line = f.readline()
while line :
    print line
    line = f.readline()
f.close()



I prefer the for loop the best, and will be great when 2.1 is released
(with the xreadlines function).

HTH,
-D




From stephen_purcell at yahoo.com  Tue Feb 13 18:39:18 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 13 Feb 2001 18:39:18 +0100
Subject: while loop with f.readline()
In-Reply-To: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR>; from chrisa@ASPATECH.COM.BR on Tue, Feb 13, 2001 at 03:29:00PM -0300
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR>
Message-ID: <20010213183918.A4918@freedom.puma-ag.com>

Chris Richard Adams wrote:
> I'm trying to create a loop that reads through a simple file with
> strings on each line.
> 
> f=open('/tmp/xferlog', 'r')
> while f.readline():
>  line = f.readline()
>  print line


In the code above, the line returned from 'f.readline()' in the 'while'
condition is simply thrown away, so only every second line is printed.

You probably mean:

  while 1:
    line = f.readline()
    if not line: break
    print line

It may look a bit strange, but it's a Python idiom and is the best way to
do this kind of input loop.

> f=open('/tmp/xferlog', 'r')
> while f.readline():
>  line = f.readline()
>  print line
>  f.close()
> 
> How do i code the loop to know that it should be executed after the
> loop...not after each iteration???

Put the 'close' call in the appropriate block -- the blocks are contiguous
areas of equal indentation. You don't want the 'close' call in the block
internal to the loop, or it will be executed each time through the loop.

You want:

  while 1:
    line = f.readline()
    if not line: break
    print line
  f.close()


-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From chris.gonnerman at usa.net  Wed Feb 14 02:06:25 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Tue, 13 Feb 2001 19:06:25 -0600
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR>
Message-ID: <00b801c09622$623886a0$a100000a@local>

----- Original Message ----- 
From: "Chris Richard Adams" <chrisa at ASPATECH.COM.BR>
Subject: while loop with f.readline()
>

>I'm trying to create a loop that reads through a simple file with
>strings on each line.
>
>f=open('/tmp/xferlog', 'r')
>while f.readline():
> line = f.readline()
> print line

I have always been a fan of the FileReader class:

class FileReader:
    def __init__(self, file):
        self.file = file
    def next(self):
        self.line = self.file.readline()
        return self.line

if __name__ == '__main__':

    f = FileReader(open("c:/autoexec.bat"))

    while f.next():
         print f.line,







From zdweeb at hotmail.com  Wed Feb 14 02:32:53 2001
From: zdweeb at hotmail.com (William Voll)
Date: Wed, 14 Feb 2001 01:32:53 GMT
Subject: while loop with f.readline()
References: <mailman.982085180.24537.python-list@python.org>
Message-ID: <B6AF4945.F6E9%zdweeb@hotmail.com>

in article mailman.982085180.24537.python-list at python.org, Chris Richard
Adams at chrisa at ASPATECH.COM.BR wrote on 2/13/01 1:29 PM:

> I'm trying to create a loop that reads through a simple file with
> strings on each line.
> 
> f=open('/tmp/xferlog', 'r')
> while f.readline():
> line = f.readline()
> print line
> 
> Problem 1: Although I know there are five lines in the file - it only
> prints the 3rd and 4th line.  Is my usage above correct for this - I'd
> expect to see all lines.
> 
> Problem 2: How do define the range of the while loop: for example if I
> place f.close() in the list of items wto go through in the loop:
> 
> 
> f=open('/tmp/xferlog', 'r')
> while f.readline():
> line = f.readline()
> print line
> f.close()
> 
> How do i code the loop to know that it should be executed after the
> loop...not after each iteration???
> 
> Thanks!
> 
This should get the job done.

f=open('/tmp/xferlogt','r').readlines()
for l in f:
    print l

--zdweeb



From phlip_cpp at my-deja.com  Wed Feb 14 06:15:14 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 14 Feb 2001 05:15:14 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org>
Message-ID: <96d492$88c@dispatch.concentric.net>

Proclaimed Chris Gonnerman from the mountaintops:

> I have always been a fan of the FileReader class:
> 
> class FileReader:
>     def __init__(self, file):
>         self.file = file
>     def next(self):
>         self.line = self.file.readline()
>         return self.line
> 
> if __name__ == '__main__':
> 
>     f = FileReader(open("c:/autoexec.bat"))
> 
>     while f.next():
>          print f.line,

Came to the thread late can't read all posts don't know if this was covered 
so sue me but...

...what's wrong with this?

        for z in f.readlines():
                print z

That's how I done it since I was a chile'. 

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  http://www.deja.com/my/pb.xp?member_name=phlip_cpp  --


From jepler at inetnebr.com  Wed Feb 14 15:17:22 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 14 Feb 2001 14:17:22 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net>
Message-ID: <slrn98l4rh.2s8.jepler@potty.housenet>

On 14 Feb 2001 05:15:14 GMT, Phlip
 <phlip_cpp at my-deja.com> wrote:
>Came to the thread late can't read all posts don't know if this was covered 
>so sue me but...
>
>...what's wrong with this?
>
>        for z in f.readlines():
>                print z
>
>That's how I done it since I was a chile'. 

If the contents of f are larger than memory, this won't work.  If the
contents of f are large compared to memory, then it won't work very well.

This is one reason that 2.1 will include "xreadlines", which will work
just like the above:
        for z in f.xreadlines():
                print z
but will never hold the entire file in memory.

Jeff


From fredrik at pythonware.com  Thu Feb 15 08:11:46 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Feb 2001 07:11:46 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net> <slrn98l4rh.2s8.jepler@potty.housenet>
Message-ID: <SiLi6.17192$AH6.2345642@newsc.telia.net>

Jeff Epler wrote:
> This is one reason that 2.1 will include "xreadlines", which will work
> just like the above:
>         for z in f.xreadlines():
>                 print z
> but will never hold the entire file in memory.

never?  not even if it's a single byte? ;-)

Cheers /F




From jepler at inetnebr.com  Thu Feb 15 14:22:39 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Thu, 15 Feb 2001 13:22:39 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net> <slrn98l4rh.2s8.jepler@potty.housenet> <SiLi6.17192$AH6.2345642@newsc.telia.net>
Message-ID: <slrn98nm0u.3v2.jepler@potty.housenet>

>Jeff Epler wrote:
>> [xreadlines] never hold the entire file in memory.
>
On Thu, 15 Feb 2001 07:11:46 GMT, Fredrik Lundh
 <fredrik at pythonware.com> wrote:
>never?  not even if it's a single byte? ;-)

I thought about making a qualification like this, but decided nobody would
feel the need to be such a smart-arse.

I'd say, "xreadlines will never hold more than a few KB of the file
in memory", but that's wrong if you have a file with a multi-megabyte line.

So, as you can see, I'm at a loss to describe just what xreadlines does,
but I'm sure it's a really wonderful thing.

Jeff


From jafo-pythonlist at tummy.com  Fri Feb 16 01:04:19 2001
From: jafo-pythonlist at tummy.com (Sean Reifschneider)
Date: Thu, 15 Feb 2001 17:04:19 -0700
Subject: while loop with f.readline()
In-Reply-To: <slrn98nm0u.3v2.jepler@potty.housenet>; from jepler@inetnebr.com on Thu, Feb 15, 2001 at 01:22:39PM +0000
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net> <slrn98l4rh.2s8.jepler@potty.housenet> <SiLi6.17192$AH6.2345642@newsc.telia.net> <slrn98nm0u.3v2.jepler@potty.housenet>
Message-ID: <20010215170419.L23577@tummy.com>

On Thu, Feb 15, 2001 at 01:22:39PM +0000, Jeff Epler wrote:
>I'd say, "xreadlines will never hold more than a few KB of the file
>in memory", but that's wrong if you have a file with a multi-megabyte line.

Say "xreadlines()" will never hold more than a single line of the file in
memory.  Note the implications if you feed it a file with multi-megabyte
line(s).

Sean
-- 
 I find that a great part of the information I have was acquired by looking
 up something and finding something else on the way.  -- Franklin P. Adams
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From jepler at inetnebr.com  Fri Feb 16 01:44:03 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Thu, 15 Feb 2001 18:44:03 -0600
Subject: while loop with f.readline()
In-Reply-To: <20010215170419.L23577@tummy.com>
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net> <slrn98l4rh.2s8.jepler@potty.housenet> <SiLi6.17192$AH6.2345642@newsc.telia.net> <slrn98nm0u.3v2.jepler@potty.housenet> <20010215170419.L23577@tummy.com>
Message-ID: <20010215184403.A8822@potty.housenet>

On Thu, Feb 15, 2001 at 05:04:19PM -0700, Sean Reifschneider wrote:
> On Thu, Feb 15, 2001 at 01:22:39PM +0000, Jeff Epler wrote:
> >I'd say, "xreadlines will never hold more than a few KB of the file
> >in memory", but that's wrong if you have a file with a multi-megabyte line.
> 
> Say "xreadlines()" will never hold more than a single line of the file in
> memory.  Note the implications if you feed it a file with multi-megabyte
> line(s).

Well, the full explanation is that it uses 'readlines(sizehint)'
internally, and it never holds more than one chunk as returned by
readlines().

So
	for line in file.xreadlines(): pass
has the memory behavior of
	while 1:
		lines = file.readlines(sizehint)
		if not lines: break
		for line in lines: pass
but the flow control of 
	for line in file.readlines(): pass

Jeff
PS I don't recall what value of sizehint is used in 2.1, it was changed
when the patch was integreated, after some benchmarking on python-dev



From db3l at fitlinxx.com  Sat Feb 17 00:33:12 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 16 Feb 2001 18:33:12 -0500
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net> <slrn98l4rh.2s8.jepler@potty.housenet> <SiLi6.17192$AH6.2345642@newsc.telia.net> <slrn98nm0u.3v2.jepler@potty.housenet> <20010215170419.L23577@tummy.com> <mailman.982284324.28766.python-list@python.org>
Message-ID: <u8zn6cifb.fsf@ctwd0143.fitlinxx.com>

Jeff Epler <jepler at inetnebr.com> writes:

> PS I don't recall what value of sizehint is used in 2.1, it was changed
> when the patch was integreated, after some benchmarking on python-dev

It looks like 8K in my recently updated copy from CVS.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From piet at cs.uu.nl  Thu Feb 22 15:59:26 2001
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 22 Feb 2001 15:59:26 +0100
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net> <slrn98l4rh.2s8.jepler@potty.housenet> <SiLi6.17192$AH6.2345642@newsc.telia.net> <slrn98nm0u.3v2.jepler@potty.housenet> <mailman.982281918.19055.python-list@python.org>
Message-ID: <usnl6high.fsf@cs.uu.nl>

>>>>> Sean Reifschneider <jafo-pythonlist at tummy.com> (SR) writes:

SR> On Thu, Feb 15, 2001 at 01:22:39PM +0000, Jeff Epler wrote:
>> I'd say, "xreadlines will never hold more than a few KB of the file
>> in memory", but that's wrong if you have a file with a multi-megabyte line.

SR> Say "xreadlines()" will never hold more than a single line of the file in
SR> memory.  Note the implications if you feed it a file with multi-megabyte
SR> line(s).

If you really want to be smart-assed you should note that that isn't
entirely true: there will probably also be some buffer, which may contain
more than one line. :=(
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From phlip_cpp at my-deja.com  Wed Feb 14 15:10:34 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 14 Feb 2001 14:10:34 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org>
Message-ID: <96e3kq$3g5@dispatch.concentric.net>

Proclaimed Chris Gonnerman from the mountaintops:

> I have always been a fan of the FileReader class:
> 
> class FileReader:
>     def __init__(self, file):
>         self.file = file
>     def next(self):
>         self.line = self.file.readline()
>         return self.line
> 
> if __name__ == '__main__':
> 
>     f = FileReader(open("c:/autoexec.bat"))
> 
>     while f.next():
>          print f.line,

Came to the thread late can't read all posts don't know if this was covered 
so sue me but...

...what's wrong with this?

        for z in f.readlines():
                print z

That's how I done it since I was a chile'. 

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  http://www.deja.com/my/pb.xp?member_name=phlip_cpp  --


From db3l at fitlinxx.com  Wed Feb 14 20:18:03 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 14 Feb 2001 14:18:03 -0500
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96e3kq$3g5@dispatch.concentric.net>
Message-ID: <upuglcbv8.fsf@ctwd0143.fitlinxx.com>

Phlip <phlip_cpp at my-deja.com> writes:

> Came to the thread late can't read all posts don't know if this was covered 
> so sue me but...
> 
> ...what's wrong with this?
> 
>         for z in f.readlines():
>                 print z
> 
> That's how I done it since I was a chile'. 

Nothing "wrong" - but realize that doing it that way will read the
entire file into memory (into the list object returned by readlines())
and then iterate over that object.

Probably not be the best idea if you're looking at large files, but
certainly doable for small to middling stuff.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From mmiller3 at iupui.edu  Thu Feb 15 18:49:11 2001
From: mmiller3 at iupui.edu (Michael A. Miller)
Date: 15 Feb 2001 12:49:11 -0500
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org>
Message-ID: <874rxv4z1k.fsf@lumen.med.iupui.edu>

    > From: "Chris Richard Adams"

    > I'm trying to create a loop that reads through a simple
    > file with strings on each line.

Take a look at the fileinput module:

    import fileinput
    for line in fileinput.input('filename'):
        process(line)



From ssthapa at classes.cs.uchicago.edu  Tue Feb 13 20:14:26 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Tue, 13 Feb 2001 19:14:26 GMT
Subject: Siphon -- CPAN functionality for python
Message-ID: <slrn98j244.234.ssthapa@ntcs-ip45.uchicago.edu>

  I've registered the project with sourceforge but it won't be available
until the project is approved which should take about a day or two.  In the 
meantime, if anyone would like a early look, I'll be happy to email you 
a copy.  Right now, there are a lot of stub functions where things are 
supposed to be but the core functionality should be there and you'll be 
able to see how the code fits together.

-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From ljohnson at resgen.com  Wed Feb 14 15:45:26 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Wed, 14 Feb 2001 08:45:26 -0600
Subject: Siphon -- CPAN functionality for python
References: <slrn98j244.234.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <t8l6bo2j633lb9@corp.supernews.com>

Suchandra,

How do this project's goals coordinate with those of the Catalog SIG?

    http://www.python.org/sigs/catalog-sig/

Lyle

<ssthapa at classes.cs.uchicago.edu> wrote in message
news:slrn98j244.234.ssthapa at ntcs-ip45.uchicago.edu...
>   I've registered the project with sourceforge but it won't be available
> until the project is approved which should take about a day or two.  In
the
> meantime, if anyone would like a early look, I'll be happy to email you
> a copy.  Right now, there are a lot of stub functions where things are
> supposed to be but the core functionality should be there and you'll be
> able to see how the code fits together.
>
> --
> --------------------------------------------------------------------------
--
>                  |
> Suchandra Thapa                       | "There are only two kinds of math
.
> s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read
beyond
>               | the first sentence, and those you
>               | can not read beyond the first page."
>               |                     -C.N. Yang
> --------------------------------------------------------------------------
--




From chrisa at ASPATECH.COM.BR  Tue Feb 13 20:20:50 2001
From: chrisa at ASPATECH.COM.BR (Chris Richard Adams)
Date: Tue, 13 Feb 2001 16:20:50 -0300
Subject: Python & XML
Message-ID: <BD674CDC0A68C642B60038986FBB56DF013853@mailsrv.ASPATECH.COM.BR>

Hi all..
 
I have a tab seperated file that I would like to parse and store in XML
format.   

file example:

line1: 12:00:00 bob nassl.pdf 123.123.123.123

would like to create an xml file from this text file in the form

<upload>
	<time>12:00:00</time>
	<user>bob</user>
	<file>nassl.pdf</file>
	<ip>123.123.123.123</ip>
</upload>

Could someone advise me on how I might do this with Python...for example
is there a utility to create XML doc like this.



From paulp at ActiveState.com  Tue Feb 13 19:44:11 2001
From: paulp at ActiveState.com (Paul Prescod)
Date: Tue, 13 Feb 2001 10:44:11 -0800
Subject: Python & XML
References: <BD674CDC0A68C642B60038986FBB56DF013853@mailsrv.ASPATECH.COM.BR>
Message-ID: <3A89807B.E49CF349@ActiveState.com>

Chris Richard Adams wrote:
> 
> Hi all..
> 
> 
> 
> I have a tab seperated file that I would like to parse and store in XML
> 
> format.
> 
> file example:
> 
> line1: 12:00:00 bob nassl.pdf 123.123.123.123
> 
> would like to create an xml file from this text file in the form
> 
> <upload>
> 
>         <time>12:00:00</time>
> 
>         <user>bob</user>
> 
>         <file>nassl.pdf</file>
> 
>         <ip>123.123.123.123</ip>
> 
> </upload>

I would suggest something like this:

time, name, file, ip = line.split("\t")

print """
<upload>
         <time>%s</time>
         <user>%s</user>
         <file>%s</file>
         <ip>%s</ip>
</upload>""" % (date, name, file, ip)

I would discourage you from spending much effort looking for
XML-specific ways to do things that you probably already know how to do.
Reading an XML file is hard and should usually be done through a
specialized module. Writing XML is easy so you often don't need a
module.

 Paul Prescod



From alet at unice.fr  Wed Feb 14 11:07:44 2001
From: alet at unice.fr (Jerome Alet)
Date: Wed, 14 Feb 2001 11:07:44 +0100
Subject: Python & XML
References: <mailman.982089014.6192.python-list@python.org>
Message-ID: <3A8A58F0.A3F32D81@unice.fr>

Chris Richard Adams wrote:
> I have a tab seperated file that I would like to parse and store in XML
> format.
>
> line1: 12:00:00 bob nassl.pdf 123.123.123.123
> 
> would like to create an xml file from this text file in the form
> 
> <upload>
>         <time>12:00:00</time>
>         <user>bob</user>
>         <file>nassl.pdf</file>
>         <ip>123.123.123.123</ip>
> </upload>

I suggest you to download my GPLed jaxml module from:

	http://cortex.unice.fr/~jerome/jaxml-2.00beta1.tar.gz

then:

--- CUT ---
import jaxml

yourfile = open("inputfile.txt")
doc = jaxml.XML_document()
for line in yourfile.readlines() :
	time, name, file, ip = line.split("\t")	# same as Paul
	doc._push()
	doc.upload().time(time).user(name).file(file).ip(ip)
	doc._pop()

doc._output("outputfile.xml")	# or: print doc
--- CUT ---

will produce exactly what you want, and is able of much more, without
boring you with the indentation and closing tags if you want to have a
human readable XML document: formatting is done automagically. This
module also includes templating facilities and other goodies.

You won't find something easier to use ! (or else just tell me)

don't be afraid of the "beta1", it should be read as "final" ;-)

Comments are very welcome.

-- 
Jerome Alet - alet at unice.fr - http://cortex.unice.fr/~jerome
Fac de Medecine de Nice        http://wwwmed.unice.fr 
Tel: (+33) 4 93 37 76 30     Fax: (+33) 4 93 53 15 15
28 Avenue de Valombrose - 06107 NICE Cedex 2 - FRANCE


From alet at unice.fr  Wed Feb 14 11:10:22 2001
From: alet at unice.fr (Jerome Alet)
Date: Wed, 14 Feb 2001 11:10:22 +0100
Subject: Python & XML
References: <mailman.982089014.6192.python-list@python.org> <3A8A58F0.A3F32D81@unice.fr>
Message-ID: <3A8A598E.7168AB0A@unice.fr>

Jerome Alet wrote:
> I suggest you to download my GPLed jaxml module from:
> 
>         http://cortex.unice.fr/~jerome/jaxml-2.00beta1.tar.gz

Sorry, this url works, but the correct url is:

	http://cortex.unice.fr/~jerome/jaxml/jaxml-2.00beta1.tar.gz

or:

	http://cortex.unice.fr/~jerome/jaxml/

if you want to read the README and NEWS (ChangeLog) files.

-- 
Jerome Alet - alet at unice.fr - http://cortex.unice.fr/~jerome
Fac de Medecine de Nice        http://wwwmed.unice.fr 
Tel: (+33) 4 93 37 76 30     Fax: (+33) 4 93 53 15 15
28 Avenue de Valombrose - 06107 NICE Cedex 2 - FRANCE


From pfenn at mmm.com  Tue Feb 13 20:33:02 2001
From: pfenn at mmm.com (pfenn at mmm.com)
Date: Tue, 13 Feb 2001 13:33:02 -0600
Subject: Controling types in PythonCOM
Message-ID: <OF9BE5ABD3.9F46D5A4-ON862569F2.006A2168@mmm.com>

Hello,

I'm working with an ill behaved type library, where the ODL for a function
specifies a variant integer, but by mucking around in Visual C++, I've
determined that the function really needs a VT_I2.  It appears that
PythonCOM is attempting to pass a VT_I4 which causes an exception.  I am
using the makepy generated  interface, and to make sure I'm doing this
right, I'll show my code:

from win32com.client import Dispatch
xrf = Dispatch('XRaw.Xraw.1')  # Invoked non-dynamically, right?

This is the makepy generated interface for the problem function.  I tried
changing that first (12,17) to a (2,0) after looking at some functions that
did have shorts for parameters but that didn't help.

# The method GetChromatogram is actually a property, but must be used as a
method to correctly pass the arguments
def GetChromatogram(self, FilterIndex=defaultNamedOptArg,
TraceSelector=defaultNamedOptArg, TimeRange=defaultNamedOptArg,
Delay=defaultNamedOptArg):
          """property Chromatogram"""
          return self._ApplyTypes_(0x1e, 2, (12, 0), ((12, 17), (12, 17),
(12, 17), (12, 17)), 'GetChromatogram', None,FilterIndex, TraceSelector,
TimeRange, Delay)

In case I'm interpreting my investigations incorrectly, here's the error I
get in Python when I call this function:

Traceback (most recent call last):
  File "C:\Python20\Pythonwin\pywin\framework\scriptutils.py", line 301, in
RunScript
    exec codeObject in __main__.__dict__
  File "C:\Python20\xcalibur.py", line 146, in ?
    chro3 = df.MIC(1)
  File "C:\Python20\xcalibur.py", line 131, in MIC
    chrom = self.msdet.GetChromatogram(FilterIndex=mode)
  File "c:
\python20\win32com\gen_py\2DDE1FB3-A60F-11D2-BD82-0060978EEB9Cx0x1x0.py",
line 1415, in GetChromatogram
    return self._ApplyTypes_(0x1e, 2, (12, 0), ((12, 17), (12, 17), (12,
17), (12, 17)), 'GetChromatogram', None,FilterIndex, TraceSelector,
TimeRange, Delay)
  File "C:\Python20\win32com\client\__init__.py", line 334, in _ApplyTypes_
    return self._get_good_object_(apply(self._oleobj_.InvokeTypes, (dispid,
0, wFlags, retType, argTypes) + args), user, resultCLSID)
com_error: (-2147352567, 'Exception occurred.', (0, None, None, None, 0,
-2147473646), None)

If I wrap the type library with ATL in Visual C++, the following line
generates a run time error:
IXChromatogramPtr pChrom = pMSDetector->GetChromatogram((long)
1,vtMissing,vtMissing,vtMissing);
but changing the type cast to (short) causes no error to occur.

Thanks ahead of time,

Tom





From MarkH at ActiveState.com  Wed Feb 14 00:05:15 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 13 Feb 2001 23:05:15 GMT
Subject: Controling types in PythonCOM
References: <mailman.982092877.21294.python-list@python.org>
Message-ID: <3A89BCC9.2000807@ActiveState.com>

pfenn at mmm.com wrote:


> I'm working with an ill behaved type library, where the ODL for a function
> specifies a variant integer, but by mucking around in Visual C++, I've
> determined that the function really needs a VT_I2.  It appears that
> PythonCOM is attempting to pass a VT_I4 which causes an exception.  I am
> using the makepy generated  interface, and to make sure I'm doing this
> right, I'll show my code:

Well that object sucks :)  It should be capable of pulling an int out of any of the variant integer types.

> from win32com.client import Dispatch
> xrf = Dispatch('XRaw.Xraw.1')  # Invoked non-dynamically, right?

Assuming you have makepy pre-generated, yes.

> This is the makepy generated interface for the problem function.  I tried
> changing that first (12,17) to a (2,0) after looking at some functions that
> did have shorts for parameters but that didn't help.

The 12 means VT_VARIANT.  The 17 is PARAMFLAG_FOPT | PARAMFLAG_FIN

So you probably should change it to (2,1) - ie (VT_I2, PARAMFLAG_FIN).  You can't specificy FTOPT with a non variant arg.

By passing (2,0) you _may_ have ended up with a byref int (as some code that sees no flags (ie, 0) gets conservative.)

> com_error: (-2147352567, 'Exception occurred.', (0, None, None, None, 0,
> -2147473646), None)

This is certainly the COM object itself raising the error, not Python.

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From bill.xu at newlix.com  Tue Feb 13 20:40:33 2001
From: bill.xu at newlix.com (Bill Xu)
Date: Tue, 13 Feb 2001 19:40:33 GMT
Subject: What is wrong in my pulldom.startElement()?
Message-ID: <R4gi6.1682$HG.377319@news4.rdc1.on.home.com>

Hi, Gang:

Will u please tell me how to use the pulldom.startElement() in the Python
XML DOM?
 Following is my simple test program and result.
 Will u please tell me how to correct this bug?
"AttributeError: 'None' object has no attribute 'createElement'";

Thanks in advance!

Bill

=====================================================
#!/usr/bin/python
#

import os, sys, string, StringIO
import xml.dom
from xml.dom import minidom
from xml.dom.pulldom import PullDOM
import traceback



# Save the new message
def test_meth():
  print "Am i here"
  builder = PullDOM()
  builder.startElement("SET", {"PATH" : "a path"})
  builder.endElement("SET")
  print "Am i here"
  confirm(builder)
 #result = self.Print(builder)

def confirm(test, testname = "Test"):
    if test:
        print "Passed " + testname
    else:
        print "Failed " + testname
        raise Exception

print "? am i herer"
test_meth()

=======================================================
Output:

? am i herer
Am i here
Traceback (innermost last):
  File "./testdom.py", line 34, in ?
    test_meth()
  File "./testdom.py", line 20, in test_meth
    builder.startElement("SET", {"PATH" : "a path"})
  File "/usr/lib/python1.5/site-packages/xml/dom/pulldom.py", line 85, in
startElement
    node = self.document.createElement(name)
AttributeError: 'None' object has no attribute 'createElement'




From loewis at informatik.hu-berlin.de  Mon Feb 19 19:48:24 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 19 Feb 2001 19:48:24 +0100
Subject: What is wrong in my pulldom.startElement()?
References: <R4gi6.1682$HG.377319@news4.rdc1.on.home.com>
Message-ID: <p6qd7cemruv.fsf@informatik.hu-berlin.de>

"Bill Xu" <bill.xu at newlix.com> writes:

> Will u please tell me how to use the pulldom.startElement() in the Python
> XML DOM?
>  Following is my simple test program and result.
>  Will u please tell me how to correct this bug?
> "AttributeError: 'None' object has no attribute 'createElement'";

I think you are using startElement itself quite correctly. However,
you have to emulate exactly a SAX parser, i.e. first you need to
invoke startDocument.

Please note that pulldom is undocumented. I don't know what you are
attempting to do, but using pulldom is most likely the wrong way.

Regards,
Martin



From afonso at mac.com  Tue Feb 13 21:06:51 2001
From: afonso at mac.com (Afonso Salcedo)
Date: Tue, 13 Feb 2001 20:06:51 +0000
Subject: Help: NumTut and Macs...
Message-ID: <B6AF445B.13F%afonso@mac.com>

I'm having tons of trouble running the NumTut package that comes with the
MacPython distribution.

Every time I try even the simplest test sample:

>>> from NumTut import *
>>> view(greece)

Everything hangs...

I can't seem to understand what exactly is going on. As the view() would be
useful to my work, is there any suggestion on what I should do?

Thanks,
Joe



From rtrocca at libero.it  Thu Feb 15 08:26:28 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Thu, 15 Feb 2001 07:26:28 GMT
Subject: Help: NumTut and Macs...
References: <B6AF445B.13F%afonso@mac.com>
Message-ID: <1eouamj.1tldmrir6uaw0N%rtrocca@libero.it>

Personal Experience:
view does not work with macPython.
I developed a solution for personal use that allows the user to
visualize an 2d NumPy array as a GrayScale Image or a RGB Image in an
IDE window. If interested I can send you the code.

Riccardo

Afonso Salcedo <afonso at mac.com> wrote:

> I'm having tons of trouble running the NumTut package that comes with the
> MacPython distribution.
> 
> Every time I try even the simplest test sample:
> 
> >>> from NumTut import *
> >>> view(greece)
> 
> Everything hangs...
> 
> I can't seem to understand what exactly is going on. As the view() would be
> useful to my work, is there any suggestion on what I should do?
> 
> Thanks,
> Joe


From chrisa at ASPATECH.COM.BR  Tue Feb 13 21:23:31 2001
From: chrisa at ASPATECH.COM.BR (Chris Richard Adams)
Date: Tue, 13 Feb 2001 17:23:31 -0300
Subject: Parsing a line with Python
Message-ID: <BD674CDC0A68C642B60038986FBB56DF013857@mailsrv.ASPATECH.COM.BR>

Can anyone suggest a technique where I could parse through a line in a
file and select individual strings? For example, I could process this
email and put string in a list or temp storage.

Thanks..in advance



From tjreedy at udel.edu  Tue Feb 13 21:29:59 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 13 Feb 2001 15:29:59 -0500
Subject: Parsing a line with Python
References: <mailman.982092035.18085.python-list@python.org>
Message-ID: <96c578$32k$1@news.udel.edu>

"Chris Richard Adams" <chrisa at ASPATECH.COM.BR> wrote in message
news:mailman.982092035.18085.python-list at python.org...
Can anyone suggest a technique where I could parse through a line in a
file and select individual strings? For example, I could process this
email and put string in a list or temp storage.

Strings have match(), find(), and split() methods.  There is also the re
(relational expression) module for more complicated searches.





From lreilly at cs.strath.ac.uk  Tue Feb 13 21:36:40 2001
From: lreilly at cs.strath.ac.uk (Lee Reilly CS1997)
Date: Tue, 13 Feb 2001 20:36:40 +0000
Subject: breasking out of a while loop (in an if statement)
Message-ID: <3A899AD8.B4182CA4@cs.strath.ac.uk>

Hi, I wonder if anyone could be kind enough to point out what I am doing
wrong here (please see code below).

The problem arises in the last 5 lines of code. 'result' is a database
query result (type string). If it's empty then it has size==2 and if not
then it's >=3. If it is empty then I have placed a return value after
it. However, if it is not then I would like it to go back around the
while loop. The code *looks* okay...

-=---=-=--=-==-=-=-=-==
while (j!=0):
    if (radiobutton=='comma'):
        tempString = string.split(text[i-1], ",")
    else:
        tempString = string.split(text[i-1], "\t")

    matric = string.rstrip(string.lstrip(tempString[0]))
    SQL = "select * from STUDENTS where matric= '" + matric + "'"
    result=db_conn.query(SQL)
    result = len(result)
    if (result>=3):
       return "Student with matric value " + matric + " already exists
in the database"
       # break
    else:
        j=j-1
-=---=-=--=-==-=-=-=-==

=> Error Type: IndexError
=> Error Value: list index out of range

I'm using Python with Zope BTW.

Can anyone see my problem? Any help would be greatly appreciated ;-)

- Best regards,
Lee Reilly


From hgg9140 at seanet.com  Wed Feb 14 01:43:17 2001
From: hgg9140 at seanet.com (Harry George)
Date: 13 Feb 2001 16:43:17 -0800
Subject: breasking out of a while loop (in an if statement)
References: <3A899AD8.B4182CA4@cs.strath.ac.uk>
Message-ID: <m3k86uw0uy.fsf@wilma.localdomain>

You don't indicate how "text" is initialized, nor why you are using
"i" to index it and "j" as your loop counter.  Probably something
funny there, because the "index out of range" is either those accesses
or the resulting tempString array access.

As for the loop itself, "j>0" might be safer than "j!=0", because you
might come in with a negative number.

Next, I'd print out the result value to see what it *really* is --
there may be conditions where it doesn't act the way you expect. 

Lee Reilly CS1997 <lreilly at cs.strath.ac.uk> writes:

> Hi, I wonder if anyone could be kind enough to point out what I am doing
> wrong here (please see code below).
> 
> The problem arises in the last 5 lines of code. 'result' is a database
> query result (type string). If it's empty then it has size==2 and if not
> then it's >=3. If it is empty then I have placed a return value after
> it. However, if it is not then I would like it to go back around the
> while loop. The code *looks* okay...
> 
> -=---=-=--=-==-=-=-=-==
> while (j!=0):
>     if (radiobutton=='comma'):
>         tempString = string.split(text[i-1], ",")
>     else:
>         tempString = string.split(text[i-1], "\t")
> 
>     matric = string.rstrip(string.lstrip(tempString[0]))
>     SQL = "select * from STUDENTS where matric= '" + matric + "'"
>     result=db_conn.query(SQL)
>     result = len(result)
>     if (result>=3):
>        return "Student with matric value " + matric + " already exists
> in the database"
>        # break
>     else:
>         j=j-1
> -=---=-=--=-==-=-=-=-==
> 
> => Error Type: IndexError
> => Error Value: list index out of range
> 
> I'm using Python with Zope BTW.
> 
> Can anyone see my problem? Any help would be greatly appreciated ;-)
> 
> - Best regards,
> Lee Reilly

-- 
Harry George
hgg9140 at seanet.com


From db3l at fitlinxx.com  Wed Feb 14 02:18:45 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 13 Feb 2001 20:18:45 -0500
Subject: breasking out of a while loop (in an if statement)
References: <3A899AD8.B4182CA4@cs.strath.ac.uk>
Message-ID: <u66ieoydm.fsf@ctwd0143.fitlinxx.com>

Lee Reilly CS1997 <lreilly at cs.strath.ac.uk> writes:

> Hi, I wonder if anyone could be kind enough to point out what I am doing
> wrong here (please see code below).
(...)
> -=---=-=--=-==-=-=-=-==
> while (j!=0):
>     if (radiobutton=='comma'):
>         tempString = string.split(text[i-1], ",")
>     else:
>         tempString = string.split(text[i-1], "\t")
> 
>     matric = string.rstrip(string.lstrip(tempString[0]))
>     SQL = "select * from STUDENTS where matric= '" + matric + "'"
>     result=db_conn.query(SQL)
>     result = len(result)
>     if (result>=3):
>        return "Student with matric value " + matric + " already exists
> in the database"
>        # break
>     else:
>         j=j-1
> -=---=-=--=-==-=-=-=-==
> 
> => Error Type: IndexError
> => Error Value: list index out of range
> 
> I'm using Python with Zope BTW.
> 
> Can anyone see my problem? Any help would be greatly appreciated ;-)

It would help to have the full traceback since that would highlight
the line where the actual error is occuring.  Also, it looks like this
code depends on surrounding code (e.g., where is "i" or "text" set?),
so having more detail about that surrounding code might help.

But given the error, and assuming that the error is actually within
this specific code, then it would seem to me to either be arising from
your reference to "text[i-1]" or "tempString[0]".

In the former case, you should verify that "text" has the contents you
expect and that it has enough entries in any pass through the loop as
indexed by "i-1".

I think the latter case is safe, since from what I can see, using
string.split with a dividing character will always at least produce a
single entry in the list, although it may be an empty string.  But you
might want to verify it in this actual code as well, since at least in
Python 1.5.2, string.split does return an empty list (e.g., if I split
an empty string on the default separator) in some cases, which could
then yield the IndexError when you try to access element 0.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From bshapiro at funnygarbage.com  Tue Feb 13 22:03:04 2001
From: bshapiro at funnygarbage.com (Ben Shapiro)
Date: Tue, 13 Feb 2001 16:03:04 -0500
Subject: Please Help:  COM Events
Message-ID: <96c7de$mb9$1@bob.news.rcn.net>

Hi all
I am having a problem with COM events.
I am using the JabberCOM library.  This library contains a bunch of classes
such as JabberSession and JabberMsg.
I create the object with:   JabberCtl =
win32com.client.DispatchWithEvents('JabberCOM.JabberSession',JabberEvts)
the JabberEvts class is registered to handle events for this object....

The problem is this.... in the JabberEvts Class i have the following method
that gets triggered every time a message arrives:
def OnMessage(self, Msg, Tag):
        print "I made it this far ---" #---this works
        print Msg.MsgType #--- never executes

In this method, Msg is an object of type JabberMsg
when i do a print Msg, python reports that it is an IDispatch

Seems to me that when the event handler is called, the argument passed in
loses it's type as a JabberMsg and becomes a generic Dispatch object
so i can't call properties or methods on it.. Is there any way to cast the
argument back to a JabberMsg object?

thx for any help

Ben




From MarkH at ActiveState.com  Wed Feb 14 00:07:29 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 13 Feb 2001 23:07:29 GMT
Subject: Please Help:  COM Events
References: <96c7de$mb9$1@bob.news.rcn.net>
Message-ID: <3A89BD4F.3040708@ActiveState.com>

Ben Shapiro wrote:


> Seems to me that when the event handler is called, the argument passed in
> loses it's type as a JabberMsg and becomes a generic Dispatch object
> so i can't call properties or methods on it.. Is there any way to cast the
> argument back to a JabberMsg object?

You should be able to say:

Msg = win32com.client.Dispatch(Msg)

However, event handlers are supposed to do this for you automatically.  Hrm.  No time to investigate ATM - I am busy adding support for vtable implemented interfaces to win32com ;-)

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From urner at alumni.princeton.edu  Tue Feb 13 22:04:51 2001
From: urner at alumni.princeton.edu (Kirby Urner)
Date: Tue, 13 Feb 2001 13:04:51 -0800
Subject: math through programming
Message-ID: <c88j8ts3gvtt0itgcpc4gp5omcfvh2ntna@4ax.com>

[ 

below is my most recent post to math-learn, clearly a 
response to something earlier -- but with enough standalone
content to give the idea.  

Earlier posts in the same thread include Python source 
code for computing Pascal's Triangle and comparing the 
results with balls randomly falling, pachinko-style, 
through a triangulated grid of pins (some of you may have 
seen the well-known science museum demo showing the bell 
curved result).

Thanks again to Tim Peters for cluing me re 'Concrete 
Mathematics' which I went out and grabbed some months ago
(if edu-sig had a search feature, I'd go find where he 
first posted about it, as an easier intro than Knuth's 
"telegraphic" 'Art of Computer Programming' -- and indeed, 
that's how this text evolved, from notes around the 
mathematical preliminaries leading into Knuth's magnum 
opus).

I'm cross-posting this here as a snap-shot of my ongoing
effort to combine the math learning process with a 
strong programming element, to create a hybrid more 
adapted to life in the 21st century.  The mass-education
juggernaut has too much inertia to steer in this new 
direction any time soon, but there's always the hope 
that more alert passengers will jump overboard and swim 
for their lives :-D (good to have this alternative more
seaworthy craft in the vicinity, given the more 
conventional curriculum simply isn't).

complete thread: 
http://www.mathforum.com/epigone/math-learn/snoipumkhee 

Kirby

]



Subject:      [math-learn] Re: Intro to Probability
Author:       pdx4d at teleport.com
Date:         Tue, 13 Feb 2001 16:33:03 -0000


> So the short way to write (n-k+1)! is just n!/(n-k)! as
                            ^^^^^^^^^^^^^^^^^^^^^^^^^^

> that cancels all terms except n(n-1)(n-2)..(n-k+1).

Right, a typo, thanks to a close reader for catching that.  What I 
wanted to express with (n-k+1)! were the k terms in n(n-1)...(n-k+1), 
but of course these aren't equivalent. What I need is notation to 
expressing a "falling factorial" running for k terms, and, indeed, 
such notation is offered in 'Concrete Mathematics' by Graham Knuth 
and Patashnik (Addison-Wesley, 1994), pp 47-48.

The notation can't easily be written in ascii unfortunately, but
"n to the m falling" is written [n to the m-underbar].  With "n to
the m rising" = [n to the m_overbar].  "These functions are also
called 'falling factorial powers' and 'rising factorial powers',
since they are closely related to the factorial function n! =
n(n-1)...(1).  In fact n! = [n to the n-underbar] = [1 to the n-
overbar]."

So anyway, what I should have written was [n to the k-underbar].

This book 'Concrete Mathematics' I've mentioned before on math-teach,
as exemplary of what I'm aiming for at the college level with my 
K-12 curriculum writing.  It's used for computer science students 
a lot, but it's also just a fine introduction to a lot of concrete 
(not too abstract) mathematics, including discrete probability, 
binomial coefficients and elementary number theory.

I'm also trying to boost geometry content, by moving 'Beyond 
Flatland' (off the plane) -- which links to a 'Beyond Calculators'
push as well (as I haven't seen many rotating polyhedra on 
calculator screens -- though maybe these have appeared in Japan
by this time, a market typically ahead of the USA's in the 
calculator department).

As I wrote to a math educator recently by email:

   You may remember from emails past that I'm one of those 
   looking for ways to incorporate more of Bucky Fuller's 
   contributions into early math ed.  I also push a "math
   through programming" approach, and converge it with the
   Fuller-informed spatial geometry.  "Beyond Flatland" and 
   "Beyond Calculators" summarize my two pronged strategy 
   to revamp math ed pedagogy.

   I've just put up a new essay, well-illustrated, which 
   looks at the "Beyond Flatland" approach:
   http://www.inetarena.com/~pdx4d/ocn/overview.html

   And here's a review of my "Beyond Calculators" initiative:
   http://www.oreillynet.com/pub/a/python/2000/10/04/pythonnews.html

Kirby




From sandysj at asme.org  Wed Feb 14 17:48:20 2001
From: sandysj at asme.org (Jeff Sandys)
Date: Wed, 14 Feb 2001 16:48:20 GMT
Subject: math through programming
References: <c88j8ts3gvtt0itgcpc4gp5omcfvh2ntna@4ax.com>
Message-ID: <3A8AB6D4.3216FC3E@asme.org>

I am cross posting this to comp.lang.logo to expose the 
Logo community to python and "Computer Programming for 
Everyone" (CP4E).

My view of Logo is a mathematical playground.  I still 
believe that it is a better language for young students 
(I like working with 9 year old and above), and it 
embraces different learning styles with a cleaner and 
more forgiving syntax.

The advantage of python is object programming, which 
some Logo languages have, and a skill that can be used 
later.  While the learning processes developed with Logo 
translate to other programming languages and to thinking 
in general, I have never seen any commercial applications 
or job openings for Logo.

What python is missing is a course outline that is age 
appropriate and accommodates different learning styles. 
I am eager to try python on some middle school students 
in an after school programming club, but without a plan 
I will stick to tried and proven Logo.

Thanks,
Jeff Sandys

Kirby Urner wrote:
> 
> [
> 
> below is my most recent post to math-learn, clearly a
> response to something earlier -- but with enough standalone
> content to give the idea.
> 
> Earlier posts in the same thread include Python source
> code for computing Pascal's Triangle and comparing the
> results with balls randomly falling, pachinko-style,
> through a triangulated grid of pins (some of you may have
> seen the well-known science museum demo showing the bell
> curved result).
> 
> Thanks again to Tim Peters for cluing me re 'Concrete
> Mathematics' which I went out and grabbed some months ago
> (if edu-sig had a search feature, I'd go find where he
> first posted about it, as an easier intro than Knuth's
> "telegraphic" 'Art of Computer Programming' -- and indeed,
> that's how this text evolved, from notes around the
> mathematical preliminaries leading into Knuth's magnum
> opus).
> 
> I'm cross-posting this here as a snap-shot of my ongoing
> effort to combine the math learning process with a
> strong programming element, to create a hybrid more
> adapted to life in the 21st century.  The mass-education
> juggernaut has too much inertia to steer in this new
> direction any time soon, but there's always the hope
> that more alert passengers will jump overboard and swim
> for their lives :-D (good to have this alternative more
> seaworthy craft in the vicinity, given the more
> conventional curriculum simply isn't).
> 
> complete thread:
> http://www.mathforum.com/epigone/math-learn/snoipumkhee
> 
> Kirby
> 
> ]
> 
> Subject:      [math-learn] Re: Intro to Probability
> Author:       pdx4d at teleport.com
> Date:         Tue, 13 Feb 2001 16:33:03 -0000
> 
> > So the short way to write (n-k+1)! is just n!/(n-k)! as
>                             ^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> > that cancels all terms except n(n-1)(n-2)..(n-k+1).
> 
> Right, a typo, thanks to a close reader for catching that.  What I
> wanted to express with (n-k+1)! were the k terms in n(n-1)...(n-k+1),
> but of course these aren't equivalent. What I need is notation to
> expressing a "falling factorial" running for k terms, and, indeed,
> such notation is offered in 'Concrete Mathematics' by Graham Knuth
> and Patashnik (Addison-Wesley, 1994), pp 47-48.
> 
> The notation can't easily be written in ascii unfortunately, but
> "n to the m falling" is written [n to the m-underbar].  With "n to
> the m rising" = [n to the m_overbar].  "These functions are also
> called 'falling factorial powers' and 'rising factorial powers',
> since they are closely related to the factorial function n! =
> n(n-1)...(1).  In fact n! = [n to the n-underbar] = [1 to the n-
> overbar]."
> 
> So anyway, what I should have written was [n to the k-underbar].
> 
> This book 'Concrete Mathematics' I've mentioned before on math-teach,
> as exemplary of what I'm aiming for at the college level with my
> K-12 curriculum writing.  It's used for computer science students
> a lot, but it's also just a fine introduction to a lot of concrete
> (not too abstract) mathematics, including discrete probability,
> binomial coefficients and elementary number theory.
> 
> I'm also trying to boost geometry content, by moving 'Beyond
> Flatland' (off the plane) -- which links to a 'Beyond Calculators'
> push as well (as I haven't seen many rotating polyhedra on
> calculator screens -- though maybe these have appeared in Japan
> by this time, a market typically ahead of the USA's in the
> calculator department).
> 
> As I wrote to a math educator recently by email:
> 
>    You may remember from emails past that I'm one of those
>    looking for ways to incorporate more of Bucky Fuller's
>    contributions into early math ed.  I also push a "math
>    through programming" approach, and converge it with the
>    Fuller-informed spatial geometry.  "Beyond Flatland" and
>    "Beyond Calculators" summarize my two pronged strategy
>    to revamp math ed pedagogy.
> 
>    I've just put up a new essay, well-illustrated, which
>    looks at the "Beyond Flatland" approach:
>    http://www.inetarena.com/~pdx4d/ocn/overview.html
> 
>    And here's a review of my "Beyond Calculators" initiative:
>    http://www.oreillynet.com/pub/a/python/2000/10/04/pythonnews.html
> 
> Kirby


From jsteve17 at tampabay.rr.com  Tue Feb 13 22:32:15 2001
From: jsteve17 at tampabay.rr.com (Jeff Stephens)
Date: Tue, 13 Feb 2001 21:32:15 GMT
Subject: Changing Fonts in IDLE-0.5
Message-ID: <3A89A78F.C4C6B70A@tampabay.rr.com>

I am just starting out with Python and was trying to use idle, but the
fonts
are so small it's giving me a headache.  I followed the instructions in
the
help menu and tried editing the 'EditorWindow.py' file to no avail. I
commented
out the first line and uncommented the second as recommended: no change.
I tried
changing the sizes of both fonts to 10, 12: no change.  The only thing I
can figure is
that idle is using some default that is not reflected by either of these
lines. Can someone
help before I go blind.  Thanks.

Regards,
Jeff Stephens



From DanielK at jBASE.com  Tue Feb 13 23:39:51 2001
From: DanielK at jBASE.com (Daniel Klein)
Date: Tue, 13 Feb 2001 14:39:51 -0800
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com>
Message-ID: <%Oii6.238$Vf.9579@typhoon.aracnet.com>

Jeff,

You got me curious about this so after a minute of digging, I found two
files, 'config-unix.txt' and 'config-win.txt' in the '/python20/tools/idle'
(or C:\python20\tools\idle) directory. It is in these files (depending on
your platform) where the font size can be changed.

Hope this helps,

Daniel Klein
Portland OR USA

"Jeff Stephens" <jsteve17 at tampabay.rr.com> wrote in message
news:3A89A78F.C4C6B70A at tampabay.rr.com...
> I am just starting out with Python and was trying to use idle, but the
> fonts
> are so small it's giving me a headache.  I followed the instructions in
> the
> help menu and tried editing the 'EditorWindow.py' file to no avail. I
> commented
> out the first line and uncommented the second as recommended: no change.
> I tried
> changing the sizes of both fonts to 10, 12: no change.  The only thing I
> can figure is
> that idle is using some default that is not reflected by either of these
> lines. Can someone
> help before I go blind.  Thanks.
>
> Regards,
> Jeff Stephens
>




From jsteve17 at tampabay.rr.com  Wed Feb 14 02:51:57 2001
From: jsteve17 at tampabay.rr.com (Jeff Stephens)
Date: Wed, 14 Feb 2001 01:51:57 GMT
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com>
Message-ID: <3A89E46C.A081293F@tampabay.rr.com>

Daniel Klein wrote:

> Jeff,
>
> You got me curious about this so after a minute of digging, I found two
> files, 'config-unix.txt' and 'config-win.txt' in the '/python20/tools/idle'
> (or C:\python20\tools\idle) directory. It is in these files (depending on
> your platform) where the font size can be changed.
>
> Hope this helps,
>
> Daniel Klein
> Portland OR USA
>

I'm running on Linux, specifically, Redhat 6.2.  I don't have either of these
files.  Here are the
files in my idle-0.5 directory:

AutoExpand.py        IdleHistory.pyc        Separator.py
AutoExpand.pyc       IdlePrefs.py           StackViewer.py
AutoIndent.py        IdlePrefs.pyc          TODO.txt
AutoIndent.pyc       MultiScrolledLists.py  ToolTip.py
Bindings.py          MultiStatusBar.py      TreeWidget.py
Bindings.pyc         MultiStatusBar.pyc     UndoDelegator.py
CallTipWindow.py     NEWS.txt               UndoDelegator.pyc
CallTips.py          ObjectBrowser.py       WidgetRedirector.py
CallTips.pyc         OldStackViewer.py      WidgetRedirector.pyc
ChangeLog            OutputWindow.py        WindowList.py
ClassBrowser.py      OutputWindow.pyc       WindowList.pyc
ColorDelegator.py    PathBrowser.py         ZoomHeight.py
ColorDelegator.pyc   Percolator.py          ZoomHeight.pyc
Debugger.py          Percolator.pyc         __init__.py
Delegator.py         PyParse.py             eventparse.py
Delegator.pyc        PyParse.pyc            extend.py
EditorWindow.py      PyShell.py             extend.pyc
EditorWindow.pyc     PyShell.pyc            extend.txt
FileList.py          README.txt             help.txt
FileList.pyc         ReplaceDialog.py       idle.bat
FormatParagraph.py   ScriptBinding.py       idle.py
FormatParagraph.pyc  ScriptBinding.pyc      idle.pyw
FrameViewer.py       ScrolledList.py        idlever.py
GrepDialog.py        SearchBinding.py       idlever.pyc
IOBinding.py         SearchBinding.pyc      keydefs.py
IOBinding.pyc        SearchDialog.py        keydefs.pyc
Icons                SearchDialogBase.py    pyclbr.py
IdleHistory.py       SearchEngine.py        testcode.py

Any other ideas?


>
> "Jeff Stephens" <jsteve17 at tampabay.rr.com> wrote in message
> news:3A89A78F.C4C6B70A at tampabay.rr.com...
> > I am just starting out with Python and was trying to use idle, but the
> > fonts
> > are so small it's giving me a headache.  I followed the instructions in
> > the
> > help menu and tried editing the 'EditorWindow.py' file to no avail. I
> > commented
> > out the first line and uncommented the second as recommended: no change.
> > I tried
> > changing the sizes of both fonts to 10, 12: no change.  The only thing I
> > can figure is
> > that idle is using some default that is not reflected by either of these
> > lines. Can someone
> > help before I go blind.  Thanks.
> >
> > Regards,
> > Jeff Stephens
> >



From dan at eevolved.com  Wed Feb 14 03:48:42 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 14 Feb 2001 02:48:42 GMT
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com>
Message-ID: <emmi6.243735$JT5.8705881@news20.bellglobal.com>

try doing

# grep "font" * 

inside your idle/ directory

Dan


From jsteve17 at tampabay.rr.com  Wed Feb 14 04:17:27 2001
From: jsteve17 at tampabay.rr.com (Jeff Stephens)
Date: Wed, 14 Feb 2001 03:17:27 GMT
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com> <emmi6.243735$JT5.8705881@news20.bellglobal.com>
Message-ID: <3A89F875.BD37A431@tampabay.rr.com>

Dan Parisien wrote:

> try doing
>
> # grep "font" *
>
> inside your idle/ directory
>
> Dan

Already did.  The only entries for font are in EditorWindow.py and
EditorWindow.pyc.  The latter file can't be viewed in
an editor.  I suspect it's some sort of compiled version of the .py
file.



From dan at eevolved.com  Wed Feb 14 06:41:11 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 14 Feb 2001 05:41:11 GMT
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com> <emmi6.243735$JT5.8705881@news20.bellglobal.com> <3A89F875.BD37A431@tampabay.rr.com>
Message-ID: <XToi6.232692$f36.9382275@news20.bellglobal.com>

Jeff Stephens wrote:
> Already did.  The only entries for font are in EditorWindow.py and
> EditorWindow.pyc.  The latter file can't be viewed in
> an editor.  I suspect it's some sort of compiled version of the .py
> file.
> 
Yep. It's the byte-compiled python source (it's faster for python to load).

A mad-hack you could pull is changing line 152:

text['font'] = edconf.get('font-name'), edconf.get('font-size')

with

text['font'] = "courier", "12"

I hope it works :)

Dan


From martin.franklin at westerngeco.com  Wed Feb 14 09:33:09 2001
From: martin.franklin at westerngeco.com (Martin Franklin)
Date: Wed, 14 Feb 2001 08:33:09 +0000
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com> <emmi6.243735$JT5.8705881@news20.bellglobal.com> <3A89F875.BD37A431@tampabay.rr.com> <XToi6.232692$f36.9382275@news20.bellglobal.com>
Message-ID: <3A8A42C5.A4A0A56C@westerngeco.com>

With idle 0.5 you should take a look around line 141 in 
EditorWindow.py


change this.....

if sys.platform[:3] == 'win':
    text['font'] = ("lucida console", 8)


to just this.....


text['font'] = ("Courier ", 24)


24 may be a little too big but you get the idea.

You should consider upgrading to python 2.0 and idle 0.6

Martin.



Dan Parisien wrote:
> 
> Jeff Stephens wrote:
> > Already did.  The only entries for font are in EditorWindow.py and
> > EditorWindow.pyc.  The latter file can't be viewed in
> > an editor.  I suspect it's some sort of compiled version of the .py
> > file.
> >
> Yep. It's the byte-compiled python source (it's faster for python to load).
> 
> A mad-hack you could pull is changing line 152:
> 
> text['font'] = edconf.get('font-name'), edconf.get('font-size')
> 
> with
> 
> text['font'] = "courier", "12"
> 
> I hope it works :)
> 
> Dan
> --
> http://mail.python.org/mailman/listinfo/python-list



From jsteve17 at tampabay.rr.com  Wed Feb 14 19:20:54 2001
From: jsteve17 at tampabay.rr.com (Jeff Stephens)
Date: Wed, 14 Feb 2001 18:20:54 GMT
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com> <emmi6.243735$JT5.8705881@news20.bellglobal.com> <3A89F875.BD37A431@tampabay.rr.com> <XToi6.232692$f36.9382275@news20.bellglobal.com> <mailman.982139716.13195.python-list@python.org>
Message-ID: <3A8ACC32.5CE1C9B4@tampabay.rr.com>

Martin Franklin wrote:

> With idle 0.5 you should take a look around line 141 in
> EditorWindow.py
>
> change this.....
>
> if sys.platform[:3] == 'win':
>     text['font'] = ("lucida console", 8)
>
> to just this.....
>
> text['font'] = ("Courier ", 24)
>
> 24 may be a little too big but you get the idea.
>
> You should consider upgrading to python 2.0 and idle 0.6
>
> Martin.
>

Well, this seems to be trying to work, except that I think that the syntax for
selecting the font must be off.  Here is the error
I get after commenting out the "if sys.platform...." statement:

Traceback (innermost last):
  File "./idle.py", line 2, in ?
    import PyShell
  File "./PyShell.py", line 15, in ?
    from EditorWindow import EditorWindow, fixwordbreaks
  File "./EditorWindow.py", line 142
    text['font'] = ("times adobe", 14)
    ^
SyntaxError: invalid syntax

Maybe I should try Python 2.0.

>
> Dan Parisien wrote:
> >
> > Jeff Stephens wrote:
> > > Already did.  The only entries for font are in EditorWindow.py and
> > > EditorWindow.pyc.  The latter file can't be viewed in
> > > an editor.  I suspect it's some sort of compiled version of the .py
> > > file.
> > >
> > Yep. It's the byte-compiled python source (it's faster for python to load).
> >
> > A mad-hack you could pull is changing line 152:
> >
> > text['font'] = edconf.get('font-name'), edconf.get('font-size')
> >
> > with
> >
> > text['font'] = "courier", "12"
> >
> > I hope it works :)
> >
> > Dan
> > --
> > http://mail.python.org/mailman/listinfo/python-list



From db3l at fitlinxx.com  Wed Feb 14 21:45:32 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 14 Feb 2001 15:45:32 -0500
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com> <emmi6.243735$JT5.8705881@news20.bellglobal.com> <3A89F875.BD37A431@tampabay.rr.com> <XToi6.232692$f36.9382275@news20.bellglobal.com> <mailman.982139716.13195.python-list@python.org> <3A8ACC32.5CE1C9B4@tampabay.rr.com>
Message-ID: <uelx1gfir.fsf@ctwd0143.fitlinxx.com>

Jeff Stephens <jsteve17 at tampabay.rr.com> writes:

> Well, this seems to be trying to work, except that I think that the
> syntax for selecting the font must be off.  Here is the error I get
> after commenting out the "if sys.platform...." statement:
> 
> Traceback (innermost last):
>   File "./idle.py", line 2, in ?
>     import PyShell
>   File "./PyShell.py", line 15, in ?
>     from EditorWindow import EditorWindow, fixwordbreaks
>   File "./EditorWindow.py", line 142
>     text['font'] = ("times adobe", 14)
>     ^
> SyntaxError: invalid syntax

You might want to include a snippet of what the file actually looks
like after your changes, including a few of the surrounding lines -
there may be an indentation problem depending on what precisely you
commented out.  Make sure that your new line is at the same
indentation level as the surrounding lines (and to be safe only use
spaces, don't try to mix in tabs).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From greg at cosc.canterbury.ac.nz  Fri Feb 16 03:46:33 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 16 Feb 2001 15:46:33 +1300
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com> <emmi6.243735$JT5.8705881@news20.bellglobal.com> <3A89F875.BD37A431@tampabay.rr.com> <XToi6.232692$f36.9382275@news20.bellglobal.com> <mailman.982139716.13195.python-list@python.org> <3A8ACC32.5CE1C9B4@tampabay.rr.com>
Message-ID: <3A8C9489.6A25D804@cosc.canterbury.ac.nz>

Jeff Stephens wrote:
> 
>   File "./EditorWindow.py", line 142
>     text['font'] = ("times adobe", 14)
>     ^
> SyntaxError: invalid syntax

Looks like an indentation problem. Make sure you
indent the line exactly the same way as it
was previously.

You sound like you're new to Python, so you
need to be aware that indentation is syntactically
significant in Python.

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From pod at merlin.net.au  Wed Feb 14 12:40:26 2001
From: pod at merlin.net.au (PoD)
Date: Wed, 14 Feb 2001 22:10:26 +1030
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com>
Message-ID: <3A8A6EAA.9E749052@merlin.net.au>

Jeff Stephens wrote:
> 
> Daniel Klein wrote:
> 
> > Jeff,
> >
> > You got me curious about this so after a minute of digging, I found two
> > files, 'config-unix.txt' and 'config-win.txt' in the '/python20/tools/idle'
> > (or C:\python20\tools\idle) directory. It is in these files (depending on
> > your platform) where the font size can be changed.
> >
> > Hope this helps,
> >
> > Daniel Klein
> > Portland OR USA
> >
> 
> I'm running on Linux, specifically, Redhat 6.2.  I don't have either of these
> files.  Here are the
> files in my idle-0.5 directory:
> 
<Snipped>

Here are the config files from my idle directory from Python-2.0, you
never know, they may work.
-------------- next part --------------
# IDLE reads several config files to determine user preferences.  This 
# file is the default config file.  When IDLE starts, it will look in
# the following four files in order:
#     config.txt                      the default config file
#     config-[win/unix/mac].txt       the generic platform config file
#     config-[sys.platform].txt       the specific platform config file
#     ~/.idle                         the user config file
# XXX what about Windows?
#
# The last definition of each option is used.  For example, you can
# override the default window size (80x24) by defining width and
# height options in the EditorWindow section of your ~/.idle file
#
# IDLE extensions can be enabled and disabled by adding them to one of
# the config files.  To enable an extension, create a section with the
# same name as the extension, e.g. the [ParenMatch] section below.  To
# disable an extension, either remove the section or add the 'enable'
# option with the value 0.  

[EditorWindow]
width= 80
height= 24
# fonts defined in config-[win/unix].txt

[Colors]
normal-foreground= black
normal-background= white
# These color types are not explicitly defined= sync, todo, stdin
keyword-foreground= #ff7700
comment-foreground= #dd0000
string-foreground= #00aa00
definition-foreground= #0000ff
hilite-foreground= #000068
hilite-background= #006868
break-foreground= #ff7777
hit-foreground= #ffffff
hit-background= #000000
stdout-foreground= blue
stderr-foreground= red
console-foreground= #770000
error-background= #ff7777
cursor-background= black

[SearchBinding]

[AutoIndent]

[AutoExpand]

[FormatParagraph]

[ZoomHeight]

[ScriptBinding]

[CallTips]

[ParenMatch]
enable= 0
style= expression
flash-delay= 500
bell= 1
hilite-foreground= black
hilite-background= #43cd80
-------------- next part --------------
[EditorWindow]
font-name= courier
font-size= 12

From DanielK at jBASE.com  Wed Feb 14 00:07:06 2001
From: DanielK at jBASE.com (Daniel Klein)
Date: Tue, 13 Feb 2001 15:07:06 -0800
Subject: SocketServer inconsistencies
Message-ID: <ycji6.241$Vf.9600@typhoon.aracnet.com>

I am attempting to write some socket code which must be platform indepent
and the following behavior indicates that I can't use the SocketServer
module as I had hoped to do.

On Linux using Python 1.5.2, 'import SocketServer' correctly imports into
the global namespace. On Windows, using Python 2.0, the same command returns

>>> import SocketServer
Traceback (innermost last):
  File "<pyshell#0>", line 1, in ?
    import SocketServer
NameError: Case mismatch for module name SocketServer
(filename C:\DanielK\python\socketserver.py)

...however when I try 'import socketserver' (note the case of the command),
it actually _starts_  the server, ie

>>> import socketserver
Waiting for a connection...

I know I'm using different version on the different platforms but could
someone please advise how to use the SocketServer module on Windows please?
Has the implementation changed from 1.5.2 to 2.0?

Please forgive my naivety if the answer is obvious but this is a serious
problem for me. If I can't use the SocketServer module on Windows, I have to
go back to the drawing board and see what I would need to do to use the
lower level 'socket' module (which will require a great deal more work,
especially since I am new to network programming). :^(

Thanks for listening, and hopefully for an answer,

Daniel Klein





From spam at spam.com  Wed Feb 14 01:00:26 2001
From: spam at spam.com (Penfold)
Date: Wed, 14 Feb 2001 00:00:26 -0000
Subject: SocketServer inconsistencies
References: <ycji6.241$Vf.9600@typhoon.aracnet.com>
Message-ID: <t8jisec2vrr24@corp.supernews.com>

There is no *inconsistency* between the platforms.  It seems to me that
you've taken the somewhat dubious step
of naming the test script that you are using "socketserver.py", or at least
one of *your* modules is named that ...
note your error

> NameError: Case mismatch for module name SocketServer
> (filename C:\DanielK\python\socketserver.py)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

When you import modules, python searches for them on its path (sys.path,
$PYTHONPATH etc).  It looks in the directory of the script it is running
before looking in the library.  Now on linux this is fine ... it looks in
the scripts directory for SocketServer and never finds it (as it is called
socketserver which is a different name on *nix).  On windows however it
strikes a problem.  To Windows, there is no case sensitivity in filenames,
so to all intents and purposes it cant tell the difference between
SocketServer --- so it imports *your* module (not the library one).  But
then its smart enought to figure out that its not quite making sense ... Now
if you import socketserver it of course just runs any code in *your*
socketserver program so its your own fault if thats gone and done something
! [learn about the __name__ == '__main__': pydiom]

So ... solution, call your modules something better than socketserver ... on
windows choosing a module name the same as library module when case is
ignored is just asking for problems.  On linux of course you would have had
a similar problem if you had called your module SocketServer.py [would you
call one of your own modules sys.py or os.py or string.py ?? i hope not ;-)]

Have a nice day

D
"Daniel Klein" <DanielK at jBASE.com> wrote in message
news:ycji6.241$Vf.9600 at typhoon.aracnet.com...
> I am attempting to write some socket code which must be platform indepent
> and the following behavior indicates that I can't use the SocketServer
> module as I had hoped to do.
>
> On Linux using Python 1.5.2, 'import SocketServer' correctly imports into
> the global namespace. On Windows, using Python 2.0, the same command
returns
>
> >>> import SocketServer
> Traceback (innermost last):
>   File "<pyshell#0>", line 1, in ?
>     import SocketServer
> NameError: Case mismatch for module name SocketServer
> (filename C:\DanielK\python\socketserver.py)
>
> ...however when I try 'import socketserver' (note the case of the
command),
> it actually _starts_  the server, ie
>
> >>> import socketserver
> Waiting for a connection...
>
> I know I'm using different version on the different platforms but could
> someone please advise how to use the SocketServer module on Windows
please?
> Has the implementation changed from 1.5.2 to 2.0?
>
> Please forgive my naivety if the answer is obvious but this is a serious
> problem for me. If I can't use the SocketServer module on Windows, I have
to
> go back to the drawing board and see what I would need to do to use the
> lower level 'socket' module (which will require a great deal more work,
> especially since I am new to network programming). :^(
>
> Thanks for listening, and hopefully for an answer,
>
> Daniel Klein
>
>
>




From DanielK at jBASE.com  Wed Feb 14 01:18:45 2001
From: DanielK at jBASE.com (Daniel Klein)
Date: Tue, 13 Feb 2001 16:18:45 -0800
Subject: SocketServer inconsistencies
References: <ycji6.241$Vf.9600@typhoon.aracnet.com> <t8jisec2vrr24@corp.supernews.com>
Message-ID: <Jfki6.246$Vf.9666@typhoon.aracnet.com>

Big slap on forehead!!!! :^(

This is one lesson I will NEVER forget.

Thanks to you and /F I am now a happy parrot instead of a dead one. :^)

Dan

"I shot an error into the air. It's still going everywhere."


"Penfold" <spam at spam.com> wrote in message
news:t8jisec2vrr24 at corp.supernews.com...
> There is no *inconsistency* between the platforms.  It seems to me that
> you've taken the somewhat dubious step
> of naming the test script that you are using "socketserver.py", or at
least
> one of *your* modules is named that ...
> note your error
>
> > NameError: Case mismatch for module name SocketServer
> > (filename C:\DanielK\python\socketserver.py)
>                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> When you import modules, python searches for them on its path (sys.path,
> $PYTHONPATH etc).  It looks in the directory of the script it is running
> before looking in the library.  Now on linux this is fine ... it looks in
> the scripts directory for SocketServer and never finds it (as it is called
> socketserver which is a different name on *nix).  On windows however it
> strikes a problem.  To Windows, there is no case sensitivity in filenames,
> so to all intents and purposes it cant tell the difference between
> SocketServer --- so it imports *your* module (not the library one).  But
> then its smart enought to figure out that its not quite making sense ...
Now
> if you import socketserver it of course just runs any code in *your*
> socketserver program so its your own fault if thats gone and done
something
> ! [learn about the __name__ == '__main__': pydiom]
>
> So ... solution, call your modules something better than socketserver ...
on
> windows choosing a module name the same as library module when case is
> ignored is just asking for problems.  On linux of course you would have
had
> a similar problem if you had called your module SocketServer.py [would you
> call one of your own modules sys.py or os.py or string.py ?? i hope not
;-)]
>
> Have a nice day
>
> D
> "Daniel Klein" <DanielK at jBASE.com> wrote in message
> news:ycji6.241$Vf.9600 at typhoon.aracnet.com...
> > I am attempting to write some socket code which must be platform
indepent
> > and the following behavior indicates that I can't use the SocketServer
> > module as I had hoped to do.
> >
> > On Linux using Python 1.5.2, 'import SocketServer' correctly imports
into
> > the global namespace. On Windows, using Python 2.0, the same command
> returns
> >
> > >>> import SocketServer
> > Traceback (innermost last):
> >   File "<pyshell#0>", line 1, in ?
> >     import SocketServer
> > NameError: Case mismatch for module name SocketServer
> > (filename C:\DanielK\python\socketserver.py)
> >
> > ...however when I try 'import socketserver' (note the case of the
> command),
> > it actually _starts_  the server, ie
> >
> > >>> import socketserver
> > Waiting for a connection...
> >
> > I know I'm using different version on the different platforms but could
> > someone please advise how to use the SocketServer module on Windows
> please?
> > Has the implementation changed from 1.5.2 to 2.0?
> >
> > Please forgive my naivety if the answer is obvious but this is a serious
> > problem for me. If I can't use the SocketServer module on Windows, I
have
> to
> > go back to the drawing board and see what I would need to do to use the
> > lower level 'socket' module (which will require a great deal more work,
> > especially since I am new to network programming). :^(
> >
> > Thanks for listening, and hopefully for an answer,
> >
> > Daniel Klein
> >
> >
> >
>
>




From fredrik at pythonware.com  Wed Feb 14 01:02:26 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Feb 2001 00:02:26 GMT
Subject: SocketServer inconsistencies
References: <ycji6.241$Vf.9600@typhoon.aracnet.com>
Message-ID: <mWji6.16710$AH6.2301782@newsc.telia.net>

Daniel Klein wrote:
> On Windows, using Python 2.0, the same command returns
>
> >>> import SocketServer
> Traceback (innermost last):
>   File "<pyshell#0>", line 1, in ?
>     import SocketServer
> NameError: Case mismatch for module name SocketServer
> (filename C:\DanielK\python\socketserver.py)

did you call your test program "socketserver.py" ?

that's a bad idea, when using a case-insensitive file
system...

Cheers /F




From jepler at inetnebr.com  Wed Feb 14 04:55:06 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 14 Feb 2001 03:55:06 GMT
Subject: SocketServer inconsistencies
References: <ycji6.241$Vf.9600@typhoon.aracnet.com>
Message-ID: <slrn98k0cp.2fe.jepler@potty.housenet>

On Tue, 13 Feb 2001 15:07:06 -0800, Daniel Klein
 <DanielK at jBASE.com> wrote:
>I am attempting to write some socket code which must be platform indepent
>and the following behavior indicates that I can't use the SocketServer
>module as I had hoped to do.

This error may mean that your Python installation is damaged, and/or
that Windows is brain damaged.

Google for "PYTHONCASEOK" and you'll find a solution to your problem.

As far as I know, the SocketServer module ought to work on Windows,
the problem is due to the case of the filename (socketserver.py vs
SocketServer.py) and the fact that the VFAT/FAT32 filesystem is
case-preserving but case-sensitive.  If you perform the suggested
web search, you can read more about it than you care to see.

Jeff


From latlong at css2.com  Wed Feb 14 00:50:27 2001
From: latlong at css2.com (LatLong)
Date: Tue, 13 Feb 2001 23:50:27 GMT
Subject: ANN: Latitude longitude database
Message-ID: <7Lji6.4747$hN2.880712@news3.cableinet.net>

Latitude longitude database of over 2.8 million locations worldwide.

Download from:
http://www.css2.com/latlong.htm

Ideal for sales and marketing applications and travel related websites.
Includes sample code.





From maderacv at pacbell.net  Wed Feb 14 01:28:34 2001
From: maderacv at pacbell.net (Louie Fimbrez)
Date: Wed, 14 Feb 2001 00:28:34 GMT
Subject: Introduction to Spanish Multimedia Website
Message-ID: <Siki6.7756$bG4.104342@news.pacbell.net>

Madera CineVideo with 15 years of experience would like to announce the opening of its new Website http://www.spanishmultimedia.com, one of the largest distributors of Spanish Educational Videos from Math, to History, Language Arts and Parenting.  In Entertainment we offer something for the whole family, a wide selection of both the old classics and the new releases.  Checkout our Website and let us know how you like it.  All comments are greatly appreciated.  mailto:maderacv at pacbell.net.



From steve.waterbury at gsfc.nasa.gov  Wed Feb 14 02:31:26 2001
From: steve.waterbury at gsfc.nasa.gov (Steve Waterbury)
Date: Tue, 13 Feb 2001 20:31:26 -0500
Subject: Introduction to Spanish Multimedia Website
References: <E14SqZ4-0002Dm-00@mail.python.org>
Message-ID: <3A89DFEE.C207A651@gsfc.nasa.gov>

python-list-request at python.org wrote:

> Subject: Introduction to Spanish Multimedia Website

Well we weren't expecting the Introduction to the Spanish 
Multimedia Website!!

:^)

                                           oo _\o
                                            \/\ \
                                              /
____________________________________________ oo _________________
"Sometimes you're the windshield; sometimes you're the bug."
- Knopfler



From dan at eevolved.com  Wed Feb 14 03:11:11 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 14 Feb 2001 02:11:11 GMT
Subject: Introduction to Spanish Multimedia Website
References: <E14SqZ4-0002Dm-00@mail.python.org> <mailman.982114395.13502.python-list@python.org>
Message-ID: <3Pli6.115170$Pm2.2181515@news20.bellglobal.com>

Steve Waterbury wrote:

> python-list-request at python.org wrote:
> 
> > Subject: Introduction to Spanish Multimedia Website
> 
> Well we weren't expecting the Introduction to the Spanish
> Multimedia Website!!

hahahahaha

good one :)
Dan


From m.wilson at bibliocraft.com  Wed Feb 14 04:39:43 2001
From: m.wilson at bibliocraft.com (Mark Wilson)
Date: Wed, 14 Feb 2001 03:39:43 -0000
Subject: Introduction to Spanish Multimedia Website
Message-ID: <021e01c09637$c7c63fc0$2c824ac3@mwdell1>

>> Subject: Introduction to Spanish Multimedia Website
>
>Well we weren't expecting the Introduction to the Spanish 
>Multimedia Website!!

_Nobody_ _expects_ the Introduction to the Spanish 
Multimedia Website!!

:->




From sholden at holdenweb.com  Wed Feb 14 13:38:07 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 14 Feb 2001 07:38:07 -0500
Subject: Introduction to Spanish Multimedia Website
References: <mailman.982122194.11337.python-list@python.org>
Message-ID: <I2vi6.9228$GA5.126098@e420r-atl1.usenetserver.com>

"Mark Wilson" <m.wilson at bibliocraft.com> wrote in message
news:mailman.982122194.11337.python-list at python.org...
> >> Subject: Introduction to Spanish Multimedia Website
> >
> >Well we weren't expecting the Introduction to the Spanish
> >Multimedia Website!!
>
> _Nobody_ _expects_ the Introduction to the Spanish
> Multimedia Website!!
>
> :->
>
There are two reasons for this. First, it's Spanish. Second, it's
Multimedia. Third, it's a Website. There are _three_ reasons for this ...

rgeards
 Steve




From sdm7g at virginia.edu  Wed Feb 14 01:34:37 2001
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 13 Feb 2001 19:34:37 -0500 (EST)
Subject: Seen on comp.lang.scheme... computer languages in comp.sci.curriculum
Message-ID: <Pine.NXT.4.21.0102131931560.268-100000@localhost.virginia.edu>

FYI:

Forwarded from comp.lang.scheme
--------------------

Dear Colleagues -- 

ACM's and IEEE's joint task on curriculum development has released its
"ironman draft" proposal for a new core curriculum in computer
science. The
draft allocates six hours to the area of programming languages, relegating
it to third-class status. I have written a letter of protest and ask for
your support. Please visit my Web site where I have posted the letter and
where you can find some additional information on the topic:

   http://www.cs.rice.edu/~matthias/cc2001.html

Regards -- Matthias Felleisen






From dan at eevolved.com  Wed Feb 14 03:32:48 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 14 Feb 2001 02:32:48 GMT
Subject: crazy lambdas
Message-ID: <k7mi6.115181$Pm2.2184753@news20.bellglobal.com>

I want to do this

class Event:
        def __init__(self):
                self.callbacks = []

        def trigger(self):
                for i in self.callbacks:
                        i()

e = Event()

person = "me"
dict = {person:1}

e.callbacks.append(lambda d=dict, p=person: del d[p])

# .... later

e.trigger() #deletes
---
those who use lambdas alot will notice my mistake. I, however, do not know 
what I am doing wrong. I think it has something to do with the fact 'del 
d[p]' doesn't have a value...

Is it possible to do what I want (make a custom one-time-use function that 
is called much later and deletes a key from a dictionary)?

thanks,
Dan


From johann at physics.berkeley.edu  Wed Feb 14 04:23:02 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 13 Feb 2001 19:23:02 -0800
Subject: crazy lambdas
References: <k7mi6.115181$Pm2.2184753@news20.bellglobal.com>
Message-ID: <mthf1yuew9.fsf@astron.berkeley.edu>

Dan Parisien writes:
> those who use lambdas alot will notice my mistake. I, however, do not know 
> what I am doing wrong. I think it has something to do with the fact 'del 
> d[p]' doesn't have a value...

Yep.  "del d[p]" is a statement, not an expression.  Take a look at
the built-in function delattr.  That should do what you want.

--J

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From sdm7g at virginia.edu  Wed Feb 14 04:32:13 2001
From: sdm7g at virginia.edu (Steven D. Majewski)
Date: Tue, 13 Feb 2001 22:32:13 -0500 (EST)
Subject: crazy lambdas
In-Reply-To: <k7mi6.115181$Pm2.2184753@news20.bellglobal.com>
Message-ID: <Pine.A32.3.90.1010213222411.14204B-100000@elvis.med.Virginia.EDU>

On Wed, 14 Feb 2001, Dan Parisien wrote:

> I want to do this
> 
> e.callbacks.append(lambda d=dict, p=person: del d[p])
> 
> # .... later
> 
> e.trigger() #deletes
> ---
> those who use lambdas alot will notice my mistake. I, however, do not know 
> what I am doing wrong. I think it has something to do with the fact 'del 
> d[p]' doesn't have a value...

No -- the problem is that 'lambda' expressions take an expression
and 'del' is a statement in the python grammar. 

You'll get the same error if you try: 'lambda x: import sys' --
The problem is 'del' , not 'd[p]' . 

> Is it possible to do what I want (make a custom one-time-use function that 
> is called much later and deletes a key from a dictionary)?


I think the more pythonic way to do this would be to subclass
UserDict and make a method that deletes the keys. 


---|  Steven D. Majewski   (804-982-0831)  <sdm7g at Virginia.EDU>  |---
---|  Department of Molecular Physiology and Biological Physics  |---
---|  University of Virginia             Health Sciences Center  |---
---|  P.O. Box 10011            Charlottesville, VA  22906-0011  |---
		"All operating systems want to be unix, 
		 All programming languages want to be lisp." 




From scarblac at pino.selwerd.nl  Wed Feb 14 11:22:26 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 14 Feb 2001 10:22:26 GMT
Subject: crazy lambdas
References: <k7mi6.115181$Pm2.2184753@news20.bellglobal.com>
Message-ID: <slrn98kmvh.554.scarblac@pino.selwerd.nl>

Dan Parisien <dan at eevolved.com> wrote in comp.lang.python:
> e.callbacks.append(lambda d=dict, p=person: del d[p])
> 
> # .... later
> 
> e.trigger() #deletes
> ---
> those who use lambdas alot will notice my mistake. I, however, do not know 
> what I am doing wrong. I think it has something to do with the fact 'del 
> d[p]' doesn't have a value...

The problem is that it is a statement, not a function. Note that you don't
use parentheses after del either, just like with print, import, exec, etc.
Assignment isn't a function either.

> Is it possible to do what I want (make a custom one-time-use function that 
> is called much later and deletes a key from a dictionary)?

Make the function with a normal def. Make another function that creates
those functions so you don't have to do it all the time...

def make_deleter(dict, person):
   def deleter(dict=dict, person=person):
      del dict[person]
   return deleter
   
e.callbacks.append(make_deleter(dict, person))

-- 
Remco Gerlich


From hgg9140 at seanet.com  Wed Feb 14 04:21:41 2001
From: hgg9140 at seanet.com (Harry George)
Date: 13 Feb 2001 19:21:41 -0800
Subject: XML Schema?
Message-ID: <m3ofw62blm.fsf@wilma.localdomain>

Anyone have a python XML Schema parser/validator?  I thought I saw
comments that it wasn't being done yet as part of xml-sig.  Of course,
we don't actually need an XML Schema validator inpython (java or C++
renditions would do fine), but there is a social cachet to it, so
maybe worth the effort.

Assuming it is an open task, here is an approach.  Anyone see holes in
this, besides it being a humongous task?

1. Get the specs from OASIS-->W3C.

2. Get test cases (for schemas and for instances) There are a few
   cases at xml-conf, but I think a lot more will be needed.  So I'll
   need to generate them, and that suggests a case generator, plus of
   course a test driver.  I have the testcase generator and driver
   done.

3. XML Schema is basically a regular expression problem, with nodes as
   the "characters".  So we can use classical lexer algorithms:
   regexpr --> NFA --> DFA.  The hassles may be at the leaf nodes,
   where XML Schema has lots of special cases.  I don't knbow if there
   are non-re constraints in the specs, but if so I'd apply them after
   the initial pass.

4. Given that state machine, run schemas through the parser until it can
   build machines from valid schemas and detect invalid ones.

5. Given a sound state machine, run instance test cases through the
   package until it is passing valid instances and detecting invalid
   ones.

6. This would probably be an iterative enhancement exercise, once the
   state machine engine was in place.

I have a lex-workalike I wrote in Modula-2, which I'll use as the
start point.  Probably could use a SAX input approach ("next node"
instead of "next char"), maybe with 1 lookahead.



-- 
Harry George
hgg9140 at seanet.com


From adaminthedomaindevtty.net  Wed Feb 14 08:00:31 2001
From: adaminthedomaindevtty.net (Adam Logghe)
Date: Tue, 13 Feb 2001 23:00:31 -0800
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain>
Message-ID: <3a8a2b13$1_1@news.nwlink.com>

I haven't seen a Python schema validator yet.

I would think that if anyone would know it would be the XML-SIG mailing list
and the 4suite.org people.

4suite has the most complete standard XML toolset in Python I've seen yet,
I've been VERY pleased.

I suspect that Schema is not far enough along yet for serious
implementation??

Adam
adam in the domain devtty.net


"Harry George" <hgg9140 at seanet.com> wrote in message
news:m3ofw62blm.fsf at wilma.localdomain...
> Anyone have a python XML Schema parser/validator?  I thought I saw
> comments that it wasn't being done yet as part of xml-sig.  Of course,
> we don't actually need an XML Schema validator inpython (java or C++
> renditions would do fine), but there is a social cachet to it, so
> maybe worth the effort.
>
> Assuming it is an open task, here is an approach.  Anyone see holes in
> this, besides it being a humongous task?
>
> 1. Get the specs from OASIS-->W3C.
>
> 2. Get test cases (for schemas and for instances) There are a few
>    cases at xml-conf, but I think a lot more will be needed.  So I'll
>    need to generate them, and that suggests a case generator, plus of
>    course a test driver.  I have the testcase generator and driver
>    done.
>
> 3. XML Schema is basically a regular expression problem, with nodes as
>    the "characters".  So we can use classical lexer algorithms:
>    regexpr --> NFA --> DFA.  The hassles may be at the leaf nodes,
>    where XML Schema has lots of special cases.  I don't knbow if there
>    are non-re constraints in the specs, but if so I'd apply them after
>    the initial pass.
>
> 4. Given that state machine, run schemas through the parser until it can
>    build machines from valid schemas and detect invalid ones.
>
> 5. Given a sound state machine, run instance test cases through the
>    package until it is passing valid instances and detecting invalid
>    ones.
>
> 6. This would probably be an iterative enhancement exercise, once the
>    state machine engine was in place.
>
> I have a lex-workalike I wrote in Modula-2, which I'll use as the
> start point.  Probably could use a SAX input approach ("next node"
> instead of "next char"), maybe with 1 lookahead.
>
>
>
> --
> Harry George
> hgg9140 at seanet.com




From uche at ogbuji.net  Tue Feb 20 01:53:54 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Tue, 20 Feb 2001 00:53:54 GMT
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3a8a2b13$1_1@news.nwlink.com>
Message-ID: <3A91C021.C6C3B9BA@ogbuji.net>

Adam Logghe wrote:
> 
> I haven't seen a Python schema validator yet.
> 
> I would think that if anyone would know it would be the XML-SIG mailing list
> and the 4suite.org people.
> 
> 4suite has the most complete standard XML toolset in Python I've seen yet,
> I've been VERY pleased.
> 
> I suspect that Schema is not far enough along yet for serious
> implementation??

We've just ben talking about this on XML-SIG.

The main problem seems to be that it's pretty hard to find someone who
likes XML Schemas enough to begin implementing it.  I personally dislike
the spec.  I prefer Schematron, TREX, RELAX, HOOK, etc.

However, given the number of other specs beginning to depend on XSchema,
I don't suppose we'll have much choice.  Probably post 4Suite 1.0.

For now Henry Thomson's XSV package is the only option for Pythoneers
looking for XSchemas.  (No wonder he's an XSchema fan: he's on the
working group).

Thanks for the kind words about 4Suite, though.  Don't miss the 0.10.2
release we just put out.  *Many* bug-fixes.


-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From uche at ogbuji.net  Wed Feb 14 15:00:40 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Wed, 14 Feb 2001 14:00:40 GMT
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain>
Message-ID: <3A8A8F8E.C1D27F0B@ogbuji.net>

Harry George wrote:
> 
> Anyone have a python XML Schema parser/validator?  I thought I saw
> comments that it wasn't being done yet as part of xml-sig.  Of course,
> we don't actually need an XML Schema validator inpython (java or C++
> renditions would do fine), but there is a social cachet to it, so
> maybe worth the effort.

I'm not personally a fan of XML Schemas, but I think this would be a
very worth-while project.  You'd probably get plenty of help as well.

> Assuming it is an open task, here is an approach.  Anyone see holes in
> this, besides it being a humongous task?
> 
> 1. Get the specs from OASIS-->W3C.
> 
> 2. Get test cases (for schemas and for instances) There are a few
>    cases at xml-conf, but I think a lot more will be needed.  So I'll
>    need to generate them, and that suggests a case generator, plus of
>    course a test driver.  I have the testcase generator and driver
>    done.
> 
> 3. XML Schema is basically a regular expression problem, with nodes as
>    the "characters".

Hmm.  I wouldn't go this far.  The most basic parts of the content model
are so, but the entire data-type system and parts of the content model
need a different approach than regular grammar.

>    So we can use classical lexer algorithms:
>    regexpr --> NFA --> DFA.  The hassles may be at the leaf nodes,
>    where XML Schema has lots of special cases.  I don't knbow if there
>    are non-re constraints in the specs, but if so I'd apply them after
>    the initial pass.

Interesting approach.

> 4. Given that state machine, run schemas through the parser until it can
>    build machines from valid schemas and detect invalid ones.
> 
> 5. Given a sound state machine, run instance test cases through the
>    package until it is passing valid instances and detecting invalid
>    ones.
> 
> 6. This would probably be an iterative enhancement exercise, once the
>    state machine engine was in place.
> 
> I have a lex-workalike I wrote in Modula-2, which I'll use as the
> start point.  Probably could use a SAX input approach ("next node"
> instead of "next char"), maybe with 1 lookahead.

Just to note: LT-XML supposedly has a Python interface and an XSchemas
validator.  I still think your effort would be worth-while, especially
given your fresh approach.

http://www.ltg.ed.ac.uk/software/xml/


-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From hgg9140 at cola.ca.boeing.com  Wed Feb 14 18:58:53 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Wed, 14 Feb 2001 17:58:53 GMT
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net>
Message-ID: <xqx3ddhqh7m.fsf@cola.ca.boeing.com>

Thanks for the pointer.  I didn't find LTXML in my initial literature
search.  Given that it exists, I don't see much reason to continue on
my effort.  Maybe the xml-conf site could use the testcase generator.

Another possibility I considered was to do a python binding to the
apache Xerces "C++".  Do you know if anyone has done that?  That would
hook into IBM's significant C++/Java XML-oriented releases.

I'm not a fan of Schema either, but it sure is being hyped to the
local decision makers -- so I need a python treatment.  The whole XML
world has migrated from "It is deliberately simple so all languages can
play" to "Let's complexify it so those pesky GPL guys can't keep up."

Uche Ogbuji <uche at ogbuji.net> writes:

> Harry George wrote:
> > 
> > Anyone have a python XML Schema parser/validator?  I thought I saw
> > comments that it wasn't being done yet as part of xml-sig.  Of course,
> > we don't actually need an XML Schema validator inpython (java or C++
> > renditions would do fine), but there is a social cachet to it, so
> > maybe worth the effort.
> 
> I'm not personally a fan of XML Schemas, but I think this would be a
> very worth-while project.  You'd probably get plenty of help as well.
> 
> > Assuming it is an open task, here is an approach.  Anyone see holes in
> > this, besides it being a humongous task?
> > 
> > 1. Get the specs from OASIS-->W3C.
> > 
> > 2. Get test cases (for schemas and for instances) There are a few
> >    cases at xml-conf, but I think a lot more will be needed.  So I'll
> >    need to generate them, and that suggests a case generator, plus of
> >    course a test driver.  I have the testcase generator and driver
> >    done.
> > 
> > 3. XML Schema is basically a regular expression problem, with nodes as
> >    the "characters".
> 
> Hmm.  I wouldn't go this far.  The most basic parts of the content model
> are so, but the entire data-type system and parts of the content model
> need a different approach than regular grammar.
> 
> >    So we can use classical lexer algorithms:
> >    regexpr --> NFA --> DFA.  The hassles may be at the leaf nodes,
> >    where XML Schema has lots of special cases.  I don't knbow if there
> >    are non-re constraints in the specs, but if so I'd apply them after
> >    the initial pass.
> 
> Interesting approach.
> 
> > 4. Given that state machine, run schemas through the parser until it can
> >    build machines from valid schemas and detect invalid ones.
> > 
> > 5. Given a sound state machine, run instance test cases through the
> >    package until it is passing valid instances and detecting invalid
> >    ones.
> > 
> > 6. This would probably be an iterative enhancement exercise, once the
> >    state machine engine was in place.
> > 
> > I have a lex-workalike I wrote in Modula-2, which I'll use as the
> > start point.  Probably could use a SAX input approach ("next node"
> > instead of "next char"), maybe with 1 lookahead.
> 
> Just to note: LT-XML supposedly has a Python interface and an XSchemas
> validator.  I still think your effort would be worth-while, especially
> given your fresh approach.
> 
> http://www.ltg.ed.ac.uk/software/xml/
> 
> 
> -- 
> Uche Ogbuji
> Personal:   uche at ogbuji.net		http://uche.ogbuji.net
> Work:       uche.ogbuji at fourthought.com	http://Fourthought.com

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From rtexier at elikya.com  Thu Feb 15 12:24:38 2001
From: rtexier at elikya.com (Romuald Texier)
Date: Thu, 15 Feb 2001 11:24:38 +0000
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com>
Message-ID: <96gap7$6qj$1@news.irisa.fr>

Did you take a look at http://4suite.org/ ?

Regards.

Romuald Texier.

Harry George wrote:

> Thanks for the pointer.  I didn't find LTXML in my initial literature
> search.  Given that it exists, I don't see much reason to continue on
> my effort.  Maybe the xml-conf site could use the testcase generator.
> 
> Another possibility I considered was to do a python binding to the
> apache Xerces "C++".  Do you know if anyone has done that?  That would
> hook into IBM's significant C++/Java XML-oriented releases.
> 
> I'm not a fan of Schema either, but it sure is being hyped to the
> local decision makers -- so I need a python treatment.  The whole XML
> world has migrated from "It is deliberately simple so all languages can
> play" to "Let's complexify it so those pesky GPL guys can't keep up."
> 
> Uche Ogbuji <uche at ogbuji.net> writes:
> 
> > Harry George wrote:
> > > 
> > > Anyone have a python XML Schema parser/validator?  I thought I saw
> > > comments that it wasn't being done yet as part of xml-sig.  Of course,
> > > we don't actually need an XML Schema validator inpython (java or C++
> > > renditions would do fine), but there is a social cachet to it, so
> > > maybe worth the effort.
> > 
> > I'm not personally a fan of XML Schemas, but I think this would be a
> > very worth-while project.  You'd probably get plenty of help as well.
> > 
> > > Assuming it is an open task, here is an approach.  Anyone see holes in
> > > this, besides it being a humongous task?
> > > 
> > > 1. Get the specs from OASIS-->W3C.
> > > 
> > > 2. Get test cases (for schemas and for instances) There are a few
> > >    cases at xml-conf, but I think a lot more will be needed.  So I'll
> > >    need to generate them, and that suggests a case generator, plus of
> > >    course a test driver.  I have the testcase generator and driver
> > >    done.
> > > 
> > > 3. XML Schema is basically a regular expression problem, with nodes as
> > >    the "characters".
> > 
> > Hmm.  I wouldn't go this far.  The most basic parts of the content model
> > are so, but the entire data-type system and parts of the content model
> > need a different approach than regular grammar.
> > 
> > >    So we can use classical lexer algorithms:
> > >    regexpr --> NFA --> DFA.  The hassles may be at the leaf nodes,
> > >    where XML Schema has lots of special cases.  I don't knbow if there
> > >    are non-re constraints in the specs, but if so I'd apply them after
> > >    the initial pass.
> > 
> > Interesting approach.
> > 
> > > 4. Given that state machine, run schemas through the parser until it
> > > can
> > >    build machines from valid schemas and detect invalid ones.
> > > 
> > > 5. Given a sound state machine, run instance test cases through the
> > >    package until it is passing valid instances and detecting invalid
> > >    ones.
> > > 
> > > 6. This would probably be an iterative enhancement exercise, once the
> > >    state machine engine was in place.
> > > 
> > > I have a lex-workalike I wrote in Modula-2, which I'll use as the
> > > start point.  Probably could use a SAX input approach ("next node"
> > > instead of "next char"), maybe with 1 lookahead.
> > 
> > Just to note: LT-XML supposedly has a Python interface and an XSchemas
> > validator.  I still think your effort would be worth-while, especially
> > given your fresh approach.
> > 
> > http://www.ltg.ed.ac.uk/software/xml/
> > 
> > 
> > --
> > Uche Ogbuji
> > Personal:   uche at ogbuji.net         http://uche.ogbuji.net
> > Work:       uche.ogbuji at fourthought.com     http://Fourthought.com
> 

-- 
Romuald Texier


From loewis at informatik.hu-berlin.de  Mon Feb 19 19:52:02 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 19 Feb 2001 19:52:02 +0100
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com>
Message-ID: <p6qae7imrot.fsf@informatik.hu-berlin.de>

Harry George <hgg9140 at cola.ca.boeing.com> writes:

> Another possibility I considered was to do a python binding to the
> apache Xerces "C++".  Do you know if anyone has done that?

People have considered that, but nobody has done it, AFAICT.
Note that you could chose to expose either SAX or DOM or both.

> That would hook into IBM's significant C++/Java XML-oriented
> releases.

I guess Xerces C++ would not hook very much into Java releases...

It's also open for debate what you'd gain from using that XML parser,
compared to, say, Expat.

Regards,
Martin


From hgg9140 at cola.ca.boeing.com  Mon Feb 19 21:28:41 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Mon, 19 Feb 2001 20:28:41 GMT
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com> <p6qae7imrot.fsf@informatik.hu-berlin.de>
Message-ID: <xqxofvyo1s6.fsf@cola.ca.boeing.com>


Martin von Loewis <loewis at informatik.hu-berlin.de> writes:

> Harry George <hgg9140 at cola.ca.boeing.com> writes:

> > That would hook into IBM's significant C++/Java XML-oriented
> > releases.
> 
> I guess Xerces C++ would not hook very much into Java releases...
>

What I meant was that IBM is pouring money into XML, using C++ and
Java as the proof-of-concept languages.  It would be easy for python
to fall off the leading edge if they got too far out there.  Binding
directly to the C++ library keeps that connection.

Not that I approve of all this churning of standards.  XML was meant
to be simple -- and it was until business people figured out they
could get lockin with complex standards.
 
> It's also open for debate what you'd gain from using that XML parser,
> compared to, say, Expat.
> 

Expat has XML Schema and UDDI?

> Regards,
> Martin

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From loewis at informatik.hu-berlin.de  Tue Feb 20 14:02:21 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 20 Feb 2001 14:02:21 +0100
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com> <p6qae7imrot.fsf@informatik.hu-berlin.de> <xqxofvyo1s6.fsf@cola.ca.boeing.com>
Message-ID: <p6q1ystmrs2.fsf@informatik.hu-berlin.de>

Harry George <hgg9140 at cola.ca.boeing.com> writes:

> > It's also open for debate what you'd gain from using that XML parser,
> > compared to, say, Expat.
> > 
> 
> Expat has XML Schema and UDDI?

As Paul points out: Xerces C++ does not have schema support,
either. Not sure how UDDI fits into the parsing business, though.

Regards,
Martin


From hgg9140 at cola.ca.boeing.com  Tue Feb 20 15:47:23 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Tue, 20 Feb 2001 14:47:23 GMT
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com> <p6qae7imrot.fsf@informatik.hu-berlin.de> <xqxofvyo1s6.fsf@cola.ca.boeing.com> <p6q1ystmrs2.fsf@informatik.hu-berlin.de>
Message-ID: <xqxg0h94dj8.fsf@cola.ca.boeing.com>

My concern was not about parsers per se, but a general body of work
rapidly developing and moving to either legal or defacto standard
status.  If that work is focused on one or a few languages, then other
languages have some difficulty staying in the game.  

So I'm looking for a low cost way to keep up.  One way is to bind to
libraries generated by others -- that's easier done against C/C++
libraries.  Another is to do idiomatic code conversion -- that's
probably easier done from Java to Python.

Martin von Loewis <loewis at informatik.hu-berlin.de> writes:

> Harry George <hgg9140 at cola.ca.boeing.com> writes:
> 
> > > It's also open for debate what you'd gain from using that XML parser,
> > > compared to, say, Expat.
> > > 
> > 
> > Expat has XML Schema and UDDI?
> 
> As Paul points out: Xerces C++ does not have schema support,
> either. Not sure how UDDI fits into the parsing business, though.
> 
> Regards,
> Martin

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From aleaxit at yahoo.com  Tue Feb 20 17:38:30 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 17:38:30 +0100
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com> <p6qae7imrot.fsf@informatik.hu-berlin.de> <xqxofvyo1s6.fsf@cola.ca.boeing.com> <p6q1ystmrs2.fsf@informatik.hu-berlin.de> <xqxg0h94dj8.fsf@cola.ca.boeing.com>
Message-ID: <96u6i001hfc@news2.newsguy.com>

"Harry George" <hgg9140 at cola.ca.boeing.com> wrote in message
news:xqxg0h94dj8.fsf at cola.ca.boeing.com...
> My concern was not about parsers per se, but a general body of work
> rapidly developing and moving to either legal or defacto standard
> status.  If that work is focused on one or a few languages, then other
> languages have some difficulty staying in the game.

...unless the "other languages" are particularly apt at "fitting into
diverse ecological niches" -- Python does that well:-).

> So I'm looking for a low cost way to keep up.  One way is to bind to
> libraries generated by others -- that's easier done against C/C++
> libraries.

Easier than what...?  Definitely not easier than Java/Python
integration (with Jython), so I must be misreading you...?

> Another is to do idiomatic code conversion -- that's
> probably easier done from Java to Python.

Among the easiest things in the world (definitely easier than
stealing candy from a baby, even not considering the moral
implications of this latter act) is using Java classes from
the Jython version of Python.  If, as I originally read your
message, such use is what you desire, then I don't understand
your concern.

If, OTOH, XML use from Python is actually your main thrust,
I'd stay with 4Suite.  But that's a personal choice -- there's
just SO much current/modern/leading-edge stuff out there for
Java, that Jython may be the best choice for cross-platform
work, just as Python + win32com probably is for Windows-only
work for similar reasons.

Anyway, my key point is that the ease of "extending and
embedding" Jython with Java is *astounding* -- a completely
different order of magnitude from 'extending and embedding'
CPython with C or C++.  You don't have to write ONE LINE of
Java code in the knowledge that it will ever be used from
Python -- you can do all that's needed on the Python side
of things.  .NET may give us all that and more besides one
(perhaps not-too-far-off) day, but Java and Jython give a
LOT today, within the JVM's limitations (speed, possible
security issues, etc) which you'd have from Java itself as
well to some extent.


Alex






From hgg9140 at seanet.com  Wed Feb 21 01:15:38 2001
From: hgg9140 at seanet.com (Harry George)
Date: 20 Feb 2001 16:15:38 -0800
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com> <p6qae7imrot.fsf@informatik.hu-berlin.de> <xqxofvyo1s6.fsf@cola.ca.boeing.com> <p6q1ystmrs2.fsf@informatik.hu-berlin.de> <xqxg0h94dj8.fsf@cola.ca.boeing.com> <96u6i001hfc@news2.newsguy.com>
Message-ID: <m34rxoao2d.fsf@wilma.localdomain>


"Alex Martelli" <aleaxit at yahoo.com> writes:

> "Harry George" <hgg9140 at cola.ca.boeing.com> wrote in message
> news:xqxg0h94dj8.fsf at cola.ca.boeing.com...
> > My concern was not about parsers per se, but a general body of work
> > rapidly developing and moving to either legal or defacto standard
> > status.  If that work is focused on one or a few languages, then other
> > languages have some difficulty staying in the game.
> 
> ...unless the "other languages" are particularly apt at "fitting into
> diverse ecological niches" -- Python does that well:-).

Yes, python is extrodinary in that niche.  But we wouldn't be having
this discusison if it were on top of all the modules that Java is
covering.

> 
> > So I'm looking for a low cost way to keep up.  One way is to bind to
> > libraries generated by others -- that's easier done against C/C++
> > libraries.
> 
> Easier than what...?  Definitely not easier than Java/Python
> integration (with Jython), so I must be misreading you...?
> 

Jython is great -- if you want to use Java.  Personally, I don't want
to lock in to that world.  So I was looking for ways to escape the
java mindset, not support it.

> > Another is to do idiomatic code conversion -- that's
> > probably easier done from Java to Python.
> 
> Among the easiest things in the world (definitely easier than
> stealing candy from a baby, even not considering the moral
> implications of this latter act) is using Java classes from
> the Jython version of Python.  If, as I originally read your
> message, such use is what you desire, then I don't understand
> your concern.

Again, I don't want to use or directly link to java.  I want to link
to other libraries.

> 
> If, OTOH, XML use from Python is actually your main thrust,
> I'd stay with 4Suite.  But that's a personal choice -- there's
> just SO much current/modern/leading-edge stuff out there for
> Java, that Jython may be the best choice for cross-platform
> work, just as Python + win32com probably is for Windows-only
> work for similar reasons.
> 

I already use 4Suite.  As noted elsewhere, it doesn't do XSchema,
which is why I started this thread.


> Anyway, my key point is that the ease of "extending and
> embedding" Jython with Java is *astounding* -- a completely
> different order of magnitude from 'extending and embedding'
> CPython with C or C++.  You don't have to write ONE LINE of
> Java code in the knowledge that it will ever be used from
> Python -- you can do all that's needed on the Python side
> of things.  .NET may give us all that and more besides one
> (perhaps not-too-far-off) day, but Java and Jython give a
> LOT today, within the JVM's limitations (speed, possible
> security issues, etc) which you'd have from Java itself as
> well to some extent.
> 

I may have to eventually use .NET at work, but it will be a cold
day in Tuxland before I use it at home...

> 
> Alex
> 
> 
> 
> 

-- 
Harry George
hgg9140 at seanet.com


From aleaxit at yahoo.com  Wed Feb 21 10:18:36 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 10:18:36 +0100
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com> <p6qae7imrot.fsf@informatik.hu-berlin.de> <xqxofvyo1s6.fsf@cola.ca.boeing.com> <p6q1ystmrs2.fsf@informatik.hu-berlin.de> <xqxg0h94dj8.fsf@cola.ca.boeing.com> <96u6i001hfc@news2.newsguy.com> <m34rxoao2d.fsf@wilma.localdomain>
Message-ID: <97015e02mpi@news1.newsguy.com>

"Harry George" <hgg9140 at seanet.com> wrote in message
news:m34rxoao2d.fsf at wilma.localdomain...
    [snip]
> > > status.  If that work is focused on one or a few languages, then other
> > > languages have some difficulty staying in the game.
> >
> > ...unless the "other languages" are particularly apt at "fitting into
> > diverse ecological niches" -- Python does that well:-).
>
> Yes, python is extrodinary in that niche.  But we wouldn't be having
> this discusison if it were on top of all the modules that Java is
> covering.

But it *IS* -- just use Jython, hey presto, you're there.


> > > So I'm looking for a low cost way to keep up.  One way is to bind to
> > > libraries generated by others -- that's easier done against C/C++
> > > libraries.
> >
> > Easier than what...?  Definitely not easier than Java/Python
> > integration (with Jython), so I must be misreading you...?
>
> Jython is great -- if you want to use Java.

This is like saying that "CPython is great -- if you want to use C".

A total non-sequitur!  If you want to use C, use C; if you want to
use Java, use Java; Python is great if you want to AVOID using C, or
Java, as much as feasible, while still extracting benefits from
existing libraries that are written in/for those languages.

The correct statements may thus be more like:
J)    Jython is great if you want to use libraries written in/for
        Java (while eschewing use of Java itself as much as possible)
just like:
C)    CPython is great if you want to use libraries written in/for
        C/C++ (while eschewing use of C or C++ as much as possible)

The fact is that, while both statements (C) and (J) are true, (J)
is *far TRUER* -- reusing ANY library written in/for Java (while
eschewing use of Java itself) is a _snap_ with Jython, while the
analogous situation in case (C) often requires *SOME* amount of
C-level work to "wrap" the library specifically for Python use.

> Personally, I don't want
> to lock in to that world.  So I was looking for ways to escape the
> java mindset, not support it.

The Java mindset seems to me to be "one language to bind them
all", etc.  Using a language that is definitely NOT Java thus
escapes "the java mindset" (as I see it) quite effectively,
and Jython is that "language that is definitely NOT Java".

Jython runs only within the *JVM*, of course -- *THAT* is the
unavoidable pre-req for reusing libraries that also run only
within the JVM.  If/when a fully-JVM-specs-compatible JIT or
AOT (ahead-of-time) compiler for JVM-bytecode emerges, then
the libraries in question will be freed from running "within"
the JVM, and so, of course, will Jython be (without any special
effort on the part of anybody EXCEPT the authors of the
hypothetical JVM-bytecode-to-something-else compilers:-).

> Again, I don't want to use or directly link to java.  I want to link
> to other libraries.

You can wrap C-coded or C++-coded libraries with lots of
existing tools such as SWIG, although much manual work
is typically needed.  But that buys you nothing squat, when
"that work is focused on one .. language", given that said
language, in practice, *IS* Java.

With Jython, you will NOT "use or directly link to" *Java* --
you WILL use the bytecodes of the JVM, rather than those
of the dedicated Python VM that's bundled with CPython,
and that's all.

> I already use 4Suite.  As noted elsewhere, it doesn't do XSchema,
> which is why I started this thread.

But neither does the C++ version of Xerces, etc, so, why
DO you at all care about 'linking' to it?!

The (perhaps-sad) current reality is that much research and
development effort regarding bleeding-edge tools is going to
build libraries that will only run within the JVM; the silver
lining in this situation is that Jython rescues you from
having to use Java as the entry-ticket to that world of
rich, bleeding-edge libraries (you do have to be able to
*read* "just enough Java to get by", just like, to use COM
Automation effectively, you have to be able to read just
enough VB to get by, and for a similar reason -- the _docs
and examples_ you will find around will be totally slanted
to Java/VB respectively; fortunately, Java and VB, as used
in docs and examples, tend to be somewhat readable -- it's
NOT as bad as if the examples and docs were slanted to C++
or Perl!-).

If you don't NEED to use certain available-for-JVM-only
bleeding-edge libraries, then CPython appears to me to be
as well-placed as any other language to use other, non-JVM
libraries.  4Suite appears to offer, roughly, about as
much XML functionality, or more, than any other XML suite
of tools that is not JVM-connected.  Easy alternatives
include using COM, which is about as effortless with CPython
(and win32all) as using for-JVM tools with Jython; if a
for-JVM-only tool IS compatible with Microsoft's JVM, you
can probably take advantage of the latter's ability to
expose any Java class as a COM/Automation object to use
it that way (I haven't tried that, and it does not seem
very strategic, given that MS's JVM is not state-of-the-
art any more, and is slowly or not-so-slowly dying).

Near-future possibilities will include XPCOM (a COM work-
alike that was developed within Mozilla and is now well
usable from CPython, thanks to ActiveState's efforts; it
remains to be seen how many bleeding-edge components &c
will in fact be available for it!) and .NET, of which
you say...:

> I may have to eventually use .NET at work, but it will be a cold
> day in Tuxland before I use it at home...

...even if/when it becomes available for, say, Linux as
well as Windows platform, and bleeding-edge tools you
need are only available for it?  Why?

Judging from this assertion, and your total reluctance
to use the JVM (which you appear to be expressing as
a reluctance to use *Java*, which would in fact be well
justified, AND a completely different thing), you would
appear to be allergic to bytecode-based execution
environments.  In which case, it's hard to understand
why you use Python, since, in any of its forms (CPython
with the special-purpose VM, Jython with JVM, Python.NET
with MSIL...), it *DOES* rely on just such bytecode ideas!-)


Alex





From loewis at informatik.hu-berlin.de  Tue Feb 20 19:08:48 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 20 Feb 2001 19:08:48 +0100
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com> <p6qae7imrot.fsf@informatik.hu-berlin.de> <xqxofvyo1s6.fsf@cola.ca.boeing.com> <p6q1ystmrs2.fsf@informatik.hu-berlin.de> <xqxg0h94dj8.fsf@cola.ca.boeing.com>
Message-ID: <j44rxp5irz.fsf@informatik.hu-berlin.de>

Harry George <hgg9140 at cola.ca.boeing.com> writes:

> So I'm looking for a low cost way to keep up.  One way is to bind to
> libraries generated by others -- that's easier done against C/C++
> libraries.  Another is to do idiomatic code conversion -- that's
> probably easier done from Java to Python.

Integrating C libraries always was one of the major strengths of
Python. I'm sure that somebody will integrate any library which would
be useful to enough people - I just doubt that Xerces is such a
library.

Python is particularly good at integrating things that come from
different origins. So if it turns out that IBM UDDI support is great,
then somebody will wrap it. You could still ignore the SAX part if you
think it sucks; I'd personally favour the small Expat library over a
heavy C++ parser any time.

Regards,
Martin


From paulp at ActiveState.com  Mon Feb 19 21:28:28 2001
From: paulp at ActiveState.com (Paul Prescod)
Date: Mon, 19 Feb 2001 12:28:28 -0800
Subject: XML Schema?
References: <m3ofw62blm.fsf@wilma.localdomain> <3A8A8F8E.C1D27F0B@ogbuji.net> <xqx3ddhqh7m.fsf@cola.ca.boeing.com>
Message-ID: <3A9181EC.D8DE0125@ActiveState.com>

Harry George wrote:
> 
> ...
> Another possibility I considered was to do a python binding to the
> apache Xerces "C++".  Do you know if anyone has done that?  That would
> hook into IBM's significant C++/Java XML-oriented releases.

Xereces C++ does not have a schema engine yet.

> I'm not a fan of Schema either, but it sure is being hyped to the
> local decision makers -- so I need a python treatment.  The whole XML
> world has migrated from "It is deliberately simple so all languages can
> play" to "Let's complexify it so those pesky GPL guys can't keep up."

That's one cynical reading of the situation. I think it is more a case
of getting more competing interests into a room than we had in the early
days. You need a much more disciplined moderator/editor in that
situation than you do when nobody cares.

-- 
Vote for Your Favorite Python & Perl Programming  
Accomplishments in the first Active Awards! 
http://www.ActiveState.com/Awards



From gReGlAnDrUm at earthlink.net  Wed Feb 14 04:27:39 2001
From: gReGlAnDrUm at earthlink.net (greg Landrum)
Date: Tue, 13 Feb 2001 19:27:39 -0800
Subject: Python20 and .pth files
Message-ID: <3A89FB2B.C1CA8A5E@earthlink.net>

This is the second time I've been bitten by this problem in two days, so I
figured that I'd post about it and see what the community consensus is.  If
the community consensus is that I'm an idiot, I'll live with that.  :-)

My system:
I'm running python20 on Win2K (using ActivePython, if that makes any
difference).  I didn't muck about with any of the installation settings, so
python lives in c:/python20

Short form of the problem:
It appears that some "advanced" python applications ignore .pth files at
times.

Specific instances of this happening:
Numeric Python likes to be installed in c:/python20.  It creates a Numeric
subdirectory and puts a Numeric.pth file in c:/python20.  Most of the time
this is just fine.

However... one of the applications I'm working on creates Python COM
servers which import Numeric.  These work fine when I call them from Python
applications (so none of my unit tests broke when I upgraded to 2.0, so I
never noticed the problem).  However, when I create one of these COM
objects from within VB, I get an ImportError for Numeric.  After much
shrieking and pulling of my hair (I'm giving a demo of these objects on
Friday, so I was not pleased to discover that they were mysteriously
broken), I worked out a solution: add c:/Python20/Numeric to my PYTHONPATH
environment variable.  Now everything is happy.

I had noticed similar problems when I tried to use my mod_python based web
stuff, these are also fixed by ensuring that c:/python20/Numeric is in
sys.path.

The Question:
Was it an accident that everything just worked before or has something
changed?  I know old versions of Numeric used to have an installer that may
or may not have set registry keys to help out with this problem.
Are .pth files being deprecated so that we (the users) should start
badgering developers who still ship modules requiring them?

I'm now looking forward to the first time I try to use PIL in one of these
cases and discover that it doesn't work either.  Luckily this time I'll
know how to work around the problem.

-greg


From MarkH at ActiveState.com  Wed Feb 14 23:12:35 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 14 Feb 2001 22:12:35 GMT
Subject: Python20 and .pth files
References: <3A89FB2B.C1CA8A5E@earthlink.net>
Message-ID: <3A8B01F3.1000605@ActiveState.com>

greg Landrum wrote:


> However... one of the applications I'm working on creates Python COM
> servers which import Numeric.  These work fine when I call them from Python
> applications (so none of my unit tests broke when I upgraded to 2.0, so I
> never noticed the problem).  However, when I create one of these COM
> objects from within VB, I get an ImportError for Numeric.  After much
> shrieking and pulling of my hair (I'm giving a demo of these objects on
> Friday, so I was not pleased to discover that they were mysteriously
> broken), I worked out a solution: add c:/Python20/Numeric to my PYTHONPATH
> environment variable.  Now everything is happy.

I believe this is a problem in certain embedding situations.  Specifically, I believe it is because Python does not know its "home", so doesnt know where to load the .pth files from.

> The Question:
> Was it an accident that everything just worked before or has something
> changed?  I know old versions of Numeric used to have an installer that may
> or may not have set registry keys to help out with this problem.
> Are .pth files being deprecated so that we (the users) should start
> badgering developers who still ship modules requiring them?

I dont believe this would ever have worked.  As you suggest, it is quite possible that only new versions of Numeric have .pth files.

This is on my "to-do" list - having Python know its correct home in embedded situations would solve a few other problems too.

Mark.


-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From thomas.heller at ion-tof.com  Thu Feb 15 08:43:22 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 15 Feb 2001 08:43:22 +0100
Subject: Python20 and .pth files
References: <3A89FB2B.C1CA8A5E@earthlink.net> <3A8B01F3.1000605@ActiveState.com>
Message-ID: <96g1ar$kuoss$1@ID-59885.news.dfncis.de>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message news:3A8B01F3.1000605 at ActiveState.com...
> greg Landrum wrote:
>
>
> > However... one of the applications I'm working on creates Python COM
> > servers which import Numeric.  These work fine when I call them from Python
> > applications (so none of my unit tests broke when I upgraded to 2.0, so I
> > never noticed the problem).  However, when I create one of these COM
> > objects from within VB, I get an ImportError for Numeric.  After much
> > shrieking and pulling of my hair (I'm giving a demo of these objects on
> > Friday, so I was not pleased to discover that they were mysteriously
> > broken), I worked out a solution: add c:/Python20/Numeric to my PYTHONPATH
> > environment variable.  Now everything is happy.
>
> I believe this is a problem in certain embedding situations.  Specifically, I believe it is because Python does not know its
"home", so doesnt know where to load the .pth files from.

Are you sure site.py is imported at all? There PYTHONPATH is extended by the contents
of the .pth files.

>
> > The Question:
> > Was it an accident that everything just worked before or has something
> > changed?  I know old versions of Numeric used to have an installer that may
> > or may not have set registry keys to help out with this problem.
> > Are .pth files being deprecated so that we (the users) should start
> > badgering developers who still ship modules requiring them?
>
> I dont believe this would ever have worked.  As you suggest, it is quite possible that only new versions of Numeric have .pth
files.
>
New versions have a .pth file (since they are distributed via distutils),
old versions used registry settings.

> This is on my "to-do" list - having Python know its correct home in embedded situations would solve a few other problems too.
>
This is a bug in getpathp.c, see
http://sourceforge.net/bugs/?func=detailbug&bug_id=131064&group_id=5470


> Mark.
>
Mark, please look at the bug. Otherwise you are forcing my to submit a patch ;-)

Thomas




From rogersd at bigfoot.com  Wed Feb 14 05:31:58 2001
From: rogersd at bigfoot.com (Dave Rogers)
Date: Tue, 13 Feb 2001 23:31:58 -0500
Subject: Python/ASP   Docs/Questions
Message-ID: <3A8A0A3E.A1EF40E9@bigfoot.com>

I'm a newcomer to this group, but I'm working on a fairly large and exciting web site using Win2K/Python/ASP (win32 135/Python2.0)

I cannot find any docs regarding the COM Scripting object that it uses.

Can anyone (Mark?) refer me to proper docs for it? - I've checked ActiveState and the net in general with no luck.

My two main problems are:

1) I would like for the COM Scripting object to not cache modules for me as I develop my code.  I see the following conceivable solutions:

a) There is some COM method I can call that will turn it off
b) There is some process I can kill or restart (I tried WWW Publishing service) that will knock everything out of memory
c) I can do manual reloads everywhere which are not desirable because I have some open source code that does "from foo import foo" where I cannot reload (foo) and I don't want to mess with the open source code because then I mess up the open source project
d) Reboot the PC

2) I want to set cookies easily using Response.  I would like to go Response.Cookies('name') = 'bob' (for a simple cookie), but instead I am forced to go

import Cookie
c = Cookie.Cookie()
...
Response.AddHeader("Set-Cookie", c[12:])
(or something similar to that)

I know Cookies is a collection - I seem to have this problem setting ASP Object values in general as I have to go Session.SetValue to set session variables.

I really appreciate any help people can give me - these seem like fairly common issues, yet beyond the newbie level.

P.S. One of the neatest things I've seen is Chuck Esterbrook's MiddleKit which was just released under Webware (http://webware.sourceforge.net).  What it does is allow you to map your python objects to a database for design and run time access.  What this means is that you can relate your objects in a 1 to 1 or 1 to many relationship and the MiddleKit will service all of your SQL needs - you do NOT write SQL or use recordsets.  Once you acquire one or more objects from the database, you can follow their object references to other database objects.  In fact I have abandoned ADO in favor of this middle layer which resides in between the database and my web site.  I probably didn't do it justice but you'll just have to check it out.

Thanks,
--
Dave Rogers



From kelly.kranabetter at gems6.gov.bc.ca  Wed Feb 14 20:07:55 2001
From: kelly.kranabetter at gems6.gov.bc.ca (KellyK)
Date: Wed, 14 Feb 2001 19:07:55 GMT
Subject: Python/ASP   Docs/Questions
References: <3A8A0A3E.A1EF40E9@bigfoot.com>
Message-ID: <3a8ad64e.222776145@nntp.gov.bc.ca>

On Tue, 13 Feb 2001 23:31:58 -0500, Dave Rogers <rogersd at bigfoot.com>
wrote:

>1) I would like for the COM Scripting object to not cache modules for me as I develop my code.  I see the following conceivable solutions:

I had this same problem. I ended up making one ASP script that did a
series of reload()'s so when I wanted to reload the modules I was
working on I would view that one page. It had the advantage of having
all the reloads in one place instead of sprinkled throughout many
pages.


From syver at NOSPAMcyberwatcher.com  Thu Feb 15 13:51:20 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Thu, 15 Feb 2001 13:51:20 +0100
Subject: Python/ASP   Docs/Questions
References: <3A8A0A3E.A1EF40E9@bigfoot.com>
Message-ID: <8V3k6.2528$X_2.64439@news1.oke.nextra.no>

"Dave Rogers" <rogersd at bigfoot.com> wrote in message
> My two main problems are:
> b) There is some process I can kill or restart (I tried WWW Publishing
service) that will knock everything out of memory
net stop w3svc works fine on my box (and then start of course)

You can also push the unload button in the application properties.


> 2) I want to set cookies easily using Response.  I would like to go
Response.Cookies('name') = 'bob' (for a simple cookie), but instead I am
forced to go

Here is an example of using the cookies collection:
    Response.Cookies[name] = value
    # the root for this application
    Response.Cookies(name).path = "/v2/" + path + "/"
    import time
    ctime = time.mktime([2033, 11, 8, 12, 4, 14, 2, 313, 0])
    Response.Cookies(name).Expires = cwutils.MakeComTime(ctime)

> P.S. One of the neatest things I've seen is Chuck Esterbrook's MiddleKit
which was just released under Webware (http://webware.sourceforge.net).
What it does is allow you to map your python objects to a database for
design and run time access.  What this means is that you can relate your
objects in a 1 to 1 or 1 to many relationship and the MiddleKit will service
all of your SQL needs - you do NOT write SQL or use recordsets.  Once you
acquire one or more objects from the database, you can follow their object
references to other database objects.  In fact I have abandoned ADO in favor
of this middle layer which resides in between the database and my web site.
I probably didn't do it justice but you'll just have to check it out.
>
> Thanks,
> --
> Dave Rogers
>




From graham at nospam.microtonal.co.uk  Thu Feb 15 21:16:00 2001
From: graham at nospam.microtonal.co.uk (Graham Breed)
Date: Thu, 15 Feb 2001 20:16 +0000 (GMT Standard Time)
Subject: Python/ASP   Docs/Questions
References: <3A8A0A3E.A1EF40E9@bigfoot.com>
Message-ID: <memo.20010215201630.10531A@gbreed.compulink.co.uk>

In article <3A8A0A3E.A1EF40E9 at bigfoot.com>, rogersd at bigfoot.com (Dave 
Rogers) wrote:

> 2) I want to set cookies easily using Response.  I would like to go 
> Response.Cookies('name') = 'bob' (for a simple cookie), but instead I 
> am forced to go
> 
> import Cookie
> c = Cookie.Cookie()
> ...
> Response.AddHeader("Set-Cookie", c[12:])
> (or something similar to that)

Here's an example that sets a cookie iff it isn't already set.

cookieName = 'example cookie'
cookie = str(Request.Cookies(cookieName))
# needs to be converted to a string to evaluate as false
if not cookie:
  Response.Cookies[cookieName]='cookie read'
  c = Response.Cookies(cookieName)
  c.expires = pywintypes.Time(time.time()+600)

> I know Cookies is a collection - I seem to have this problem setting 
> ASP Object values in general as I have to go Session.SetValue to set 
> session variables.

Yes, it's because Python doesn't allow you to assign to the result of a 
function.

Not that it's a problem with session variables, because they're evil 
anyway.

               Graham


From gReGlAnDrUm at earthlink.net  Wed Feb 14 06:09:37 2001
From: gReGlAnDrUm at earthlink.net (greg Landrum)
Date: Tue, 13 Feb 2001 21:09:37 -0800
Subject: A PythonCOM tidbit
Message-ID: <3A8A1311.7088EDB8@earthlink.net>

This is probably abundantly clear to everyone who actually thinks hard
about pythonCOM, but I just ran into problems with it, so I figure I'll
point it out...

I was feeling particularly uncreative whilst creating the names of my
source files for COM servers.  So I had two different COM servers Foo.Obj
and Bar.Obj implemented in files called COMServer.py (yeah, yeah, I was
*asking* for it) which lived in different directories (Foo and Bar).  These
were both registered and worked just fine.

Until I tried to use them in succession on Excel. Then I started getting
AttributeErrors.  I'd create an instance of Foo.Obj, work with it for a
bit, then try and create an instance of Bar.Obj.  BAM! AttributeError for
BarServer (the class defining Bar.Obj).  Or I could work with Bar.Obj for a
while and then try creating a Foo.Obj.  Instant AttributeError for
FooServer (the class defining Foo.Obj).

After spending a while trying to figure out what the heck was going on,
frantically flipping through PPW32 (once again: kudos to Mark and Andy for
this fantastic tome), and generally getting all worked up, it hit me... 
when I first create a Foo.Obj, pythoncom is importing a module called
COMServer from directory Foo. Later, when I try to create Bar.Obj,
pythoncom tries to import COMServer, but it must see that a module called
COMServer is already loaded, so it doesn't load the new one.

The solution to this problem is, of course, to not use names quite so
stupid as COMServer.

-greg


From shredwheat at mediaone.net  Wed Feb 14 07:09:55 2001
From: shredwheat at mediaone.net (Pete Shinners)
Date: Wed, 14 Feb 2001 06:09:55 GMT
Subject: ANN: pygame-0.9
Message-ID: <Tipi6.10103$_D.572878@typhoon.we.rr.com>

Pygame, Python Game Development.
(A wrapper for the SDL and SDL daughter libraries)


Pygame 0.9 released on Feb 12th, 2001
Pygame homepage:  http://pygame.seul.org
SDL homepage:     http://www.libsdl.org
Maintainer:       Pete Shinners (pete at shinners.org)


Pygame is a set of python modules written to help create games in Python.
The package is wrapped around SDL and several of the SDL daughter libraries.
SDL is a cross-platform media library created by Loki Games. SDL is currently
used for many commercial and opensource game projects. This pygame binding
takes SDL's C interface and turns it into a more pythonic library with
objects, exceptions, and some python utility modules.

This version 0.9, solidifies the package, as well as adding new features
like line drawing, defineable cursors, and shared image data. At this point
the library api has settled into a solid state. The coming plans call
for a move to 1.0 within a couple of weeks. Now is a great time to examine
the project and send in feedback.


A tally list of pygame features (note, many are also SDL features);

    Cross platform, with multiple backends (x11, fbcon, svgalib, etc)
    Fullscreen or windowed accelerated graphics
    Image loading from popular formats; jpg, bmp, gif, png, etc
    Access image pixel data as Numeric Python arrays
    Blitting with colorkey and alpha support
    Multiple mixing sound channels
    Streaming music data; mp3, mod, wav, etc
    CDROM support with audio playback
    Joystick support
    Truetype font rendering, with antialiasing
    Multithreaded support


The pygame package also comes with complete documentation, tutorials,
and many playable examples. The pygame mailing list has an active
list of users, which makes getting started much more accessable.

Currently installable from source, RPM, or Windows Installer.

Pygame is distributed under the LGPL licese, which is the same
license used by many its dependencies. There are several types of
distributions for download; RPM, Precompiled Win32, and Source.





From jschmitt at vmlabs.com  Wed Feb 14 07:13:14 2001
From: jschmitt at vmlabs.com (John Schmitt)
Date: Tue, 13 Feb 2001 22:13:14 -0800
Subject: passing object methods to os.path.walk ?
Message-ID: <008F0A63472BD311AF9800104BCD102561CC85@minirex.vmlabs.com>

I did something like this:

class foo:
	def __init__(self):
		os.path.walk( self.stuff, self.slurp, self.otherstuff )
	def slurp(self):
		#some code here

and I got a 'TypeError: too many arguments; expected 3, got 4' message from
ntpath.py line 264 which reads

	func( arg, top, names )

I fixed this by moving the slurp() function out of class foo and made it a
function outside of any class.  I don't get what's going on here.  Is there
an entry in TFM for this?

BTW, Python 2.0 on Win2000

John



From porter at et.byu.edu  Wed Feb 14 07:15:20 2001
From: porter at et.byu.edu (C. Porter Bassett)
Date: Tue, 13 Feb 2001 23:15:20 -0700 (MST)
Subject: Python profiler
In-Reply-To: <008F0A63472BD311AF9800104BCD102561CC85@minirex.vmlabs.com>
Message-ID: <Pine.GHP.4.21.0102132313530.26849-100000@don.et.byu.edu>

In the profiler module, function sort_stats, what exactly is the
difference between "time" and "cumulative" as arguments?  I looked in the
docs, but it didn't explain it very well.  






From fredrik at pythonware.com  Wed Feb 14 12:07:33 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Feb 2001 11:07:33 GMT
Subject: passing object methods to os.path.walk ?
References: <mailman.982131193.13060.python-list@python.org>
Message-ID: <VFti6.13386$Qb7.2125493@newsb.telia.net>

John Schmitt wrote:
> I did something like this:
>
> class foo:
> def __init__(self):
> os.path.walk( self.stuff, self.slurp, self.otherstuff )
> def slurp(self):
> #some code here

does "something like" mean:

    def slurp(arg, top, names):
        # correct if slurp is a function, wrong if it's a method

or

    def slurp(self, arg, top, names):
        # wrong if slurp is a function, correct if it's a method

> and I got a 'TypeError: too many arguments; expected 3, got 4' message from
> ntpath.py line 264 which reads
>
> func( arg, top, names )

os.path.walk calls the callback with three arguments.  If you use a
bound instance method as callback, Python inserts the instance (self)
first in the argument list.

Cheers /F




From chris at collegewafer.com  Wed Feb 14 10:18:16 2001
From: chris at collegewafer.com (Chris Baker)
Date: Wed, 14 Feb 2001 04:18:16 -0500
Subject: Hi, Need 6"-8" Wafer Special! Si, SOI, ZnO, MEMS & More!
Message-ID: <200102140918.DAA22252@ionet.net>

Hi,

Let CollegeWafer.com take care of your wafer needs!  Save time and $$$ It's a 
FREE service to you!

Visit http://www.collegewafer.com or call 800-713-9375, Fax 888-832-0340 or email 
chris at collegewafer.com


We have AIN, Fused Silica, Pyrex, GaN, SiC, ZnO of course Si, SOI, Ultra-Thin 
Si, Ge, InP, ZnSe, ZnO, ZnS, GaP, Infrared and so much more, including:

Custom orders for Prime wafers, Test wafers, or Coin Roll wafers. 

5" N<100> 5-10 ohm-cm.   24-25.9mils
5" N <100> 10-20 ohm-cm. 24-26mils 
5" P<100> 5-10 ohm-cm.   24-25.9mils 
5" P<100> 10-20 ohm-cm.  24-25.9mils
6" N<100> 1-5 ohm-cm.    25-27.9mils 
6" N<100> 10-20 ohm-cm.  585-725?m  
6" N<100> 10-20 ohm-cm.  >650?m     
6" N<100> 20-30 ohm-cm.  625-725?m
6" N<100> 20-30 ohm-cm.  25-27.5mils
6" P<100> 1-10 ohm-cm.   625-725?m 
6" P<100> 10-20 ohm-cm.  625-725?m 
6" P<100> 2-15 ohm-cm.   625-725?m 
6" P<100> >30 ohm-cm.    25-27.5mils 
8" P<100> 1-10 ohm-cm.   675-775?m 
8" P<100> 10-20 ohm-cm.  675-775?m 
8" P<100> 10-20 ohm-cm.  700-750?m 
8" P<100> 10-20 ohm-cm.  700-750?m 
12" P <100> 1-100 ohm-cm.750-800?m 

Polishing and Reclaim:
We can polish test and prime wafers at better      than SEMI standard specifications.

Need Wafer inspection services?
Test your wafers for type, resistivity,                     thickness, TIR, STIR, 
TTV, FPD and cleanliness           analysis.

How about:
Cleaning, Lapping, Etching, Edge round,           graphic printouts (2D & 3D), 
Certificate of      Conformance, Certificate of Analysis,     Polishing, Reclaim 
Services, Laser-cutting and           Oxide Services 

Ask about In-House Oxide Services: 

Thermally grown oxide for 2"-8" wafers ranging in thickness from 500-100,000 
Angstroms.

http://www.collegewafer.com/contact_us/Remove/remove.html



11ada22











From latlong at css2.com  Wed Feb 14 10:34:31 2001
From: latlong at css2.com (LatLong)
Date: Wed, 14 Feb 2001 09:34:31 GMT
Subject: ANN: Latitude longitude database
Message-ID: <Hisi6.5549$hN2.1055418@news3.cableinet.net>

Latitude longitude database of over 2.8 million locations worldwide.

Download from:
http://www.css2.com/latlong.htm

Ideal for sales and marketing applications and travel related websites.
Includes sample code.





From un4e at rzstud1.rz.uni-karlsruhe.de  Wed Feb 14 10:51:07 2001
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 14 Feb 2001 09:51:07 GMT
Subject: Getting an instance's class name?
Message-ID: <96dkeb$fcj$1@news.rz.uni-karlsruhe.de>

Hi,

can anyone tell me how to obtain the name(s) of the class(es) that
constitute an instance's type?

Thanks in advance,
		--Bj?rn


From smigowsky at dspace.de  Wed Feb 14 11:08:41 2001
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Wed, 14 Feb 2001 11:08:41 +0100
Subject: AW: Getting an instance's class name?
Message-ID: <84257D042AA7D411B3F700D0B7DB9B7C13A93B@newsintern.dspace.de>

Hi,

try 


>>> class A:
... 	pass
... 
>>> C = A()
>>> C.__class__
<class __main__.A at eb27c8>
>>> C.__class__.__name__
'A'
>>> 

  Stefan

-----Urspr?ngliche Nachricht-----
Von: Bjoern Giesler [mailto:un4e at rzstud1.rz.uni-karlsruhe.de]
Gesendet: Mittwoch, 14. Februar 2001 10:51
An: python-list at python.org
Betreff: Getting an instance's class name?


Hi,

can anyone tell me how to obtain the name(s) of the class(es) that
constitute an instance's type?

Thanks in advance,
		--Bj?rn
-- 
http://mail.python.org/mailman/listinfo/python-list



From stephen_purcell at yahoo.com  Wed Feb 14 11:31:29 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 14 Feb 2001 11:31:29 +0100
Subject: Getting an instance's class name?
References: <96dkeb$fcj$1@news.rz.uni-karlsruhe.de>
Message-ID: <20010214113129.A9420@freedom.puma-ag.com>

Bjoern Giesler wrote:
> can anyone tell me how to obtain the name(s) of the class(es) that
> constitute an instance's type?
> 

With the '__class__' attribute of the instance, and the '__name__' attribute
of the class:

>>> p = Exception()
>>> p.__class__.__name__
'Exception'
>>> 

Note that you'll get an attribute error for objects that are not class
instances. For non-class instances, a handy way to get the type name is:

>>> p = 1
>>> type(p).__name__
'int'
>>>


-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From un4e at rzstud1.rz.uni-karlsruhe.de  Wed Feb 14 15:50:21 2001
From: un4e at rzstud1.rz.uni-karlsruhe.de (Bjoern Giesler)
Date: 14 Feb 2001 14:50:21 GMT
Subject: Getting an instance's class name?
References: <96dkeb$fcj$1@news.rz.uni-karlsruhe.de> <mailman.982146676.10265.python-list@python.org>
Message-ID: <96e5vd$d98$1@news.rz.uni-karlsruhe.de>

Hi,

thanks to all... sorry, that wasn't too smart of me :-)

Regards,
		--Bj?rn


From alberik at gmx.de  Wed Feb 14 11:08:19 2001
From: alberik at gmx.de (Thomas Off)
Date: Wed, 14 Feb 2001 11:08:19 +0100
Subject: PHP und remote files
Message-ID: <96dlf3$kms4b$1@ID-66991.news.dfncis.de>

Wie kann ich remote files, in denen wiederum PHP-Code steckt in meine
PHP-Files einbinden, so dass der Code aus dem remote file auch verarbeitet
wird?

MfG Thomas Off




From alberik at gmx.de  Wed Feb 14 11:10:29 2001
From: alberik at gmx.de (Thomas Off)
Date: Wed, 14 Feb 2001 11:10:29 +0100
Subject: PHP und remote files
References: <96dlf3$kms4b$1@ID-66991.news.dfncis.de>
Message-ID: <96dljn$kbuk0$1@ID-66991.news.dfncis.de>

Sorry, falsche Newsgroup!

Thomas Off

Thomas Off <alberik at gmx.de> schrieb in im Newsbeitrag:
96dlf3$kms4b$1 at ID-66991.news.dfncis.de...
> Wie kann ich remote files, in denen wiederum PHP-Code steckt in meine
> PHP-Files einbinden, so dass der Code aus dem remote file auch verarbeitet
> wird?
>
> MfG Thomas Off
>
>




From mix77 at usa.net  Wed Feb 14 11:58:40 2001
From: mix77 at usa.net (Mix)
Date: Wed, 14 Feb 2001 12:58:40 +0200
Subject: Image Manipulation
Message-ID: <3A8A64E0.EF55A2CA@usa.net>

How does one  rotate an image (like  gif,jpg,etc) using Python (cgi).
Also,
how can one resize an image? What I am trying to do is load an image
in a browser (Netscape, say) and the allow the user to rotate and resize

the image to fit in a window so as print it.






From rtrocca at libero.it  Thu Feb 15 08:26:26 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Thu, 15 Feb 2001 07:26:26 GMT
Subject: Image Manipulation
References: <3A8A64E0.EF55A2CA@usa.net>
Message-ID: <1eou9rm.1l3ednv1uee4e8N%rtrocca@libero.it>

Try using PIL from pythonware. URL should be www.pythonware.com

Mix <mix77 at usa.net> wrote:

> How does one  rotate an image (like  gif,jpg,etc) using Python (cgi).
> Also,
> how can one resize an image? What I am trying to do is load an image
> in a browser (Netscape, say) and the allow the user to rotate and resize
> 
> the image to fit in a window so as print it.


From bk at xk7.com  Wed Feb 14 12:12:20 2001
From: bk at xk7.com (Burkhard Kloss)
Date: Wed, 14 Feb 2001 11:12:20 -0000
Subject: Type Hierarchies
Message-ID: <982149142.9745@master.nyc.kbcfp.com>

Is there a way to check whether an object is a sequence? I'm not interested
in the specific type, just that it qualifies as a sequence, so it could be
built-in or user defined sequence. The language spec talks about type
hierarchies, but if there's any way to access that concept in code I must
have missed it.

Obviously I can try a call, and catch the exception:

def is_seq(a):
    try:
        len(a)
        return 1
    except:
        return 0

which works:

assert is_seq( [] )
assert is_seq( () )
assert is_seq( "" )
assert not is_seq( 1 )

but isn't very elegant.  It's probably also not very fast...

Am I missing something obvious here? Are there better ways?

Thanks,

    Burkhard






From stephen_purcell at yahoo.com  Wed Feb 14 12:40:02 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 14 Feb 2001 12:40:02 +0100
Subject: Type Hierarchies
In-Reply-To: <982149142.9745@master.nyc.kbcfp.com>; from bk@xk7.com on Wed, Feb 14, 2001 at 11:12:20AM -0000
References: <982149142.9745@master.nyc.kbcfp.com>
Message-ID: <20010214124002.A13418@freedom.puma-ag.com>

Burkhard Kloss wrote:
> Is there a way to check whether an object is a sequence? I'm not interested
> in the specific type, just that it qualifies as a sequence, so it could be
> built-in or user defined sequence. The language spec talks about type
> hierarchies, but if there's any way to access that concept in code I must
> have missed it.

There's no real way to do this. As I spotted at the end of one of Tim Peters'
posts on this topic,

 "in-Python-what-you-want-to-do-with-an-object-is-more-important-than-
  what-you-call-it"


> Obviously I can try a call, and catch the exception:
> 
> def is_seq(a):
>     try:
>         len(a)
>         return 1
>     except:
>         return 0

This kind of thing is probably the best that you can hope for. The most
pythonic way (shocking though it may sound to some) is to throw the object
you hope is a sequence into the function that will use it, and test that in
ordinary use of your system the function always works correctly.

-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From erno-news at erno.iki.fi  Wed Feb 14 13:13:06 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 14 Feb 2001 14:13:06 +0200
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com>
Message-ID: <kuitmdwjhp.fsf@lasipalatsi.fi>

In article <982149142.9745 at master.nyc.kbcfp.com>, "Burkhard Kloss"
<bk at xk7.com> writes:

| def is_seq(a):
|     try:
|         len(a)
|         return 1
|     except:
|         return 0

len() also works on mappings.

  -- erno


From sholden at holdenweb.com  Wed Feb 14 13:30:00 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 14 Feb 2001 07:30:00 -0500
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com>
Message-ID: <6Xui6.9193$GA5.125501@e420r-atl1.usenetserver.com>

"Burkhard Kloss" <bk at xk7.com> wrote in message
news:982149142.9745 at master.nyc.kbcfp.com...
> Is there a way to check whether an object is a sequence? I'm not
interested
> in the specific type, just that it qualifies as a sequence, so it could be
> built-in or user defined sequence. The language spec talks about type
> hierarchies, but if there's any way to access that concept in code I must
> have missed it.
>
> Obviously I can try a call, and catch the exception:
>
> def is_seq(a):
>     try:
>         len(a)
>         return 1
>     except:
>         return 0
>
> which works:
>
> assert is_seq( [] )
> assert is_seq( () )
> assert is_seq( "" )
> assert not is_seq( 1 )
>
> but isn't very elegant.  It's probably also not very fast...
>
> Am I missing something obvious here? Are there better ways?
>
> Thanks,
>
>     Burkhard
>
Well, this might cause you to think again:

>>> tbl = {1:'one', 2:'two'}
>>> len(tbl)
2

Sorry.

regards
 Steve




From teep at inet.co.th  Wed Feb 14 14:02:46 2001
From: teep at inet.co.th (Prateep Siamwalla)
Date: Wed, 14 Feb 2001 20:02:46 +0700
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com>
Message-ID: <96dvfs$rbq$1@news.inet.co.th>

Hi, after browsing through D.Beazley's Python Essential Reference (p 20), I
believe this is what you are looking for ... (apologies if I'm way off)

import types

def am_i_a_sequence(ob):
    return type(ob) in
[types.StringType,types.ListType,types.TupleType,types.XRangeType]

hope this helps,
['t','e','e','p']



Burkhard Kloss <bk at xk7.com> wrote in message
news:982149142.9745 at master.nyc.kbcfp.com...
> Is there a way to check whether an object is a sequence? I'm not
interested
> in the specific type, just that it qualifies as a sequence, so it could be
> built-in or user defined sequence. The language spec talks about type
> hierarchies, but if there's any way to access that concept in code I must
> have missed it.
>
> Obviously I can try a call, and catch the exception:
>
> def is_seq(a):
>     try:
>         len(a)
>         return 1
>     except:
>         return 0
>
> which works:
>
> assert is_seq( [] )
> assert is_seq( () )
> assert is_seq( "" )
> assert not is_seq( 1 )
>
> but isn't very elegant.  It's probably also not very fast...
>
> Am I missing something obvious here? Are there better ways?
>
> Thanks,
>
>     Burkhard
>
>
>
>




From bk at xk7.com  Wed Feb 14 16:35:51 2001
From: bk at xk7.com (Burkhard Kloss)
Date: Wed, 14 Feb 2001 15:35:51 -0000
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com> <96dvfs$rbq$1@news.inet.co.th>
Message-ID: <982164952.565065@master.nyc.kbcfp.com>

Thanks to all the people who reminded me that len() also works on mappings
...

> def am_i_a_sequence(ob):
>     return type(ob) in
> [types.StringType,types.ListType,types.TupleType,types.XRangeType]

Yes, this works great for the built-in types - but not for user defined
types, which is specifically what I'm after.

> ['t','e','e','p']
:)

I guess concretely, the problem I'm trying to solve is iterating over a
sequence which may contain sequences or items, e.g.

[1, [2,3], 4, 5, [6,7,8,9]]

and I was hoping to write something along the lines of

for item in sequence:
    if is_sequence (item):
        for it in sequence:
            process (it)
    else:
        process (item)

Which IMHO looks marginally cleaner than

for item in sequence:
    try:
        for it in sequence:
            process (it)
    except:
        process (item)

Although I guess what I should really do is

for item in flatten(sequence):
    process (item)

Hmmm.... now how do I effectively flatten a sequence ?




From cjc26 at nospam.cornell.edu  Wed Feb 14 21:26:55 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Wed, 14 Feb 2001 20:26:55 GMT
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com> <96dvfs$rbq$1@news.inet.co.th> <982164952.565065@master.nyc.kbcfp.com>
Message-ID: <jSBi6.134946$o91.14037937@typhoon.nyroc.rr.com>

* Burkhard Kloss <bk at xk7.com> menulis:
| 
| I guess concretely, the problem I'm trying to solve is iterating over a
| sequence which may contain sequences or items, e.g.
| 
| [1, [2,3], 4, 5, [6,7,8,9]]
| 
| and I was hoping to write something along the lines of
| 
| for item in sequence:
|     if is_sequence (item):
|         for it in sequence:
|             process (it)
|     else:
|         process (item)

Well, 'for' uses the __getitem__ method to iterate through the items of
its sequence, so what you could do is check if the object has a
__getitem__ method:

for item in sequence:
    if hasattr(item, "__getitem__"):
        for it in item:
            process(it)
    else:
        process(item)

You'll run into problems though if the object implements a __getitem__
which doesn't ever throw an IndexError, because then the loop will go on
forever.

Also, I just tried this out on built-in sequences (lists and tuples),
and apparently they don't have __getitem__()...?


| Which IMHO looks marginally cleaner than
| 
| for item in sequence:
|     try:
|         for it in sequence:
|             process (it)
|     except:
|         process (item)

This is kind of unsafe, because the "except:" will catch any exception,
not just TypeError (which is what gets raised when you try to iterate
over a non-sequence).


-- 
Cliff Crawford                 http://www.people.cornell.edu/pages/cjc26/
"huh?  Are you all there?  You missing some key parts up there?  Becoming  
a English major or an minor?   Do you need a annual report?  Are you slow
today, eh?  Did someone sever you parietal lobe?" - Mark


From me at nospam.net  Thu Feb 15 07:13:34 2001
From: me at nospam.net (Scottie)
Date: Wed, 14 Feb 2001 22:13:34 -0800
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com> <96dvfs$rbq$1@news.inet.co.th> <982164952.565065@master.nyc.kbcfp.com>
Message-ID: <t8mssnkmq16n28@corp.supernews.com>

Well, I'd suggest you think of what you mean by "flatten".  It has the
same problem that something like "deepCopy" does: who knows when
you cross the boundary between interface and implementation.
If you have a list of numbers, and it turns out that my implementation
of rationals responds to "len," do you intend to iterate ove the interior
of my implementation?  How about a new version of "long" which
allows access to the individual 16-bit elements of its representation?

I guess what I am saying is that "recurse through sequences and
sequence-like objects may be a surprisingly ill-defined operation.

-Scott David Daniels
Scott.Daniels at Acm.Org

"Burkhard Kloss" <bk at xk7.com> wrote in message
news:982164952.565065 at master.nyc.kbcfp.com...
> Thanks to all the people who reminded me that len() also works on mappings
> ...
>
> > def am_i_a_sequence(ob):
> >     return type(ob) in
> > [types.StringType,types.ListType,types.TupleType,types.XRangeType]
>
> Yes, this works great for the built-in types - but not for user defined
> types, which is specifically what I'm after.
>
> > ['t','e','e','p']
> :)
>
> I guess concretely, the problem I'm trying to solve is iterating over a
> sequence which may contain sequences or items, e.g.
>
> [1, [2,3], 4, 5, [6,7,8,9]]
>
> and I was hoping to write something along the lines of
>
> for item in sequence:
>     if is_sequence (item):
>         for it in sequence:
>             process (it)
>     else:
>         process (item)
>
> Which IMHO looks marginally cleaner than
>
> for item in sequence:
>     try:
>         for it in sequence:
>             process (it)
>     except:
>         process (item)
>
> Although I guess what I should really do is
>
> for item in flatten(sequence):
>     process (item)
>
> Hmmm.... now how do I effectively flatten a sequence ?
>
>




From bk at xk7.com  Thu Feb 15 19:53:58 2001
From: bk at xk7.com (Burkhard Kloss)
Date: Thu, 15 Feb 2001 18:53:58 -0000
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com> <96dvfs$rbq$1@news.inet.co.th> <982164952.565065@master.nyc.kbcfp.com> <t8mssnkmq16n28@corp.supernews.com>
Message-ID: <982263239.671435@master.nyc.kbcfp.com>

> I guess what I am saying is that "recurse through sequences and
> sequence-like objects may be a surprisingly ill-defined operation.
In general, yes - but thankfully in my particular application context its is
well defined.  Thanks for pointingit out though.




From porter at et.byu.edu  Mon Feb 26 05:29:56 2001
From: porter at et.byu.edu (C. Porter Bassett)
Date: Sun, 25 Feb 2001 21:29:56 -0700 (MST)
Subject: Recursive Directory Contents
In-Reply-To: <t8mssnkmq16n28@corp.supernews.com>
References: <982149142.9745@master.nyc.kbcfp.com> <96dvfs$rbq$1@news.inet.co.th> <982164952.565065@master.nyc.kbcfp.com> <t8mssnkmq16n28@corp.supernews.com>
Message-ID: <Pine.GHP.4.21.0102252128140.29016-100000@don.et.byu.edu>

Is there a common module out there that will recursively record the
contents of a directory?  I want to use it to do tests to see what files
have been modified since last check.






From brian at sweetapp.com  Mon Feb 26 06:44:34 2001
From: brian at sweetapp.com (Brian Quinlan)
Date: Sun, 25 Feb 2001 21:44:34 -0800
Subject: Recursive Directory Contents
In-Reply-To: <Pine.GHP.4.21.0102252128140.29016-100000@don.et.byu.edu>
Message-ID: <HPEJKHHCMJHPMNLDAPLMAEOCCCAA.brian@sweetapp.com>

Try looking at the "walk" function in "os.path". 

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of C. Porter Bassett
Sent: Sunday, February 25, 2001 8:30 PM
To: python-list at python.org
Cc: python-list at python.org
Subject: Recursive Directory Contents


Is there a common module out there that will recursively record the
contents of a directory?  I want to use it to do tests to see what files
have been modified since last check.



-- 
http://mail.python.org/mailman/listinfo/python-list




From f8dy at yahoo.com  Mon Feb 26 17:24:48 2001
From: f8dy at yahoo.com (Mark Pilgrim)
Date: Mon, 26 Feb 2001 11:24:48 -0500
Subject: Recursive Directory Contents
References: <982149142.9745@master.nyc.kbcfp.com> <96dvfs$rbq$1@news.inet.co.th> <982164952.565065@master.nyc.kbcfp.com> <t8mssnkmq16n28@corp.supernews.com> <mailman.983161805.496.python-list@python.org>
Message-ID: <97dvtb$ovqli$1@ID-77331.news.dfncis.de>

The os.path module has a wonderful method called "walk", which takes a root
directory, a function, and an arbitrary argument.  It recursively walks
through the root directory, calling the function with 3 parameters: <your
arbitrary argument>, <current directory>, <list of files in current
directory>.  For example, this script will find all the files anywhere
within the "/music" directory that have been modified since Feb 15, 2001,
store them (with fully qualified paths) in modifiedFilesWithPaths, and then
print them.

-----

import os
import time

def mtimeChecker(arg, dirname, fileList):
  (mtime, modifiedList) = arg
  for f in [os.path.join(dirname, file) for file in fileList]:
    if os.path.getmtime(f) > mtime:
      modifiedList.append(f)

modifiedFilesWithPaths = []
os.path.walk('/music', mtimeChecker, \
  (time.mktime((2001, 2, 15, 0, 0, 0, 0, 0, 0)), \
  modifiedFilesWithPaths))
for f in modifiedFilesWithPaths:
  print f

-----

-M
You're smart; why haven't you learned Python yet?
http://diveintopython.org/

"C. Porter Bassett" <porter at et.byu.edu> wrote in message
news:mailman.983161805.496.python-list at python.org...
> Is there a common module out there that will recursively record the
> contents of a directory?  I want to use it to do tests to see what files
> have been modified since last check.
>
>
>
>




From andrew at andrewcooke.free-online.co.uk  Wed Feb 14 15:32:37 2001
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Wed, 14 Feb 2001 14:32:37 +0000
Subject: Type Hierarchies
References: <982149142.9745@master.nyc.kbcfp.com>
Message-ID: <3A8A9705.7F10944B@andrewcooke.free-online.co.uk>

Often you can avoid testing types by making sure that all types that you
use implement whatever you require.  Rather than, for example, testing
for the type of something so that you can infer whether or not to use
len, it might be easier to include something that implements __len__
somewhere in the superclasses fo whatever you are manipulating.  DOn't
forget that Python has multiple inheritance, so this isn't hard (a lot
easier than Java!).

I don't know if it helps in your case, but it's a pretty standard way to
avoid tests with OO languages.  Sorry if I'm saying the obvious.

Andrew

Burkhard Kloss wrote:
> 
> Is there a way to check whether an object is a sequence? I'm not interested
> in the specific type, just that it qualifies as a sequence, so it could be
> built-in or user defined sequence. The language spec talks about type
> hierarchies, but if there's any way to access that concept in code I must
> have missed it.
> 
> Obviously I can try a call, and catch the exception:
> 
> def is_seq(a):
>     try:
>         len(a)
>         return 1
>     except:
>         return 0
> 
> which works:
> 
> assert is_seq( [] )
> assert is_seq( () )
> assert is_seq( "" )
> assert not is_seq( 1 )
> 
> but isn't very elegant.  It's probably also not very fast...
> 
> Am I missing something obvious here? Are there better ways?
> 
> Thanks,
> 
>     Burkhard


From bk at xk7.com  Wed Feb 14 12:20:16 2001
From: bk at xk7.com (Burkhard Kloss)
Date: Wed, 14 Feb 2001 11:20:16 -0000
Subject: overloading on type?
Message-ID: <982149618.95754@master.nyc.kbcfp.com>

One design problem I run across frequently in Python is how to nicely deal
with different types of arguments.  For example, writing a date class you'd
often have overloaded constructors in a statically typed language like C++:

class date
{
public:
    date (const date &);
    date (const string &);
    date (long);

in python, I catch myself writing things along the lines of

class date:
    def __init__ (self, arg):
        if type(arg) == types.InstanceType:
            ....
        elif type(arg) == types.IntType:
            ...

which works, but just seems so *wrong*.

Maybe I'm just having a conceptual block here as a result of my upbringing
(I'm a longtime and unrepentant C++ user and lover), but surely in a
language as nice as python there must be a better way to do this? :)
Overloading obviously doesn't fit into the conceptual framework of the
language.

Thoughts, anyone?

    Burkhard






From stephen_purcell at yahoo.com  Wed Feb 14 12:47:10 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 14 Feb 2001 12:47:10 +0100
Subject: overloading on type?
In-Reply-To: <982149618.95754@master.nyc.kbcfp.com>; from bk@xk7.com on Wed, Feb 14, 2001 at 11:20:16AM -0000
References: <982149618.95754@master.nyc.kbcfp.com>
Message-ID: <20010214124710.B13418@freedom.puma-ag.com>

Burkhard Kloss wrote:
> in python, I catch myself writing things along the lines of
> 
> class date:
>     def __init__ (self, arg):
>         if type(arg) == types.InstanceType:
>             ....
>         elif type(arg) == types.IntType:
>             ...
> 
> which works, but just seems so *wrong*.


There are cases where this sort of type check is the best way to make a
clear API. A solution might be:

  class Date:
     def __init__(self, millis): # construct with int or long
	 self.millis = millis

  def parse_date(str):
      millis = ...
      return Date(millis)

This is quite common in the standard libraries. There are many module-level
functions that 'return an object of type ...', and the mechanism for
constructing such objects directly is deliberately obscured.

Modules form a convenient namespace for factory methods.

-Steve
         
-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From oconnor at bioreason.com  Wed Feb 14 16:53:58 2001
From: oconnor at bioreason.com (Jay O'COnnor)
Date: Wed, 14 Feb 2001 08:53:58 -0700
Subject: overloading on type?
References: <982149618.95754@master.nyc.kbcfp.com>
Message-ID: <3A8AAA14.EF207BA7@bioreason.com>


Burkhard Kloss wrote:

> One design problem I run across frequently in Python is how to nicely deal
> with different types of arguments.  For example, writing a date class you'd
> often have overloaded constructors in a statically typed language like C++:
>
> class date
> {
> public:
>     date (const date &);
>     date (const string &);
>     date (long);
>
> in python, I catch myself writing things along the lines of
>
> class date:
>     def __init__ (self, arg):
>         if type(arg) == types.InstanceType:
>             ....
>         elif type(arg) == types.IntType:
>             ...
>
> which works, but just seems so *wrong*.
>
> Maybe I'm just having a conceptual block here as a result of my upbringing
> (I'm a longtime and unrepentant C++ user and lover), but surely in a
> language as nice as python there must be a better way to do this? :)
> Overloading obviously doesn't fit into the conceptual framework of the
> language.

One approach you can use is called "Double Dispatch".  This will allow you to
resolve to a specific method from a general method based on the type of the
argument; even in dynamically bound languages susch as Python and Smalltalk.

The approach is to get the argument invovled in the decision as well. Your
receiver object calls back to the argument object telling it to do something
with itself.  The argument object, in turn, calls back to the first object with
a specific message based on it's own type (which it already knows)

Here's an example using a simple DrawingContext and Shape hierarchy

                         #!/usr/bin/python
                         #covarient.py

                         class DrawingContext:
                                 def drawShape (self, aShape):
                                         aShape.drawOn(self)

                                 def drawCircle (self, aCircle):
                                         print "Drawing circle at " +
aCircle.printLocation()

                                 def drawSquare (self, aSquare):
                                         print "Drawing square at " +
aSquare.printLocation()

                         class Shape:
                                 def __init__(self, x=0, y=0):
                                         self.x = x
                                         self.y = y

                                 def printLocation (self):
                                         return "X = " + `self.x` + " Y= " +
`self.y`

                                 def drawOn(self, aDrawingContext):
                                         #Abstract method
                                         pass

                         class Circle(Shape):

                                 def __init__ (self, x=0, y=0, radius=0):
                                         Shape.__init__(self, x,y)
                                         self.radius = radius

                                 def printLocation (self):
                                         return Shape.printLocation(self) + "
Radius = " + `self.radius`

                                 def drawOn (self, aDrawingContext):
                                         aDrawingContext.drawCircle(self)

                         class Square(Shape):

                                 def __init__ (self, x=0, y=0, sideLength=0):
                                         Shape.__init__(self, x,y)
                                         self.sideLength = sideLength

                                 def printLocation (self):
                                         return Shape.printLocation(self) + "
Side Length = " + `self.sideLength`

                                 def drawOn (self, aDrawingContext):
                                         aDrawingContext.drawSquare(self)

                         # Test script starts here

                         dc = DrawingContext()
                         circle = Circle (x=100,y=200,radius=5)
                         square = Square (x=10,y=150, sideLength=10)

                         dc.drawShape( circle)
                         dc.drawShape (square)

and here's the output

                          [jay at altaica misc_junk]$ ./covarient.py
                         Drawing circle at X = 100 Y= 200 Radius = 5
                         Drawing square at X = 10 Y= 150 Side Length = 10

(Note: it's called covarient.py because I took it from some sample code I
posted in another message forum about covarients in Java)

The way it works is that the DrawingContext object must draw the shape, but it
draws it in a different manner based on what kind of shape it is.   So it tells
the given shape to draw on itself (the canvas)  The shape then calls back to
the DrawingContext with the message for the specific drawing operation
(drawCircle(), etc)

This works very well at resolving from an abstract interface to a specific
operation based on the tyhpes of both objects.

It does, however, require that you be able to add code to both partners of the
operation.  If this is not possible, I usually use an approach of building a
Dictionary   where the keys are the type and the values are the methods to
call.  I thjink this looks a little cleaner than having  an 'if' conditional.

        import types

        class Date:
                def __init__(self, fromObject)
                    initializer =initOperations[type(fromObject)]
                    initializer (self, fromObject)

                def initializeFromInt(self, anInt):
                    pass
                def initializeFromInstance (self, anInstance):
                    pass
        initOperations = {
            types.IntType:Date.initializeFromInt,
            types.InstanceType: Date.initializeFromInstance}

Note: I haven't tested that, but you get the idea.

I consider this second approach a not very good one and only appropriate if I
need to make such a decisision either using primitive types or third party
objects that I can't add the douple dispatch support to


Take care,

Jay O'Connor
joconor at cybermesa.com

Python Language Discussion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython



From froydnj at rose-hulman.edu  Thu Feb 15 06:20:46 2001
From: froydnj at rose-hulman.edu (Nathan Froyd)
Date: 15 Feb 2001 05:20:46 GMT
Subject: overloading on type?
References: <982149618.95754@master.nyc.kbcfp.com>
Message-ID: <slrn98mpg3.ecu.froydnj@froyd3.laptop.rose-hulman.edu>

In article <982149618.95754 at master.nyc.kbcfp.com>, Burkhard Kloss wrote:
>Maybe I'm just having a conceptual block here as a result of my upbringing
>(I'm a longtime and unrepentant C++ user and lover), but surely in a
>language as nice as python there must be a better way to do this? :)
>Overloading obviously doesn't fit into the conceptual framework of the
>language.

You might try Multimethod.py:

http://www.sff.net/people/neelk/open-source/

It might not do *exactly* what you want (I don't know if you can
specialize on built-in datatypes and I don't know what the speed is
like), but it's probably a good starting point.
-- 
</nathan>  froydnj at rose-hulman.edu  |  http://www.rose-hulman.edu/~froydnj/

Yes, God had a deadline.  So He wrote it all in Lisp.


From neelk at alum.mit.edu  Fri Feb 16 17:11:03 2001
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 16 Feb 2001 16:11:03 GMT
Subject: overloading on type?
References: <982149618.95754@master.nyc.kbcfp.com> <slrn98mpg3.ecu.froydnj@froyd3.laptop.rose-hulman.edu>
Message-ID: <slrn98qo2g.4b8.neelk@alum.mit.edu>

On 15 Feb 2001 05:20:46 GMT, Nathan Froyd <froydnj at rose-hulman.edu> wrote:
>In article <982149618.95754 at master.nyc.kbcfp.com>, Burkhard Kloss wrote:
>>Maybe I'm just having a conceptual block here as a result of my upbringing
>>(I'm a longtime and unrepentant C++ user and lover), but surely in a
>>language as nice as python there must be a better way to do this? :)
>>Overloading obviously doesn't fit into the conceptual framework of the
>>language.
>
>You might try Multimethod.py:
>
>http://www.sff.net/people/neelk/open-source/
>
> It might not do *exactly* what you want (I don't know if you can
> specialize on built-in datatypes and I don't know what the speed is
> like), but it's probably a good starting point.

Yes, you can specialize on C types. Unlike C++ overloading, this does
real multiple dispatch -- the overloading is resolved based on the
*runtime* class of the methods, not the static type. Thes performance
overhead is about what you'd expect from wrapping something with a
__call__ method.


Neel


From gregm at iname.com  Wed Feb 14 12:54:18 2001
From: gregm at iname.com (Greg McFarlane)
Date: Wed, 14 Feb 2001 22:54:18 +1100
Subject: ANNOUNCE: Pmw megawidgets 0.8.5
Message-ID: <E14T0Vm-0003tx-00@holiday>

			Pmw megawidgets

			   Pmw.0.8.5

    * Python 2.0 release
    * Pmw moved to SourceForge

    Pmw new home:   http://pmw.sourceforge.net/
    Pmw maintainer: Greg McFarlane <gregm at iname.com>


A new release of Pmw is out.  This is mainly a bug fix release
for python 2.0, fixing a problem with Pmw.Counter caused by a
change in the ascii representation of a long (no 'L').

Also, the home for Pmw development has moved to SourceForge.
The Pmw sources and documentation are under CVS control and
releases can be downloaded from there.

I would like to thank Graham Dumpleton for allowing Pmw to
be hosted at his web site for so long.  I hope SourceForge
will match the stability of Pmw's old site.

Other changes in this release are:

  - Minor fixes to tests for Tk 8.3.

  - Fixed bug in Pmw.ScrolledFrame when given invalid flex options.

  - Added pmw2 font scheme, since the font used for balloon text with
    pmw1 is too small on Linux.

  - Removed syntax coloring from code window in demos.  It did not
    look good and the pattern matching was not always correct.

  - Changed font size used for demos to 12 for Unix, since 14 looked
    too big under Linux.

For more information on Pmw, see the (new) home page at
  http://pmw.sourceforge.net/

You can fetch the entire distribution from
  http://download.sourceforge.net/pmw/Pmw.0.8.5.tar.gz

If you have any comments, enhancements or new contributions, please
contact me (gregm at iname.com).

=====================================================================

What is Pmw?

    Pmw is a GUI toolkit for building high-level compound widgets in
    Python using the Tkinter module.  These compound widgets, called
    'megawidgets', are constructed using other Tkinter widgets or Pmw
    megawidgets as component parts.  It promotes consistent look and
    feel within and between graphical applications, is highly
    configurable and is easy to use.

    Pmw consists of:

	- A few base classes, providing a foundation for building
	  megawidgets.

	- A library of flexible and extensible megawidgets built on
	  the base classes, such as buttonboxes, notebooks,  
	  comboboxes, selection widgets, paned widgets, scrolled
	  widgets and dialog windows.

	- A lazy importer/dynamic loader which is automatically
	  invoked when Pmw is first imported.  This gives unified
	  access to all Pmw classes and functions through the Pmw.
	  prefix.  It also speeds up module loading time by only
	  importing Pmw sub-modules when needed.

	- Reference documentation, consisting of complete listings of
	  megawidget options, methods and components.  Full
	  descriptions are also available for all the base classes and
	  several other megawidget classes.  Descriptions of the other
	  megawidgets will be released soon.

	- A test framework and tests for Pmw megawidgets.

	- A slick demonstration of the megawidgets.

	- An interface to the BLT busy, graph and vector commands.

    The interface to Pmw megawidgets is similar to basic Tk widgets,
    so it is easy for developers to include both megawidgets and basic
    Tk widgets in their graphical applications.  In addition, Pmw
    megawidgets may themselves be extended, using either inheritance
    or composition.

    The use of the Pmw megawidgets replaces common widget combinations
    with higher level abstractions.  This simplifies code, making it
    more readable and maintainable.  The ability to extend Pmw
    megawidgets enables developers to create new megawidgets based on
    previous work.

-- 
Greg McFarlane  Really Good Software Pty Ltd  Sydney Australia  gregm at iname.com



From rytrom at zoote.com  Wed Feb 14 13:13:22 2001
From: rytrom at zoote.com (Roman Rytov)
Date: Wed, 14 Feb 2001 14:13:22 +0200
Subject: Difference btw JPython and Jython?
Message-ID: <newscache$j4yq8g$kca$1@lnews.actcom.co.il>

Is there any difference btw them or it has as much common things as Java and
JavaScript:-))




From sholden at holdenweb.com  Wed Feb 14 14:12:47 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 14 Feb 2001 08:12:47 -0500
Subject: Difference btw JPython and Jython?
References: <newscache$j4yq8g$kca$1@lnews.actcom.co.il>
Message-ID: <dzvi6.9279$GA5.129069@e420r-atl1.usenetserver.com>

"Roman Rytov" <rytrom at zoote.com> wrote in message
news:newscache$j4yq8g$kca$1 at lnews.actcom.co.il...
> Is there any difference btw them or it has as much common things as Java
and
> JavaScript:-))
>
Jython is the name for the current implementation of what used to be
JPython.

This latter name is trademarked by CNRI, Guido's former employer.

regards
 Steve




From hamish_lawson at yahoo.co.uk  Wed Feb 14 13:44:59 2001
From: hamish_lawson at yahoo.co.uk (=?iso-8859-1?q?Hamish=20Lawson?=)
Date: Wed, 14 Feb 2001 12:44:59 +0000 (GMT)
Subject: Suggestion: Python tools book
Message-ID: <20010214124459.18461.qmail@web217.mail.yahoo.com>

I'd welcome a book that covered the more popular third-party toolkits
for Python, plus those toolkits (*) in the standard library that could
use more coverage than given in the library reference. Perhaps the
various toolkit authors could contribute a respective chapter. Does the
community think this kind of book is needed, and what toolkits should
be included? Here's my stab at a list of candidates:

   wxPython
   DB-API
   Python Imaging Library
   Numerical Python
   mxDateTime
   mod_python / mod_snake
   HTMLgen
   mxTextTools
   ReportLab
   PyXML*
   TkInter*

Hamish Lawson


=====
Hamish Lawson  hamish_lawson at yahoo.co.uk

____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From sandipan at vsnl.com  Wed Feb 14 14:17:32 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Wed, 14 Feb 2001 18:47:32 +0530
Subject: Suggestion: Python tools book
References: <20010214124459.18461.qmail@web217.mail.yahoo.com>
Message-ID: <005101c09688$7dd1e440$020000c3@node02>

Hear, hear !

Also:
SSL socket stuff
freeze/packaging

Sandipan

----- Original Message ----- 
From: "Hamish Lawson" <hamish_lawson at yahoo.co.uk>
To: <python-list at python.org>
Sent: Wednesday, February 14, 2001 6:14 PM
Subject: Suggestion: Python tools book


> I'd welcome a book that covered the more popular third-party toolkits
> for Python, plus those toolkits (*) in the standard library that could
> use more coverage than given in the library reference. Perhaps the
> various toolkit authors could contribute a respective chapter. Does the
> community think this kind of book is needed, and what toolkits should
> be included? Here's my stab at a list of candidates:
> 
>    wxPython
>    DB-API
>    Python Imaging Library
>    Numerical Python
>    mxDateTime
>    mod_python / mod_snake
>    HTMLgen
>    mxTextTools
>    ReportLab
>    PyXML*
>    TkInter*
> 
> Hamish Lawson
> 
> 
> =====
> Hamish Lawson  hamish_lawson at yahoo.co.uk
> 
> ____________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
> or your free @yahoo.ie address at http://mail.yahoo.ie
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From ljohnson at resgen.com  Wed Feb 14 15:53:36 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Wed, 14 Feb 2001 08:53:36 -0600
Subject: Suggestion: Python tools book
References: <mailman.982154801.10484.python-list@python.org>
Message-ID: <t8l6r1gfeks1ba@corp.supernews.com>

Hamish,

There's a new book out from SAMS called "Python Developer's Handbook", by
Andre Lessa. It appears to cover a lot of these topics that you mentioned
and so you might want to check it out. I personally have not read this one
and so this isn't an endorsement of the book!

Lyle

"Hamish Lawson" <hamish_lawson at yahoo.co.uk> wrote in message
news:mailman.982154801.10484.python-list at python.org...
> I'd welcome a book that covered the more popular third-party toolkits
> for Python, plus those toolkits (*) in the standard library that could
> use more coverage than given in the library reference. Perhaps the
> various toolkit authors could contribute a respective chapter. Does the
> community think this kind of book is needed, and what toolkits should
> be included? Here's my stab at a list of candidates:
>
>    wxPython
>    DB-API
>    Python Imaging Library
>    Numerical Python
>    mxDateTime
>    mod_python / mod_snake
>    HTMLgen
>    mxTextTools
>    ReportLab
>    PyXML*
>    TkInter*
>
> Hamish Lawson
>
>
> =====
> Hamish Lawson  hamish_lawson at yahoo.co.uk
>
> ____________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
> or your free @yahoo.ie address at http://mail.yahoo.ie
>




From pj at sgi.com  Wed Feb 14 21:01:46 2001
From: pj at sgi.com (Paul Jackson)
Date: 14 Feb 2001 20:01:46 GMT
Subject: Suggestion: Python tools book
References: <mailman.982154801.10484.python-list@python.org> <t8l6r1gfeks1ba@corp.supernews.com>
Message-ID: <96eo7a$43422$1@fido.engr.sgi.com>

|> There's a new book out from SAMS called "Python Developer's Handbook", by
|> Andre Lessa. It appears to cover a lot of these topics that you mentioned

Read the reviews of this book on amazon.com before buying it.
I for one found the book rather lightweight (well, heavy by the
pound, light by the technical depth), and ended up returning
my copy.

Some of the Amazon reviews are more positive than mine, but
all the reviews seem to agree on what are the strengths and
weaknesses of this book, so you should get a good idea of
whether it fits your needs.

-- 
                          I won't rest till it's the best ...
                          Manager, Linux System Software
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From rickylee at americasm01.nt.com  Wed Feb 14 14:30:19 2001
From: rickylee at americasm01.nt.com (Lee, Rick)
Date: Wed, 14 Feb 2001 08:30:19 -0500
Subject: StringType add operation seems every inefficient
Message-ID: <3A8A886B.BB7329F6@americasm01.nt.com>

If I do something like this:

s = ''
for i in listOfStrings:
    s = s + i

Once listOfStrings is say 10000 members of 100 bytes, it takes forever
(ie. minutes) for the "for" loop to finish.  This is running on a 700MHz
WinNT machine, Python 2.0

Contrast with:

s = ''
blocks = []
for i in listOfStrings:
    s = s + i
    if len(s) > 4000
        blocks.append(s); s = ''

This takes a blink of the eye to finish, so it is faster than the first
piece of code by between 100 and 1000 times.

Is this relative inefficiency in the string add operation expected?

I eventually have to produce a big long string of approx. 10MB.  What is
the fastest way to do it without having to write C-extensions?

- Rick Lee







From jeremy at alum.mit.edu  Tue Feb 13 16:00:14 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Tue, 13 Feb 2001 10:00:14 -0500 (EST)
Subject: StringType add operation seems every inefficient
In-Reply-To: <3A8A886B.BB7329F6@americasm01.nt.com>
References: <3A8A886B.BB7329F6@americasm01.nt.com>
Message-ID: <14985.19454.873485.617135@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "RL" == Lee, Rick <rickylee at americasm01.nt.com> writes:

  RL> If I do something like this: 
  RL> s = '' 
  RL> for i in listOfStrings:
  RL>     s = s + i

  RL> Once listOfStrings is say 10000 members of 100 bytes, it takes
  RL> forever (ie. minutes) for the "for" loop to finish.  This is
  RL> running on a 700MHz WinNT machine, Python 2.0

  RL> Contrast with:

  RL> s = '' 
  RL> blocks = [] 
  RL> for i in listOfStrings:
  RL>     s = s + i if len(s) > 4000
  RL>         blocks.append(s); s = ''

  RL> This takes a blink of the eye to finish, so it is faster than
  RL> the first piece of code by between 100 and 1000 times.

  RL> Is this relative inefficiency in the string add operation
  RL> expected?

It is expected, because the string add version is doing something
quite different than accumulator version.  The 's = s + i' line is
making a new string for each iteration of the loop.  To make the new
string, the interpreter allocates a new string object and copies the
contents of s and i into it.

It may help you to see what is happening if you unroll the loop by
hand: 

    s0 = ''
    s1 = s0 + listOfStrings[0]
    s2 = s1 + listOfStrings[1]
#     ~= s0 + listOfStrings[0] + listOfStrings[1]
    s3 = s2 + listOfStrings[2]
#     ~= s0 + listOfStrings[0] + listOfStrings[1] + listOfStrings[2]

In the end, that's O(N) string allocations and O(N**2) string copies,
where N is the length of listOfStrings.

  RL> I eventually have to produce a big long string of approx. 10MB.
  RL> What is the fastest way to do it without having to write
  RL> C-extensions?

The second technique you show -- using a list to accumulate the
individual strings -- is a standard Python idiom.  At the end, you do
"".join(blocks) or somesuch to produce a single string.  This approach
does exactly one copy for each constituent string.

I believe someone also suggested using cStringIO, which is probably
just as fast.  

Jeremy




From jhauser at ifm.uni-kiel.de  Wed Feb 14 14:46:11 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 14 Feb 2001 14:46:11 +0100
Subject: StringType add operation seems every inefficient
References: <3A8A886B.BB7329F6@americasm01.nt.com>
Message-ID: <873ddh1ios.fsf@lisboa.ifm.uni-kiel.de>

"Lee, Rick" <rickylee at americasm01.nt.com> writes:

> If I do something like this:
> 
> s = ''
> for i in listOfStrings:
>     s = s + i
> 
This always generates a new string object, so there is a lot of memory
copies involved.

> Once listOfStrings is say 10000 members of 100 bytes, it takes forever
> (ie. minutes) for the "for" loop to finish.  This is running on a 700MHz
> WinNT machine, Python 2.0
> 
> Contrast with:
> 
> s = ''
> blocks = []
> for i in listOfStrings:
>     s = s + i
>     if len(s) > 4000
>         blocks.append(s); s = ''
> 
if you have the string parts already in a list and then do a
string.join(listOfStrings) this should be much faster. (In Python2.0 this
should be ''.join(list))

HTH,
__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From fuess at att.net  Wed Feb 14 14:48:31 2001
From: fuess at att.net (David Fuess)
Date: Wed, 14 Feb 2001 13:48:31 GMT
Subject: StringType add operation seems every inefficient
References: <3A8A886B.BB7329F6@americasm01.nt.com>
Message-ID: <ts2l8tsi30ko1lqdghll57fnlac03bdci2@4ax.com>

Don't really know if it was expected, but the presence of the StringIO
module in both Python and "C" versions indicates that it was.

For really fast performance in string concatenation, use 
cStringIO, for a somewhat slower but more feature rich implementation
use StringIO.

Dave

On Wed, 14 Feb 2001 08:30:19 -0500, "Lee, Rick"
<rickylee at americasm01.nt.com> wrote:

>If I do something like this:
>
>s = ''
>for i in listOfStrings:
>    s = s + i
>
>Once listOfStrings is say 10000 members of 100 bytes, it takes forever
>(ie. minutes) for the "for" loop to finish.  This is running on a 700MHz
>WinNT machine, Python 2.0
>
>Contrast with:
>
>s = ''
>blocks = []
>for i in listOfStrings:
>    s = s + i
>    if len(s) > 4000
>        blocks.append(s); s = ''
>
>This takes a blink of the eye to finish, so it is faster than the first
>piece of code by between 100 and 1000 times.
>
>Is this relative inefficiency in the string add operation expected?
>
>I eventually have to produce a big long string of approx. 10MB.  What is
>the fastest way to do it without having to write C-extensions?
>
>- Rick Lee
>
>
>
>



From rwklee at home.com  Wed Feb 14 23:29:26 2001
From: rwklee at home.com (Rick Lee)
Date: Wed, 14 Feb 2001 22:29:26 GMT
Subject: StringType add operation seems every inefficient
References: <3A8A886B.BB7329F6@americasm01.nt.com>
Message-ID: <3A8B06C6.281511E3@home.com>

Thank you all for your suggestions.  I will try out both join() and
cStringIO.

"Lee, Rick" wrote:

> If I do something like this:
>
> s = ''
> for i in listOfStrings:
>     s = s + i
>
> Once listOfStrings is say 10000 members of 100 bytes, it takes forever
> (ie. minutes) for the "for" loop to finish.  This is running on a 700MHz
> WinNT machine, Python 2.0
>
> Contrast with:
>
> s = ''
> blocks = []
> for i in listOfStrings:
>     s = s + i
>     if len(s) > 4000
>         blocks.append(s); s = ''
>
> This takes a blink of the eye to finish, so it is faster than the first
> piece of code by between 100 and 1000 times.
>
> Is this relative inefficiency in the string add operation expected?
>
> I eventually have to produce a big long string of approx. 10MB.  What is
> the fastest way to do it without having to write C-extensions?
>
> - Rick Lee



From rytrom at zootec.com  Wed Feb 14 15:26:14 2001
From: rytrom at zootec.com (Roman Rytov)
Date: Wed, 14 Feb 2001 16:26:14 +0200
Subject: Developer's tools?
Message-ID: <newscache$1a4r8g$xng$1@lnews.actcom.co.il>

Is there any recommended enviroment for developing in Jython or Python that
allows tracing, debuging and doing all regular programmer's stuff?





From fredrik at pythonware.com  Wed Feb 14 16:43:20 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 14 Feb 2001 15:43:20 GMT
Subject: Developer's tools?
References: <newscache$1a4r8g$xng$1@lnews.actcom.co.il>
Message-ID: <sIxi6.13396$Qb7.2127394@newsb.telia.net>

Roman Rytov wrote:
> Is there any recommended enviroment for developing in Jython or Python that
> allows tracing, debuging and doing all regular programmer's stuff?

start here:
http://www.faqts.com/knowledge_base/view.phtml/aid/6433/fid/199

Cheers /F




From trentm at ActiveState.com  Wed Feb 14 20:12:22 2001
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 14 Feb 2001 11:12:22 -0800
Subject: Developer's tools?
In-Reply-To: <sIxi6.13396$Qb7.2127394@newsb.telia.net>; from fredrik@pythonware.com on Wed, Feb 14, 2001 at 03:43:20PM +0000
References: <newscache$1a4r8g$xng$1@lnews.actcom.co.il> <sIxi6.13396$Qb7.2127394@newsb.telia.net>
Message-ID: <20010214111222.J19947@ActiveState.com>

On Wed, Feb 14, 2001 at 03:43:20PM +0000, Fredrik Lundh wrote:
> Roman Rytov wrote:
> > Is there any recommended enviroment for developing in Jython or Python that
> > allows tracing, debuging and doing all regular programmer's stuff?
> 
> start here:
> http://www.faqts.com/knowledge_base/view.phtml/aid/6433/fid/199
> 


You can also add Komodo to that list. This is a cross-platform scripting
language IDE based on Mozilla:

    http://www.activestate.com/Products/Komodo/index.html

Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com



From bill.xu at newlix.com  Wed Feb 14 15:43:02 2001
From: bill.xu at newlix.com (Bill Xu)
Date: Wed, 14 Feb 2001 14:43:02 GMT
Subject: Anybody know about pulldom in PyXML?
Message-ID: <WPwi6.5804$HG.1720272@news4.rdc1.on.home.com>

Hi, Gang:

Will u please tell me how to use the pulldom.startElement() in the Python
XML DOM (PyXml 0.6.3, Python 1.5, Redhat Linux 6.2) ?
 Following is my simple test program and result.
 Will u please tell me how to correct this bug?
"AttributeError: 'None' object has no attribute 'createElement'";

Thanks in advance!

Bill

=====================================================
#!/usr/bin/python
#

import os, sys, string, StringIO
import xml.dom
from xml.dom import minidom
from xml.dom.pulldom import PullDOM
import traceback



# Save the new message
def test_meth():
  print "Am i here"
  builder = PullDOM()
  builder.startElement("SET", {"PATH" : "a path"})
  builder.endElement("SET")
  print "Am i here"
  confirm(builder)
 #result = self.Print(builder)

def confirm(test, testname = "Test"):
    if test:
        print "Passed " + testname
    else:
        print "Failed " + testname
        raise Exception

print "? am i herer"
test_meth()

=======================================================
Output:

? am i herer
Am i here
Traceback (innermost last):
  File "./testdom.py", line 34, in ?
    test_meth()
  File "./testdom.py", line 20, in test_meth
    builder.startElement("SET", {"PATH" : "a path"})
  File "/usr/lib/python1.5/site-packages/xml/dom/pulldom.py", line 85, in
startElement
    node = self.document.createElement(name)
AttributeError: 'None' object has no attribute 'createElement'






From dan.rolander at marriott.com  Wed Feb 14 15:56:34 2001
From: dan.rolander at marriott.com (Dan Rolander)
Date: Wed, 14 Feb 2001 09:56:34 -0500
Subject: Help! Problems with win32wnet.WNetAddConnection2
Message-ID: <004801c09696$53636ea0$11260340@yin>

I am using the following function to break any existing connections between
a local drive and a network share and create a new one. It works just fine
on Windows 2000, but on Windows 98 both WNetCancelConnection2 and
WNetAddConnection2 return the following pywintypes.api_error:

     'The network request is not supported.'

This should work on Windows 98 shouldn't it? What am I doing wrong?

def map(drive, unc, username=None, password=None, persistent='no'):
    if persistent == 'yes':
        flags = win32netcon.CONNECT_UPDATE_PROFILE
    else:
        flags = 0
    drive_type = win32file.GetDriveType(drive)
    if drive_type == win32con.DRIVE_REMOTE:
        win32wnet.WNetCancelConnection2(drive, flags, 1)
    win32wnet.WNetAddConnection2(
        win32netcon.RESOURCETYPE_DISK,
        drive,
        unc,
        None,
        username,
        password,
        flags
        )






From aahz at panix.com  Wed Feb 14 16:15:47 2001
From: aahz at panix.com (Aahz Maruch)
Date: 14 Feb 2001 07:15:47 -0800
Subject: Google?
Message-ID: <96e7f3$b75$1@panix6.panix.com>

Anybody here work for Google (or know somebody who does)?  I'm thinking
of applying for a job there, but I'd like to know something about the
working conditions.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

'Gender' isn't a definition, it's a heuristic.  --Aahz


From olk at olk.co.kr  Wed Feb 14 16:21:37 2001
From: olk at olk.co.kr (Online Korea)
Date: Thu, 15 Feb 2001 00:21:37 +0900
Subject: =?ks_c_5601-1987?B?cHl0aG9uLWxpc3S01CC+yLPnx8+8vL/kPyA=?=
Message-ID: <3W-POP3-ADBDUfdfdhN00000d39@3w-pop3-ad.korea.com>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010215/d79a9af3/attachment.htm>

From zope at thewebsons.com  Wed Feb 14 16:36:57 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Wed, 14 Feb 2001 07:36:57 -0800
Subject: Link Exchange Robot
Message-ID: <5.0.2.1.0.20010214073626.00a05280@thewebsons.com>

Hi;
There's a great little proggie I use to scout out potential link exchange 
partners for my clients called Zeus.
<disclaimer>
I have no financial interest in Zeus.
</disclaimer>
*However*, there are some facilities in this proggie I don't care for: 
facilities designed to slow one down (and presumably keep the company from 
being known as a developer of devices for spammers, not that I advocate 
spamming). I also have a Linux box with a T-1 line that could kick serious 
butt, and this (of course) is built for Windoze. That, plus the fact that I 
now find I need a whole bunch of these robots and I'd like to legally skate 
on the licencing fee :) Which (finally) brings me to my question: Does 
anyone know of an open source program that does the same thing?
TIA,
BenO




From bshapiro at funnygarbage.com  Wed Feb 14 17:03:44 2001
From: bshapiro at funnygarbage.com (Ben Shapiro)
Date: Wed, 14 Feb 2001 11:03:44 -0500
Subject: COM 'Catastrophic failure' ????
Message-ID: <96ea87$b2d$1@bob.news.rcn.net>

Anyone know what could be causing the 'Catastrophic failure'??
I am using an OCX as the media control

thx for your help

C:\code\3LP\PyFileunder>mediaplayer.py
Traceback (most recent call last):
  File "C:\code\3LP\PyFileunder\MediaPlayer.py", line 39, in ?
    t = MediaPlayer()
  File "C:\code\3LP\PyFileunder\MediaPlayer.py", line 22, in __init__
    p = w.Open("c:\\code\\3LP\\pyfileunder\\music\\test.mp3")
  File
"c:\python20\win32com\gen_py\2DF09BC5-B050-11D2-81C1-00C0DFEAA961x0x1x0.py",
line 75, in Open
    return self._oleobj_.InvokeTypes(0x33, LCID, 1, (3, 0), ((8, 0),),URL)
pywintypes.com_error: (-2147418113, 'Catastrophic failure', None, None)




From MarkH at ActiveState.com  Wed Feb 14 23:17:21 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 14 Feb 2001 22:17:21 GMT
Subject: COM 'Catastrophic failure' ????
References: <96ea87$b2d$1@bob.news.rcn.net>
Message-ID: <3A8B0312.1080906@ActiveState.com>

Ben Shapiro wrote:

> Anyone know what could be causing the 'Catastrophic failure'??
> I am using an OCX as the media control
> 
> thx for your help
> 
> C:\code\3LP\PyFileunder>mediaplayer.py
> Traceback (most recent call last):
>   File "C:\code\3LP\PyFileunder\MediaPlayer.py", line 39, in ?
>     t = MediaPlayer()
>   File "C:\code\3LP\PyFileunder\MediaPlayer.py", line 22, in __init__
>     p = w.Open("c:\\code\\3LP\\pyfileunder\\music\\test.mp3")
>   File
> "c:\python20\win32com\gen_py\2DF09BC5-B050-11D2-81C1-00C0DFEAA961x0x1x0.py",
> line 75, in Open
>     return self._oleobj_.InvokeTypes(0x33, LCID, 1, (3, 0), ((8, 0),),URL)
> pywintypes.com_error: (-2147418113, 'Catastrophic failure', None, None)

I would guess that the media player wants a real OCX container. 
win32com doesnt support that.  Pythonwin does tho - see the 
pywin\demos\ocx samples and see if that helps it work.

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From lutz at rmi.net  Wed Feb 14 17:07:00 2001
From: lutz at rmi.net (Mark Lutz)
Date: Wed, 14 Feb 2001 09:07:00 -0700
Subject: Python training next week, Denver
Message-ID: <008601c096a0$2ff51ae0$dc73fea9@vaio>

There are still some seats available in the upcoming 
Kaivo Python training session described here:

http://www.kaivo.com

Kaivo classes are open to the public, and accept
individual and group enrollments.

--Mark Lutz  (http://www.rmi.net/~lutz)




From phd at phd.russ.ru  Wed Feb 14 17:20:33 2001
From: phd at phd.russ.ru (Oleg Broytmann)
Date: 14 Feb 2001 10:20:33 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Feb 14)
Message-ID: <63B38717DB69B0D6.7FB13CCE0B8ED7BA.330DC4E9A51B2B9B@lp.airnews.net>

    Google.com buys Deja archives. All links change!
        http://groups.google.com/groups?q=group:comp.lang.python&start=30&hl=en&lr=&safe=off&rnum=35&seld=926663091&ic=1

    Suchandra Thapa starts a long thread on creating a CPAN-like
    service for Python
        http://groups.google.com/groups?hl=en&lr=&safe=off&th=6202938a19f67d9b&thn=CPAN+functionality+for+python&ic=1&seekd=927216290

    Steven D. Arnold announces the release of pure-Python encryption tools
        http://groups.google.com/groups?q=group:comp.lang.python&start=20&hl=en&lr=&safe=off&rnum=30&seld=926624071&ic=1

    Barry Pederson releases a Python-powered QuakeWorld dedicated server
        http://groups.google.com/groups?hl=en&lr=&safe=off&ic=1&th=c01ded3a01dfc324&thn=QuakeWorld+Python+1.0&seekd=926538529

    In a thread dedicated to boolean COM properties Mark Hammond
    explains PyTrue and PyFalse
        http://groups.google.com/groups?hl=en&lr=&safe=off&ic=1&th=1cd0ed3943dead44&thn=Setting+boolean+COM+properties&seekd=926499672


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org
      Notice especially the master FAQ
          http://www.python.org/doc/FAQ.html

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
          http://deja.com/group/comp.lang.python.announce

      Andrew Kuchling writes marvelous summaries twice a month of
      the action on the python-dev mailing list, where the future
      of Python is truly determined
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collect Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m  Wed Feb 14 17:41:01 2001
From: k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m (Keith Ray)
Date: Wed, 14 Feb 2001 16:41:01 GMT
Subject: python interpreter in smalltalk
References: <9579gr$g0jjg$1@ID-65212.news.dfncis.de> <V1Ud6.568$nD1.12282@nnrp1.sbc.net> <95abs4$g1nic$1@ID-65212.news.dfncis.de> <95coka$hp3$1@nnrp1.deja.com> <k1e2i3t4h5r6a7y-84C34A.18574802022001@news> <7MGJOov0nvx+VW3xYgkmsmMz2tDQ@4ax.com>
Message-ID: <k1e2i3t4h5r6a7y-53B089.08410214022001@news>

In article <7MGJOov0nvx+VW3xYgkmsmMz2tDQ at 4ax.com>, sma at 3plus4.de wrote:

> On Sat, 03 Feb 2001 02:57:36 GMT, Keith Ray
> <k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote:
> 
> >Has anyone tried to write a Python interpreter in Smalltalk?
> 
> I tried this, although using Squeak Smalltalk not Dolphin Smalltalk.
> I never finished the project but there's a partially working parser
> which I could share.
> 
> bye
> --
> Stefan Matthias Aust____Truth until Paradox!____________________
>      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de

I have the desire to see Python running inside Squeak 
<http://www.squeak.org/>.

I was planning to start porting from the Jython sources, since those are 
already in an object oriented language (Java).  If you could, please 
email your parser to me and it will help me learn how Squeak could host 
a Python language. 

C. Keith Ray
remove spaces to get my email address
  c k e i t h r a y @ h o m e . c o m

-- 
       <http://homepage.mac.com/keithray/resume.html>


From marten at phoenix-edv.netzservice.de  Thu Feb 15 08:55:24 2001
From: marten at phoenix-edv.netzservice.de (Marten Feldtmann)
Date: Thu, 15 Feb 2001 08:55:24 +0100
Subject: python interpreter in smalltalk
References: <9579gr$g0jjg$1@ID-65212.news.dfncis.de> <V1Ud6.568$nD1.12282@nnrp1.sbc.net> <95abs4$g1nic$1@ID-65212.news.dfncis.de> <95coka$hp3$1@nnrp1.deja.com> <k1e2i3t4h5r6a7y-84C34A.18574802022001@news> <7MGJOov0nvx+VW3xYgkmsmMz2tDQ@4ax.com> <k1e2i3t4h5r6a7y-53B089.08410214022001@news>
Message-ID: <3A8B8B6C.A19DC08D@phoenix-edv.netzservice.de>


Keith Ray schrieb:
> 
> In article <7MGJOov0nvx+VW3xYgkmsmMz2tDQ at 4ax.com>, sma at 3plus4.de wrote:
> 
> > On Sat, 03 Feb 2001 02:57:36 GMT, Keith Ray
> > <k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote:
> >
> > >Has anyone tried to write a Python interpreter in Smalltalk?
> >
> > I tried this, although using Squeak Smalltalk not Dolphin Smalltalk.
> > I never finished the project but there's a partially working parser
> > which I could share.
> >
> > bye
> > --
> > Stefan Matthias Aust____Truth until Paradox!____________________
> >      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de
> 
> I have the desire to see Python running inside Squeak
> <http://www.squeak.org/>.
> 

 Why not call Python via it's C interface ?

 Marten


From k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m  Thu Feb 15 18:37:56 2001
From: k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m (Keith Ray)
Date: Thu, 15 Feb 2001 17:37:56 GMT
Subject: python interpreter in smalltalk
References: <9579gr$g0jjg$1@ID-65212.news.dfncis.de> <V1Ud6.568$nD1.12282@nnrp1.sbc.net> <95abs4$g1nic$1@ID-65212.news.dfncis.de> <95coka$hp3$1@nnrp1.deja.com> <k1e2i3t4h5r6a7y-84C34A.18574802022001@news> <7MGJOov0nvx+VW3xYgkmsmMz2tDQ@4ax.com> <k1e2i3t4h5r6a7y-53B089.08410214022001@news> <3A8B8B6C.A19DC08D@phoenix-edv.netzservice.de>
Message-ID: <k1e2i3t4h5r6a7y-32D413.09375615022001@news>

In article <3A8B8B6C.A19DC08D at phoenix-edv.netzservice.de>, Marten 
Feldtmann <marten at phoenix-edv.netzservice.de> wrote:

> Keith Ray schrieb:
> > 
> > In article <7MGJOov0nvx+VW3xYgkmsmMz2tDQ at 4ax.com>, sma at 3plus4.de wrote:
> > 
> > > On Sat, 03 Feb 2001 02:57:36 GMT, Keith Ray
> > > <k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote:
> > >
> > > >Has anyone tried to write a Python interpreter in Smalltalk?
> > >
> > > I tried this, although using Squeak Smalltalk not Dolphin Smalltalk.
> > > I never finished the project but there's a partially working parser
> > > which I could share.
> > >
> > > bye
> > > --
> > > Stefan Matthias Aust____Truth until Paradox!____________________
> > >      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de
> > 
> > I have the desire to see Python running inside Squeak
> > <http://www.squeak.org/>.
> > 
> 
>  Why not call Python via it's C interface ?

Because that wouldn't allow writing Python subclasses of Smalltalk 
classes and vice versa. Both Python and Squeak use bytecodes, and run in 
a kind of VM. Squeak can call out to C code, but C code can't call into 
Squeak -- the hooks are not there (yet).

Squeak happens to be implemented "in itself" -- that is, its virtual 
machine was written and debugged in a subset of normal Smalltalk (I 
think they call it 'Slang'). Then a translator (also written in 
Smalltalk) converts the Slang code into C code, which is compiled and 
becomes the real virtual machine. The original Slang-based virtual 
machine can also be executed in Squeak -- so you have an "emulated" 
virtual machine running (very slowly) inside a real virtual machine.

If Squeak can be made to handle multiple sets of byte-codes types, then 
it could transparently handle Python, Java, and Smalltalk in its 
"universal VM". 

Or the JPython route -- implement Python within the host language/vm... 
Jython allows interoperating with Java the way I want Python to 
interoperate with Squeak Smalltalk.

Squeak can reconstuct the source code of a method from the byte-code 
(does not include comments and probably not the local variable names), 
it is possible that a combined Squeak/Python vm could let you choose to 
view a method in Smalltalk syntax or in Python syntax.

Squeak has been ported to many platforms, including some that don't have 
an operating system (ported to the "bare metal"). It has really good 
MacOS support (my preferred platform).

-- 
       <http://homepage.mac.com/keithray/resume.html>


From k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m  Fri Feb 16 03:32:23 2001
From: k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m (Keith Ray)
Date: Fri, 16 Feb 2001 02:32:23 GMT
Subject: python interpreter in smalltalk
References: <9579gr$g0jjg$1@ID-65212.news.dfncis.de> <V1Ud6.568$nD1.12282@nnrp1.sbc.net> <95abs4$g1nic$1@ID-65212.news.dfncis.de> <95coka$hp3$1@nnrp1.deja.com> <k1e2i3t4h5r6a7y-84C34A.18574802022001@news> <7MGJOov0nvx+VW3xYgkmsmMz2tDQ@4ax.com> <k1e2i3t4h5r6a7y-53B089.08410214022001@news> <3A8C189F.B3040253@canis.uiuc.edu>
Message-ID: <k1e2i3t4h5r6a7y-1BA723.18322115022001@news>

In article <3A8C189F.B3040253 at canis.uiuc.edu>, Les Tyrrell 
<tyrrell at canis.uiuc.edu> wrote:

> You might consider an approach similar to that used in
> Frost ( http://oasis.canis.uiuc.edu:8080/Frost ) where
> bytecodes for Java are translated to Smalltalk bytecodes.
> Frost also parses Java source code, but when it does it
> goes straight to VisualWorks bytecodes.
> 
> One of the things I'd like to consider for Frost would be
> accomodating languages other than Java, especially languages
> like Python or Ruby.  Not because I want to do Java, Python,
> or Ruby development, but because I do NOT want to do Java,
> Python, or Ruby development, but also do not want to have to
> manually re-implement any useful code written in those languages
> in Smalltalk.
> 
> It's also the sort of thing that hopefully puts positive pressures
> on the VM designers, perhaps along the lines mentioned in the
> "Universal VM" thread.
> 
> - les

Thanks for the URL... and it's open source!

-- 
       <http://homepage.mac.com/keithray/resume.html>


From sma at 3plus4.de  Sun Feb 18 09:33:11 2001
From: sma at 3plus4.de (Stefan Matthias Aust)
Date: Sun, 18 Feb 2001 09:33:11 +0100
Subject: python interpreter in smalltalk
References: <9579gr$g0jjg$1@ID-65212.news.dfncis.de> <V1Ud6.568$nD1.12282@nnrp1.sbc.net> <95abs4$g1nic$1@ID-65212.news.dfncis.de> <95coka$hp3$1@nnrp1.deja.com> <k1e2i3t4h5r6a7y-84C34A.18574802022001@news> <7MGJOov0nvx+VW3xYgkmsmMz2tDQ@4ax.com> <k1e2i3t4h5r6a7y-53B089.08410214022001@news>
Message-ID: <CYiPOj7XcM5XXhPxfIA3+zQuo=sQ@4ax.com>

On Wed, 14 Feb 2001 16:41:01 GMT, Keith Ray
<k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote:

>> I tried this, although using Squeak Smalltalk not Dolphin Smalltalk.
>> I never finished the project but there's a partially working parser
>> which I could share.
>> 
>> bye
>> --
>> Stefan Matthias Aust____Truth until Paradox!____________________
>>      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de
>
>I have the desire to see Python running inside Squeak 
><http://www.squeak.org/>.

Keith,

I've up my sources for an unfinished python parser to
www.3plus4.de/squeak.  Feel free to use is a foundation for a real
Python system.  I think, this wouldn't be that difficult.  When I
started, I mostly ignored the runtime system but reading the Python
specification I did find anything which couldn't implemented in
Smalltalk.

bye
--
Stefan Matthias Aust____Truth until Paradox!____________________
     Jobs? ==> jobs at baltic-online.de   www.baltic-online.de


From k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m  Mon Feb 19 03:10:47 2001
From: k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m (Keith Ray)
Date: Mon, 19 Feb 2001 02:10:47 GMT
Subject: python interpreter in smalltalk
References: <9579gr$g0jjg$1@ID-65212.news.dfncis.de> <V1Ud6.568$nD1.12282@nnrp1.sbc.net> <95abs4$g1nic$1@ID-65212.news.dfncis.de> <95coka$hp3$1@nnrp1.deja.com> <k1e2i3t4h5r6a7y-84C34A.18574802022001@news> <7MGJOov0nvx+VW3xYgkmsmMz2tDQ@4ax.com> <k1e2i3t4h5r6a7y-53B089.08410214022001@news> <CYiPOj7XcM5XXhPxfIA3+zQuo=sQ@4ax.com>
Message-ID: <k1e2i3t4h5r6a7y-E75930.18104818022001@news>

In article <CYiPOj7XcM5XXhPxfIA3+zQuo=sQ at 4ax.com>, sma at 3plus4.de wrote:

> On Wed, 14 Feb 2001 16:41:01 GMT, Keith Ray
> <k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote:
> 
> >> I tried this, although using Squeak Smalltalk not Dolphin Smalltalk.
> >> I never finished the project but there's a partially working parser
> >> which I could share.
> >> 
> >> bye
> >> --
> >> Stefan Matthias Aust____Truth until Paradox!____________________
> >>      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de
> >
> >I have the desire to see Python running inside Squeak 
> ><http://www.squeak.org/>.
> 
> Keith,
> 
> I've up my sources for an unfinished python parser to
> www.3plus4.de/squeak.  Feel free to use is a foundation for a real
> Python system.  I think, this wouldn't be that difficult.  When I
> started, I mostly ignored the runtime system but reading the Python
> specification I did find anything which couldn't implemented in
> Smalltalk.
> 
> bye
> --
> Stefan Matthias Aust____Truth until Paradox!____________________
>      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de

Thanks!

( By the way the link named "Python" shows up as the invalid link 
"<http://www.3plus4.de/squeak/www.python.org>". )

-- 
       <http://homepage.mac.com/keithray/resume.html>


From LauraL  Wed Feb 14 17:44:43 2001
From: LauraL (LauraL)
Date: 14 Feb 2001 16:44:43 GMT
Subject: Suggestion: Python tools book
References: <mailman.982156696.19387.python-list@python.org>
Message-ID: <96eclr01m74@news2.newsguy.com>

Hi,
You will find a lot of these tools covered in the new (2nd) edition of Programming Python, due out early March.  We are also in the planning stages on a Cookbook that will hit on these topics.
Laura
LLewin at oreilly.com

> Hear, hear !
> 
> Also:
> SSL socket stuff
> freeze/packaging
> 
> Sandipan
> 
> ----- Original Message ----- 
> From: "Hamish Lawson" <hamish_lawson at yahoo.co.uk>
> To: <python-list at python.org>
> Sent: Wednesday, February 14, 2001 6:14 PM
> Subject: Suggestion: Python tools book
> 
> 
> > I'd welcome a book that covered the more popular third-party toolkits
> > for Python, plus those toolkits (*) in the standard library that could
> > use more coverage than given in the library reference. Perhaps the
> > various toolkit authors could contribute a respective chapter. Does the
> > community think this kind of book is needed, and what toolkits should
> > be included? Here's my stab at a list of candidates:
> > 
> >    wxPython
> >    DB-API
> >    Python Imaging Library
> >    Numerical Python
> >    mxDateTime
> >    mod_python / mod_snake
> >    HTMLgen
> >    mxTextTools
> >    ReportLab
> >    PyXML*
> >    TkInter*
> > 
> > Hamish Lawson
> > 
> > 
> > =====
> > Hamish Lawson  hamish_lawson at yahoo.co.uk


==================================
Posted via http://nodevice.com
Linux Programmer's Site


From dvoitenko at qode.com  Wed Feb 14 17:51:18 2001
From: dvoitenko at qode.com (Voitenko, Denis)
Date: Wed, 14 Feb 2001 11:51:18 -0500
Subject: Working with DNS via Python
Message-ID: <0FF17853157CD31197E100902784B34B7A1F2E@FILESERVER1>

Could someone point me to some functions or material that'd explain how to
work with DNS functions (nslookup, etc) via Python?

I'd be interested in sumulating something like this:

Default Server:  localhost
Address:  127.0.0.1

> set q=mx
> sony.com
Server:  localhost
Address:  127.0.0.1

Non-authoritative answer:
sony.com        preference = 10, mail exchanger = mail5.fw-bc.sony.com
sony.com        preference = 10, mail exchanger = mail4.fw-sj.sony.com
sony.com        preference = 10, mail exchanger = mail4.fw-bc.sony.com

Denis

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010214/4c7fb2f0/attachment.htm>

From phd at phd.pp.ru  Wed Feb 14 17:58:29 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 14 Feb 2001 19:58:29 +0300 (MSK)
Subject: Working with DNS via Python
In-Reply-To: <0FF17853157CD31197E100902784B34B7A1F2E@FILESERVER1>
Message-ID: <Pine.LNX.4.30.0102141955310.10721-100000@fep132.fep.ru>

On Wed, 14 Feb 2001, Voitenko, Denis wrote:
> Could someone point me to some functions or material that'd explain how to
> work with DNS functions (nslookup, etc) via Python?

   The first and most strightforward idea - just use os.popen and run
nslookup! :)

   Also there is dnslib - it is not in standard library, it is in Demo/dns
subdirectory in source distribution.

   Also there is adns (asynchronous dns lib in C); Andy Dustman wrote
python wrapper for it.

   (Russian? Why not to ask question on Russian Python list?)

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From dvoitenko at qode.com  Wed Feb 14 18:13:46 2001
From: dvoitenko at qode.com (Voitenko, Denis)
Date: Wed, 14 Feb 2001 12:13:46 -0500
Subject: Working with DNS via Python
Message-ID: <0FF17853157CD31197E100902784B34B7A1F30@FILESERVER1>

>> work with DNS functions (nslookup, etc) via Python?
>The first and most strightforward idea - just use os.popen and run
nslookup! :)

If I wanted to do that I'd use Bash.


>Also there is dnslib - it is not in standard library, it is in Demo/dns
subdirectory in source distribution.

Is it documented anywhere?

>Also there is adns (asynchronous dns lib in C); Andy Dustman wrote
python wrapper for it.

Where do I get it?

>(Russian? Why not to ask question on Russian Python list?)

If I do, only Russians will be able to use my findings. Kinda makes it
unfair, eh?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010214/437d7075/attachment.htm>

From phd at phd.pp.ru  Wed Feb 14 18:14:57 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 14 Feb 2001 20:14:57 +0300 (MSK)
Subject: Working with DNS via Python
In-Reply-To: <0FF17853157CD31197E100902784B34B7A1F30@FILESERVER1>
Message-ID: <Pine.LNX.4.30.0102142013300.10721-100000@fep132.fep.ru>

On Wed, 14 Feb 2001, Voitenko, Denis wrote:
> >Also there is dnslib - it is not in standard library, it is in Demo/dns
> subdirectory in source distribution.
>
> Is it documented anywhere?

   No.

> >Also there is adns (asynchronous dns lib in C); Andy Dustman wrote
> python wrapper for it.
>
> Where do I get it?

   http://dustman.net/andy/python/

> >(Russian? Why not to ask question on Russian Python list?)
>
> If I do, only Russians will be able to use my findings. Kinda makes it
> unfair, eh?

   Absoluteley no. And this makes converstaion easy for people who don't
know English well.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From cg at schlund.de  Wed Feb 14 18:34:17 2001
From: cg at schlund.de (Carsten Gaebler)
Date: Wed, 14 Feb 2001 18:34:17 +0100
Subject: Working with DNS via Python
References: <mailman.982169356.5480.python-list@python.org>
Message-ID: <3A8AC199.50A4B5BB@schlund.de>

> 
> Could someone point me to some functions or material that'd explain how
> to work with DNS functions (nslookup, etc) via Python?
> 

I was looking for such functions today, too, and found this on Parnassus:

http://www.interlink.com.au/anthony/python/dns.tar.gz

It works. There is a bug in Lib.py: index.append() takes a 2-tuple
argument, not two single arguments.

cg.


From phd at phd.pp.ru  Wed Feb 14 18:46:12 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 14 Feb 2001 20:46:12 +0300 (MSK)
Subject: Working with DNS via Python
In-Reply-To: <3A8AC199.50A4B5BB@schlund.de>
Message-ID: <Pine.LNX.4.30.0102142045490.10721-100000@fep132.fep.ru>

On Wed, 14 Feb 2001, Carsten Gaebler wrote:
> It works. There is a bug in Lib.py: index.append() takes a 2-tuple
> argument, not two single arguments.

   It is not a bug - it's a change from 1.5.2 to 1.6.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From michael at stroeder.com  Thu Feb 15 07:26:07 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 15 Feb 2001 07:26:07 +0100
Subject: Working with DNS via Python
References: <mailman.982169356.5480.python-list@python.org> <3A8AC199.50A4B5BB@schlund.de>
Message-ID: <3A8B767F.BE8236EB@stroeder.com>

Carsten Gaebler wrote:
> 
> > Could someone point me to some functions or material that'd explain how
> > to work with DNS functions (nslookup, etc) via Python?
> 
> I was looking for such functions today, too, and found this on Parnassus:
> 
> http://www.interlink.com.au/anthony/python/dns.tar.gz

Note: I patched this to work under Python 1.6+ and I've added
support for SRV RR's.

I would like to see something like this in the standard lib. It
seems to be well designed. Even I understood very quickly how to add
a new RR type. ;-)

Ciao, Michael.


From rozen at rgelpc254.rgv.hp.com  Wed Feb 14 18:17:19 2001
From: rozen at rgelpc254.rgv.hp.com (Don Rozenberg)
Date: 14 Feb 2001 09:17:19 -0800
Subject: [ANN] PAGE 0.1 - A drag and drop GUI generator for
Message-ID: <80lhf1xchgg.fsf@rgelpc254.rgv.hp.com>

PAGE is a tool which helps to create GUI for Python programs.  It uses
Tk and Tix widgets.  PAGE is not an end-all, be-all tool, but rather
one that attempts to ease the burden on the Python programmer. It is
aimed at the user who will put up with a less-than-general GUI
capability in order to get an easily generated GUI. It does not build
an entire application but rather is aimed at building a single GUI
window.   PAGE is built on version 1.2.2 of Visual Tcl, version
8.1.1 of Tix, and Tcl/Tk version 8.3.2. It works with Python 2.0.
This is an early version which run on Linux and Windows/NT. It
supports many Tk widgets and a subset of the Tix widget set.  The user
will have to acquire and install Tix 8.1.1 and tcl8.3.2 to use this
program.
It is released under the GPL.

The PAGE web site is

 http://www.geocities.com/page_python

-- 
Don Rozenberg
rozen at rgv.hp.com
408-343-6266


From franze at cs.unibo.it  Wed Feb 14 18:32:25 2001
From: franze at cs.unibo.it (Francesco Franze)
Date: 14 Feb 2001 17:32:25 GMT
Subject: linking with python20.dll
Message-ID: <96eff9$o5d$1@le.cs.unibo.it>

i tried to link a new extension module for activepython (2.0)
it seems some symbols (like _Py_NoneStruct) are missing in python20.lib
i rebuilt the lib file and added the apparently missing symbols
the dll now is created (with cygnus gcc or ms visual studio c++)
linking with python15 generates a lib conflict...

any idea?

f.


From dsh8290 at rit.edu  Wed Feb 14 19:29:55 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 14 Feb 2001 13:29:55 -0500
Subject: pth information
Message-ID: <20010214132954.A17229@harmony.cs.rit.edu>

I was searching through the documentation on python.org, but couldn't
find anything describing the usage and syntax for .pth files.  All I
managed to find was a brief mention of their existence in some
extension modules in section 3.23 of the Library Reference.

Could someone point me in the right direction? (Or explain here if you
feel inclined to do so)

Thanks,
-D




From phd at phd.pp.ru  Wed Feb 14 19:38:30 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 14 Feb 2001 21:38:30 +0300 (MSK)
Subject: pth information
In-Reply-To: <20010214132954.A17229@harmony.cs.rit.edu>
Message-ID: <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru>

On Wed, 14 Feb 2001, D-Man wrote:
> I was searching through the documentation on python.org, but couldn't
> find anything describing the usage and syntax for .pth files.  All I

   http://www.python.org/doc/essays/packages.html

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From dsh8290 at rit.edu  Wed Feb 14 21:43:36 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 14 Feb 2001 15:43:36 -0500
Subject: pth information
In-Reply-To: <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru>; from phd@phd.pp.ru on Wed, Feb 14, 2001 at 09:38:30PM +0300
References: <20010214132954.A17229@harmony.cs.rit.edu> <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru>
Message-ID: <20010214154336.A17557@harmony.cs.rit.edu>

On Wed, Feb 14, 2001 at 09:38:30PM +0300, Oleg Broytmann wrote:
| On Wed, 14 Feb 2001, D-Man wrote:
| > I was searching through the documentation on python.org, but couldn't
| > find anything describing the usage and syntax for .pth files.  All I
| 
|    http://www.python.org/doc/essays/packages.html

Thanks!
-D

|            Programmers don't die, they just GOSUB without RETURN.

BTW,  I think this is kind of cute.




From dsh8290 at rit.edu  Wed Feb 14 22:20:46 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 14 Feb 2001 16:20:46 -0500
Subject: pth information
In-Reply-To: <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru>; from phd@phd.pp.ru on Wed, Feb 14, 2001 at 09:38:30PM +0300
References: <20010214132954.A17229@harmony.cs.rit.edu> <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru>
Message-ID: <20010214162046.A17719@harmony.cs.rit.edu>

On Wed, Feb 14, 2001 at 09:38:30PM +0300, Oleg Broytmann wrote:
| On Wed, 14 Feb 2001, D-Man wrote:
| > I was searching through the documentation on python.org, but couldn't
| > find anything describing the usage and syntax for .pth files.  All I
| 
|    http://www.python.org/doc/essays/packages.html

This did a good job of explaining how packages are supposed to work
and how packages differ and interact with modules.  It also explained
what .pth files are for and that .pth files shouldn't be used unless
absolutely necessary.

Ok, but I still want to know what such a file should look like.  For
example, on one of my systems, I have Win2k, Python 2.0, and wxPython
2.2.5.  Python is installed in d:\apps\python20 wxPython is in
d:\lib\python_modules\wxPython.  Right now I have
d:\lib\python_modules in my PYTHONPATH environment variable.  If I
want to use a wxPython.pth file instead, how would I do that?   Should
it go in d:\apps\python20\lib?  What should it contain?  Simply the
path to d:\lib\python_modules\wxPython on a line by itself?

Thanks,
-D




From tuttledon at hotmail.com  Wed Feb 14 23:02:50 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Wed, 14 Feb 2001 22:02:50 GMT
Subject: pth information
References: <20010214132954.A17229@harmony.cs.rit.edu> <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru> <mailman.982185680.5800.python-list@python.org>
Message-ID: <egDi6.75870$p8.19410163@typhoon.southeast.rr.com>

"D-Man" <dsh8290 at rit.edu> wrote in message
news:mailman.982185680.5800.python-list at python.org...
>
> Ok, but I still want to know what such a file should look like.  For
> example, on one of my systems, I have Win2k, Python 2.0, and wxPython
> 2.2.5.  Python is installed in d:\apps\python20 wxPython is in
> d:\lib\python_modules\wxPython.  Right now I have
> d:\lib\python_modules in my PYTHONPATH environment variable.  If I
> want to use a wxPython.pth file instead, how would I do that?   Should
> it go in d:\apps\python20\lib?  What should it contain?  Simply the
> path to d:\lib\python_modules\wxPython on a line by itself?

I was wondering about this just a couple of days ago.  From what I could
find out the .pth should contain a one line path to the folder where the
module resides.

In your case the "wxPython.pth" would contain this line:
d:\lib\python_modules\wxPython

If you had installed it in the \Python20 folder, all you would need is the
folder name (path relative to \Python20):
    wxPython.pth
    wxPython

Question:  Where does Python look first?  The PythonPath or for a .pth file?

Hope this helps,
Don






From paul.moore at uk.origin-it.com  Thu Feb 15 15:33:13 2001
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Thu, 15 Feb 2001 15:33:13 +0100
Subject: pth information
References: <20010214132954.A17229@harmony.cs.rit.edu> <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru> <mailman.982185680.5800.python-list@python.org> <egDi6.75870$p8.19410163@typhoon.southeast.rr.com>
Message-ID: <TuiLOsaCjHXUN6Br6IrUjYrDWJx9@4ax.com>

On Wed, 14 Feb 2001 23:02:50 +0100, "Don Tuttle"
<tuttledon at hotmail.com> wrote:

>Question:  Where does Python look first?  The PythonPath or for a .pth file?

The standard module site.py is loaded first. You can read the source
of this for the definitive answer, but basically, site.py searches the
sirectory sys.prefix (and maybe sys.exec_prefix) for files whose names
end in ".pth". All such files are read, and each line is treated as
the name of a directory (relative to the directory containing the .pth
file, if it isn't an absolute name already) to be added at the end of
sys.path.

All this happens at startup, and from then on, sys.path is just used
as normal.

Hope this helps,
Paul.



From cerutti at together.net  Thu Feb 15 16:48:06 2001
From: cerutti at together.net (Neil Cerutti)
Date: 15 Feb 2001 15:48:06 GMT
Subject: pth information
References: <20010214132954.A17229@harmony.cs.rit.edu> <Pine.LNX.4.30.0102142138130.10721-100000@fep132.fep.ru> <mailman.982185680.5800.python-list@python.org>
Message-ID: <slrn98nv0g.7s.cerutti@fiad06.norwich.edu>

D-Man posted:
>On Wed, Feb 14, 2001 at 09:38:30PM +0300, Oleg Broytmann wrote:
>| On Wed, 14 Feb 2001, D-Man wrote:
>| > I was searching through the documentation on python.org, but
>couldn't | > find anything describing the usage and syntax for
>.pth files.  All I
>| 
>|    http://www.python.org/doc/essays/packages.html
>
>This did a good job of explaining how packages are supposed to
>work and how packages differ and interact with modules.  It also
>explained what .pth files are for and that .pth files shouldn't
>be used unless absolutely necessary.
>
>Ok, but I still want to know what such a file should look like.
>For example, on one of my systems, I have Win2k, Python 2.0, and
>wxPython 2.2.5.  Python is installed in d:\apps\python20
>wxPython is in d:\lib\python_modules\wxPython.  Right now I have
>d:\lib\python_modules in my PYTHONPATH environment variable.  If
>I want to use a wxPython.pth file instead, how would I do that?
>Should it go in d:\apps\python20\lib?  What should it contain?
>Simply the path to d:\lib\python_modules\wxPython on a line by
>itself?

On Windows, I gather that it should be a one-line text file:

lib/python_modules/wxPython

The path should be relative to your Python installation directory
in Windows.

-- 
Neil Cerutti <cerutti at together.net>


From fluxent at yahoo.com  Wed Feb 14 19:30:03 2001
From: fluxent at yahoo.com (fluxent at yahoo.com)
Date: Wed, 14 Feb 2001 18:30:03 -0000
Subject: tweaking ADO RecordSet tuple?
Message-ID: <96eirb+ast4@eGroups.com>

context/architecture: Python, ASP, MsSQL. Sending query to MsSQL, 
getting back RecordSet, outputting HTML table.

Am I correct in understanding that to work with a RecordSet in 
a "normal" (to me) manner, I need to:
- transpose it, since GetRows returns columns, not rows
- convert from unicode and dateobjects to useful stuff

Is there some code sitting someplace that people are usually using 
for this purpose? Or are so few people using ADO that it hasn't 
bubbled up?





From KatMouse at gmx.de  Wed Feb 14 20:59:16 2001
From: KatMouse at gmx.de (Eduard Hiti)
Date: Wed, 14 Feb 2001 20:59:16 +0100
Subject: tweaking ADO RecordSet tuple?
References: <mailman.982175492.29997.python-list@python.org>
Message-ID: <96eo4a$h2r$02$1@news.t-online.com>

I'm using ADO with Python just fine.
COM Unicode usage was always handled very well by win32com.

Date values can be converted to 'time' module format simply by using int():

    import pywintypes, time

    t = pywintypes.Time(time.time())     # this is the COM date type in
Python
    time.strftime("%Y-%m-%d %H:%M:%S", time.localtime( int(t) ) )

This seems to be a little messy, since time.time() returns a float, so
conversion with int() will lose information. What is worse, using float()
will return wrong values.

pywintypes.Time has a 'Format' method, too:

    t.Format("%m%d%Y")

which will return a string representation of the date.

I heard that the mxDateTime module has excellent support for COM date types,
but I didn't need it, so can't comment on that module.

Greetings,

    Eduard


----- Original Message -----
From: <fluxent at yahoo.com>
Newsgroups: comp.lang.python
To: <python-list at cwi.nl>
Sent: Wednesday, February 14, 2001 7:30 PM
Subject: tweaking ADO RecordSet tuple?


| context/architecture: Python, ASP, MsSQL. Sending query to MsSQL,
| getting back RecordSet, outputting HTML table.
|
| Am I correct in understanding that to work with a RecordSet in
| a "normal" (to me) manner, I need to:
| - transpose it, since GetRows returns columns, not rows
| - convert from unicode and dateobjects to useful stuff
|
| Is there some code sitting someplace that people are usually using
| for this purpose? Or are so few people using ADO that it hasn't
| bubbled up?
|
|
|





From fuess at att.net  Thu Feb 15 04:23:27 2001
From: fuess at att.net (David Fuess)
Date: Thu, 15 Feb 2001 03:23:27 GMT
Subject: tweaking ADO RecordSet tuple?
References: <mailman.982175492.29997.python-list@python.org> <96eo4a$h2r$02$1@news.t-online.com>
Message-ID: <r2im8t4i08qokdk1o4jfn5i5k9j9tdkkqu@4ax.com>

On Wed, 14 Feb 2001 20:59:16 +0100, "Eduard Hiti" <KatMouse at gmx.de>
wrote:

General comment: With COM and ADO the number of variable types and
situations you encounter are so wide and varied that each one seems to
be a special case. I have managed to control the beast by restricting
the variable types created in my databases. But we don't always have
that luxury ...

>I'm using ADO with Python just fine.
>COM Unicode usage was always handled very well by win32com.

But you may still want to convert them to latin1 if you intend to
perform string operations on the data items. Many string functions
fail when you feed them certain unicode characters (like u-umlaut).

>
>Date values can be converted to 'time' module format simply by using int():
>
>    import pywintypes, time
>
>    t = pywintypes.Time(time.time())     # this is the COM date type in
>Python
>    time.strftime("%Y-%m-%d %H:%M:%S", time.localtime( int(t) ) )
>

I've had a lot of trouble with this because the datum may be different
between the database server and the UNIX time functions. On SQL Server
7 the datum is 1/1/1900. Instead I use the .Format() function of the
database time object to format the time into a string. It takes the
same arguments as strftime. (This is the method duscussed next :-) )

>This seems to be a little messy, since time.time() returns a float, so
>conversion with int() will lose information. What is worse, using float()
>will return wrong values.
>
>pywintypes.Time has a 'Format' method, too:
>
>    t.Format("%m%d%Y")
>
>which will return a string representation of the date.
>
>I heard that the mxDateTime module has excellent support for COM date types,
>but I didn't need it, so can't comment on that module.
>
>Greetings,
>
>    Eduard
>
>
>----- Original Message -----
>From: <fluxent at yahoo.com>
>Newsgroups: comp.lang.python
>To: <python-list at cwi.nl>
>Sent: Wednesday, February 14, 2001 7:30 PM
>Subject: tweaking ADO RecordSet tuple?
>
>
>| context/architecture: Python, ASP, MsSQL. Sending query to MsSQL,
>| getting back RecordSet, outputting HTML table.
>|
>| Am I correct in understanding that to work with a RecordSet in
>| a "normal" (to me) manner, I need to:
>| - transpose it, since GetRows returns columns, not rows
>| - convert from unicode and dateobjects to useful stuff
>|
>| Is there some code sitting someplace that people are usually using
>| for this purpose? Or are so few people using ADO that it hasn't
>| bubbled up?
>|
>|
>|
>
>



From marco at atmosp.physics.utoronto.ca  Wed Feb 14 19:48:24 2001
From: marco at atmosp.physics.utoronto.ca (marco at atmosp.physics.utoronto.ca)
Date: 14 Feb 2001 18:48:24 GMT
Subject: I don't understand popen2 :(
Message-ID: <96ejto$4q3$1@news.netmar.com>

Hello All,

I'm a python newbie, happy so far, but somewhat bewildered :(

My problem is with popen2.popen3. It sometimes gets stuck when
reading the process std_out/err. Here is an example:

>>> (po, pi, pe) = popen2.popen3('ssh computer "ls"')
>>> po_out = po.read()
>>> pe_err = pe.read()
>>> pe_err
''
>>> po_out
[Directory listing OK]

Alternatively (notice the read order):

>>> (po, pi, pe) = popen2.popen3('ssh computer "ls"')
>>> pe_err = pe.read()
>>> po_out = po.read()
>>> pe_err
''
>>> po_out
[Directory listing OK]

Fine. Now I try something slightly different:

>>> (po, pi, pe) = popen2.popen3('ssh computer "find /home/marco"')
>>> po_out = po.read()
>>> pe_err = pe.read()
>>> pe_err
''
>>> po_out
[snip find results]

which is fine, BUT if I do:

>>> (po, pi, pe) = popen2.popen3('ssh computer "find /home/marco"')
>>> pe_err = pe.read()

[freezes!! I then Ctrl-C]

Traceback (innermost last):
  File "<stdin>", line 1, in ?
KeyboardInterrupt

>>> po_out = po.read()
>>> po_out
[snip find results]

So, in this last case if I pe.read() first
my script freezes, and I have to Ctrl-C. Sure, I
could always po.read() first, but my biggest problem
is that sometimes it's the other way around! That is,
performing po.read() first causes a freeze! Unfortunately,
I cannot reliably reproduce this last behaviour, but here
is the result from a script I had to Ctrl-C:

prompt> ./myprog.py

[frozen -- Ctrl-C]

Traceback (innermost last):
  File "./myprog.py", line 1245, in ?
    get_files()
  File "./myprog.py", line 1102, in get_files
    po_out = po.read()
KeyboardInterrupt

line #
 1101     (po, pi, pe) = popen2.popen3(scp_from_remote_IP)
 1102     po_out = po.read()
 1103     pe_err = pe.read()

Sadly, the scp operation *does* seem to work most of the time,
why does it sometimes get stuck? Thoughts? Ideas?

I'm using:
Python 1.5.2 (#1, Feb  1 2000, 16:32:16)  [GCC egcs-2.91.66 19990314/Linux
(egcs- on linux-i386)]

Thanks for any help!
(Oh, and please CC me a reply if possible -- Deja's undergoing
transition pains to Google right now)

 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse at newsone.net


From spam at spam.com  Wed Feb 14 23:10:03 2001
From: spam at spam.com (Penfold)
Date: Wed, 14 Feb 2001 22:10:03 -0000
Subject: I don't understand popen2 :(
References: <96ejto$4q3$1@news.netmar.com>
Message-ID: <t8m0p6j29c9ia3@corp.supernews.com>

This is simply the usual nightmare of dealing with pipes between processes.
The problem is simple ... both the
po_out.read() and pe_err.read() are *blocking* reads ... further more since
you used read and not read(100) eg,
they will block until they can read to the end of file (ie the whole output
of the command).

In your case when you do a po_read first it works since the comand has
produced all its output, you read it, then it dies
and you read the error stream, all well and cool.  In the latter case, you
try and read the error stream first ... but your
child process refuses to die until its output pipe has been read ... hence
it idles doing nothing and you idle since you've just
entered a blocking read on the error pipe == deadlock.  Bummer :-)  Or
something like that.

Solution ... I tend to alter the pipes to no longer be blocking ... I'm not
sure what other people do in these cases.  You
can use fctl to do this but I forget the flags you want, something like
NDELAY and O_NOBLOCK ... I have none of
my sample code to hand.  Look at man fcntl.

Alternatively, just read the pipes in a sensible order :-)

D.

<marco at atmosp.physics.utoronto.ca> wrote in message
news:96ejto$4q3$1 at news.netmar.com...
>
> Hello All,
>
> I'm a python newbie, happy so far, but somewhat bewildered :(
>
> My problem is with popen2.popen3. It sometimes gets stuck when
> reading the process std_out/err. Here is an example:
>
> >>> (po, pi, pe) = popen2.popen3('ssh computer "ls"')
> >>> po_out = po.read()
> >>> pe_err = pe.read()
> >>> pe_err
> ''
> >>> po_out
> [Directory listing OK]
>
> Alternatively (notice the read order):
>
> >>> (po, pi, pe) = popen2.popen3('ssh computer "ls"')
> >>> pe_err = pe.read()
> >>> po_out = po.read()
> >>> pe_err
> ''
> >>> po_out
> [Directory listing OK]
>
> Fine. Now I try something slightly different:
>
> >>> (po, pi, pe) = popen2.popen3('ssh computer "find /home/marco"')
> >>> po_out = po.read()
> >>> pe_err = pe.read()
> >>> pe_err
> ''
> >>> po_out
> [snip find results]
>
> which is fine, BUT if I do:
>
> >>> (po, pi, pe) = popen2.popen3('ssh computer "find /home/marco"')
> >>> pe_err = pe.read()
>
> [freezes!! I then Ctrl-C]
>
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> KeyboardInterrupt
>
> >>> po_out = po.read()
> >>> po_out
> [snip find results]
>
> So, in this last case if I pe.read() first
> my script freezes, and I have to Ctrl-C. Sure, I
> could always po.read() first, but my biggest problem
> is that sometimes it's the other way around! That is,
> performing po.read() first causes a freeze! Unfortunately,
> I cannot reliably reproduce this last behaviour, but here
> is the result from a script I had to Ctrl-C:
>
> prompt> ./myprog.py
>
> [frozen -- Ctrl-C]
>
> Traceback (innermost last):
>   File "./myprog.py", line 1245, in ?
>     get_files()
>   File "./myprog.py", line 1102, in get_files
>     po_out = po.read()
> KeyboardInterrupt
>
> line #
>  1101     (po, pi, pe) = popen2.popen3(scp_from_remote_IP)
>  1102     po_out = po.read()
>  1103     pe_err = pe.read()
>
> Sadly, the scp operation *does* seem to work most of the time,
> why does it sometimes get stuck? Thoughts? Ideas?
>
> I'm using:
> Python 1.5.2 (#1, Feb  1 2000, 16:32:16)  [GCC egcs-2.91.66 19990314/Linux
> (egcs- on linux-i386)]
>
> Thanks for any help!
> (Oh, and please CC me a reply if possible -- Deja's undergoing
> transition pains to Google right now)
>
>  -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the
eb  -----
>   http://newsone.net/ -- Free reading and anonymous posting to 60,000+
groups
>    NewsOne.Net prohibits users from posting spam.  If this or other posts
> made through NewsOne.Net violate posting guidelines, email
abuse at newsone.net




From marco at atmosp.physics.utoronto.ca  Fri Feb 16 06:23:26 2001
From: marco at atmosp.physics.utoronto.ca (marco at atmosp.physics.utoronto.ca)
Date: 16 Feb 2001 05:23:26 GMT
Subject: I don't understand popen2 :(
References: <96ejto$4q3$1@news.netmar.com> <t8m0p6j29c9ia3@corp.supernews.com>
Message-ID: <96idge$97v$1@news.netmar.com>

In article <t8m0p6j29c9ia3 at corp.supernews.com>, Penfold <spam at spam.com>
writes:
>This is simply the usual nightmare of dealing with pipes between processes.
[snip]

Thanks for the explanation, I suspected something along those lines...

>Solution ... I tend to alter the pipes to no longer be blocking ... I'm not
>sure what other people do in these cases.  You
>can use fctl to do this but I forget the flags you want, something like
>NDELAY and O_NOBLOCK ... I have none of
>my sample code to hand.  Look at man fcntl.

Good tip! I've started looking into this, but if anyone has some
examples it would be a great help (I would assume people run into
these problems all the time? Unfortunately my Python books -- the
O'Reilly ones -- make nary a mention of fcntl.)

>Alternatively, just read the pipes in a sensible order :-)

Well, I tried to do that, and I think I got it working.
However, it doesn't seem to be completely reliable, i.e.
I have to run popen2.popen3 many times and once in a while
it will get stuck, even if the order is right 99% of the time :(

Thanks for your help!

 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse at newsone.net


From donn at oz.net  Fri Feb 16 08:20:54 2001
From: donn at oz.net (Donn Cave)
Date: 16 Feb 2001 07:20:54 GMT
Subject: I don't understand popen2 :(
References: <96ejto$4q3$1@news.netmar.com> <t8m0p6j29c9ia3@corp.supernews.com> <96idge$97v$1@news.netmar.com>
Message-ID: <96ikcm$c1l$0@216.39.151.169>

Quoth marco at atmosp.physics.utoronto.ca:
| In article <t8m0p6j29c9ia3 at corp.supernews.com>, Penfold <spam at spam.com>
| writes:
| >This is simply the usual nightmare of dealing with pipes between processes.
| [snip]
|
| Thanks for the explanation, I suspected something along those lines...
|
|> Solution ... I tend to alter the pipes to no longer be blocking ... I'm not
|> sure what other people do in these cases.  You
|> can use fctl to do this but I forget the flags you want, something like
|> NDELAY and O_NOBLOCK ... I have none of
|> my sample code to hand.  Look at man fcntl.
|
| Good tip! I've started looking into this, but if anyone has some
| examples it would be a great help (I would assume people run into
| these problems all the time? Unfortunately my Python books -- the
| O'Reilly ones -- make nary a mention of fcntl.)

Non-blocking may not help at all.  I personally prefer to always
leave things in a blocking state and use select();  that's essentially
similar to non-blocking, though a little better documented, so it's
just as likely to be of no use.  Both of these techniques apply to
UNIX file descriptors (see file object fileno()), and if you're going
that way I would also use the UNIX I/O functions - posix.read(),
write() etc. and stay away from the file object.  The UNIX read()
will incidentally return what's in the pipe, instead of blocking
for more input as the file object read() does.


|> Alternatively, just read the pipes in a sensible order :-)
|
| Well, I tried to do that, and I think I got it working.
| However, it doesn't seem to be completely reliable, i.e.
| I have to run popen2.popen3 many times and once in a while
| it will get stuck, even if the order is right 99% of the time :(

Some applications of this just won't work.  The most common problem
is when you want to leave both the input and output units open, and
the process on the other end of the pipe mysteriously fails to write
output you expected.  That's because its output is block buffered,
and it won't be actually written to the pipe until the buffer is
full or the application exits.  You can't control that from the
other end of the pipe.  If you can close the other process' input
pipe, then it may exit and flush its buffers.  The other problem,
and this one select() can help, is when you're blocking on one
pipe and the other process is blocking on the other, either waiting
to write more data in a full pipe or waiting to read.

	Donn Cave, donn at oz.net


From nstalkie at tvd.be  Wed Feb 14 20:09:44 2001
From: nstalkie at tvd.be (mannaert sammy)
Date: Wed, 14 Feb 2001 19:09:44 GMT
Subject: mailtje van sammy
Message-ID: <3A8C48D6.AB561557@tvd.be>

dag esmi,

zoals daarnet beloofd aan de telefoon :
... mijn cdlijst.

als ge iets wilt lenen, ge laat maar iets weten ... 

sammy.
-------------- next part --------------
apoptygma berzerk	7
blackhouse		holy war
calva y nada		el peste perverso lleva mi peluca
clair obscur		a collection of isolated tracks 1982 - 1988
clair obscur		play
collapse		section
covenant		sequencer
cubanate		oxyacetelyne
current 93		all dolled up like christ (2CD)
current 93		crooked crosses for a noddy god
current 93		earth covers earth
current 93		faust 
cybertec project        cybertec project
daniel mensche		static burn
das ich			die propheten
das ich			feuer
das ich			satanische verse
de fabriek		save the planet	
death in june		discriminate (2CD)
death in june		operation hummingbird
death in june		rose clouds of holocaust
death in june 		something is coming (2CD) 
death in june 		take care and control
death in june		the wall of sacrifice
der prager handgriff	arglistige tauschung
diary of dreams		cholymelan
die form		ad infinitum
die form		silent order re versions (CDS)
die form		suspiria de profundis
die form                vicious circles / the best of
dissecting table	music for performance 'dead body and me'
dive			live action (CDS)
dorsetshire		das letzte gefecht
einsturzende neubauten	zeichnungen des patienten o.t.
engelsstaub		ignis fatuus: irrlichter
esplendor geometrico	1980 - 1982 (2CD)
esplendor geometrico	1983 - 1987
esplendor geometrico	balearic rhythms
esplendor geometrico    kosmos kino
esplendor geometrico	nador
esplendor geometrico	tarikat (2CD)
esplendor geometrico	tokyo sin fin
fire + ice		birdking
front 242		front by front
front line assembly	live wired (2CD)
goethes erben 		blau
goethes erben 		der die das
goethes erben		leben im niemandsland
goethes erben 		sitz der genade
goethes erben 		tote augen sehen leben
hybryds			cortex stimulation
hybryds			ein phallisher gott
hypnoskull		ffwd>>burnout
imminent starvation     nord
in slaughter natives    in slaughter natives
in slaughter natives	purgate my stain
klinik			black leather
london after midnight	selected scenes from the end of the world
madre del vizio		dio dio dio
maeror tri		mort aux vaches
merzbow			oersted
muslimgauze		bruto (CDS)
muslimgauze		silknoose
muslimgauze		zealot (2CD)
noisex			over and out
non			might!
noshinto		revenge of the 50ft monogroove
oomph!			der neue gott (CDS)
puissance		let us lead
qntal 			II 
raison d'etre		in sadness, silence and solitude
relatives menschsein	gefallene engel
sol invictus		lex talionis
sonar			sonar
sopor aeternus		ehjeh asher ehjeh
sopor aeternus		ich tote mich jedesmal aufs neue ...
sopor aeternus		the inexperienced spiral traveller
sopor aeternus		todeswunsch
SPK			information overload unit
SPK			leichenschrei 
sungod			sungod
the moon lay hidden ...	a new soldier follows the path of a new king	
the moon lay hidden ... a night in fear
throbbing gristle	assume power focus
umbra et imago		gedanken eines vampirs
umbra et imago		infantile spiele
umbra et imago		traume, sex und tod
vromb			le facteur humain
yasnaia			oniro
zoviet france		what is not true
v/a			2/3
v/a			antzen anthology (2CD) (only cd 1)
v/a			bodyhorst's popshow
v/a			dance macabre sampler
v/a			electrocity vol. 5
v/a			invisible domains
v/a			jekura - deep the eternal forest
v/a			sound-line vol. 2
v/a			sound-line vol. 3
v/a			sound-line vol. 4
v/a			sound-line vol. 5
v/a			the tyranny of the beat vol. II
v/a			the tyranny of the beat vol. III
v/a			to cheer up all yer suicide parties
v/a			touched by the hand of god (2CD) (only cd 1) 
v/a			zillo clubhits 3
vnv nation 		praise the fallen
x marks the pedwalk 	freaks 
x marks the pedwalk	meshwork

From nstalkie at tvd.be  Wed Feb 14 23:47:40 2001
From: nstalkie at tvd.be (mannaert sammy)
Date: Wed, 14 Feb 2001 22:47:40 GMT
Subject: mailtje van sammy
References: <3A8C48D6.AB561557@tvd.be>
Message-ID: <3A8C7BEB.20F26277@tvd.be>

mannaert sammy wrote:
> 
> dag esmi,
> 
> zoals daarnet beloofd aan de telefoon :
> ... mijn cdlijst.
> 
> als ge iets wilt lenen, ge laat maar iets weten ...
> 
> sammy.
> 

argh; i'm sorry about this.


From dwig at advancedmp.net  Wed Feb 14 20:19:44 2001
From: dwig at advancedmp.net (Don Dwiggins)
Date: 14 Feb 2001 11:19:44 -0800
Subject: Possible bug in httplib
In-Reply-To: Harry George's message of "Wed, 14 Feb 2001 17:58:53 GMT"
Message-ID: <ur911gjhr.fsf_-_@advancedmp.net>

I've been "doing exercises" with httplib to familiarize myself with it, and
I ran into a problem that I think might be a bug.  

The HTTPConnection.request method takes an optional "body" argument, useful
for POST requests.  When I tried to use it, the server I was talking to came
back with an empty reply.  I was finally able to get the request accepted
using a sequence of putheader calls to send the Content-Type header as well
as Content-Length.  I don't know whether Content-Type is required by the
HTTP spec, but at least one server calls for it.

Can anyone offer a more authoritative answer as to whether the request
method should be sending the Content-Type header?

Thanks,

-- 
Don Dwiggins
Advanced MP Technology
dwig at advancedmp.net




From boswell at caltech.edu  Fri Feb 16 00:10:36 2001
From: boswell at caltech.edu (Dustin Boswell)
Date: Thu, 15 Feb 2001 15:10:36 -0800
Subject: Possible bug in httplib
References: <mailman.982178415.9822.python-list@python.org>
Message-ID: <3A8C61EC.5FCE3D64@caltech.edu>

I'm not sure if it'll help your problem, but you should get fresh copies
of httplib.py and urllib.py (and any others).  There are a bunch of
little bugs in the Python2.0 distribution.

Try the cvs repository....
http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/python/dist/src/?cvsroot=python

Hope that helps.

-Dustin Boswell ( boswell at caltech.edu )
Don Dwiggins wrote:

> I've been "doing exercises" with httplib to familiarize myself with it, and
> I ran into a problem that I think might be a bug.
>
> The HTTPConnection.request method takes an optional "body" argument, useful
> for POST requests.  When I tried to use it, the server I was talking to came
> back with an empty reply.  I was finally able to get the request accepted
> using a sequence of putheader calls to send the Content-Type header as well
> as Content-Length.  I don't know whether Content-Type is required by the
> HTTP spec, but at least one server calls for it.
>
> Can anyone offer a more authoritative answer as to whether the request
> method should be sending the Content-Type header?
>
> Thanks,
>
> --
> Don Dwiggins
> Advanced MP Technology
> dwig at advancedmp.net



From bobhicks.nospam at adelphia.net  Wed Feb 14 20:28:23 2001
From: bobhicks.nospam at adelphia.net (Robert L Hicks)
Date: Wed, 14 Feb 2001 19:28:23 GMT
Subject: Python should have a IDE like this...
Message-ID: <140220011428227667%bobhicks.nospam@adelphia.net>

http://vtcl.sourceforge.net/snapshot.html

This is a good GUI builder for Tcl...and would be nice to have a Python
version.

- Bob


From boud at valdyas.org  Wed Feb 14 21:29:05 2001
From: boud at valdyas.org (Boudewijn Rempt)
Date: Wed, 14 Feb 2001 21:29:05 +0100
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net>
Message-ID: <3a8aeab2$0$36171$1a0eedbc@dreader2.news.xs4all.nl>

Robert L Hicks wrote:

> http://vtcl.sourceforge.net/snapshot.html
> 
> This is a good GUI builder for Tcl...and would be nice to have a Python
> version.
> 

Well, you can get something quite close with the combination of PyQt and Qt 
Architect. Or, if you are willing to shell out money, with BlackAdder.

Boudewijn Rempt | http://www.valdyas.org


From bobhicks.nospam at adelphia.net  Wed Feb 14 21:41:57 2001
From: bobhicks.nospam at adelphia.net (Robert L Hicks)
Date: Wed, 14 Feb 2001 20:41:57 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net> <3a8aeab2$0$36171$1a0eedbc@dreader2.news.xs4all.nl>
Message-ID: <140220011541568826%bobhicks.nospam@adelphia.net>

I actually do like BlackAdder...but it has always been in the back of
my demented mind that there should be a good visual IDE just for Python
and Tkinter. The other widget sets(Pmw,Tix, etc.) would be optional.
That way it would run everywhere Python and Tkinter do.

In article <3a8aeab2$0$36171$1a0eedbc at dreader2.news.xs4all.nl>,
Boudewijn Rempt <boud at valdyas.org> wrote:

> Robert L Hicks wrote:
> 
> > http://vtcl.sourceforge.net/snapshot.html
> > 
> > This is a good GUI builder for Tcl...and would be nice to have a Python
> > version.
> > 
> 
> Well, you can get something quite close with the combination of PyQt and Qt 
> Architect. Or, if you are willing to shell out money, with BlackAdder.
> 
> Boudewijn Rempt | http://www.valdyas.org


From bs1535 at sbc.com  Wed Feb 14 22:24:26 2001
From: bs1535 at sbc.com (SNYDER, BARRON F. (AIT))
Date: Wed, 14 Feb 2001 16:24:26 -0500
Subject: Python should have a IDE like this...
Message-ID: <35BD410BA148D411A7ED00508BCFFBDA05CB2830@msgil65170u05.nbk2305.il.ameritech.com>

How about PythonWorks from Secret Labs (www.pythonware.com)? 
Where does it place on the IDE ladder?


-----Original Message-----
From: Robert L Hicks [mailto:bobhicks.nospam at adelphia.net]
Sent: Wednesday, February 14, 2001 2:42 PM
To: python-list at python.org
Subject: Re: Python should have a IDE like this...


I actually do like BlackAdder...but it has always been in the back of
my demented mind that there should be a good visual IDE just for Python
and Tkinter. The other widget sets(Pmw,Tix, etc.) would be optional.
That way it would run everywhere Python and Tkinter do.

In article <3a8aeab2$0$36171$1a0eedbc at dreader2.news.xs4all.nl>,
Boudewijn Rempt <boud at valdyas.org> wrote:

> Robert L Hicks wrote:
> 
> > http://vtcl.sourceforge.net/snapshot.html
> > 
> > This is a good GUI builder for Tcl...and would be nice to have a Python
> > version.
> > 
> 
> Well, you can get something quite close with the combination of PyQt and
Qt 
> Architect. Or, if you are willing to shell out money, with BlackAdder.
> 
> Boudewijn Rempt | http://www.valdyas.org
-- 
http://mail.python.org/mailman/listinfo/python-list



From dsh8290 at rit.edu  Wed Feb 14 22:41:33 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 14 Feb 2001 16:41:33 -0500
Subject: Python should have a IDE like this...
In-Reply-To: <140220011428227667%bobhicks.nospam@adelphia.net>; from bobhicks.nospam@adelphia.net on Wed, Feb 14, 2001 at 07:28:23PM +0000
References: <140220011428227667%bobhicks.nospam@adelphia.net>
Message-ID: <20010214164133.A17793@harmony.cs.rit.edu>

On Wed, Feb 14, 2001 at 07:28:23PM +0000, Robert L Hicks wrote:
| http://vtcl.sourceforge.net/snapshot.html
| 
| This is a good GUI builder for Tcl...and would be nice to have a Python
| version.
| 
| - Bob

As far as GUI builders go, Glade (PyGTK) and wxDesigner (wxPython) are
really nice.  Glade+libglade is a really cool way to build a GUI.
I've just started to try out wxDesigner, and it seems to be a little
odd in the way it organizes the code it generates.  It might just be
that I'm not used to it ...

-D




From bobhicks at adelphia.net  Wed Feb 14 23:52:05 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Wed, 14 Feb 2001 22:52:05 GMT
Subject: Python should have a IDE like this...
References: <mailman.982185918.6661.python-list@python.org>
Message-ID: <B6B07642.636%bobhicks@adelphia.net>

expensive...I am thinking something that goes for free...

> From: "SNYDER, BARRON F. (AIT)" <bs1535 at sbc.com>
> Organization: Baymountain
> Newsgroups: comp.lang.python
> Date: Wed, 14 Feb 2001 16:24:26 -0500
> To: python-list at python.org
> Subject: RE: Python should have a IDE like this...
> 
> 
> How about PythonWorks from Secret Labs (www.pythonware.com)?
> Where does it place on the IDE ladder?
> 
> 
> -----Original Message-----
> From: Robert L Hicks [mailto:bobhicks.nospam at adelphia.net]
> Sent: Wednesday, February 14, 2001 2:42 PM
> To: python-list at python.org
> Subject: Re: Python should have a IDE like this...
> 
> 
> I actually do like BlackAdder...but it has always been in the back of
> my demented mind that there should be a good visual IDE just for Python
> and Tkinter. The other widget sets(Pmw,Tix, etc.) would be optional.
> That way it would run everywhere Python and Tkinter do.
> 
> In article <3a8aeab2$0$36171$1a0eedbc at dreader2.news.xs4all.nl>,
> Boudewijn Rempt <boud at valdyas.org> wrote:
> 
>> Robert L Hicks wrote:
>> 
>>> http://vtcl.sourceforge.net/snapshot.html
>>> 
>>> This is a good GUI builder for Tcl...and would be nice to have a Python
>>> version.
>>> 
>> 
>> Well, you can get something quite close with the combination of PyQt and
> Qt 
>> Architect. Or, if you are willing to shell out money, with BlackAdder.
>> 
>> Boudewijn Rempt | http://www.valdyas.org
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 



From bobhicks at adelphia.net  Wed Feb 14 23:52:35 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Wed, 14 Feb 2001 22:52:35 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net> <mailman.982186937.10543.python-list@python.org>
Message-ID: <B6B07661.637%bobhicks@adelphia.net>

I am thinking just a Python/Tkinter version...so it can go wherever Python
can.

> From: D-Man <dsh8290 at rit.edu>
> Organization: Baymountain
> Newsgroups: comp.lang.python
> Date: Wed, 14 Feb 2001 16:41:33 -0500
> To: python-list at python.org
> Subject: Re: Python should have a IDE like this...
> 
> On Wed, Feb 14, 2001 at 07:28:23PM +0000, Robert L Hicks wrote:
> | http://vtcl.sourceforge.net/snapshot.html
> | 
> | This is a good GUI builder for Tcl...and would be nice to have a Python
> | version.
> | 
> | - Bob
> 
> As far as GUI builders go, Glade (PyGTK) and wxDesigner (wxPython) are
> really nice.  Glade+libglade is a really cool way to build a GUI.
> I've just started to try out wxDesigner, and it seems to be a little
> odd in the way it organizes the code it generates.  It might just be
> that I'm not used to it ...
> 
> -D
> 
> 



From wware at world.std.com  Thu Feb 15 02:46:52 2001
From: wware at world.std.com (Will Ware)
Date: Thu, 15 Feb 2001 01:46:52 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net> <mailman.982186937.10543.python-list@python.org>
Message-ID: <G8rzM4.HK8@world.std.com>

D-Man (dsh8290 at rit.edu) wrote:
> Glade+libglade is a really cool way to build a GUI.

Doesn't Glade only produce C code? I suppose you could make calls
to Python from C, but it'd be cool if a GUI builder produced Python
directly. Unless I've misunderstood Glade, and it already does this?
That would be very cool.

-- 
-----------------------------------+---------------------
 22nd century: Esperanto, geodesic | Will Ware
 domes, hovercrafts, metric system | wware at world.std.com


From jepler at inetnebr.com  Thu Feb 15 04:30:41 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Thu, 15 Feb 2001 03:30:41 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net> <mailman.982186937.10543.python-list@python.org> <G8rzM4.HK8@world.std.com>
Message-ID: <slrn98mjb0.2s8.jepler@potty.housenet>

On Thu, 15 Feb 2001 01:46:52 GMT, Will Ware
 <wware at world.std.com> wrote:
>D-Man (dsh8290 at rit.edu) wrote:
>> Glade+libglade is a really cool way to build a GUI.
>
>Doesn't Glade only produce C code? I suppose you could make calls
>to Python from C, but it'd be cool if a GUI builder produced Python
>directly. Unless I've misunderstood Glade, and it already does this?
>That would be very cool.

When using python+libglade, you read in the xml-format .glade file from
disk.  I seem to recall another package which generated Python source
code instead, but I don't know what advantages doing this has over
reading the xml format.  (For instance, you can let a user replace your
whole UI by replacing the .glade file, so long as they use the proper
widget names)

Jeff


From wware at world.std.com  Thu Feb 15 16:28:17 2001
From: wware at world.std.com (Will Ware)
Date: Thu, 15 Feb 2001 15:28:17 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net> <mailman.982186937.10543.python-list@python.org> <G8rzM4.HK8@world.std.com>
Message-ID: <G8t1n5.H6y@world.std.com>

I (wware at world.std.com) wrote:
> Doesn't Glade only produce C code?
> ...it'd be cool if a GUI builder produced Python directly.

I got curious enough to invest some time in a web search, and
found several Python programs that translate Glade's XML file
to Python code:
http://glc.sourceforge.net/
http://www.baypiggies.org/stubby.py
http://www.fcoutant.freesurf.fr/gladepyc.html

These all require pygtk to work:
http://www.daa.com.au/~james/pygtk/

There is some info on using Gnome/GTK/Glade with Python here:
http://www.baypiggies.org/10mintig.html

In my wanderings I stumbled across a very nice one-page
Vault-of-Parnassus-type thing:
http://www.rimbault.net/python/

-- 
-----------------------------------+---------------------
 22nd century: Esperanto, geodesic | Will Ware
 domes, hovercrafts, metric system | wware at world.std.com


From delza at antarcti.ca  Thu Feb 15 18:14:16 2001
From: delza at antarcti.ca (Dethe Elza)
Date: Thu, 15 Feb 2001 17:14:16 GMT
Subject: XML for UI  (Was: Python should have a IDE like this...)
References: <140220011428227667%bobhicks.nospam@adelphia.net> <mailman.982186937.10543.python-list@python.org> <G8rzM4.HK8@world.std.com>
Message-ID: <I7Ui6.101693$t94.1397680@news1.rdc1.bc.home.com>

As I understand it, Glade can output your GUI as XML, which you can tweak by
hand if desired, and read in at run-time.  It's one of the new crop of
XML-based GUI tools, but the documentation (last time I checked) was
virtually non-existent.

WxWindows is also working on an XML format for UI (also under-documented),
UIML is an attempt to create XML for UI across systems, languages, and
toolkits (it suffers from being *too* general, IMHO), XForms is the W3C
attempt to bring standard UI tools to XML/XHTML, XUL is Mozilla's XML
toolkit (one of the most featureful which is actually *implemented*), and
BML (Bean Markup Language) from IBM's alphaWorks is an XML language for
wiring together JavaBeans, which could mean building a UI in Swing with XML.

Wouldn't it be nice to have an XML dialect which you could read in in Glade,
or wxWindows, or Java, or
Tkinter, or Mozilla?  Check back soon...

--Dethe

"Will Ware" <wware at world.std.com> wrote in message
news:G8rzM4.HK8 at world.std.com...
> D-Man (dsh8290 at rit.edu) wrote:
> > Glade+libglade is a really cool way to build a GUI.
>
> Doesn't Glade only produce C code? I suppose you could make calls
> to Python from C, but it'd be cool if a GUI builder produced Python
> directly. Unless I've misunderstood Glade, and it already does this?
> That would be very cool.
>
> --
> -----------------------------------+---------------------
>  22nd century: Esperanto, geodesic | Will Ware
>  domes, hovercrafts, metric system | wware at world.std.com




From bobhicks.nospam at adelphia.net  Thu Feb 15 19:23:25 2001
From: bobhicks.nospam at adelphia.net (Robert L Hicks)
Date: Thu, 15 Feb 2001 18:23:25 GMT
Subject: XML for UI  (Was: Python should have a IDE like this...)
References: <140220011428227667%bobhicks.nospam@adelphia.net> <mailman.982186937.10543.python-list@python.org> <G8rzM4.HK8@world.std.com> <I7Ui6.101693$t94.1397680@news1.rdc1.bc.home.com>
Message-ID: <150220011323227393%bobhicks.nospam@adelphia.net>

I have little problem with any of the various suggestions that have
been made off my original posting. I say little because the topic has
strayed a little. The idea was to have an IDE with GUI builder
implented in Tkinter.

I have actually used the different implementations suggested but...they
still are using *other* widget sets (i.e. Qt, GTK+, wxPython).

Maybe it is just a "something to get to in the future"...

- Bob


In article <I7Ui6.101693$t94.1397680 at news1.rdc1.bc.home.com>, Dethe
Elza <delza at antarcti.ca> wrote:

> As I understand it, Glade can output your GUI as XML, which you can tweak by
> hand if desired, and read in at run-time.  It's one of the new crop of
> XML-based GUI tools, but the documentation (last time I checked) was
> virtually non-existent.
> 
> WxWindows is also working on an XML format for UI (also under-documented),
> UIML is an attempt to create XML for UI across systems, languages, and
> toolkits (it suffers from being *too* general, IMHO), XForms is the W3C
> attempt to bring standard UI tools to XML/XHTML, XUL is Mozilla's XML
> toolkit (one of the most featureful which is actually *implemented*), and
> BML (Bean Markup Language) from IBM's alphaWorks is an XML language for
> wiring together JavaBeans, which could mean building a UI in Swing with XML.
> 
> Wouldn't it be nice to have an XML dialect which you could read in in Glade,
> or wxWindows, or Java, or
> Tkinter, or Mozilla?  Check back soon...
> 
> --Dethe
> 
> "Will Ware" <wware at world.std.com> wrote in message
> news:G8rzM4.HK8 at world.std.com...
> > D-Man (dsh8290 at rit.edu) wrote:
> > > Glade+libglade is a really cool way to build a GUI.
> >
> > Doesn't Glade only produce C code? I suppose you could make calls
> > to Python from C, but it'd be cool if a GUI builder produced Python
> > directly. Unless I've misunderstood Glade, and it already does this?
> > That would be very cool.
> >
> > --
> > -----------------------------------+---------------------
> >  22nd century: Esperanto, geodesic | Will Ware
> >  domes, hovercrafts, metric system | wware at world.std.com


From jepler at inetnebr.com  Thu Feb 15 04:20:34 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Thu, 15 Feb 2001 03:20:34 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net>
Message-ID: <slrn98mio1.2s8.jepler@potty.housenet>

On Wed, 14 Feb 2001 19:28:23 GMT, Robert L Hicks
 <bobhicks.nospam at adelphia.net> wrote:
>http://vtcl.sourceforge.net/snapshot.html
>
>This is a good GUI builder for Tcl...and would be nice to have a Python
>version.
>
>- Bob

If vtcl just outputs tcl code, it seems that it should be possible to
execute this tcl code using tk.call("source", filename) or tk.call("eval",
string) , and then create Python/Tkinter objects corresponding to each
widget after the fact.  Something like the following could perform the latter
task.

A better, but more complicate idea, would be to modify vtcl to output Python 
code directly.  And, of course, this code below does nothing to take care of
the functions to be called via -command or bind.  Lastly, a GUI builder
written in Python might be preferable to one written in some other
language.

import Tkinter, string

class _X: pass

# Take a tree of widgets created in pure tcl and make them
# into Tkinter objects
def widget_to_tkinter(path=".", master=None, tk=Tkinter.tkinter.main_interp):
	if master: tk = master.tk
        ret = _X()
        klass = tk.call("winfo", "class", path)

        if path == ".":
                klass = "Toplevel"
        try:
                ret.__class__ = getattr(Tkinter, klass)
        except AttributeError:
                ret.__class__ = Tkinter.Widget
        if path == ".":
                ret._w = "."
                assert master is None and tk is not None
                ret.tk = tk
                ret.children = {}
        else:
                assert master is not None
                name = string.split(path, ".")[-1]
                ret._setup(master, {"name": name})
        children = ret.tk.call("winfo", "children", path)
        if children:
                children = string.split(children, " ")
                for item in children:
                        widget_to_tkinter(item, master=ret)
        return ret



From see at post.for.email  Thu Feb 15 08:09:05 2001
From: see at post.for.email (Vicegrip)
Date: Thu, 15 Feb 2001 07:09:05 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net>
Message-ID: <lgLi6.135497$KP3.37129176@news3.rdc1.on.home.com>

Take a look at Komodo..still beta... like the feel of it though. 
You can download it from Activestate for free.

(being beta, I wouldn't recommend it for major projects, but for tinkering, 
it seems robust enough)

/JP



From bobhicks at adelphia.net  Thu Feb 15 15:14:32 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Thu, 15 Feb 2001 14:14:32 GMT
Subject: Python should have a IDE like this...
References: <140220011428227667%bobhicks.nospam@adelphia.net> <lgLi6.135497$KP3.37129176@news3.rdc1.on.home.com>
Message-ID: <B6B14E49.835%bobhicks@adelphia.net>

Komodo is not a visual builder as vtcl is...it is a nice IDE though.

- Bob

> From: see at post.for.email (Vicegrip)
> Organization: Your Company
> Newsgroups: comp.lang.python
> Date: Thu, 15 Feb 2001 07:09:05 GMT
> Subject: Re: Python should have a IDE like this...
> 
> Take a look at Komodo..still beta... like the feel of it though.
> You can download it from Activestate for free.
> 
> (being beta, I wouldn't recommend it for major projects, but for tinkering,
> it seems robust enough)
> 
> /JP
> 



From insanc at cc.gatech.edu  Wed Feb 14 20:32:11 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 14 Feb 2001 19:32:11 GMT
Subject: tkinter x button override
Message-ID: <96emfr$ca5$1@solaria.cc.gatech.edu>

after creating a frame with the tkinter module, it automatically creates a
close button (x on the top right). is there anyway to override the function
of this button? i haven't found any documentation so far. thank you.

-- 
---
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From matt at mondoinfo.com  Wed Feb 14 21:00:36 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Wed, 14 Feb 2001 20:00:36 GMT
Subject: tkinter x button override
References: <96emfr$ca5$1@solaria.cc.gatech.edu>
Message-ID: <slrn98lov4.mt.matt@happy-hour.mondoinfo.com>

On 14 Feb 2001 19:32:11 GMT, Joseph Holland King
<insanc at cc.gatech.edu> wrote:

>after creating a frame with the tkinter module, it automatically creates a
>close button (x on the top right). is there anyway to override the function
>of this button? i haven't found any documentation so far. thank you.

It's your window manager that's creating the close button but Tkinter
can handle the event. The best documentation for Tkinter that I've
found is Fredrik Lundh's excellent An Introduction to Tkinter. See
(wrapped for line length):

http://www.pythonware.com/library/tkinter/introduction/
  events-and-bindings.htm

And scroll down to the section on protocols.

Regards,
Matt



From insanc at cc.gatech.edu  Wed Feb 14 21:16:07 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 14 Feb 2001 20:16:07 GMT
Subject: tkinter x button override
References: <96emfr$ca5$1@solaria.cc.gatech.edu> <slrn98lov4.mt.matt@happy-hour.mondoinfo.com>
Message-ID: <96ep27$dgc$1@solaria.cc.gatech.edu>

Matthew Dixon Cowles <matt at mondoinfo.com> wrote:
: On 14 Feb 2001 19:32:11 GMT, Joseph Holland King
: <insanc at cc.gatech.edu> wrote:

: It's your window manager that's creating the close button but Tkinter

ok i was not aware that it worked like this, that explains a lot.

: http://www.pythonware.com/library/tkinter/introduction/
:   events-and-bindings.htm

this was exactly what i was looking for. thank you. 

-- 
---
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From zope at thewebsons.com  Wed Feb 14 20:50:12 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Wed, 14 Feb 2001 11:50:12 -0800
Subject: Robots
Message-ID: <5.0.2.1.0.20010214113928.00a0cec0@thewebsons.com>

Hi;
Does anyone have a good robot script they'd like to share? I need to build 
one that searches the Web for similarly themed Web sites and grabs email 
addresses.
BenO




From gerhard.nospam at bigfoot.de  Wed Feb 14 21:31:56 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Wed, 14 Feb 2001 21:31:56 +0100
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org>
Message-ID: <3A8AEB3C.4EC31E35@bigfoot.de>

Ben Ocean wrote:
> 
> Hi;
> Does anyone have a good robot script they'd like to share? I need to build
> one that searches the Web for similarly themed Web sites and grabs email
> addresses.

Are you trying to fill a spam database? If so, you deserve a good kick
in your ass. Perhaps this would help you.

Gerhard (who hates being spammed)

PS: In the improbable case that spamming is not your intention, I
apologize. I just cannot imagine any other reason for grabbing emails
off html documents.
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From zope at thewebsons.com  Wed Feb 14 22:40:46 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Wed, 14 Feb 2001 13:40:46 -0800
Subject: Robots
In-Reply-To: <3A8AEB3C.4EC31E35@bigfoot.de>
References: <mailman.982180157.16563.python-list@python.org>
Message-ID: <5.0.2.1.0.20010214133927.00a2fa60@thewebsons.com>

At 09:31 PM 2/14/2001 +0100, you wrote:
>Ben Ocean wrote:
> >
> > Hi;
> > Does anyone have a good robot script they'd like to share? I need to build
> > one that searches the Web for similarly themed Web sites and grabs email
> > addresses.
>
>Are you trying to fill a spam database? If so, you deserve a good kick
>in your ass. Perhaps this would help you.

LOL! No, sorry that I misled you! I want to contact webmasters individually 
with individualized emails and invite them to swap links. So... any ideas 
on a script?
BenO

>Gerhard (who hates being spammed)
>
>PS: In the improbable case that spamming is not your intention, I
>apologize. I just cannot imagine any other reason for grabbing emails
>off html documents.
>--
>Sorry for the fake email, please use the real one below to reply.
>contact: g e r h a r d @ b i g f o o t . d e
>web:     http://highqualdev.com
>--
>http://mail.python.org/mailman/listinfo/python-list




From jcollin at exis.net  Thu Feb 15 16:13:33 2001
From: jcollin at exis.net (Jay Collins)
Date: Thu, 15 Feb 2001 10:13:33 -0500
Subject: Robots
In-Reply-To: <3A8AEB3C.4EC31E35@bigfoot.de>
References: <mailman.982180157.16563.python-list@python.org> <3A8AEB3C.4EC31E35@bigfoot.de>
Message-ID: <Pine.LNX.4.21.0102151012330.4310-100000@tarpon.exis.net>

On Wed, 14 Feb 2001, Gerhard [iso-8859-1] H?ring wrote:

>
>Gerhard (who hates being spammed)
>
>PS: In the improbable case that spamming is not your intention, I
>apologize. I just cannot imagine any other reason for grabbing emails
>off html documents.

The anti-spam program sam spadey(?) has a feature that will crawl a
website looking for email address so you can figure out who to write to if
they got no abuse@ or postmaster at .

Take care.



From gerhard.nospam at bigfoot.de  Thu Feb 15 00:06:39 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Thu, 15 Feb 2001 00:06:39 +0100
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org> <mailman.982186876.10323.python-list@python.org>
Message-ID: <3A8B0F7F.F31995D3@bigfoot.de>

> >Are you trying to fill a spam database? If so, you deserve a good kick
> >in your ass. Perhaps this would help you.
> 
> LOL! No, sorry that I misled you! I want to contact webmasters individually
> with individualized emails and invite them to swap links. So... any ideas
> on a script?

You are a lucky guy! ;-)

Fiddling with HTML is always fun. What about the following for a start?

Gerhard

--- attempt #1 follows ---

#!/usr/bin/env python
import re

class MailGrabber:
    def __init__( self ):
        self.mail_adrs = []
        self.mail_regex = re.compile( 'mailto:.*?"' )

    def handle_doc( self, s ):
        for result in re.findall( self.mail_regex, s ):
            result = result[ len( "mailto:" ) : -1 ]
            self.mail_adrs.append( result )            
            
grabber = MailGrabber()

f = open( 'c://tmp//namenliste.htm', 'r' )
s = f.read()
f.close()

grabber.handle_doc( s )
print grabber.mail_adrs

-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From max at alcyone.com  Thu Feb 15 02:03:53 2001
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 14 Feb 2001 17:03:53 -0800
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org> <mailman.982186876.10323.python-list@python.org>
Message-ID: <3A8B2AF9.4E2A4029@alcyone.com>

Ben Ocean wrote:

> LOL! No, sorry that I misled you! I want to contact webmasters
> individually
> with individualized emails and invite them to swap links. So... any
> ideas
> on a script?

So then, yes, you _do_ want build a spam database.  Unsolicited bulk
email is unsolicited bulk email, regardless of whether or not _you_
think it's relevant.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ War is cruelty, and you cannot refine it.
\__/ Karl Shapiro
    Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
 A highly categorized list of Web links.


From zope at thewebsons.com  Thu Feb 15 09:33:09 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Thu, 15 Feb 2001 00:33:09 -0800
Subject: Robots
In-Reply-To: <3A8B2AF9.4E2A4029@alcyone.com>
References: <mailman.982180157.16563.python-list@python.org>
 <mailman.982186876.10323.python-list@python.org>
Message-ID: <5.0.2.1.0.20010215002942.00a27d90@thewebsons.com>

At 05:03 PM 2/14/2001 -0800, you wrote:
>Ben Ocean wrote:
>
> > LOL! No, sorry that I misled you! I want to contact webmasters
> > individually
> > with individualized emails and invite them to swap links. So... any
> > ideas
> > on a script?
>
>So then, yes, you _do_ want build a spam database.  Unsolicited bulk
>email is unsolicited bulk email, regardless of whether or not _you_
>think it's relevant.

With all due respect...
Let's not start a flame war. This will be my last response to this 
nonsense. The keyword is *bulk* When it's sent individually, it's not 
*bulk*. Put that in your pipe and smoke it. I'm trying to build something 
similar to the Zeus robot at
http://cyber-robotics.com/
Now, for you purists that think that *any* use of automation to communicate 
with others is spam, well, you're using automation...
The key is you're not doing it in bulk. Neither am I. Okay???
BenO


>--
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
>/  \ War is cruelty, and you cannot refine it.
>\__/ Karl Shapiro
>     Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
>  A highly categorized list of Web links.
>--
>http://mail.python.org/mailman/listinfo/python-list




From michael at stroeder.com  Thu Feb 15 07:29:04 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Thu, 15 Feb 2001 07:29:04 +0100
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org> <mailman.982186876.10323.python-list@python.org>
Message-ID: <3A8B7730.2B3E43CD@stroeder.com>

Ben Ocean wrote:
> 
> At 09:31 PM 2/14/2001 +0100, you wrote:
> >Ben Ocean wrote:
> > >
> > > Hi;
> > > Does anyone have a good robot script they'd like to share? I need to build
> > > one that searches the Web for similarly themed Web sites and grabs email
> > > addresses.
> >
> >Are you trying to fill a spam database? If so, you deserve a good kick
> >in your ass. Perhaps this would help you.
> 
> LOL! No, sorry that I misled you! I want to contact webmasters individually
> with individualized emails and invite them to swap links.

Please add me to the opt-out list. I'm getting tired of e-mails
suggesting swapping links.

Michael.


From scarblac at pino.selwerd.nl  Thu Feb 15 18:49:31 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 15 Feb 2001 17:49:31 GMT
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org> <mailman.982186876.10323.python-list@python.org>
Message-ID: <slrn98o5hp.7rr.scarblac@pino.selwerd.nl>

Ben Ocean <zope at thewebsons.com> wrote in comp.lang.python:
> LOL! No, sorry that I misled you! I want to contact webmasters individually 
> with individualized emails and invite them to swap links. So... any ideas 
> on a script?

So, getting the emails automatically won't be much help. After all, you'll
collect many adresses from people other than the relevant webmaster, and
you'll have to look at the pages manually anyway, to decide who you should
send your request to.

Although you might make a program to find out which pages have email
adresses. If you're writing some sort of spider, check out
Tools/webchecker/webchecker.py in the source distribution, a program that
checks links on a website; you might be able to adapt it.

So what's left is finding the interesting pages. That's a very hard problem,
do you have any idea how you want to decide what constitutes a web page
similar to yours? If you're looking for certain keywords, it may be easiest
to just enter them into Google...

-- 
Remco Gerlich


From zope at thewebsons.com  Fri Feb 16 00:20:34 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Thu, 15 Feb 2001 15:20:34 -0800
Subject: Robots
In-Reply-To: <slrn98o5hp.7rr.scarblac@pino.selwerd.nl>
References: <mailman.982180157.16563.python-list@python.org>
 <mailman.982186876.10323.python-list@python.org>
Message-ID: <5.0.2.1.0.20010215151735.00a0ccd0@thewebsons.com>

At 05:49 PM 2/15/2001 +0000, you wrote:
>So, getting the emails automatically won't be much help. After all, you'll
>collect many adresses from people other than the relevant webmaster, and
>you'll have to look at the pages manually anyway, to decide who you should
>send your request to.
>
>Although you might make a program to find out which pages have email
>adresses. If you're writing some sort of spider, check out
>Tools/webchecker/webchecker.py in the source distribution, a program that
>checks links on a website; you might be able to adapt it.
>
>So what's left is finding the interesting pages. That's a very hard problem,
>do you have any idea how you want to decide what constitutes a web page
>similar to yours? If you're looking for certain keywords, it may be easiest
>to just enter them into Google...

Yeah. The way Zeus operates, you enter 100 keyword phrases and it looks for 
relevant matches. You could well be right that just hitting Google, or what 
have you, would be just as effective: I've thought the same. Maybe I'll do 
just that. If not, I should at least salt my robot-generated results with 
that. Or better yet, feed that info into my robot. Thanks!
BenO




From greg at cosc.canterbury.ac.nz  Fri Feb 16 04:06:09 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 16 Feb 2001 16:06:09 +1300
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org>
	 <mailman.982186876.10323.python-list@python.org> <mailman.982228515.536.python-list@python.org>
Message-ID: <3A8C9921.66FFC046@cosc.canterbury.ac.nz>

Ben Ocean wrote:
> 
> The keyword is *bulk*

Actually, it's not. When you send someone a piece
of unsolicited email, if they're going to be annoyed
by it, then they'll be just as annoyed regardless of
how many *other* people have got the same piece of mail.

The key word is *unsolicited*.

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From sholden at holdenweb.com  Fri Feb 16 14:46:00 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 16 Feb 2001 08:46:00 -0500
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org> <mailman.982186876.10323.python-list@python.org> <mailman.982228515.536.python-list@python.org> <3A8C9921.66FFC046@cosc.canterbury.ac.nz>
Message-ID: <neaj6.16539$GA5.259269@e420r-atl1.usenetserver.com>

"Greg Ewing" <greg at cosc.canterbury.ac.nz> wrote from the wilds of New
Zealand ...
> Ben Ocean wrote:
> >
> > The keyword is *bulk*
>
> Actually, it's not. When you send someone a piece
> of unsolicited email, if they're going to be annoyed
> by it, then they'll be just as annoyed regardless of
> how many *other* people have got the same piece of mail.
>
> The key word is *unsolicited*.
>
Yup. I get mails from people offering me link exchanges when my email
address appears on the site as the developer: I have no rights to accept or
refuse link exchanges for the sites, and get quite enough email as it is,
thank you.

It may be a low-volume project Ben is proposing, but there can be low-volume
spam as well as high-volume spam. And, of course, Ben, you'll be programming
in such refinements as a database to ensure you only ever bother each email
address once, right?

I tend to work on the basis that my web site (page, really) doesn't contain
enough of interest to make a link on it valuable. Ben's might be different,
of course.

regards
 Steve



From joseamartin at venmex.com  Wed Feb 14 21:18:33 2001
From: joseamartin at venmex.com (Jose Martin)
Date: Wed, 14 Feb 2001 16:18:33 -0400
Subject: Calling Stored Procedure in DCOracle How To ?
Message-ID: <t8lqnmiq0dfh63@corp.supernews.com>

Hi,  How can i call a stored procedure within DCOracle ?

Mi stored procedure has no parameters.


1-    I was tryed calling  dbc.callproc('myproc') but get an error about the
numbers of arguments
        tell me that  the params espected are 1 and got 0


2-   I was tryed usign dbc.procedures.myproc  but nothing happend.

3-   I was tryed using dbc.procedures.myproc.__call__() , but it is the same
thing what 1 error.


What i can do ?











From phd at phd.pp.ru  Thu Feb 15 13:12:57 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 15 Feb 2001 15:12:57 +0300 (MSK)
Subject: Calling Stored Procedure in DCOracle How To ?
In-Reply-To: <t8lqnmiq0dfh63@corp.supernews.com>
Message-ID: <Pine.LNX.4.30.0102151512320.15078-100000@fep132.fep.ru>

On Wed, 14 Feb 2001, Jose Martin wrote:
> 1-    I was tryed calling  dbc.callproc('myproc') but get an error about the
> numbers of arguments
>         tell me that  the params espected are 1 and got 0

   Try dbc.callproc('myproc', [])

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From joseamartin at mail.venmex.com  Thu Feb 15 13:56:21 2001
From: joseamartin at mail.venmex.com (Jose Martin)
Date: Thu, 15 Feb 2001 08:56:21 -0400
Subject: Calling Stored Procedure in DCOracle How To ?
References: <Pine.LNX.4.30.0102151512320.15078-100000@fep132.fep.ru>
Message-ID: <004e01c0974e$b220ef40$9101a8c0@venmex.com>

THanks Oleg.
but 

dbc.callproc('myproc', [])     <---- Second Argument must be a tuple.
Error.


I have more troubles.

I was modified my stored procedures an add a dummy parameter for calling without errors
then yes, I can call them !!!!.

But.  Another Error appears

>>> dbc.callproc('pktrans.proc_solicitudes',(1,))
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python1.5/DCOracle/ocidb.py", line 98, in callproc
    return apply(getattr(self.procedures, name),params)
  File "/usr/lib/python1.5/DCOracle/ociProc.py", line 133, in __call__
    if oci_.oexec(oc): self._error()
  File "/usr/lib/python1.5/DCOracle/ociProc.py", line 106, in _error
    raise error, (rc, oci_.OracleErrorMessage(c.lda, rc))
oci.error: (1403, 'ORA-01403: no data found\012ORA-06512: at "TRANS.PKTRANS", li
ne 54\012ORA-06512: at "TRANS.PKTRANS", line 92\012ORA-06512: at "TRANS.PKTRANS"
, line 30\012ORA-06512: at line 2')


I can run This Stored Procedure from SQLPlus and works ok without erros.

Can Help !.

I am doing a simulation of my bussines in python , i really need to call this stored procedures.

I have another ideas, but are some tricky.

I can define a dummy table with an update trigger and then this trigger can call internally to the Procedure inserted or updated
in  ProcName column of dummy table.

But may exits any way to call Stored Procedures from DCOracle ?


Thanks...






  ----- Original Message ----- 
  From: Oleg Broytmann 
  To: Jose Martin 
  Cc: python-list at python.org 
  Sent: Thursday, February 15, 2001 8:12 AM
  Subject: Re: Calling Stored Procedure in DCOracle How To ?


  On Wed, 14 Feb 2001, Jose Martin wrote:
  > 1-    I was tryed calling  dbc.callproc('myproc') but get an error about the
  > numbers of arguments
  >         tell me that  the params espected are 1 and got 0

     Try dbc.callproc('myproc', [])

  Oleg.
  ----
       Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
             Programmers don't die, they just GOSUB without RETURN.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010215/a026c8aa/attachment.htm>

From phd at phd.pp.ru  Thu Feb 15 17:07:48 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 15 Feb 2001 19:07:48 +0300 (MSK)
Subject: Calling Stored Procedure in DCOracle How To ?
In-Reply-To: <004e01c0974e$b220ef40$9101a8c0@venmex.com>
Message-ID: <Pine.LNX.4.30.0102151907000.16280-100000@fep132.fep.ru>

On Thu, 15 Feb 2001, Jose Martin wrote:
> dbc.callproc('myproc', [])     <---- Second Argument must be a tuple.

   I always used (when I used DCOracle at all) lists. But I never had
procedures without parameters.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From joseamartin at venmex.com  Thu Feb 15 14:35:59 2001
From: joseamartin at venmex.com (Jose Martin)
Date: Thu, 15 Feb 2001 09:35:59 -0400
Subject: Calling Stored Procedure in DCOracle How To ?
References: <mailman.982239257.6703.python-list@python.org>
Message-ID: <t8nngt38c04i15@corp.supernews.com>

Thanks Oleg.
I am Currently calling my procedure with a dummy param
in this form:
dbc.procedures.PKTRANS.PROC_SOLICITUDES(1)

and works ok.
but still need to define a dummy parameter for the stored procedure!.

Bye.


Oleg Broytmann <phd at phd.pp.ru> escribi? en el mensaje de noticias
mailman.982239257.6703.python-list at python.org...
> On Wed, 14 Feb 2001, Jose Martin wrote:
> > 1-    I was tryed calling  dbc.callproc('myproc') but get an error about
the
> > numbers of arguments
> >         tell me that  the params espected are 1 and got 0
>
>    Try dbc.callproc('myproc', [])
>
> Oleg.
> ----
>      Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
>            Programmers don't die, they just GOSUB without RETURN.
>
>




From MarkH at ActiveState.com  Wed Feb 14 23:22:32 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 14 Feb 2001 22:22:32 GMT
Subject: COM support for vtable based interfaces.
Message-ID: <3A8B0448.1000905@ActiveState.com>

Hi all,
	I have recently checked in a new version of win32com that allows you to 
implement _any_ COM interface decribed in a type library, either 
IDispatch based or not.  This is excellent news :)  You still can't 
_call_ these arbitary interfaces, but no doubt this will follow in good 
time.

If you are keen, please pull the latest sources from CVS and let me know 
what happens :)  win32com\readme.html has a brief example.

Mark.
-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From tuttledon at hotmail.com  Thu Feb 15 00:41:28 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Wed, 14 Feb 2001 23:41:28 GMT
Subject: COM support for vtable based interfaces.
References: <3A8B0448.1000905@ActiveState.com>
Message-ID: <IIEi6.75902$p8.19514692@typhoon.southeast.rr.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A8B0448.1000905 at ActiveState.com...
> Hi all,
> I have recently checked in a new version of win32com that allows you to
> implement _any_ COM interface decribed in a type library, either
> IDispatch based or not... If you are keen, please pull the latest sources
from CVS > Mark.

Good news Mark!   However I have a request you can pass on to Dick or
whoever.  Why isn't ActiveState releasing Python builds BETWEEN major
releases that incorporate all the changes in the CVS?  This would be a great
way to hook people into using their build. Especially us CVS impaired types
;-)

Don




From syver at NOSPAMcyberwatcher.com  Thu Feb 15 13:40:58 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Thu, 15 Feb 2001 13:40:58 +0100
Subject: COM support for vtable based interfaces.
References: <3A8B0448.1000905@ActiveState.com>
Message-ID: <r7Qi6.195$1%6.1750@news1.oke.nextra.no>

Super, this means that I can implement COM classes support the same
interface as a typical ATL generated dual-interface COM object doesn't it?
That's good news indeed! I've been needing this for a little while.

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A8B0448.1000905 at ActiveState.com...
> Hi all,
> I have recently checked in a new version of win32com that allows you to
> implement _any_ COM interface decribed in a type library, either
> IDispatch based or not.  This is excellent news :)  You still can't
> _call_ these arbitary interfaces, but no doubt this will follow in good
> time.
>
> If you are keen, please pull the latest sources from CVS and let me know
> what happens :)  win32com\readme.html has a brief example.
>
> Mark.
> --
> Nominate Your Favorite Programmers
> for the Perl and Python Active Awards!
> http://www.ActiveState.com/Awards/
>




From MarkH at ActiveState.com  Thu Feb 15 23:19:10 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 15 Feb 2001 22:19:10 GMT
Subject: COM support for vtable based interfaces.
References: <3A8B0448.1000905@ActiveState.com> <r7Qi6.195$1%6.1750@news1.oke.nextra.no>
Message-ID: <3A8C5500.6040709@ActiveState.com>

Syver Enstad wrote:

> Super, this means that I can implement COM classes support the same
> interface as a typical ATL generated dual-interface COM object doesn't it?
> That's good news indeed! I've been needing this for a little while.

It does indeed!

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From gerhard.nospam at bigfoot.de  Fri Feb 16 00:47:45 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 16 Feb 2001 00:47:45 +0100
Subject: COM support for vtable based interfaces.
References: <3A8B0448.1000905@ActiveState.com>
Message-ID: <3A8C6AA1.81CA8580@bigfoot.de>

Mark Hammond wrote:
> [...]
> If you are keen, please pull the latest sources from CVS and let me know
> what happens :)  win32com\readme.html has a brief example.

Silly question maybe, but: which CVS?? I can't find it in the Python CVS
tree at SourceForge. Or is it in a CVS server somewhere at ActiveState?

Gerhard
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From MarkH at ActiveState.com  Fri Feb 16 05:24:32 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 16 Feb 2001 04:24:32 GMT
Subject: COM support for vtable based interfaces.
References: <3A8B0448.1000905@ActiveState.com> <3A8C6AA1.81CA8580@bigfoot.de>
Message-ID: <3A8CAAA3.9030306@ActiveState.com>

Gerhard H?ring wrote:

> Mark Hammond wrote:
> 
>> [...]
>> If you are keen, please pull the latest sources from CVS and let me know
>> what happens :)  win32com\readme.html has a brief example.
> 
> 
> Silly question maybe, but: which CVS?? I can't find it in the Python CVS
> tree at SourceForge. Or is it in a CVS server somewhere at ActiveState?

See http://starship.python.net/crew/mhammond/cvs.html

Mark.
-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From dubois at users.sourceforge.net  Thu Feb 15 00:22:04 2001
From: dubois at users.sourceforge.net (Paul F. Dubois)
Date: Wed, 14 Feb 2001 23:22:04 GMT
Subject: Utility for showing script execution?
Message-ID: <wqEi6.96553$R5.4114508@news1.frmt1.sfba.home.com>

I'm looking for a utility that will take a Python .py file and execute it,
producing output that shows the "input" lines, marked with a prompt, and
then the resulting output, so that one has an output text file that
simulates what a user would see. Surely someone already did this?

Paul





From fredrik at pythonware.com  Thu Feb 15 08:45:50 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Feb 2001 07:45:50 GMT
Subject: Utility for showing script execution?
References: <wqEi6.96553$R5.4114508@news1.frmt1.sfba.home.com>
Message-ID: <OOLi6.17198$AH6.2346368@newsc.telia.net>

Paul F. Dubois wrote:
> I'm looking for a utility that will take a Python .py file and execute it,
> producing output that shows the "input" lines, marked with a prompt, and
> then the resulting output, so that one has an output text file that
> simulates what a user would see. Surely someone already did this?

of course!

# adapted from code-example-1.py in
# the eff-bot guide to the python standard library

import code
import string

SCRIPT = """
a = 5
for i in range(a):
    print i
print 'hello'
"""

chunk = ""
for line in string.split(SCRIPT, "\n"):
    chunk = chunk + line + "\n"
    co = code.compile_command(chunk, "<stdin>", "exec")
    if co:
        # got a complete statement.  execute it!
        prompt = ">>>"
        if string.count(chunk, "\n") == 1:
            print prompt, chunk,
        else:
            for line in string.split(chunk, "\n"):
                print prompt, line
                prompt = "..."
        exec co
        chunk = ""


Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From skip at mojam.com  Thu Feb 15 00:25:43 2001
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 14 Feb 2001 17:25:43 -0600 (CST)
Subject: general purpose database cleanup tools (design patterns or Python)?
Message-ID: <14987.5111.12981.891744@beluga.mojam.com>

A side effect of people using the Mojam and Musi-Cal web sites is that a
fairly complex MySQL database gets built.  There are several tables in the
database (venues, concerts, cities, etc) that can accumulate errors fairly
rapidly because the system allows anyone with concerts to publicize to add
them to our database.  Because of the size of the tables and their
interdependencies, manually correcting all but the most straightforward
errors is impossible.

I'm looking for pointers to design patterns or tools (preferably in Python)
that might be applicable to the task.  I think such tools would separate the
description of the database's structure from the rules about deleting and
merging rows, which would, in turn, be separate from the engine that
actually manipulates the database.

We have a tool now that works reasonably well but is very difficult to
maintain and enhance because none of the three functions are separate.
Before I launch into a complete redesign, I thought I'd see what already
exists.  This can't be a problem unique to our environment.

Thx,

-- 
Skip Montanaro (skip at mojam.com)
Support Mojam & Musi-Cal: http://www.musi-cal.com/sponsor.shtml
(847)971-7098



From missive at frontiernet.net  Sat Feb 17 05:03:55 2001
From: missive at frontiernet.net (Lee Harr)
Date: 17 Feb 2001 04:03:55 GMT
Subject: general purpose database cleanup tools (design patterns or Python)?
References: <mailman.982193235.2170.python-list@python.org>
Message-ID: <96kt7b$1k5q$1@node17.cwnet.frontiernet.net>

On Wed, 14 Feb 2001 17:25:43 -0600 (CST), Skip Montanaro <skip at mojam.com> wrote:
>

I can't really tell what you are wanting to do, but I get the sense that
what you are looking for is a relational database which allows you to 
maintain relational integrity.

Take a look at postgres( http://www.postgresql.org )


>A side effect of people using the Mojam and Musi-Cal web sites is that a
>fairly complex MySQL database gets built.  There are several tables in the
>database (venues, concerts, cities, etc) that can accumulate errors fairly
>rapidly because the system allows anyone with concerts to publicize to add
>them to our database.  Because of the size of the tables and their
>interdependencies, manually correcting all but the most straightforward
>errors is impossible.
>
>I'm looking for pointers to design patterns or tools (preferably in Python)
>that might be applicable to the task.  I think such tools would separate the
>description of the database's structure from the rules about deleting and
>merging rows, which would, in turn, be separate from the engine that
>actually manipulates the database.
>
>We have a tool now that works reasonably well but is very difficult to
>maintain and enhance because none of the three functions are separate.
>Before I launch into a complete redesign, I thought I'd see what already
>exists.  This can't be a problem unique to our environment.
>
>Thx,
>
>-- 
>Skip Montanaro (skip at mojam.com)
>Support Mojam & Musi-Cal: http://www.musi-cal.com/sponsor.shtml
>(847)971-7098
>


From simon at callan.demon.co.uk  Thu Feb 15 00:54:02 2001
From: simon at callan.demon.co.uk (Simon Callan)
Date: Wed, 14 Feb 2001 23:54:02 +0000
Subject: os.path.walk oddity
Message-ID: <7fceef4c4a.simon@satin.callan.demon.co.uk>

For reasons to complicated to go into here, I am having to rearrange a
source tree. Basically every .c file has to be moved into a
subdirectory called 'c', with the extension removed, while .h files
are moved to a 'h' directory. I.e
fred/jim.c -> fred.c.jim
fred/sheila.h -> fred.h.sheila

After figuring out how to stop os.path.walk crashing due to renaming
files, I ended up with the following:

=====

#!/usr/bin/env python

import os

def rename(dir, root, ext):
  new_dir = os.path.join(dir, ext)
  if not os.path.exists(new_dir):
    os.mkdir(new_dir)
  o_file = os.path.join(dir, root) + '.' + ext
  n_file = os.path.join(dir, ext, root)
  print "%s -> %s" % (o_file, n_file)
  os.rename(o_file, n_file)

def toRiscos(arg, dir, names):
  temp = names
  count = 0
  for file in temp:
    root, ext = os.path.splitext(file)
    if ext == '.c':
      del names[count]
      rename(dir, root, 'c')
    elif ext == '.h':
      del names[count]
      rename(dir, root, 'h')
    count += 1

os.path.walk('.', toRiscos, None)

=====

However, there is one problem. If I run the above on the python 2.0
source directory, it takes 3 runs of the programme to completely
convert the tree.

Can anyone explain why this is?

Simon

-- 
http://www.callan.demon.co.uk/simon/


From fredrik at pythonware.com  Thu Feb 15 09:07:26 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Feb 2001 08:07:26 GMT
Subject: os.path.walk oddity
References: <7fceef4c4a.simon@satin.callan.demon.co.uk>
Message-ID: <27Mi6.17200$AH6.2346966@newsc.telia.net>

Simon Callan wrote:
> def toRiscos(arg, dir, names):
>   temp = names

this assignment doesn't make a copy, it just adds
another reference to the "names" array.

>   count = 0
>   for file in temp:

same thing as iterating over names.

>     root, ext = os.path.splitext(file)
>     if ext == '.c':
>       del names[count]

here you're modifing the array you're looping over.
that's bad.

>       rename(dir, root, 'c')
>     elif ext == '.h':
>       del names[count]
>       rename(dir, root, 'h')
>     count += 1

two possible solutions: change the loop to

    for file in names[:]:

or get rid of the del statements.  for-in won't
loop over an item more than once anyway...

Cheers /F




From simon at callan.demon.co.uk  Thu Feb 15 20:33:31 2001
From: simon at callan.demon.co.uk (Simon Callan)
Date: Thu, 15 Feb 2001 19:33:31 +0000
Subject: os.path.walk oddity
References: <7fceef4c4a.simon@satin.callan.demon.co.uk> <27Mi6.17200$AH6.2346966@newsc.telia.net>
Message-ID: <7bca5b4d4a.simon@satin.callan.demon.co.uk>

In message <27Mi6.17200$AH6.2346966 at newsc.telia.net>
          "Fredrik Lundh" <fredrik at pythonware.com> wrote:

> Simon Callan wrote:
> > def toRiscos(arg, dir, names):
> >   temp = names
> 
> this assignment doesn't make a copy, it just adds another reference
> to the "names" array.

<mumble> I throw myself on the mercy of the court, and plead
mitigating circumstances, to wit, having just spent a lot of effort
getting a C++ program to do a deep copy under precisely these
conditions.

> two possible solutions: change the loop to
> 
>     for file in names[:]:

Bingo! This works first time, after allowing for the list changing in
size after a del.

> or get rid of the del statements.  for-in won't loop over an item
> more than once anyway...

The reason for using the del is that os.path.walk() throws a serious
wobbly when it finds that the file has moved away from under it.

Thanks for your help.

Simon

-- 
http://www.callan.demon.co.uk/simon/


From billings at eudoramail.com  Thu Feb 15 02:06:13 2001
From: billings at eudoramail.com (Scott Billings)
Date: Wed, 14 Feb 2001 19:06:13 -0600
Subject: Getting Value From Tkinter Text Object
Message-ID: <96faar$30s$1@ins21.netins.net>

I have an app, in which I need to get the entire contents of a Tkinter Text 
object. It will be of an arbitrary length.

The only option I can find that even seems to be close to what I want, is 
the get() method, which expects a starting and ending point. Starting point 
should be easy enough, but I have no way of knowing what the ending point 
will be.

Is there some other method for Text objects that I can use, that has a 
similar function to the Entry object's get() method?
-- 
-Scott-


From fuess at att.net  Thu Feb 15 04:25:21 2001
From: fuess at att.net (David Fuess)
Date: Thu, 15 Feb 2001 03:25:21 GMT
Subject: Getting Value From Tkinter Text Object
References: <96faar$30s$1@ins21.netins.net>
Message-ID: <40jm8tk0r075er31k24960slmaj74lv42k@4ax.com>

On Wed, 14 Feb 2001 19:06:13 -0600, Scott Billings
<billings at eudoramail.com> wrote:

>I have an app, in which I need to get the entire contents of a Tkinter Text 
>object. It will be of an arbitrary length.
>
>The only option I can find that even seems to be close to what I want, is 
>the get() method, which expects a starting and ending point. Starting point 
>should be easy enough, but I have no way of knowing what the ending point 
>will be.

So use END. i.e. control.get(0,END)

>
>Is there some other method for Text objects that I can use, that has a 
>similar function to the Entry object's get() method?



From fredrik at pythonware.com  Thu Feb 15 08:49:31 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Feb 2001 07:49:31 GMT
Subject: Getting Value From Tkinter Text Object
References: <96faar$30s$1@ins21.netins.net>
Message-ID: <fSLi6.17199$AH6.2346379@newsc.telia.net>

Scott Billings wrote:
> The only option I can find that even seems to be close to what I want, is
> the get() method, which expects a starting and ending point. Starting point
> should be easy enough, but I have no way of knowing what the ending point
> will be.

http://www.pythonware.com/library/tkinter/introduction/text.htm
=> concepts, patterns

    "END (or "end") corresponds to the position just after the last
    character in the buffer.

    "To fetch the text contents of the widget, use the get method:

        contents = text.get(1.0, END)

Cheers /F




From k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m  Thu Feb 15 04:23:42 2001
From: k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m (Keith Ray)
Date: Thu, 15 Feb 2001 03:23:42 GMT
Subject: Universal Virtual Machine
References: <Ryyi6.95397$t94.1295979@news1.rdc1.bc.home.com>
Message-ID: <k1e2i3t4h5r6a7y-FEE3E8.19234214022001@news>

In article <Ryyi6.95397$t94.1295979 at news1.rdc1.bc.home.com>, "Steve 
Wart" <no-unsolicited-mail-to-swart at bigserver.com> wrote:

[...]
> Whether .NET supports Smalltalk or not is not relevant. It is proprietary
> and closed (I wonder if Microsoft understands that .NET is supposed to be
> the operating system, not something you package with Windows or Windows 
> CE
> or whatever they're calling it now).
[...]
> What I am thinking of is an open source project to create a standard VM 
> for
> Smalltalk but that also supports other languages, such as Java and perl, 
> and
> more importantly, new language directions for Self-like and MOP 
> extensions
> to Smalltalk. Similar to the Transmeta architecture (except all in
> software), one could define pluggable byte code dictionaries that the VM
> could swap in as required.
> 
> This would allow one VM to work with several Smalltalk dialects, without
> requiring vendors to agree on a standard set of byte codes. The issues 
> with
> the Garbage Collector seem more difficult to me, but if Microsoft can 
> claim
> that .NET is a multi-language VM that provides GC services to any 
> language,
> than it should be possible to do this for Smalltalk as well, shouldn't 
> it?
[...]

I would like to see an open-source universal VM that runs on MacOS, 
Windows, Linux, etc... that is why I want to take the first step in that 
direction by porting a popular byte-code-compiled open-source language 
(Python/Jython) into Squeak, an open-source Smalltalk. 

See my reply to "sma at 3plus4.de":
 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
In article <7MGJOov0nvx+VW3xYgkmsmMz2tDQ at 4ax.com>, sma at 3plus4.de wrote:

> On Sat, 03 Feb 2001 02:57:36 GMT, Keith Ray
> <k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote:
> 
> >Has anyone tried to write a Python interpreter in Smalltalk?
> 
> I tried this, although using Squeak Smalltalk not Dolphin Smalltalk.
> I never finished the project but there's a partially working parser
> which I could share.
> 
> bye
> --
> Stefan Matthias Aust____Truth until Paradox!____________________
>      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de

I have the desire to see Python running inside Squeak 
<http://www.squeak.org/> + <http://www.python.org/>

I was planning to start porting from the Jython sources, since those are 
already in an object oriented language (Java).  If you could, please 
email your parser to me and it will help me learn how Squeak could host 
a Python language. 

C. Keith Ray
remove spaces to get my email address
  c k e i t h r a y @ h o m e . c o m

-- 
       <http://homepage.mac.com/keithray/resume.html>

-- 
       <http://homepage.mac.com/keithray/resume.html>


From andrew at andrewcooke.free-online.co.uk  Thu Feb 15 09:58:16 2001
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Thu, 15 Feb 2001 08:58:16 +0000
Subject: Universal Virtual Machine
References: <Ryyi6.95397$t94.1295979@news1.rdc1.bc.home.com> <k1e2i3t4h5r6a7y-FEE3E8.19234214022001@news>
Message-ID: <3A8B9A28.97BA7DEA@andrewcooke.free-online.co.uk>

Maybe repeating info here, but are you aware of C--?

http://www.cminusminus.org/index.html

Keith Ray wrote:
> I would like to see an open-source universal VM that runs on MacOS,
> Windows, Linux, etc... that is why I want to take the first step in that
> direction by porting a popular byte-code-compiled open-source language
> (Python/Jython) into Squeak, an open-source Smalltalk.


From lopemanc at swbell.net  Thu Feb 15 18:54:57 2001
From: lopemanc at swbell.net (Chris Lopeman)
Date: Thu, 15 Feb 2001 17:54:57 GMT
Subject: Universal Virtual Machine
References: <Ryyi6.95397$t94.1295979@news1.rdc1.bc.home.com> <k1e2i3t4h5r6a7y-FEE3E8.19234214022001@news> <3A8B9A28.97BA7DEA@andrewcooke.free-online.co.uk>
Message-ID: <3A8C1804.1BB614F7@swbell.net>


Andrew Cooke wrote:

> Maybe repeating info here, but are you aware of C--?

I had not heard of this before.  Its kind of funny because I knew Sun people
that would call Java C++--.


>
>
> http://www.cminusminus.org/index.html
>
> Keith Ray wrote:
> > I would like to see an open-source universal VM that runs on MacOS,
> > Windows, Linux, etc... that is why I want to take the first step in that
> > direction by porting a popular byte-code-compiled open-source language
> > (Python/Jython) into Squeak, an open-source Smalltalk.



From jschmitt at vmlabs.com  Thu Feb 15 04:46:19 2001
From: jschmitt at vmlabs.com (John Schmitt)
Date: Wed, 14 Feb 2001 19:46:19 -0800
Subject: __contains__()
Message-ID: <008F0A63472BD311AF9800104BCD102561CC8B@minirex.vmlabs.com>

I did something a little cheezy.

>>>class foo:
       def __contains__(self,n):
           return 42

>>>f = foo()

>>>3 in f
1

I thought I would get back 42.  The manual doesn't say that it must strictly
return 0 or 1, just something true or false.

I wonder what the PSU has to say about this.  How does one become a memb



From greg at cosc.canterbury.ac.nz  Fri Feb 16 04:24:14 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 16 Feb 2001 16:24:14 +1300
Subject: __contains__()
References: <mailman.982208777.26211.python-list@python.org>
Message-ID: <3A8C9D5E.9AEBECB7@cosc.canterbury.ac.nz>

John Schmitt wrote:
> 
> The manual doesn't say that it must strictly
> return 0 or 1, just something true or false.

Which is what you're doing, and it's interpreting it as
"true". It also doesn't say that "x in y" will return
the exact value returned by y's __contains()__ method,
either.

The reason is that, internally, it's being funneled
through the C version of the __contains__ method in
the typeobject, which returns a C int, not a general
Python object. The interpreter then converts that into
a canonical Python boolean value.

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From not.this at seebelow.org  Thu Feb 15 04:54:49 2001
From: not.this at seebelow.org (Grant Griffin)
Date: Wed, 14 Feb 2001 21:54:49 -0600
Subject: recursing glob
Message-ID: <3A8B5309.F5B07C85@seebelow.org>

Hi Gang,

I had a need to recurse "glob" into subdirectories today, and I ended up
coming up with something using os.walk.  It was pretty brief (only a few
lines), but it didn't seem very elegant.  (Somehow I never am happy with
stuff I do with os.walk--is it just me?)  Anyway, does anybody have any
ideas on how to recurse glob more elegantly?

BTW, why isn't subdirectory recursion an option for glob.glob, or
perhaps another function in the "glob" module.  (And how does one
suggest new functions for modules--without writing a PEP? <wink>)

if-they-gave-"glob"-a-more-attractive-name-might-be-able-to-make
   -some-friends-<wink>-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From fredrik at pythonware.com  Thu Feb 15 08:35:44 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Feb 2001 07:35:44 GMT
Subject: recursing glob
References: <3A8B5309.F5B07C85@seebelow.org>
Message-ID: <kFLi6.17197$AH6.2346155@newsc.telia.net>

Grant Griffin wrote:
> Somehow I never am happy with  stuff I do with os.walk--is it
> just me?

no.

> Anyway, does anybody have any ideas on how to recurse glob
> more elegantly?

define "more elegantly" ;-)

the following involves more library code, but is much
easier to use.

(reposted)

From: "Fredrik Lundh" <effbot at telia.com>
Subject: Re: glob
Date: Wed, 10 May 2000 09:35:43 +0200

> Is there a recursive glob?

no, but you can easily roll your own using os.path.walk
and the fnmatch module.

or you can use this iterator class, adapted from an example
in the eff-bot library guide (see below):

#
# adapted from os-path-walk-example-3.py

import os
import fnmatch

class GlobDirectoryWalker:
    # a forward iterator that traverses a directory tree

    def __init__(self, directory, pattern="*"):
        self.stack = [directory]
        self.pattern = pattern
        self.files = []
        self.index = 0

    def __getitem__(self, index):
        while 1:
            try:
                file = self.files[self.index]
                self.index = self.index + 1
            except IndexError:
                # pop next directory from stack
                self.directory = self.stack.pop()
                self.files = os.listdir(self.directory)
                self.index = 0
            else:
                # got a filename
                fullname = os.path.join(self.directory, file)
                if os.path.isdir(fullname) and not os.path.islink(fullname):
                    self.stack.append(fullname)
                if fnmatch.fnmatch(file, self.pattern):
                    return fullname

for file in GlobDirectoryWalker(".", "*.py"):
    print file

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From not.this at seebelow.org  Fri Feb 16 04:58:36 2001
From: not.this at seebelow.org (Grant Griffin)
Date: Thu, 15 Feb 2001 21:58:36 -0600
Subject: recursing glob
References: <3A8B5309.F5B07C85@seebelow.org> <kFLi6.17197$AH6.2346155@newsc.telia.net>
Message-ID: <3A8CA56C.4F1676F1@seebelow.org>

Fredrik Lundh wrote:
> 
> Grant Griffin wrote:
> > Somehow I never am happy with  stuff I do with os.walk--is it
> > just me?
> 
> no.

In retrospect, I think os.walk would be a lot friendlier if it passed
two arguments to its function instead of one.  (When I first began with
Python, I didn't realize that that "one" argument could be a list or a
tuple.)

> 
> > Anyway, does anybody have any ideas on how to recurse glob
> > more elegantly?
> 
> define "more elegantly" ;-)
> 
> the following involves more library code, but is much
> easier to use.
> 
> (reposted)
> 
> From: "Fredrik Lundh" <effbot at telia.com>
> Subject: Re: glob
> Date: Wed, 10 May 2000 09:35:43 +0200
> 
> > Is there a recursive glob?
> 
> no, but you can easily roll your own using os.path.walk
> and the fnmatch module.
> 
> or you can use this iterator class, adapted from an example
> in the eff-bot library guide (see below):
> 
> #
> # adapted from os-path-walk-example-3.py
> 
> import os
> import fnmatch
> 
> class GlobDirectoryWalker:
>     # a forward iterator that traverses a directory tree
> 
>     def __init__(self, directory, pattern="*"):
>         self.stack = [directory]
>         self.pattern = pattern
>         self.files = []
>         self.index = 0
> 
>     def __getitem__(self, index):
>         while 1:
>             try:
>                 file = self.files[self.index]
>                 self.index = self.index + 1
>             except IndexError:
>                 # pop next directory from stack
>                 self.directory = self.stack.pop()
>                 self.files = os.listdir(self.directory)
>                 self.index = 0
>             else:
>                 # got a filename
>                 fullname = os.path.join(self.directory, file)
>                 if os.path.isdir(fullname) and not os.path.islink(fullname):
>                     self.stack.append(fullname)
>                 if fnmatch.fnmatch(file, self.pattern):
>                     return fullname
> 
> for file in GlobDirectoryWalker(".", "*.py"):
>     print file

Cool!  Thanks.

I think that's definitely more elegant than what I did:

    def glob_recurse(args, dirname, names):
        """ args[0] = pattern, args[1] = file list """
        args[1] += glob.glob(os.path.join(dirname, args[0]))

and then:

    os.path.walk('.', glob_recurse, [pattern, file_names])

I think this would have been pretty easy for a more seasoned Pythoneer
to cough up, but it took me awhile.  In particular, I tried to pass it a
tuple at first, and it took me awhile to figure out the source of the
error message.

we-horses-don't-always-drink-the-water-we're-lead-to-if-it's-not
   -clearly-labeled-as-such-<wink>-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From dwig at advancedmp.net  Fri Feb 16 19:46:31 2001
From: dwig at advancedmp.net (Don Dwiggins)
Date: 16 Feb 2001 10:46:31 -0800
Subject: recursing glob
In-Reply-To: Grant Griffin's message of "Thu, 15 Feb 2001 21:58:36 -0600"
References: <3A8B5309.F5B07C85@seebelow.org>
	<kFLi6.17197$AH6.2346155@newsc.telia.net>
	<3A8CA56C.4F1676F1@seebelow.org>
Message-ID: <uelwycvp4.fsf@advancedmp.net>

Grant Griffin writes:
> Fredrik Lundh wrote:
>> 
>> Grant Griffin wrote:
>> > Somehow I never am happy with  stuff I do with os.walk--is it
>> > just me?
>> 
>> no.

> In retrospect, I think os.walk would be a lot friendlier if it passed
> two arguments to its function instead of one.  (When I first began with
> Python, I didn't realize that that "one" argument could be a list or a
> tuple.)

Maybe what's needed here is an implementation of the Visitor pattern from
the Gamma et al. "Design Patterns" book.  I did a file system Visitor in
Visual Basic (pardon my French!) as a learning exercise a while back, and
found it easy to implement and use once I grokked it (should be much nicer
in Python, though).  It's worth considering (not necessarily choosing)
Visitor whenever you need to "iterate" over a complex structure.

-- 
Don Dwiggins                    "Solvitur Ambulando"
Advanced MP Technology
dwig at advancedmp.net




From theller at python.net  Thu Feb 15 09:52:04 2001
From: theller at python.net (Thomas Heller)
Date: Thu, 15 Feb 2001 09:52:04 +0100 (CET)
Subject: [ANN] py2exe - new version
Message-ID: <982227124.3a8b98b4a366f@www.ion-tof.com>

py2exe is a distutils extension to convert python scripts into standalone
windows executables.

The URL is: http://starship.python.net/crew/theller/py2exe/

I've released version 0.2.3 of py2exe, which seems pretty stable, although
it should still be considered beta.

It has been used for creating wxPython, PyGTK, pygame, win32com client and
other standalone programs.

I would like to thank the early adopters(in alphabetical order)
Dan Rolander, Dave Brueck, Franz Geiger, Markus Gritsch,
Pete Shinners, Peter Hansen, Robin Dunn, and Wayne Izatt
for helpfull suggestions, testing and support.

Development will continue. Features planned for the next release(s):
- More aggressive way to find dependencies, hopefully making the building
  process more automatic
- COM server support (localserver, inprocserver)

Regards,

Thomas



From robin at jessikat.fsnet.co.uk  Thu Feb 15 12:16:23 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 15 Feb 2001 11:16:23 +0000
Subject: [ANN] py2exe - new version
References: <mailman.982227198.28101.python-list@python.org>
Message-ID: <4$+f4LAHq7i6Ew1H@jessikat.demon.co.uk>

In article <mailman.982227198.28101.python-list at python.org>, Thomas
Heller <theller at python.net> writes
>py2exe is a distutils extension to convert python scripts into standalone
>windows executables.
>
>The URL is: http://starship.python.net/crew/theller/py2exe/
>
>I've released version 0.2.3 of py2exe, which seems pretty stable, although
>it should still be considered beta.
>
>It has been used for creating wxPython, PyGTK, pygame, win32com client and
>other standalone programs.
>
>I would like to thank the early adopters(in alphabetical order)
>Dan Rolander, Dave Brueck, Franz Geiger, Markus Gritsch,
>Pete Shinners, Peter Hansen, Robin Dunn, and Wayne Izatt
>for helpfull suggestions, testing and support.
>
>Development will continue. Features planned for the next release(s):
>- More aggressive way to find dependencies, hopefully making the building
>  process more automatic
>- COM server support (localserver, inprocserver)
>
>Regards,
>
>Thomas
>
I've not tried this yet, but what about using the checking stuff from
Gordon McMillan's Installer. The only problems that I've been getting
using it are reasonably unusual. In particular it has problems with

package\_init__.py contains
from bingo import BINGO

as it when it sees 'from package import BINGO' it doesn't know to look
in bingo.py 

Gordon's compressed import packages are really neat and I hacked the c
stuff to clean up more aggressively so users don't see any dlls lying
about after use.
-- 
Robin Becker


From thomas.heller at ion-tof.com  Thu Feb 15 14:46:56 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 15 Feb 2001 14:46:56 +0100
Subject: [ANN] py2exe - new version
References: <mailman.982227198.28101.python-list@python.org> <4$+f4LAHq7i6Ew1H@jessikat.demon.co.uk>
Message-ID: <96gmkn$l1lp6$1@ID-59885.news.dfncis.de>

"Robin Becker" <robin at jessikat.fsnet.co.uk> wrote in message news:4$+f4LAHq7i6Ew1H at jessikat.demon.co.uk...
> I've not tried this yet, but what about using the checking stuff from
> Gordon McMillan's Installer.

It IS in fact based on the same code than Gordon's installer,
which is tools\freeze\modulefinder...


> The only problems that I've been getting
> using it are reasonably unusual. In particular it has problems with
>
> package\_init__.py contains
> from bingo import BINGO
>
> as it when it sees 'from package import BINGO' it doesn't know to look
> in bingo.py
>

... and so it has the same problems with your example. Thanks for pointing
this out! I will look into this.

> Gordon's compressed import packages are really neat and I hacked the c
> stuff to clean up more aggressively so users don't see any dlls lying
> about after use.

The reason I wrote py2exe are:
- Integration with distutils seems natural for me
- There _have_ been some problems with installer and python 2.0
  (maybe they are solved in th meantime).
- There _have_ been problems with compiled extensions contained
  in packages
- There are still (?) problems with win32all/win32com
- Writing code myself is always more fun than hacking other's code
- maybe others I don' remeber at the moment

Thomas




From slhath at home.com.nospam  Thu Feb 15 14:07:31 2001
From: slhath at home.com.nospam (Scott Hathaway)
Date: Thu, 15 Feb 2001 13:07:31 GMT
Subject: [ANN] py2exe - new version
References: <mailman.982227198.28101.python-list@python.org>
Message-ID: <nwQi6.2107$Sx5.157702@news.flash.net>

Thomas,

This is awesome!  Being new to python, I have tried Gordon's Installer and
have had a lot of trouble with it
(this is not a problem with Installer, but my lack of understanding python,
I am sure...)
I was able to create an .exe for a fairly complex project that uses xml-rpc
and about 20 class files in about 10 minutes
after the download!

Thanks so much,
Scott

"Thomas Heller" <theller at python.net> wrote in message
news:mailman.982227198.28101.python-list at python.org...
> py2exe is a distutils extension to convert python scripts into standalone
> windows executables.
>
> The URL is: http://starship.python.net/crew/theller/py2exe/
>
> I've released version 0.2.3 of py2exe, which seems pretty stable, although
> it should still be considered beta.
>
> It has been used for creating wxPython, PyGTK, pygame, win32com client and
> other standalone programs.
>
> I would like to thank the early adopters(in alphabetical order)
> Dan Rolander, Dave Brueck, Franz Geiger, Markus Gritsch,
> Pete Shinners, Peter Hansen, Robin Dunn, and Wayne Izatt
> for helpfull suggestions, testing and support.
>
> Development will continue. Features planned for the next release(s):
> - More aggressive way to find dependencies, hopefully making the building
>   process more automatic
> - COM server support (localserver, inprocserver)
>
> Regards,
>
> Thomas
>




From MailFreeZone  Thu Feb 15 10:10:59 2001
From: MailFreeZone (OpenMined)
Date: Thu, 15 Feb 2001 01:10:59 -0800
Subject: What's the "best" Python?
Message-ID: <3a8b9b18_1@news4.newsfeeds.com>

Planning to take the big plunge and invest time in a new language- Python,
yer "it"!

Prefer to learn "going in" and not in hindsight what the hidden 'gotchas'
may be.

Will eventually be programming on/for Linux systems, but will start using
Python on Win98.

*** Is any particular version or variant the wisest choice, i.e., fewest
bugs, most stable release?***

Thanks




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From thomas at xs4all.net  Thu Feb 15 11:07:27 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Thu, 15 Feb 2001 11:07:27 +0100
Subject: What's the "best" Python?
In-Reply-To: <3a8b9b18_1@news4.newsfeeds.com>; from Mail.Free.Zone@xs4all.nl on Thu, Feb 15, 2001 at 01:10:59AM -0800
References: <3a8b9b18_1@news4.newsfeeds.com>
Message-ID: <20010215110727.K4924@xs4all.nl>

On Thu, Feb 15, 2001 at 01:10:59AM -0800, OpenMined wrote:

> *** Is any particular version or variant the wisest choice, i.e., fewest
> bugs, most stable release?***

The latest release, obviously. It is the only one that will get its bugs
fixed :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From LauraL  Thu Feb 15 14:58:50 2001
From: LauraL (LauraL)
Date: 15 Feb 2001 13:58:50 GMT
Subject: What's the best Python?
References: <3a8b9b18_1@news4.newsfeeds.com>
Message-ID: <96gnaq01219@news2.newsguy.com>

Hi,
If you go to oreillynet.com, there's a new article on "alternative python distributions" (in the Python dev center) that should be helpful to you.
Laura
LLewin at oreilly.com

> Planning to take the big plunge and invest time in a new language- Python,
> yer "it"!
> 
> Prefer to learn "going in" and not in hindsight what the hidden 'gotchas'
> may be.
> 
> Will eventually be programming on/for Linux systems, but will start using
> Python on Win98.
> 
> *** Is any particular version or variant the wisest choice, i.e., fewest
> bugs, most stable release?***
> 
> Thanks
> 


==================================
Posted via http://nodevice.com
Linux Programmer's Site


From claird at starbase.neosoft.com  Fri Feb 16 15:05:48 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 16 Feb 2001 08:05:48 -0600
Subject: What's the best Python?
References: <3a8b9b18_1@news4.newsfeeds.com> <96gnaq01219@news2.newsguy.com>
Message-ID: <29F094D6FAA03AF7.460EE66A0A7774DE.DBB4260F3E0CA12D@lp.airnews.net>

In article <96gnaq01219 at news2.newsguy.com>,  <LauraL> wrote:
>Hi,
>If you go to oreillynet.com, there's a new article on "alternative python distributions" (in the Python dev center) that should be helpful to you.
			.
			.
			.
To my surprise, I agree, enough that I'll reproduce
<URL: http://www.oreillynet.com/pub/a/python/2001/02/14/pythonnews.html >
here for the convenience of those curious about it.
The surprise isn't that O'ReillyNet published some-
thing "that should be helpful to you"; it's just
that I hadn't noticed it, and it seems so perfectly
timed as an answer to the original poster.

One of the benefits of this piece is that Stephen's
a pretty responsive guy who follows this newsgroup,
and answers his e-mail.  If you have questions or
comments about what you find in his piece, he's
likely to answer them.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From aahz at panix.com  Thu Feb 15 15:42:05 2001
From: aahz at panix.com (Aahz Maruch)
Date: 15 Feb 2001 06:42:05 -0800
Subject: What's the "best" Python?
References: <3a8b9b18_1@news4.newsfeeds.com>
Message-ID: <96gprt$lif$1@panix2.panix.com>

In article <3a8b9b18_1 at news4.newsfeeds.com>, OpenMined <Mail Free Zone> wrote:
>
>*** Is any particular version or variant the wisest choice, i.e., fewest
>bugs, most stable release?***

Overall, Python 2.0 is the best bet.  Python 2.1 is in alpha, with
expected release in April.  However, if you're a Perler with a yen for
complex regexes, Python 1.5.2 may be your better choice until Python 2.1
gets released.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

'Gender' isn't a definition, it's a heuristic.  --Aahz


From rytrom at zootec.com  Thu Feb 15 11:52:43 2001
From: rytrom at zootec.com (Roman Rytov)
Date: Thu, 15 Feb 2001 12:52:43 +0200
Subject: remote debuging?
Message-ID: <newscache$b2ps8g$qql$1@lnews.actcom.co.il>

Is there any soft alllowing to debug Jython scripts remotely (Unix
preferable)





From jurgen.defurne at philips.com  Thu Feb 15 13:29:39 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Thu, 15 Feb 2001 13:29:39 +0100
Subject: Memory leak when spawning
Message-ID: <0056900015941884000002L042*@MHS>

I don't know if this piece of source is small enough, but it is probably
the smallest I can create with the same symptoms : every time
a spawn is made, and the spawned program stops, some 4k to 12k
of memory is lost (not 800 to 1200 as I said last time).

If the originating process stops, then all memory is returned, but
a program which has cron functionality should of course not display
memory loss.

This is the version of Python I use (downloaded binary)

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32

Here is the code :


Jurgen
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cron.py
Type: application/octet-stream
Size: 718 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010215/e02b83df/attachment.obj>

From MarkH at ActiveState.com  Thu Feb 15 13:46:42 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 15 Feb 2001 12:46:42 GMT
Subject: Memory leak when spawning
References: <mailman.982239616.8088.python-list@python.org>
Message-ID: <3A8BCED3.50000@ActiveState.com>

jurgen.defurne at philips.com wrote:

> I don't know if this piece of source is small enough, but it is probably
> the smallest I can create with the same symptoms 

Hrm.  Not really I am afraid.  I dont see any leaks running this code.  You should be able to trim it down to something with a tight loop (ie, very quick or no sleeping) and see the memory usage rise pretty quickly.

Also, note that your code:
    (None, None, None,
     None, None, sec,
     None, None, None) = localtime(time())

Is very very dangerous - it is changing the builtin value None!!

Mark.



From stephen_purcell at yahoo.com  Thu Feb 15 15:31:22 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Thu, 15 Feb 2001 15:31:22 +0100
Subject: Memory leak when spawning
In-Reply-To: <3A8BCED3.50000@ActiveState.com>; from MarkH@ActiveState.com on Thu, Feb 15, 2001 at 12:46:42PM +0000
References: <mailman.982239616.8088.python-list@python.org> <3A8BCED3.50000@ActiveState.com>
Message-ID: <20010215153122.A479@freedom.puma-ag.com>

Mark Hammond wrote:
> Also, note that your code:
>     (None, None, None,
>      None, None, sec,
>      None, None, None) = localtime(time())
> 
> Is very very dangerous - it is changing the builtin value None!!

Indeed. That got me thinking; is it defined how the following should work?

  (blah, blah, blah) = (1,2,3)

I get blah == 3.

I won't lose sleep over it...

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From tim_one at email.msn.com  Fri Feb 16 01:04:34 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 15 Feb 2001 19:04:34 -0500
Subject: Memory leak when spawning
In-Reply-To: <20010215153122.A479@freedom.puma-ag.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEKNIPAA.tim_one@email.msn.com>

[Steve Purcell]
> Indeed. That got me thinking; is it defined how the following
> should work?
>
>   (blah, blah, blah) = (1,2,3)
>
> I get blah == 3.

Yes, left-to-right is the defined order of sequence assignment.

> I won't lose sleep over it...

You would if you thought more about it <wink>.  For example,

>>> a = [0]*5
>>> i, a[i] = 2, 666
>>> a
[0, 0, 666, 0, 0]
>>>

Python (unlike most other languages) defines this stuff so you don't get
surprises across releaess or platforms, but some people just can't resist
the temptation to *rely* on obscure side effects.  Luckily, they're tracked
down and killed just as fast as the PSU finds t




From juanjux at yahoo.es  Sat Feb 17 05:13:29 2001
From: juanjux at yahoo.es (Juanjo =?iso-8859-1?Q?=C1lvarez?=)
Date: Sat, 17 Feb 2001 05:13:29 +0100
Subject: Memory leak when spawning
References: <mailman.982239616.8088.python-list@python.org> <3A8BCED3.50000@ActiveState.com> <mailman.982247539.6281.python-list@python.org>
Message-ID: <slrn98ruj9.6qs.juanjux@Pascual.es>

Steve Purcell dijo sobre Re: Memory leak when spawning,
en fecha: Thu, 15 Feb 2001 15:31:22 +0100

|Mark Hammond wrote:
|> Also, note that your code:
|>     (None, None, None,
|>      None, None, sec,
|>      None, None, None) = localtime(time())
|> 
|> Is very very dangerous - it is changing the builtin value None!!
|
|Indeed. That got me thinking; is it defined how the following should work?
|
|  (blah, blah, blah) = (1,2,3)

I prefer (foo, foo, foo) = (1,2,3), it's more traditional.


From chris.gonnerman at usa.net  Mon Feb 19 04:11:58 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sun, 18 Feb 2001 21:11:58 -0600
Subject: Memory leak when spawning
References: <mailman.982239616.8088.python-list@python.org> <3A8BCED3.50000@ActiveState.com> <mailman.982247539.6281.python-list@python.org> <slrn98ruj9.6qs.juanjux@Pascual.es>
Message-ID: <012f01c09a21$ba010080$a100000a@local>

----- Original Message -----
From: "Juanjo ?lvarez" <juanjux at yahoo.es>
Subject: Re: Memory leak when spawning


> Steve Purcell dijo sobre Re: Memory leak when spawning,
> en fecha: Thu, 15 Feb 2001 15:31:22 +0100
>
> |Mark Hammond wrote:
> |> Also, note that your code:
> |>     (None, None, None,
> |>      None, None, sec,
> |>      None, None, None) = localtime(time())
> |>
> |> Is very very dangerous - it is changing the builtin value None!!
> |
> |Indeed. That got me thinking; is it defined how the following should
work?
> |
> |  (blah, blah, blah) = (1,2,3)
>
> I prefer (foo, foo, foo) = (1,2,3), it's more traditional.

To be traditional, shouldn't it be:

(spam, spam, spam) = (1, 2, 3)

pythonically-yr's  --  Chris.





From piet at cs.uu.nl  Wed Feb 21 15:27:03 2001
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 21 Feb 2001 15:27:03 +0100
Subject: None (was Re: Memory leak when spawning)
References: <mailman.982239616.8088.python-list@python.org> <3A8BCED3.50000@ActiveState.com>
Message-ID: <u1yssi020.fsf_-_@cs.uu.nl>

>>>>> Mark Hammond <MarkH at ActiveState.com> (MH) writes:

MH> jurgen.defurne at philips.com wrote:
>> I don't know if this piece of source is small enough, but it is probably
>> the smallest I can create with the same symptoms

MH> Hrm. Not really I am afraid. I dont see any leaks running this code.
MH> You should be able to trim it down to something with a tight loop (ie,
MH> very quick or no sleeping) and see the memory usage rise pretty
MH> quickly.

MH> Also, note that your code:
MH>     (None, None, None,
MH>      None, None, sec,
MH>      None, None, None) = localtime(time())

MH> Is very very dangerous - it is changing the builtin value None!!

Wouldn't this be a good reason to make None a constant rather than a
built-in identifier?
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From aahz at panix.com  Thu Feb 15 15:44:39 2001
From: aahz at panix.com (Aahz Maruch)
Date: 15 Feb 2001 06:44:39 -0800
Subject: Memory leak when spawning
References: <mailman.982239616.8088.python-list@python.org>
Message-ID: <96gq0n$m1c$1@panix2.panix.com>

In article <mailman.982239616.8088.python-list at python.org>,
 <jurgen.defurne at philips.com> wrote:
>
>I don't know if this piece of source is small enough, but it is
>probably the smallest I can create with the same symptoms : every time
>a spawn is made, and the spawned program stops, some 4k to 12k of
>memory is lost (not 800 to 1200 as I said last time).

Tim Peters posted recently that a change was made in Python 2.0 such
that a new GIL object was created in spawned programs.  However, this
should not affect the spawning program, so I don't know what's causing
your problem.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

'Gender' isn't a definition, it's a heuristic.  --Aahz


From jurgen.defurne at philips.com  Thu Feb 15 14:19:16 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Thu, 15 Feb 2001 14:19:16 +0100
Subject: Memory leak when spawning
Message-ID: <0056900015944653000002L032*@MHS>

I thought that I could use None, like 'undef' in Perl.

Any suggestions on that ?

I'll implement the change and try again.

Thanks.

Jurgen




MarkH at ActiveState.com@SMTP at python.org on 15/02/2001 14:04:39
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	Re: Memory leak when spawning
Classification:	

jurgen.defurne at philips.com wrote:

> I don't know if this piece of source is small enough, but it is probably
> the smallest I can create with the same symptoms

Hrm.  Not really I am afraid.  I dont see any leaks running this code.  You should be able to trim it down to something with a tight loop (ie, very quick or no sleeping) and see the memory usage rise pretty quickly.

Also, note that your code:
    (None, None, None,
     None, None, sec,
     None, None, None) = localtime(time())

Is very very dangerous - it is changing the builtin value None!!

Mark.

--
http://mail.python.org/mailman/listinfo/python-list





From erik.wilsher at iname.com  Thu Feb 15 15:25:28 2001
From: erik.wilsher at iname.com (erik.wilsher at iname.com)
Date: Thu, 15 Feb 2001 14:25:28 -0000
Subject: Memory leak when spawning
In-Reply-To: <0056900015944653000002L032*@MHS>
Message-ID: <96goso+2cnv@eGroups.com>

--- In python-list at y..., jurgen.defurne at p... wrote:
> Any suggestions on that ?

    secs = localtime(time())[5]

    EW




From tim_one at email.msn.com  Fri Feb 16 01:06:48 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 15 Feb 2001 19:06:48 -0500
Subject: Memory leak when spawning
In-Reply-To: <0056900015944653000002L032*@MHS>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEKNIPAA.tim_one@email.msn.com>

[jurgen.defurne at philips.com]
> I thought that I could use None, like 'undef' in Perl.

You can, but it will rebind the name None.  There is no "don't care"
assignment target in Python.

> Any suggestions on that ?

Use the name undef, if you like:

    undef, middle, undef = range(3)

jurgen also works:

    jurgen, middle, jurgen = range(3)

However, don't even think about using "tim" for this purpose <wink>.




From chris.gonnerman at usa.net  Fri Feb 16 03:23:09 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 15 Feb 2001 20:23:09 -0600
Subject: Memory leak when spawning
References: <0056900015944653000002L032*@MHS>
Message-ID: <00a401c097bf$722be180$a100000a@local>

----- Original Message ----- 
From: <jurgen.defurne at philips.com>
Subject: Re: Memory leak when spawning
>

>I thought that I could use None, like 'undef' in Perl.
>
>Any suggestions on that ?
>

What's wrong with:

sec = localtime(time())[5]







From jurgen.defurne at philips.com  Thu Feb 15 15:08:36 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Thu, 15 Feb 2001 15:08:36 +0100
Subject: Memory leak when spawning
Message-ID: <0056900015947222000002L022*@MHS>

Allright, I think this should definitely be the smallest program I can come
up with. I tested it, memory leaks like a rocket going up.

#

# External modules
import os
import sys

from time import *

if len(sys.argv) == 1:
    while 1:
	# Doesn't matter if P_NOWAIT or P_DETACH, tried 'em both
	os.spawnl(os.P_DETACH, 'd:\\python20\\python', 'python', 'cron.py', 'check')
elif sys.argv[1] == 'check':
    pass



Jurgen
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cron.py
Type: application/octet-stream
Size: 288 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010215/30ca414c/attachment.obj>

From MarkH at ActiveState.com  Thu Feb 15 23:20:32 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 15 Feb 2001 22:20:32 GMT
Subject: Memory leak when spawning
References: <mailman.982248256.9121.python-list@python.org>
Message-ID: <3A8C5552.90605@ActiveState.com>

jurgen.defurne at philips.com wrote:

> Allright, I think this should definitely be the smallest program I can come
> up with. I tested it, memory leaks like a rocket going up.

Not for me :(  What version of Python?  Maybe it has been fixed in 2.1?

Mark.



From Gareth.McCaughan at pobox.com  Fri Feb 16 18:52:00 2001
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Fri, 16 Feb 2001 17:52:00 +0000
Subject: Memory leak when spawning
References: <mailman.982282096.19931.python-list@python.org>
Message-ID: <slrn98qq60.bp.Gareth.McCaughan@g.local>

Tim Peters wrote:

>                                             There is no "don't care"
> assignment target in Python.
> 
> > Any suggestions on that ?
> 
> Use the name undef, if you like:

I like "_".

    >>> _,_,x,_,_ = 1,2,3,4,5
    >>> x
    3

Easy to type, visually mnemonic, and using _ as a disposable
variable fits well with its use in interactive mode.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construc


From mwh21 at cam.ac.uk  Fri Feb 16 19:30:07 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Feb 2001 18:30:07 +0000
Subject: Memory leak when spawning
References: <mailman.982282096.19931.python-list@python.org> <slrn98qq60.bp.Gareth.McCaughan@g.local>
Message-ID: <m3itmaeb0w.fsf@atrus.jesus.cam.ac.uk>

Gareth.McCaughan at pobox.com (Gareth McCaughan) writes:

> Tim Peters wrote:
> 
> >                                             There is no "don't care"
> > assignment target in Python.
> > 
> > > Any suggestions on that ?
> > 
> > Use the name undef, if you like:
> 
> I like "_".

So do I, but the i18n folks tend to shout at you.

Cheers,
M.

-- 
  The PROPER way to handle HTML postings is to cancel the article,
  then hire a hitman to kill the poster, his wife and kids, and fuck
  his dog and smash his computer into little bits. Anything more is
  just extremism.                 -- attributed to Paul Tomblin on ASR


From Gareth.McCaughan at pobox.com  Sun Feb 18 20:00:07 2001
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Sun, 18 Feb 2001 19:00:07 +0000
Subject: Memory leak when spawning
References: <mailman.982282096.19931.python-list@python.org> <slrn98qq60.bp.Gareth.McCaughan@g.local> <m3itmaeb0w.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn9906tn.3ts.Gareth.McCaughan@g.local>

Michael Hudson wrote:

> Gareth.McCaughan at pobox.com (Gareth McCaughan) writes:
> 
> > Tim Peters wrote:
> > 
> > >                                             There is no "don't care"
> > > assignment target in Python.
> > > 
> > > > Any suggestions on that ?
> > > 
> > > Use the name undef, if you like:
> > 
> > I like "_".
> 
> So do I, but the i18n folks tend to shout at you.

Oh. Why? (I'm shamefully ignorant of i18n issues...)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construc


From mwh21 at cam.ac.uk  Mon Feb 19 01:20:49 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Feb 2001 00:20:49 +0000
Subject: Memory leak when spawning
References: <mailman.982282096.19931.python-list@python.org> <slrn98qq60.bp.Gareth.McCaughan@g.local> <m3itmaeb0w.fsf@atrus.jesus.cam.ac.uk> <slrn9906tn.3ts.Gareth.McCaughan@g.local>
Message-ID: <m366i7v7z2.fsf@atrus.jesus.cam.ac.uk>

Gareth.McCaughan at pobox.com (Gareth McCaughan) writes:

> Michael Hudson wrote:
> 
> > Gareth.McCaughan at pobox.com (Gareth McCaughan) writes:
> > 
> > > Tim Peters wrote:
> > > 
> > > >                                             There is no "don't care"
> > > > assignment target in Python.
> > > > 
> > > > > Any suggestions on that ?
> > > > 
> > > > Use the name undef, if you like:
> > > 
> > > I like "_".
> > 
> > So do I, but the i18n folks tend to shout at you.
> 
> Oh. Why? (I'm shamefully ignorant of i18n issues...)

'cause they do things like this:

import gettext ; _ = gettext.gettext

...

print _("This is a foo message!")

to get translated strings.

I personally think prolog got there first, but I can see their point.

Cheers,
M.

-- 
  I'm sorry, was my bias showing again? :-) 
                                      -- William Tanksley, 13 May 2000


From Gareth.McCaughan at pobox.com  Tue Feb 20 20:34:12 2001
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: Tue, 20 Feb 2001 19:34:12 +0000
Subject: Memory leak when spawning
References: <mailman.982282096.19931.python-list@python.org> <slrn98qq60.bp.Gareth.McCaughan@g.local> <m3itmaeb0w.fsf@atrus.jesus.cam.ac.uk> <slrn9906tn.3ts.Gareth.McCaughan@g.local> <m366i7v7z2.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <slrn995hlk.108.Gareth.McCaughan@g.local>

Michael Hudson wrote:

[me, him, me, him:]
> > > > I like "_".
> > > 
> > > So do I, but the i18n folks tend to shout at you.
> > 
> > Oh. Why? (I'm shamefully ignorant of i18n issues...)
> 
> 'cause they do things like this:
> 
> import gettext ; _ = gettext.gettext
> 
> ...
> 
> print _("This is a foo message!")
> 
> to get translated strings.

Ohhh! I see. Thanks.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construc


From bryan at eevolved.com  Thu Feb 15 13:35:48 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Thu, 15 Feb 2001 12:35:48 GMT
Subject: Gauging system load
Message-ID: <E2Qi6.237705$f36.9512658@news20.bellglobal.com>

Out of curiosity, does anyone know of a module or a neat trick that will 
permit me to gauge the system load at a given time? For example, in a 
utopian dream world:

import sys, time

# A percentage value
THRESHOLD = 20

while 1:
  if sys.cpuActivity <  THRESHOLD:
    # Do some heavy work
    break
  else
    time.sleep(10)


I've thought of maybe spawning a thread and gauging the amount of time it 
takes for it to be serviced, but that's pretty crude and innaccurate.

Any tips?
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"I want to know God's thoughts; the rest are details."-- Einstein



From annis at biostat.wisc.edu  Thu Feb 15 16:17:16 2001
From: annis at biostat.wisc.edu (William Annis)
Date: 15 Feb 2001 09:17:16 -0600
Subject: Gauging system load
References: <E2Qi6.237705$f36.9512658@news20.bellglobal.com>
Message-ID: <jtsnlgt1qb.fsf@wazor.biostat.wisc.edu>

Bryan Mongeau <bryan at eevolved.com> writes:

> Out of curiosity, does anyone know of a module or a neat trick that will 
> permit me to gauge the system load at a given time?

        What OS?  I have no idea how you'd do this on a PC or a Mac,
but on any unix flavor you should be able to parse the output from
uptime.  And if you're running Solaris, you can grab my PyKstat
package and rip the information right out of the kernel with no
parsing at all.

        http://www.biostat.wisc.edu/~annis/creations/pykstat.html

Under Linux, you can parse the contents of /proc/loadavg.

-- 
William Annis - System Administrator - Biomedical Computing Group
annis at biostat.wisc.edu                       PGP ID:1024/FBF64031
Mi parolas Esperanton - La Internacian Lingvon  www.esperanto.org


From jafo at tummy.com  Thu Feb 15 23:59:57 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Thu, 15 Feb 2001 15:59:57 -0700
Subject: Gauging system load
In-Reply-To: <jtsnlgt1qb.fsf@wazor.biostat.wisc.edu>; from annis@biostat.wisc.edu on Thu, Feb 15, 2001 at 09:17:16AM -0600
References: <E2Qi6.237705$f36.9512658@news20.bellglobal.com> <jtsnlgt1qb.fsf@wazor.biostat.wisc.edu>
Message-ID: <20010215155957.F23577@tummy.com>

On Thu, Feb 15, 2001 at 09:17:16AM -0600, William Annis wrote:
>but on any unix flavor you should be able to parse the output from
>uptime.  And if you're running Solaris, you can grab my PyKstat

Uptime isn't that good of a metric.  It only represents the number of
processes waiting to run, and is often highly affected by disc I/O.
If are doing a CPU-intensive task on Linux, you can process /proc/stat
to find out what % of CPU utilization is going on.  Probably want to
check on disc IO too, if that's important.

Sean
-- 
 George Washington was first in war, first in peace -- and first to
 have his birthday juggled to make a long weekend.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From norman at earth.ox.ac.uk  Thu Feb 15 13:47:37 2001
From: norman at earth.ox.ac.uk (Norman Charnley)
Date: Thu, 15 Feb 2001 12:47:37 +0000
Subject: PIL under Windows
Message-ID: <3A8BCFE9.729F19B4@earth.ox.ac.uk>

I'd like to run Python Imaging Library under Win95. I've already
installed the Python 2.0 distribution without problem, in C:\Python20
(along with Pmw, Blt, Numeric and Medusa).

However, when looking for PIL to install, I followed the links from the
Python homepage and I ended up at Pythonworks, which offers another
almost complete Python distribution, which installed under C:\Py20, but
without Tools (e.g. Idle). 

Is there an easy way to reconcile the two? - preferably by amalgamating
them under C:\python20, since various other links and registry settings
are already associated with that directory...

Thanks,

Norman

 =================================================
 Dr. Norman Charnley
 Department of Earth Sciences
 University of Oxford
 Oxford OX1 3PR, UK.
==================================================


From clee at gnwy100.wuh.wustl.edu  Thu Feb 15 18:18:13 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 15 Feb 2001 11:18:13 -0600
Subject: PIL under Windows
References: <3A8BCFE9.729F19B4@earth.ox.ac.uk>
Message-ID: <7u4rxvlvai.fsf@gnwy100.wuh.wustl.edu>

>>>>> "Norman" == Norman Charnley <norman at earth.ox.ac.uk> writes:

    Norman> However, when looking for PIL to install, I followed the links
    Norman> from the Python homepage and I ended up at Pythonworks, which
    Norman> offers another almost complete Python distribution, which
    Norman> installed under C:\Py20, but without Tools (e.g. Idle).

    Norman> Is there an easy way to reconcile the two? - preferably by
    Norman> amalgamating them under C:\python20, since various other links
    Norman> and registry settings are already associated with that
    Norman> directory...

If you've already downloaded py20 from pythonware, you can copy all PIL
related files over from c:\py20 to the equivalent place in your c:\Python20
directory. (I believe the files are something like PIL.pth, the entire PIL
subdirectory, _imaging.dll and _tkimaging.dll.)

If you haven't already downloaded py20 and you don't want all the extra's
it gives, you might want to grab the one of the win32 versions available at
http://piddle.sourceforge.net/PIL/ and unzip it under your c:\python20
folder.

-chris


From olafb+usenet at pvv.org  Thu Feb 15 14:53:51 2001
From: olafb+usenet at pvv.org (Olaf Trygve Berglihn)
Date: 15 Feb 2001 14:53:51 +0100
Subject: httplib + FakeSocket: how not to download whole file in memory
Message-ID: <dn1g0hgnjbk.fsf@lizard.pvv.ntnu.no>

I'm using httplib to download files over https. However, the
FakeSocket-construction loads the whole file in memory and returns it
as a StringIO-object.

Does anyone have a suggestion how to avoid this, i.e. get a normal
filepointer construction from the makefile-method and not grabbing all
my memory.

Olaf
-- 
Olaf Trygve Berglihn <olafb+usenet at pvv.org>


From gmunsey at adobe.com  Thu Feb 15 15:37:24 2001
From: gmunsey at adobe.com (Grant Munsey)
Date: Thu, 15 Feb 2001 06:37:24 -0800
Subject: Problems with "a+" open mode on v2.0 Windows?
Message-ID: <FQRi6.2413$C92.256142@news.pacbell.net>

On the win32 BeOpen Python 2.0 I have a strange thing happening.
(800 Mhz Win2k Pro)

The code below (abstracted from my actual app) dies randomly with
an io error 13 trying to open hdr1fname. This error happens about 20% of
the time. Not always on the same file name.

If I go and look at the file every thing is fine ... no access errors.
It's as if out1fn.close() is delayed some and the next call can't reopen the
file
for some reason.

Any clue to what's going on?

filelist = <10s of file names>

for i in filelist:
        parseFileEx(i)

def parsefileEx(hdr1fname):
        out1fn = open(hdr1fname, "a+")
        out1fn.write("blah\n");
        ... lots more writing ...
        out1fn.close()






From emile at fenx.com  Thu Feb 15 15:39:58 2001
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 15 Feb 2001 06:39:58 -0800
Subject: Missing Post
Message-ID: <96gpqe$l29g1$1@ID-11957.news.dfncis.de>

Over the last couple of days I've noticed a significant drop
off in posts, something like one half to one third of what
I've come to expect.  I'm also seeing replies to post I'm
not seeing.

Anyone else have this going on?

I'm pulling from news.cis.dfn.de.

Thanks,

--

Emile van Sebille
emile at fenx.com
-------------------






From MailFreeZone  Thu Feb 15 18:13:16 2001
From: MailFreeZone (OpenMined)
Date: Thu, 15 Feb 2001 09:13:16 -0800
Subject: Missing Post
References: <96gpqe$l29g1$1@ID-11957.news.dfncis.de>
Message-ID: <3a8c0c16_1@news4.newsfeeds.com>

Yes-- seeing replies to my post of last night (Thanks!) but not the
original.  My news server should be mentioned in the header- not the same
one.

Somebody's server is being naughty.

"Emile van Sebille" <emile at fenx.com> wrote in message
news:96gpqe$l29g1$1 at ID-11957.news.dfncis.de...
> Over the last couple of days I've noticed a significant drop
> off in posts, something like one half to one third of what
> I've come to expect.  I'm also seeing replies to post I'm
> not seeing.
>
> Anyone else have this going on?
>
> I'm pulling from news.cis.dfn.de.
>
> Thanks,
>
> --
>
> Emile van Sebille
> emile at fenx.com
> -------------------
>
>
>
>




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From Burkhard.Kayser at marconi.com  Thu Feb 15 17:29:35 2001
From: Burkhard.Kayser at marconi.com (Burkhard Kayser)
Date: Thu, 15 Feb 2001 17:29:35 +0100
Subject: WaitCommEvent
Message-ID: <96h007$kri$1@newsfeed.pit.comms.marconi.com>

Does anyone know, where I can find the documentation for this function ?

It is used in the example win32comport_demo.py.

rc, mask = WaitCommEvent(self.handle, overlapped)

What are the possible return values of rc and mask ?





From ws-news at gmx.at  Thu Feb 15 22:24:50 2001
From: ws-news at gmx.at (Werner Schiendl)
Date: Thu, 15 Feb 2001 22:24:50 +0100
Subject: WaitCommEvent
References: <96h007$kri$1@newsfeed.pit.comms.marconi.com>
Message-ID: <982272544.13628@newsmaster-04.atnet.at>

Could not find a python related doc, but corresponding to the microsoft docs
at
http://msdn.microsoft.com/library/default.asp?URL=/library/psdk/winbase/comm
un_4bck.htm

i would _guess_ as following

rc ... return code
mask ... what is returned in lpEvtMask

hth
werner

Burkhard Kayser <Burkhard.Kayser at marconi.com> wrote in message
news:96h007$kri$1 at newsfeed.pit.comms.marconi.com...
> Does anyone know, where I can find the documentation for this function ?
>
> It is used in the example win32comport_demo.py.
>
> rc, mask = WaitCommEvent(self.handle, overlapped)
>
> What are the possible return values of rc and mask ?
>
>
>
>




From MarkH at ActiveState.com  Thu Feb 15 23:23:35 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 15 Feb 2001 22:23:35 GMT
Subject: WaitCommEvent
References: <96h007$kri$1@newsfeed.pit.comms.marconi.com>
Message-ID: <3A8C5609.7010103@ActiveState.com>

Burkhard Kayser wrote:

> Does anyone know, where I can find the documentation for this function ?
> 
> It is used in the example win32comport_demo.py.
> 
> rc, mask = WaitCommEvent(self.handle, overlapped)
> 
> What are the possible return values of rc and mask ?

msdn.microsoft.com

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From mwh21 at cam.ac.uk  Thu Feb 15 17:56:54 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Thu, 15 Feb 2001 16:56:54 +0000 (GMT)
Subject: python-dev summary 2001-02-01 - 2001-02-15
Message-ID: <Pine.LNX.4.10.10102151655490.1491-100000@localhost.localdomain>

It is with some trepidation that I post:

 This is a summary of traffic on the python-dev mailing list between
 Feb 1 and Feb 14 2001.  It is intended to inform the wider Python
 community of ongoing developments.  To comment, just post to
 python-list at python.org or comp.lang.python in the usual way. Give
 your posting a meaningful subject line, and if it's about a PEP,
 include the PEP number (e.g. Subject: PEP 201 - Lockstep iteration)
 All python-dev members are interested in seeing ideas discussed by
 the community, so don't hesitate to take a stance on a PEP if you
 have an opinion.

 This is the first python-dev summary written by Michael Hudson.
 Previous summaries were written by Andrew Kuchling and can be found
 at:

   <http://www.amk.ca/python/dev/>

 New summaries will probably appear at:

  <http://starship.python.net/crew/mwh/summaries/>

 When I get round to it.

  Posting distribution (with apologies to mbm)

  Number of articles in summary: 498

   80 |                 ]|[                                    
      |                 ]|[                                    
      |                 ]|[                                    
      |                 ]|[                                    
      |     ]|[         ]|[                                    
   60 |     ]|[         ]|[                                    
      |     ]|[         ]|[                                    
      |     ]|[         ]|[                                    
      |     ]|[         ]|[ ]|[                                
      |     ]|[         ]|[ ]|[                                
   40 |     ]|[         ]|[ ]|[         ]|[             ]|[    
      |     ]|[ ]|[ ]|[ ]|[ ]|[         ]|[             ]|[    
      |     ]|[ ]|[ ]|[ ]|[ ]|[         ]|[             ]|[    
      | ]|[ ]|[ ]|[ ]|[ ]|[ ]|[         ]|[         ]|[ ]|[    
      | ]|[ ]|[ ]|[ ]|[ ]|[ ]|[         ]|[         ]|[ ]|[ ]|[
   20 | ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[     ]|[ ]|[     ]|[ ]|[ ]|[
      | ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[     ]|[ ]|[     ]|[ ]|[ ]|[
      | ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[     ]|[ ]|[     ]|[ ]|[ ]|[
      | ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[
      | ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[ ]|[
    0 +-029-067-039-037-080-048-020-009-040-021-008-030-043-027
       Thu 01| Sat 03| Mon 05| Wed 07| Fri 09| Sun 11| Tue 13|
           Fri 02  Sun 04  Tue 06  Thu 08  Sat 10  Mon 12  Wed 14

 A fairly busy fortnight on python-dev, falling just short of five
 hundred articles.  Much of this is making ready for the Python 2.1
 release, but people's horizons are beginning to rise above the
 present.


   * Python 2.1a2 *
 
 Python 2.1a2 was released on Feb. 2.  One of the more controversial
 changes was the disallowing of "from module import *" at anything
 other than module level; this restriction was weakened after some
 slightly heated discussion on comp.lang.python.

  <http://mail.python.org/pipermail/python-dev/2001-February/012815.html>

 It is possible that non-module-level "from module import *" will
 produce some kind of warning in Python 2.1 but this code has not yet
 been written.


   * Performance *

 Almost two weeks ago, we were talking about performance.  Michael
 Hudson posted the results of an extended benchmarking session using
 Marc-Andre Lemburg's pybench suite:

  <http://mail.python.org/pipermail/python-dev/2001-January/012330.html>

 to which the conclusion was that python 2.1 will be marginally
 slower than python 2.0, but it's not worth shouting about.  The use
 of Vladimir Marangoz's obmalloc patch in some of the benchmarks
 sparked a discussion about whether this patch should be incorporated
 into Python 2.1.  There was support from many for adding it on an
 opt-in basis, since when nothing has happened...


   * Imports on case-insensitive file systems *

 There was quite some discussion about how to handle imports on a
 case-insensitive file system (eg. on Windows).  I didn't follow the
 details, but Tim Peters is on the case (sorry), so I'm confident it
 will get sorted out.


   * Sets & iterators *

 The Sets discussion rumbled on, moving into areas of syntax.  The
 syntax:

   for key:value in dict:

 was proposed.  Discussion went round and round for a while and moved
 on to more general iteration constructs, prompting Ka-Ping Yee to
 write a PEP entitled "iterators":

  <http://python.sourceforge.net/peps/pep-0234.html>

 Please comment!

 Greg Wilson announced that BOFs for both sets and iterators have
 been arranged at the python9 conference in March:

  <http://mail.python.org/pipermail/python-dev/2001-February/012824.html>


   * Stackless Python in Korea *

 Christian Tismer gave a presentation on stackless python to over 700
 Korean pythonistas:

  <http://mail.python.org/pipermail/python-dev/2001-February/012601.html>

 I think almost everyone was amazed and delighted to find that Python
 has such a fan base.  Next stop, the world!

   * string methodizing the standard library *

 Eric Raymond clearly got bored one evening and marched through the
 standard library, converting almost all uses of the string module to
 use to equivalent string method.  


   * Python's release schedule *

 Skip Montanaro raised some concerns about Python's accelerated
 release schedule, and it was pointed out that the default Python for
 both debian unstable and Redhat 7.1 beta was still 1.5.2.  Have
 *you* upgraded to Python 2.0?  If not, why not?


   * Unit testing (again) *

 The question of replacing Python's hoary old regrtest-driven test
 suite with something more modern came up again.  Andrew Kuchling
 enquired whether the issue was to be decided by voting or BDFL fiat:

  <http://mail.python.org/pipermail/python-dev/2001-February/012830.html>

 Guido obliged:

  <http://mail.python.org/pipermail/python-dev/2001-February/012831.html>

 There was then some discussion of what changes people would like to
 see made in the standard-Python-unit-testing-framework-elect
 (PyUnit) before they would be happy with it.

Cheers,
M.




From kalle at gnupung.net  Thu Feb 15 18:34:42 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Thu, 15 Feb 2001 18:34:42 +0100
Subject: python-dev summary 2001-02-01 - 2001-02-15
In-Reply-To: <Pine.LNX.4.10.10102151655490.1491-100000@localhost.localdomain>; from mwh21@cam.ac.uk on Thu, Feb 15, 2001 at 04:56:54PM +0000
References: <Pine.LNX.4.10.10102151655490.1491-100000@localhost.localdomain>
Message-ID: <20010215183442.A437@apone.network.loc>

Sez Michael Hudson:
> It is with some trepidation that I post:
> 
>  This is a summary of traffic on the python-dev mailing list between
>  Feb 1 and Feb 14 2001.  It is intended to inform the wider Python
[...]

Thanks!

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010215/08dc7a4d/attachment.pgp>

From dalke at acm.org  Thu Feb 15 18:39:10 2001
From: dalke at acm.org (Andrew Dalke)
Date: Thu, 15 Feb 2001 10:39:10 -0700
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <96h49t$68f$1@slb1.atl.mindspring.net>

Michael Hudson wrote:
[summary of python-dev]

Thanks for the summary.  In case you are wondering, it was
helpful to me to have it.  As a result I just:

  - sent mail to one of my clients saying they'll need
    to audit their code for "from spam import *" problems
    before upgrading to 2.1 (they use that construct in
    about a dozen places).

  - sent mail to one of the other projects I'm on talking
    about planning a standardization on PyUnit for regression
    testing.  (We use two; a modified regrtest and a different
    XP-based tester.)

Also, you asked:
> Have *you* upgraded to Python 2.0?  If not, why not?

I have for my own stuff.  In fact, I just last week
installed Python 2.1 from CVS.  Doing so reminded my why
I don't like upgrading Python.  There was a change in
the C level API which required a rebuild of all of my
extensions.  Many of them are from RPMs but as I don't
use them all that often (eg, the gtk modules) I decided
only to rebuild the modules needed for my own immediate
development.

One of my clients (the same one mentioned above) is still
using 1.5.2.  They are using it still because it works,
because they weren't sure if the 2.0 would have any
problems (it being a ".0" release) and because they didn't
want to reinstall all of their extensions (PIL, Numeric,
wxWindows, etc.) on their windows boxes.

                    Andrew
                    dalke at acm.org





From loredo at spacenet.tn.cornell.edu  Thu Feb 15 22:42:16 2001
From: loredo at spacenet.tn.cornell.edu (Tom Loredo)
Date: Thu, 15 Feb 2001 16:42:16 -0500
Subject: Rebuilding extensions (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <96h49t$68f$1@slb1.atl.mindspring.net>
Message-ID: <3A8C4D38.CF757D47@spacenet.tn.cornell.edu>

Hi folks-

Michael, thank you very much for continuing the tradition that Andrew K.
started.  He did a great service for us non-dev types, and your
1st go at it was great, too.

Andrew D. notes:
> In fact, I just last week
> installed Python 2.1 from CVS.  Doing so reminded my why
> I don't like upgrading Python.  There was a change in
> the C level API which required a rebuild of all of my
> extensions.

This is a big nuisance for me.  I'm curious about what to expect in 
this regard.  Will every .X revision of Python require rebuilding
extensions, or is it required only when there are changes to
the C API?  If the latter, how likely is it that the C API will
*not* change between revisions?  I do a lot of scientific
computing with Python, on Solaris and MacOS boxes, and rely
heavily on C extensions for numerical computing.

Thanks,
Tom Loredo


From dalke at acm.org  Fri Feb 16 08:08:02 2001
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 16 Feb 2001 00:08:02 -0700
Subject: Rebuilding extensions (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <96h49t$68f$1@slb1.atl.mindspring.net> <3A8C4D38.CF757D47@spacenet.tn.cornell.edu>
Message-ID: <96ipv6$hug$1@nntp9.atl.mindspring.net>

Tom Loredo wrote:
>I'm curious about what to expect in
>this regard.  Will every .X revision of Python require rebuilding
>extensions, or is it required only when there are changes to
>the C API?  If the latter, how likely is it that the C API will
>*not* change between revisions?

Looking at modsupport.h gives an idea of the frequency of change

http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/~checkout~/python/dist/src/Inc
lude/modsupport.h?rev=2.36&content-type=text/plain&cvsroot=python


>   25-Jan-2001  FLD     1010    Parameters added to PyCode_New() and
>                                PyFrame_New(); Python 2.1a2
>
>   14-Mar-2000  GvR     1009    Unicode API added
>
>   3-Jan-1999 GvR 1007 Decided to change back!  (Don't reuse 1008!)
>
>   3-Dec-1998 GvR 1008 Python 1.5.2b1
>
>   18-Jan-1997 GvR 1007 string interning and other speedups


so it looks like about once a year, that is, a change almost
every release.

                    Andrew
                    dalke at acm.org





From bkc at Murkworks.com  Thu Feb 15 21:04:38 2001
From: bkc at Murkworks.com (Brad Clements)
Date: Thu, 15 Feb 2001 15:04:38 -0500
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <7xWi6.4$R4.203@news.clarkson.edu>

Thanks for doing this!


--
Brad Clements,
bkc at murkworks.com
"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:mailman.982256267.9885.python-list at python.org...
> It is with some trepidation that I post:





From aahz at panix.com  Thu Feb 15 21:16:40 2001
From: aahz at panix.com (Aahz Maruch)
Date: 15 Feb 2001 12:16:40 -0800
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <96hdf8$qnt$1@panix6.panix.com>

In article <mailman.982256267.9885.python-list at python.org>,
Michael Hudson  <mwh21 at cam.ac.uk> wrote:
>
> This is the first python-dev summary written by Michael Hudson.

Huzzah!
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From mjackson at wc.eso.mc.xerox.com  Thu Feb 15 21:29:26 2001
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 15 Feb 2001 20:29:26 GMT
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <96he76$f68$1@news.wrc.xerox.com>

Michael Hudson <mwh21 at cam.ac.uk> writes:

>  This is the first python-dev summary written by Michael Hudson.

Thank you!

>  was proposed.  Discussion went round and round for a while and moved
>  on to more general iteration constructs, prompting Ka-Ping Yee to
>  write a PEP entitled "iterators":
> 
>   <http://python.sourceforge.net/peps/pep-0234.html>
> 
>  Please comment!

With the possible exception of ":spam" and "eggs:" it doesn't look like
it would make code less readable, and the functionality seems good.
Mildly in favor.

>    * Python's release schedule *
> 
>  Skip Montanaro raised some concerns about Python's accelerated
>  release schedule, and it was pointed out that the default Python for
>  both debian unstable and Redhat 7.1 beta was still 1.5.2.  Have
>  *you* upgraded to Python 2.0?  If not, why not?

Ran into difficulties building on Solaris and didn't have the time to
resolve them.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Having your book made into a movie is like having
	your ox made into a bouillon cube.
				- Bill Neely




From tuttledon at hotmail.com  Fri Feb 16 05:19:42 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Fri, 16 Feb 2001 04:19:42 GMT
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <yT1j6.75408$k8.17055190@typhoon.southeast.rr.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:mailman.982256267.9885.python-list at python.org...

>  New summaries will probably appear at:
>
>   <http://starship.python.net/crew/mwh/summaries/>
>

Yeah!
<shout> JUST DON'T FORGET TO ALSO POST THEM HERE!  </shout>






From not.this at seebelow.org  Fri Feb 16 05:28:39 2001
From: not.this at seebelow.org (Grant Griffin)
Date: Thu, 15 Feb 2001 22:28:39 -0600
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <3A8CAC77.91555B93@seebelow.org>

Michael Hudson wrote:
> 
...
>  This is the first python-dev summary written by Michael Hudson.

Thanks, Michael--keep up the great work!

...
>  Discussion went round and round for a while and moved
>  on to more general iteration constructs, prompting Ka-Ping Yee to
>  write a PEP entitled "iterators":
> 
>   <http://python.sourceforge.net/peps/pep-0234.html>
> 
>  Please comment!

I went through this one, and I really like it: it seems both very
"useful" and very "beautiful".  What more could it possibly need?

However...

I regret Python's increasing loss of status as "executable
pseudo-code".  This proposal, as well as 2.0's "list comprehensions",
seem to be making Python harder for the uninitiated to read.  (To be
fair, though, as a "moderately-initiated" Pythoneer, I've recently
warmed up to list comprehensions.)

Whenever I explain my Python code to a non-Pythoneer, it's always quite
easy to explain--except for slices.  Sure, those are essential in
Python, so the principle of "if something is hard to explain, it's
probably a bad idea" doesn't hold true in that case.  But I think the
principle itself is good: Python should read more like Dutch than
Greek.  And I think this new proposal would definitely be even harder to
explain to the uninitiated than slices--especially the "k:" and ":v"
versions.

I think Python is at a crossroads where it is changing from "executable
pseudo-code" into something increasingly...well..."complex".  If
"growth" in a language means "new features", I guess complexity is
inevitable.  In all honesty, the techie in me is all in favor of that. 
But my wiser part realizes that the techie part needs to be kept in
check.  (Or, as the great Archie Bunker said, "Stifle yourself!")

So, give me this great new feature of iterators--if you can't manage to
save me from myself.

ambivalent-ly y'rs, 

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From stephen_purcell at yahoo.com  Fri Feb 16 09:51:07 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 16 Feb 2001 09:51:07 +0100
Subject: iterators (was: python-dev summary)
In-Reply-To: <3A8CAC77.91555B93@seebelow.org>; from not.this@seebelow.org on Thu, Feb 15, 2001 at 10:28:39PM -0600
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org>
Message-ID: <20010216095107.F7411@freedom.puma-ag.com>

Grant Griffin wrote:
> I think Python is at a crossroads where it is changing from "executable
> pseudo-code" into something increasingly...well..."complex".  If
> "growth" in a language means "new features", I guess complexity is
> inevitable.

It's a pity. I also like things simple, because complexity makes my head hurt.

Python is really simple, which is why I love it.

This new proposal sounds 'neat', but just think about plugging the new
iterator syntax into list comprehensions like the following (stolen from the
PEP):

>>> nums = [1,2,3,4]
>>> fruit = ["Apples", "Peaches", "Pears", "Bananas"]
>>> print [(i,f) for i in nums for f in fruit if f[0] == "P" if i%2 == 1]
[(1, 'Peaches'), (1, 'Pears'), (3, 'Peaches'), (3, 'Pears')]

Sprinkle a few ':' iterator magic symbols in there, and it doesn't look like
Python any more. Of course, if the intention is to win over the Perl
community...

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From dalke at acm.org  Fri Feb 16 09:54:14 2001
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 16 Feb 2001 01:54:14 -0700
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org>
Message-ID: <96ipv7$hug$2@nntp9.atl.mindspring.net>

Grant Griffin:
>However...
>
>I regret Python's increasing loss of status as "executable
>pseudo-code".  This proposal, as well as 2.0's "list comprehensions",
>seem to be making Python harder for the uninitiated to read.  (To be
>fair, though, as a "moderately-initiated" Pythoneer, I've recently
>warmed up to list comprehensions.)
>[more comments removed]

I admit to a "Me too!" here.

I work in computational biology and chemisty.  This means
I work with developers (people with a lot of programming
background and some science background), scientists (the
opposite) and people in various states in between (I'm a
physicist who became a developer).

I've pushed Python partially because it is a language
which developers enjoy *and* so do scientists.  I compare
it to Perl which developers enjoy but researchers don't.
(As a generality - there are exceptions.)  Another example
is Tcl, which developers often find tedious but researchers
use just fine.  (Mind you, I used Tcl in the pre-8.0 days.)

The goal of a researcher is to do science, and they will
put up with a lot to do it.  A computational scientist
will often learn a dozen different ad hoc scripting languages
needed by the various programs they use.  However, there
is a certain level of complexity beyond which few researchers
will learn.  In general, the more obvious the better, as with
Tcl.

On the other hand, developers will learn about all the
intricacies of a language, and attempt to use the appropriate
feature for a given task.  Sometimes a feature is used
for the shear elegance or novelity of a construct, at the
detriment of clarity.

For example, I'll bet list comprehensions will be a
problem when my clients start using Python 2.0.  It is
syntax driven and implicit, rather than using named
functions with an explicit loop.  That's why I plan to
use list comprehensions rarely, and only for a single
level of comprehension.

I see similar problems with ">>" and now with "k:v" iterators.
These all mean there will be multiple ways to do the same
task.  Especially worrying is that the best ways will
be the most obscure.

This reminds me of a story Gerald Weinberg tells about a
"possum burger" company where a consultant comes in and says
the company can save a large amount of money by reducing the
number of sesame seeds on the bun by 1, with no measureable
change in taste.  The problem is, enough unnoticable changes
do make a difference.  Eventually people will go elsewhere
for possum burgers.

Explicit is better, so in another branch of this thread
I elaborate on a suggestion for how to make explicit
iterators (via functions & classes in a new module).

                    Andrew Dalke
                    dalke at acm.org





From erno-news at erno.iki.fi  Fri Feb 16 10:55:00 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 16 Feb 2001 11:55:00 +0200
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net>
Message-ID: <kud7cjuf4b.fsf@lasipalatsi.fi>

In article <96ipv7$hug$2 at nntp9.atl.mindspring.net>, "Andrew Dalke"
<dalke at acm.org> writes:

| Grant Griffin:
|| However...
|| 
|| I regret Python's increasing loss of status as "executable
|| pseudo-code".  This proposal, as well as 2.0's "list comprehensions",
|| seem to be making Python harder for the uninitiated to read.  (To be
|| fair, though, as a "moderately-initiated" Pythoneer, I've recently
|| warmed up to list comprehensions.)
|| [more comments removed]

| I admit to a "Me too!" here.

me three!

actually, the first time i saw the proposal for

for key:value in dict:
    ...

i thought it was a joke. ok, it's not obscure, but i
don't see the point either.

list comprehensions also look pretty obscure to me (compared to the
rest of python's syntax), and don't do a lot that you can't do with
map/filter now that we have lexical scoping.

  -- erno


From jeremy at alum.mit.edu  Fri Feb 16 02:10:28 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Thu, 15 Feb 2001 20:10:28 -0500 (EST)
Subject: iterators (was: python-dev summary)
In-Reply-To: <kud7cjuf4b.fsf@lasipalatsi.fi>
References: <mailman.982256267.9885.python-list@python.org>
	<3A8CAC77.91555B93@seebelow.org>
	<96ipv7$hug$2@nntp9.atl.mindspring.net>
	<kud7cjuf4b.fsf@lasipalatsi.fi>
Message-ID: <14988.32260.786366.394218@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "EK" == Erno Kuusela <erno-news at erno.iki.fi> writes:

  EK> list comprehensions also look pretty obscure to me (compared to
  EK> the rest of python's syntax), and don't do a lot that you can't
  EK> do with map/filter now that we have lexical scoping.

I have the opposite opinion.  I think the list comprehension syntax is
often clearer and more intuitive than the equivalent map or filter
expression.  I expect more non- and novice programmers will be
familiar with set construction notation than with higher-order
functions like map.

Jeremy



From erno-news at erno.iki.fi  Fri Feb 16 17:42:09 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 16 Feb 2001 18:42:09 +0200
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net> <kud7cjuf4b.fsf@lasipalatsi.fi> <mailman.982340500.13146.python-list@python.org>
Message-ID: <ku7l2qvau6.fsf@lasipalatsi.fi>

In article <mailman.982340500.13146.python-list at python.org>, Jeremy
Hylton <jeremy at alum.mit.edu> writes:

| Erno Kuusela <erno-news at erno.iki.fi> writes:
|| list comprehensions also look pretty obscure to me (compared to
|| the rest of python's syntax), and don't do a lot that you can't
|| do with map/filter now that we have lexical scoping.

| I have the opposite opinion.  I think the list comprehension syntax is
| often clearer and more intuitive than the equivalent map or filter
| expression.  I expect more non- and novice programmers will be
| familiar with set construction notation than with higher-order
| functions like map.

i suspect both map/filter + lambda and list comprehensions are
a little confusing for the beginner at first, and require some
pondering. but list comprehensions seem like an unnecessary
additional way to spell the same thing... at least map/filter
are pretty straightforward constructions that build on the
basic concept of function.

also, i find the lack of delimiters between tha parts in
a list comprehension a little un-aesthetic.
[a*b   for b in c   if d % 2]
     ^            ^
  -- erno


From jeremy at alum.mit.edu  Fri Feb 16 18:00:31 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 16 Feb 2001 12:00:31 -0500 (EST)
Subject: iterators (was: python-dev summary)
In-Reply-To: <ku7l2qvau6.fsf@lasipalatsi.fi>
References: <mailman.982256267.9885.python-list@python.org>
	<3A8CAC77.91555B93@seebelow.org>
	<96ipv7$hug$2@nntp9.atl.mindspring.net>
	<kud7cjuf4b.fsf@lasipalatsi.fi>
	<mailman.982340500.13146.python-list@python.org>
	<ku7l2qvau6.fsf@lasipalatsi.fi>
Message-ID: <14989.23727.53276.346479@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "EK" == Erno Kuusela <erno-news at erno.iki.fi> writes:

  EK> i suspect both map/filter + lambda and list comprehensions are a
  EK> little confusing for the beginner at first, and require some
  EK> pondering. but list comprehensions seem like an unnecessary
  EK> additional way to spell the same thing... at least map/filter
  EK> are pretty straightforward constructions that build on the basic
  EK> concept of function.

As I said before, I think we just don't agree.  List comprehensions
build on the basic notions of loops and ifs.  I think these are
simpler than higher-order functions.

  EK> also, i find the lack of delimiters between tha parts in a list
  EK> comprehension a little un-aesthetic.  [a*b for b in c if d % 2]
  EK>      ^ ^

I agree with you here, but I don't think anyone had a good solution.
Don't remember what was discussed, as we never got around to writing
the PEP.

Jeremy
       





From donn at u.washington.edu  Fri Feb 16 19:55:27 2001
From: donn at u.washington.edu (Donn Cave)
Date: 16 Feb 2001 18:55:27 GMT
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org>  <3A8CAC77.91555B93@seebelow.org>  <96ipv7$hug$2@nntp9.atl.mindspring.net>  <kud7cjuf4b.fsf@lasipalatsi.fi>  <mailman.982340500.13146.python-list@python.org>  <ku7l2qvau6.fsf@lasipalatsi.fi> <mailman.982342971.24194.python-list@python.org>
Message-ID: <96jt2v$ogk$1@nntp6.u.washington.edu>

Quoth Jeremy Hylton <jeremy at alum.mit.edu>:
| >>>>> "EK" == Erno Kuusela <erno-news at erno.iki.fi> writes:
|
| EK> i suspect both map/filter + lambda and list comprehensions are a
| EK> little confusing for the beginner at first, and require some
| EK> pondering. but list comprehensions seem like an unnecessary
| EK> additional way to spell the same thing... at least map/filter
| EK> are pretty straightforward constructions that build on the basic
| EK> concept of function.
|
| As I said before, I think we just don't agree.  List comprehensions
| build on the basic notions of loops and ifs.  I think these are
| simpler than higher-order functions.

The basic notions are simpler, but the notation that puts them
together isn't obvious.  Until I have had a chance to look that
up and use them for something enough to commit that syntax to
memory, I won't understand the list comprehension code I see.
That is not hypothetical, I really don't understand them -
believe I have seen it explained and understood the explanation,
but with no need for such a thing, I have since forgotten.

If it's a big step towards something useful, fine.  I have heard
claims that they're faster, and that's a priority I can understand.
But every time some cool notion gets added to Python, it adds to
the amount that we all eventually have to know.  I can't indefinitely
postpone my education in list comprehensions without marginalizing
myself, just as someone who comes along and learns them right away
will still have to learn all the alternatives too.  Does that add
to our ability to write software, or subtract from it?

	Donn Cave, donn at u.washington.edu


From erno-news at erno.iki.fi  Fri Feb 16 20:22:43 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 16 Feb 2001 21:22:43 +0200
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net> <kud7cjuf4b.fsf@lasipalatsi.fi> <mailman.982340500.13146.python-list@python.org> <ku7l2qvau6.fsf@lasipalatsi.fi> <mailman.982342971.24194.python-list@python.org> <96jt2v$ogk$1@nntp6.u.washington.edu>
Message-ID: <ku3ddev3ek.fsf@lasipalatsi.fi>

In article <96jt2v$ogk$1 at nntp6.u.washington.edu>, Donn Cave
<donn at u.washington.edu> writes:

| I have heard claims that they're faster, and that's a priority I can
| understand.

actually, they are slower (than map + python function). but don't
think that should be important.

| But every time some cool notion gets added to Python, it adds to
| the amount that we all eventually have to know.  I can't indefinitely
| postpone my education in list comprehensions without marginalizing
| myself, just as someone who comes along and learns them right away
| will still have to learn all the alternatives too.  Does that add
| to our ability to write software, or subtract from it?

well said.

  -- erno


From tim.one at home.com  Fri Feb 16 21:27:11 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 15:27:11 -0500
Subject: iterators (was: python-dev summary)
In-Reply-To: <96jt2v$ogk$1@nntp6.u.washington.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEOGIPAA.tim.one@home.com>

[Donn Cave]
> The basic notions are simpler, but the notation that puts them
> together isn't obvious.  Until I have had a chance to look that
> up and use them for something enough to commit that syntax to
> memory, I won't understand the list comprehension code I see.

Here, guess what this does:

    print [x**2 for x in range(10)]

Bingo -- you're now educated <wink>.

> ...
> I have heard claims that they're faster, and that's a priority I
> can understand.

They *could* be quicker, but in the current implementation are usually
slower.  The primary intent is to be clearer than the alternatives.

> But every time some cool notion gets added to Python, it adds to
> the amount that we all eventually have to know.

As well as increase the set of available tools.  That was true in 0.9.6
too -- these are tradeoffs.

> I can't indefinitely postpone my education in list comprehensions
> without marginalizing myself, just as someone who comes along and
> learns them right away will still have to learn all the alternatives
> too.  Does that add to our ability to write software, or subtract
> from it?

I think it obvious that it does both.  Guido approves of ideas that, after
consideration, he believes will be a net win; he rejects ideas that he
believes will be a net loss (and no suggestion is without *some* value!).
You won't agree with him every time; and neither will I, although I pretend
that I do to keep universal harmony flowing on c.l.py <wink>.

that's-why-he-doesn't-want-anyone-reading-python-dev<wink>-ly y'rs  - tim




From amk at mira.erols.com  Sat Feb 17 02:07:57 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 17 Feb 2001 01:07:57 GMT
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org>  <3A8CAC77.91555B93@seebelow.org>  <96ipv7$hug$2@nntp9.atl.mindspring.net>  <kud7cjuf4b.fsf@lasipalatsi.fi>  <mailman.982340500.13146.python-list@python.org>  <ku7l2qvau6.fsf@lasipalatsi.fi> <mailman.982342971.24194.python-list@python.org> <96jt2v$ogk$1@nntp6.u.washington.edu>
Message-ID: <slrn98rkib.ik.amk@mira.erols.com>

On 16 Feb 2001 18:55:27 GMT, Donn Cave <donn at u.washington.edu> wrote:
>If it's a big step towards something useful, fine.  I have heard

Agreed.  Some time ago I argued that once Unicode was added to Python
1.6, we should just stamp the language definition "done" and leave it
alone.  Sure, we can add little tweaky features like iterators, but
will they enable a startling new application domain?  Will they
attract more users?  In both cases the answer is likely "no", so we
should just write off the new features -- they might be nice, but why
bother making books obsolete and making the code more bloated?

I think I'll repeat the argument for 2.1: call the language as defined
by the 2.1 docs the final version, and leave it alone.  Obviously you
could still work on a faster interpreter, fix bugs in it, etc.  This
means interfaces won't get implemented, which is unfortunate, but if
you let in "just *one* more feature" it'll never stop.

--amk



From donn at oz.net  Sat Feb 17 06:56:25 2001
From: donn at oz.net (Donn Cave)
Date: 17 Feb 2001 05:56:25 GMT
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org>  <3A8CAC77.91555B93@seebelow.org>  <96ipv7$hug$2@nntp9.atl.mindspring.net>  <kud7cjuf4b.fsf@lasipalatsi.fi>  <mailman.982340500.13146.python-list@python.org>  <ku7l2qvau6.fsf@lasipalatsi.fi> <mailman.982342971.24194.python-list@python.org> <96jt2v$ogk$1@nntp6.u.washington.edu> <slrn98rkib.ik.amk@mira.erols.com>
Message-ID: <96l3q9$718$0@216.39.151.169>

Quoth amk at mira.erols.com (A.M. Kuchling):
| On 16 Feb 2001 18:55:27 GMT, Donn Cave <donn at u.washington.edu> wrote:
| >If it's a big step towards something useful, fine.  I have heard
|
| Agreed.  Some time ago I argued that once Unicode was added to Python
| 1.6, we should just stamp the language definition "done" and leave it
| alone.  Sure, we can add little tweaky features like iterators, but
| will they enable a startling new application domain?  Will they
| attract more users?  In both cases the answer is likely "no", so we
| should just write off the new features -- they might be nice, but why
| bother making books obsolete and making the code more bloated?
|
| I think I'll repeat the argument for 2.1: call the language as defined
| by the 2.1 docs the final version, and leave it alone.  Obviously you
| could still work on a faster interpreter, fix bugs in it, etc.  This
| means interfaces won't get implemented, which is unfortunate, but if
| you let in "just *one* more feature" it'll never stop.

I can't tell how serious you are, but that appeals to me.  There's
a lot to 2.1.  It's a big language with a lot of things to learn,
and everything you need to write great software.  It would be a
crime to load it down with unnecessary gimmicks, and yet it's all
too common for any software system to go just that way.  Maybe
universal.

Still, I think I wouldn't be in favor of a constitutional prohibition
on new features.  I can't say there's no room for Python to evolve,
it feels too much like the people who a century ago were convinced
that most everything of consequence was already known to science.
I would like to find the antidote for creeping featurism, though.
For Python, I guess that antidote is the BDFL, as Tim Peters observed
in another followup, and that works for me - it would be paradoxical
to say Python is the perfect language but we can't trust its creator's
judgement.  And since it's his baby anyway, who cares what we think.
Yet the development environment around Python has changed in recent
years, and I guess that's where we came in.

One thing I think would be worth some thought is a positive expression
of what we're thinking when we say "no thanks".  I mean, opposition
to features is the negative complement to a positive idea about Python,
about its elegance and power the way it is.  We don't get that out as
often or as articulately as maybe we could.

	Donn Cave, donn at oz.net


From tim_one at email.msn.com  Sat Feb 17 07:53:11 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Feb 2001 01:53:11 -0500
Subject: Language change (was RE: iterators (was: python-dev summary))
In-Reply-To: <96l3q9$718$0@216.39.151.169>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEABJAAA.tim_one@email.msn.com>

[Donn Cave]
> ...
> Still, I think I wouldn't be in favor of a constitutional prohibition
> on new features.

Then, unlike Andrew's, your future may be a happy one <wink>.

> I can't say there's no room for Python to evolve, it feels too much
> like the people who a century ago were convinced that most everything
> of consequence was already known to science.

Good example:  nobody predicted any particular problem at the start due to
types and classes occupying distinct universes in Python (can only create a
class in Python, can only create a type in C).  In retrospect, that was
braindead.  Healing that split would be a *simplification* of the language,
although, ya, it would probably have some visible consequences.

> I would like to find the antidote for creeping featurism, though.
> For Python, I guess that antidote is the BDFL, as Tim Peters observed
> in another followup, and that works for me - it would be paradoxical
> to say Python is the perfect language but we can't trust its creator's
> judgement.  And since it's his baby anyway, who cares what we think.

Guido does!  But he doesn't have time to argue with 10,000 people on every
point, and (this won't be a surprise to grownups <wink>) sometimes he
couldn't even if he did have time:  Python isn't entirely the end result of
an objective deductive process.  Many of the best ideas were adopted simply
because they "felt like" a good idea at the time.  Sometimes that backfires.
But that's rare, and that's what makes Guido a better language designer than
his critics (including me, on the one occasion I disagreed with him <wink>).

> Yet the development environment around Python has changed in recent
> years, and I guess that's where we came in.

What changed is that moving to SourceForge multiplied the number of people
who *could* commit changes from one (Guido) to dozens (Guido's army of
mindless drones).  Most of the flurry of changes checked in for 2.0 were
things Guido wanted for years, but didn't have the personal bandwidth to do
all by himself.

> One thing I think would be worth some thought is a positive expression
> of what we're thinking when we say "no thanks".  I mean, opposition
> to features is the negative complement to a positive idea about Python,
> about its elegance and power the way it is.  We don't get that out as
> often or as articulately as maybe we could.

You know that we're heading towared 100 rejected patches on SoureForge?
That's heading on 100 ideas people wanted badly enough to actually do the
hard work of implementing them (and that's impressive!) -- but they got
tossed in the bit bucket anyway.  People who think the floodgates have
opened are simply wrong about that (although a couple years' worth of backed
up ideas got implemented quickly).  Here's the latest one Guido rejected:

http://sourceforge.net/patch/
    index.php?func=detailpatch&patch_id=103693&group_id=5470

That's for a tiny, fully backward-compatible change, and one that's been
asked for several times on c.l.py over the years.  If I were the submitter,
I would find Guido's rejection comment unconvincing.  A good mental exercise
is to think of some change you *like*, and dream up a rejection comment you
would be satisfied with -- but realizing that in the real world, it's not
going to end with "QED".

it-won't-end-with-"bugger-off"-either-but-that's-how-it-often-
    gets-read-ly y'rs  - tim




From robin at jessikat.fsnet.co.uk  Sat Feb 17 10:37:05 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 17 Feb 2001 09:37:05 +0000
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <96l3q9$718$0@216.39.151.169>
 <mailman.982392863.25850.python-list@python.org>
Message-ID: <MA+i3XABZkj6EwoH@jessikat.fsnet.co.uk>

In article <mailman.982392863.25850.python-list at python.org>, Tim Peters
<tim_one at email.msn.com> writes
....
>opened are simply wrong about that (although a couple years' worth of backed
>up ideas got implemented quickly).  Here's the latest one Guido rejected:
>
>http://sourceforge.net/patch/
>    index.php?func=detailpatch&patch_id=103693&group_id=5470
>
>That's for a tiny, fully backward-compatible change, and one that's been
>asked for several times on c.l.py over the years.  If I were the submitter,
...
The language would have been smaller if only lists were allowed, but
that would make a lot of things unsafe/more interesting.
-cults of personality are always wrong-ly yrs- 
Robin Becker


From not.this at seebelow.org  Mon Feb 19 05:35:54 2001
From: not.this at seebelow.org (Grant Griffin)
Date: Sun, 18 Feb 2001 22:35:54 -0600
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org>
Message-ID: <3A90A2AA.21FB5D05@seebelow.org>

Tim Peters wrote:
> 
> [Donn Cave]
...
> > Yet the development environment around Python has changed in recent
> > years, and I guess that's where we came in.
> 
> What changed is that moving to SourceForge multiplied the number of people
> who *could* commit changes from one (Guido) to dozens (Guido's army of
> mindless drones).  Most of the flurry of changes checked in for 2.0 were
> things Guido wanted for years, but didn't have the personal bandwidth to do
> all by himself.

OK, I've enjoyed and appreciated almost all of those--thanks!.

But obviously that creates a kind of pressure.  (Put, conversely, I
guess new features previously had to be _really_ useful for Guido to
find time to add them.)

> > One thing I think would be worth some thought is a positive expression
> > of what we're thinking when we say "no thanks".  I mean, opposition
> > to features is the negative complement to a positive idea about Python,
> > about its elegance and power the way it is.  We don't get that out as
> > often or as articulately as maybe we could.

Guilty as charged!  If my implied compliment wasn't received here,
consider it given!  I have said this before: Genius is the art of
simplicity, and Python is a work of great simplicity.

We in the Peanut Gallery fuss about new features the way folks would
fuss if Bach suddenly decided to add some notes to "Tocatta and Fugue in
D Minor".  (Then again, nothing's perfect. <wink>)

> You know that we're heading towared 100 rejected patches on SoureForge?
> That's heading on 100 ideas people wanted badly enough to actually do the
> hard work of implementing them (and that's impressive!) -- but they got
> tossed in the bit bucket anyway.  People who think the floodgates have
> opened are simply wrong about that (although a couple years' worth of backed
> up ideas got implemented quickly). 

I guess that tends to create a certain impression...

> Here's the latest one Guido rejected:
> 
> http://sourceforge.net/patch/
>     index.php?func=detailpatch&patch_id=103693&group_id=5470
> 
> That's for a tiny, fully backward-compatible change, and one that's been
> asked for several times on c.l.py over the years.  If I were the submitter,
> I would find Guido's rejection comment unconvincing.

Maybe he should have just said, "There should be one-- and preferably
only one --obvious way to do it." <wink> 

I guess some of us are guilty of over-subscribing to the "Zen of Python"
principles (at http://www.python.org/doc/Humor.html#zen, if you haven't
read 'em, Tim)  Perhaps we shouldn't adhere to these rigidly (except for
the ones which have a lot of wiggle room <wink>), but it's tempting to
do so: they make so darn much sense.  (Also, having done Perl, some of
us see what can happen when a language's central organizing principle is
"there's more than one way to do it"---what the heck kindda central
organizing principle is that?!)

I have never seen any statement from Guido to the effect of endorsing
your principles, but if he doesn't endorse them, at least the things
that "felt like a good idea at the time" to him seem to be mostly
explainable by them in retrospect.

Anyway, my point is that there certainly are forces of change and
"featurism", which come up squarely against one of Python's central
features: its lack of central features.  (Ironic, ain't it? <wink>)

Or, as a great bot once said (more or less), "those who succeed in the
language design business do so by choosing a design philosophy and
sticking to it."

which-is-yet-another-statement-you-probably-wish-we-literal
   -minded-folks-had-never-heard-<wink>-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From jkraska1 at san.rr.com  Mon Feb 19 06:54:20 2001
From: jkraska1 at san.rr.com (Courageous)
Date: Mon, 19 Feb 2001 05:54:20 GMT
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org>
Message-ID: <sid19t4icqn1lad77eblpnincphdbdr1c6@4ax.com>

>Anyway, my point is that there certainly are forces of change and
>"featurism", which come up squarely against one of Python's central
>features: its lack of central features.  (Ironic, ain't it? <wink>)

:-)

This is one of Python's true redeeming values; I hope and pray
that it doesn't get lost in the woods over the years.


C//





From othello at javanet.com  Wed Feb 21 02:08:20 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 20 Feb 2001 20:08:20 -0500
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <sid19t4icqn1lad77eblpnincphdbdr1c6@4ax.com>
Message-ID: <3A931503.366F2B7E@javanet.com>

Courageous wrote:

> >Anyway, my point is that there certainly are forces of change and
> >"featurism", which come up squarely against one of Python's central
> >features: its lack of central features.  (Ironic, ain't it? <wink>)
>
> :-)
>
> This is one of Python's true redeeming values; I hope and pray
> that it doesn't get lost in the woods over the years.
>

I absolutely agree that there should be a strong resistance toward
creeping featurism, but think there should be an equally strong will
to improve the language and not reject ideas out of hand.

Examples of "good" directions for changes:
-- change things that surprise uses (like unexpected integer math)
-- change things that cause users to write awkward or kludgy code (like having
to
    set-up a loop counter when indexing over a list -- which PEP 234 proposes
    to fix)
-- change things that aren't polymorphic (like eval taking only dictionary
objects
    instead of general mapping objects and like the tuple.count() which GvR
rejected).
-- do tough things like unifying the type/class distinction (it will be a bear,
but life
    will be better afterwards).
-- change difficult to use tools (like improving the debugger or adding
interactive help).
-- change things that prevent people from using their favorite tricks (like
adding nested
    scopes so that the functional programmers can go wild).
-- implement things which improve performance (like xreadlines() or
stacklessness).
-- add more examples to the documentation

We can have language improvement without creeping featurism.

Raymond

"In theory, there is no difference between theory and practice.
 In practice, there is." -- Yogi Berra



From donn at oz.net  Wed Feb 21 08:19:55 2001
From: donn at oz.net (Donn Cave)
Date: 21 Feb 2001 07:19:55 GMT
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <sid19t4icqn1lad77eblpnincphdbdr1c6@4ax.com> <3A931503.366F2B7E@javanet.com>
Message-ID: <96vq6r$9b4$0@216.39.151.169>

Quoth Raymond Hettinger <othello at javanet.com>:
| Courageous wrote: (quoting someone else)
|>| Anyway, my point is that there certainly are forces of change and
|>| "featurism", which come up squarely against one of Python's central
|>| features: its lack of central features.  (Ironic, ain't it? <wink>)
|>
|> :-)
|>
|> This is one of Python's true redeeming values; I hope and pray
|> that it doesn't get lost in the woods over the years.


| I absolutely agree that there should be a strong resistance toward
| creeping featurism, but think there should be an equally strong will
| to improve the language and not reject ideas out of hand.
|
| Examples of "good" directions for changes:
[... various perennial favorites ...]
| We can have language improvement without creeping featurism.

Right - let's pile on every "good" thing we can think of!

Changes like this aren't rejected "out of hand".  Per an early
post in this thread, the process may not always be any more rigorous
than the designer's hunch, but that hunch is what makes or breaks
the language.  If not everything has been accepted, I think the
reports of wholesale patch rejection could be a little deceiving,
depending on what all those rejected patches were actually proposing
to do.

But you can count on it, some of those changes you like will indeed
not be deemed worthy, and there doesn't need to be a reason that says
"this feature is actually bad, and here's what breaks".  Any change
is disruptive, any extra feature adds to the burden of learning and
using the language.  If you can't make a strong enough case for the
benefit of a new feature, then its value won't outweigh that burden.

Many of the notions you think are good directions aren't compelling
to me.  In fact they fall into several common classes of uncompelling:

1. "This will be easier for programmers who aren't programmers."
   - Who is the legitimate voice of this class of people?  Do you
     or anyone have evidence that your theories are going to pay
     off in practice?  Take the integer arithmetic thing - could
     there be a big can of worms waiting for the unwary user of
     the alternative, which I suppose would be floating point?
     If the innocent programmer is surprised by integer truncation,
     what about 1.0 != 1.0?
   - Are real programmers supposed to stick with Python as it sells
     them out?  Will a non-programmer constituency pan out for the
     new Python?  I overstate this as a conflict, but there is an
     element of that.  If the implication is that I should accept
     changes that are bad for me, because they're good for someone
     else who doesn't actually use Python and may never really have
     all that much use for any programming language, well, I'm not
     sure that makes total sense to me.
2. "This makes things more consistent."
   - Who cares?  Is Python's general level of inconsistency a big
     problem?  Some changes in that direction could conceivably make
     programming significantly more productive, but it has to be
     spelled out.  More consistent is not compelling per se.
3. "This alleviates awkwardness, kludginess, etc."
   - Who cares?  If we're talking about programming problems that
     really hurt at a software engineering level, for example force
     you to write the same thing in several places when you could
     have done it once given the feature, then that's interesting,
     but again, the onus is on the proponent to sell that.  The
     more usual case is that this "it's kludgy!" reaction is as
     vacuous is as it is emphatic.

Now I'm not saying the things you propose are altogether without
value, and I might even agree about one or two if I looked at it
hard enough.  The key is that the benefit must be clear and practical.

	Donn Cave, donn at oz.net
-------
| Examples of "good" directions for changes:
| -- change things that surprise uses (like unexpected integer math)
| -- change things that cause users to write awkward or kludgy code (like having
| to
|     set-up a loop counter when indexing over a list -- which PEP 234 proposes
|     to fix)
| -- change things that aren't polymorphic (like eval taking only dictionary
| objects
|     instead of general mapping objects and like the tuple.count() which GvR
| rejected).
| -- do tough things like unifying the type/class distinction (it will be a bear,
| but life
|     will be better afterwards).
| -- change difficult to use tools (like improving the debugger or adding
| interactive help).
| -- change things that prevent people from using their favorite tricks (like
| adding nested
|     scopes so that the functional programmers can go wild).
| -- implement things which improve performance (like xreadlines() or
| stacklessness).
| -- add more examples to the documentation


From donn at oz.net  Mon Feb 19 08:13:19 2001
From: donn at oz.net (Donn Cave)
Date: 19 Feb 2001 07:13:19 GMT
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org>
Message-ID: <96qh2f$962$1@216.39.151.169>

Quoth Grant Griffin <not.this at seebelow.org>:

| Guilty as charged!  If my implied compliment wasn't received here,
| consider it given!  I have said this before: Genius is the art of
| simplicity, and Python is a work of great simplicity.

I don't know how to say this, but I think it isn't just the language -
I'm afraid we're simple too.  Even though we didn't create it, we who
use the language are an important part of the phenomenon.  Our simple-
minded, numb tolerance for expressions like "while 1", our willingness
to write out "self" just to disambiguate namespaces, our uncritical
acceptance of incomplete OO-ness, make it possible for a language
like Python to thrive.  What are we, techno-peasants or something?

	Donn Cave, donn at oz.net


From aahz at panix.com  Mon Feb 19 08:28:16 2001
From: aahz at panix.com (Aahz Maruch)
Date: 18 Feb 2001 23:28:16 -0800
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <96qh2f$962$1@216.39.151.169>
Message-ID: <96qhug$446$1@panix2.panix.com>

In article <96qh2f$962$1 at 216.39.151.169>, Donn Cave  <donn at oz.net> wrote:
>
>What are we, techno-peasants or something?

Absolutely.  I regularly call myself a technophilic Luddite.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From amk at mira.erols.com  Tue Feb 20 02:08:27 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 20 Feb 2001 01:08:27 GMT
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <96qh2f$962$1@216.39.151.169> <96qhug$446$1@panix2.panix.com>
Message-ID: <slrn993hnd.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>

On 18 Feb 2001 23:28:16 -0800, Aahz Maruch <aahz at panix.com> wrote:
>Absolutely.  I regularly call myself a technophilic Luddite.

Never used that combination of words before, but it's a good one.
We should write a manifesto!

--amk



From aahz at panix.com  Tue Feb 20 02:15:48 2001
From: aahz at panix.com (Aahz Maruch)
Date: 19 Feb 2001 17:15:48 -0800
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <96qh2f$962$1@216.39.151.169> <96qhug$446$1@panix2.panix.com> <slrn993hnd.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>
Message-ID: <96sgg4$a33$1@panix6.panix.com>

In article <slrn993hnd.hb.amk at 207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>,
A.M. Kuchling <akuchlin at mems-exchange.org> wrote:
>On 18 Feb 2001 23:28:16 -0800, Aahz Maruch <aahz at panix.com> wrote:
>>
>>Absolutely.  I regularly call myself a technophilic Luddite.
>
>Never used that combination of words before, but it's a good one.
>We should write a manifesto!

Sure, as long as it's accessible with Lynx.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Laws that restrict speech are an abomination for a simple reason.
They're like a poison gas; they seem to be a good weapon to vanquish
your enemy, but the wind has a way of shifting."  -- Ira Glasser


From sholden at holdenweb.com  Wed Feb 28 21:23:32 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 28 Feb 2001 15:23:32 -0500
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <96qh2f$962$1@216.39.151.169> <96qhug$446$1@panix2.panix.com>
Message-ID: <4kdn6.4166$jv4.258966@e420r-atl2.usenetserver.com>

"Aahz Maruch" <aahz at panix.com> wrote in message
news:96qhug$446$1 at panix2.panix.com...
> In article <96qh2f$962$1 at 216.39.151.169>, Donn Cave  <donn at oz.net> wrote:
> >
> >What are we, techno-peasants or something?
>
> Absolutely.  I regularly call myself a technophilic Luddite.

Well, yes, but then you also regularly call yourself an androgynous poly
kinky vanilla queer het, so I'm unsure how much weight to attach to this
statement,

regards
 Steve





From tim.one at home.com  Wed Feb 28 21:57:29 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 28 Feb 2001 15:57:29 -0500
Subject: Language change (was RE: iterators (was: python-dev summary))
In-Reply-To: <4kdn6.4166$jv4.258966@e420r-atl2.usenetserver.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCENAJCAA.tim.one@home.com>

[Donn Cave]
> What are we, techno-peasants or something?

[Aahz]
> Absolutely.  I regularly call myself a technophilic Luddite.

[Steve Holden]
> Well, yes, but then you also regularly call yourself an androgynous poly
> kinky vanilla queer het, so I'm unsure how much weight to attach to this
> statement,

I've met Aahz -- trust me, they both fit.  Well, I'm not so sure about the
"vanilla" part, now that you mention it.

but-i'm-just-an-old-bot-who-am-i-to-judge-ly y'rs  - tim




From not.this at seebelow.org  Mon Feb 19 21:47:15 2001
From: not.this at seebelow.org (Grant Griffin)
Date: 19 Feb 2001 12:47:15 -0800
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <96qh2f$962$1@216.39.151.169>
Message-ID: <96s0oj02nsq@drn.newsguy.com>

In article <96qh2f$962$1 at 216.39.151.169>, Donn says...
>
>Quoth Grant Griffin <not.this at seebelow.org>:
>
>| Guilty as charged!  If my implied compliment wasn't received here,
>| consider it given!  I have said this before: Genius is the art of
>| simplicity, and Python is a work of great simplicity.
>
>I don't know how to say this, but I think it isn't just the language -
>I'm afraid we're simple too.  Even though we didn't create it, we who
>use the language are an important part of the phenomenon.

Good point.  I personally think of myself as "simple" but somewhat "wise". 
(Then again, aren't the wisest people mostly pretty simple anyway? <wink>)  The
wisdom comes in knowing one's limitations--and in actively seeking things like
Python that help.

>Our simple-
>minded, numb tolerance for expressions like "while 1",

I recently figured out (I think) that one the *really* bad thing about do/until
in Python is that it doesn't work with Python's block style.  (In fact, if you
think about it, even in C/C++ do/until a little strange in being the only thing
that has an expression at the _end_ of the block!)

If language design is a compromise, and this compromise frees me from braces,
I'm a happy camper.

>our willingness
>to write out "self" just to disambiguate namespaces,...

If that were the _only_ reason, it might not be so great.  But you're talking to
the guy who actually _likes_ "self.", as a replacement for the "m_" notation he
uses in C++ to mark class member variables.  (True, I sometimes wake up
screaming in the middle of the night about those extra three characters Python's
convention makes me type.  But then I tell myself that I could have used "m." if
I had really wanted to. <wink>)

>...our uncritical
>acceptance of incomplete OO-ness,

But some of us _like_ a mixed OO/procedural paradigm.  (Although "math.sin(x)"
sure does read a little fishy...)

>make it possible for a language
>like Python to thrive.  What are we, techno-peasants or something?

<grin>  I prefer to think of us more as thirsty horses being lead to water.

which-is-not-to-say-we'll-all-drink-*everything*-we're
   -lead-to-<wink>-ly y'rs

=g2

_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com



From mwh21 at cam.ac.uk  Mon Feb 19 22:32:27 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Feb 2001 21:32:27 +0000
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <96qh2f$962$1@216.39.151.169> <96s0oj02nsq@drn.newsguy.com>
Message-ID: <m3ofvytl3o.fsf@atrus.jesus.cam.ac.uk>

Grant Griffin <not.this at seebelow.org> writes:

> If that were the _only_ reason, it might not be so great.  But
> you're talking to the guy who actually _likes_ "self.", as a
> replacement for the "m_" notation he uses in C++ to mark class
> member variables.  (True, I sometimes wake up screaming in the
> middle of the night about those extra three characters Python's
> convention makes me type.  But then I tell myself that I could have
> used "m." if I had really wanted to. <wink>)

I still have the mental scars from trying to understand code that
named *local* variables m_blah.  Explicit *is* better than implicit.

Cheers,
M.

-- 
  please realize that the Common  Lisp community is more than 40 
  years old.  collectively, the community has already been where 
  every clueless newbie  will be going for the next three years.  
  so relax, please.                     -- Erik Naggum, comp.lang.lisp


From hinsen at cnrs-orleans.fr  Mon Feb 19 11:52:35 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 19 Feb 2001 11:52:35 +0100
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net> <kud7cjuf4b.fsf@lasipalatsi.fi> <mailman.982340500.13146.python-list@python.org> <ku7l2qvau6.fsf@lasipalatsi.fi> <mailman.982342971.24194.python-list@python.org> <96jt2v$ogk$1@nntp6.u.washington.edu> <slrn98rkib.ik.amk@mira.erols.com>
Message-ID: <m3elwvt05o.fsf@chinon.cnrs-orleans.fr>

amk at mira.erols.com (A.M. Kuchling) writes:

> I think I'll repeat the argument for 2.1: call the language as defined
> by the 2.1 docs the final version, and leave it alone.  Obviously you
> could still work on a faster interpreter, fix bugs in it, etc.  This
> means interfaces won't get implemented, which is unfortunate, but if
> you let in "just *one* more feature" it'll never stop.

This seems a bit strict, but I agree with the general idea. Adding a
new feature that opens up new possibilities (like safety in the case
of interfaces) is OK, but adding syntactic sugar like list
comprehensions mostly bloats the language definition. Simplicity is
one of Python's attractions. If you want a complicated language, try
Fortran, Ada, or C++.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From sholden at holdenweb.com  Wed Feb 28 21:38:30 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 28 Feb 2001 15:38:30 -0500
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org>  <3A8CAC77.91555B93@seebelow.org>  <96ipv7$hug$2@nntp9.atl.mindspring.net>  <kud7cjuf4b.fsf@lasipalatsi.fi>  <mailman.982340500.13146.python-list@python.org>  <ku7l2qvau6.fsf@lasipalatsi.fi> <mailman.982342971.24194.python-list@python.org> <96jt2v$ogk$1@nntp6.u.washington.edu> <slrn98rkib.ik.amk@mira.erols.com>
Message-ID: <Godn6.4167$jv4.259816@e420r-atl2.usenetserver.com>

"A.M. Kuchling" <amk at mira.erols.com> wrote in message
news:slrn98rkib.ik.amk at mira.erols.com...
> On 16 Feb 2001 18:55:27 GMT, Donn Cave <donn at u.washington.edu> wrote:
> >If it's a big step towards something useful, fine.  I have heard
>
> Agreed.  Some time ago I argued that once Unicode was added to Python
> 1.6, we should just stamp the language definition "done" and leave it
> alone.  Sure, we can add little tweaky features like iterators, but
> will they enable a startling new application domain?  Will they
> attract more users?  In both cases the answer is likely "no", so we
> should just write off the new features -- they might be nice, but why
> bother making books obsolete and making the code more bloated?
>
> I think I'll repeat the argument for 2.1: call the language as defined
> by the 2.1 docs the final version, and leave it alone.  Obviously you
> could still work on a faster interpreter, fix bugs in it, etc.  This
> means interfaces won't get implemented, which is unfortunate, but if
> you let in "just *one* more feature" it'll never stop.
>
No, really, have just one more feature -- they're wafer thin!

Apart from leacing a lot of pretty competent programmers with less to do, I
can't think of many disadvantages to freezing the specification. This, I
presume, was originally the thinking behind Py3K.

It seems to me that the future may cause the Python applications base to
slowly crumble, as incompatible features which "break very little code" are
added.

but-then-i'm-just-an-old-fart-luddite-ly y'rs  - steve





From kevin.digweed at dial.pipex.com  Sun Feb 18 20:26:19 2001
From: kevin.digweed at dial.pipex.com (Kevin Digweed)
Date: Sun, 18 Feb 2001 19:26:19 +0000
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net> <kud7cjuf4b.fsf@lasipalatsi.fi>
Message-ID: <3A9021DB.D18A19D4@dial.pipex.com>

Erno Kuusela wrote:
> 
> In article <96ipv7$hug$2 at nntp9.atl.mindspring.net>, "Andrew Dalke"
> <dalke at acm.org> writes:
> 
> | Grant Griffin:
> || However...
> ||
> || I regret Python's increasing loss of status as "executable
> || pseudo-code".  This proposal, as well as 2.0's "list comprehensions",
> || seem to be making Python harder for the uninitiated to read.  (To be
> || fair, though, as a "moderately-initiated" Pythoneer, I've recently
> || warmed up to list comprehensions.)
> || [more comments removed]
> 
> | I admit to a "Me too!" here.
> 
> me three!
> 
> actually, the first time i saw the proposal for
> 
> for key:value in dict:
>     ...
> 
> i thought it was a joke. ok, it's not obscure, but i
> don't see the point either.

Sadly, I don't Python as much as I used to, but as someone who has
started reading the newsgroup again since just after 2.0 was released,
this PEP did come as something of a surprise to me. I don't think that
the 'for x:y in dict' syntax bothered me so much as the 'for :y' and
'for x:' variants. It's too subtle. I saw a post recently where someone
did "(None, None, None, ThingIWant, OtherThingIWant, None) =
funcreturning6_tuple();". I confess to doing something similar when I
first started Python. I expected None to be something that, given it's
name, was something of a black hole. The solution is to assign all those
things you don't want to a dummy name which you can subsequently
'del'ete.
I think that the proposed extension to the 'for' syntax should be
limited to *only* the "x:y" variety. If you don't want the left or right
component, you assign it to a dummy name and 'del'ete it if neccessary.

OTOH, if the other variations on the 'for' syntax were to be allowed, I
think that "for (,,, ThingIWant, OtherthingIWant,,) =
funcreturning6_tuple();" should also be allowed. (BTW, I'm being
sarcastic).

I guess that what I'm saying is that I don't think the *absense* of a
word/keyword should be treated as significant.

Kev.


From jepler at inetnebr.com  Mon Feb 19 02:59:24 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Mon, 19 Feb 2001 01:59:24 GMT
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net> <kud7cjuf4b.fsf@lasipalatsi.fi> <3A9021DB.D18A19D4@dial.pipex.com>
Message-ID: <slrn990vfr.c3f.jepler@potty.housenet>

On Sun, 18 Feb 2001 19:26:19 +0000, Kevin Digweed
 <kevin.digweed at dial.pipex.com> wrote:
>I guess that what I'm saying is that I don't think the *absense* of a
>word/keyword should be treated as significant.

... except in the case of slices, right?

x[1:2]
x[:2]
x[1:]
x[:]

Jeff


From dwig at advancedmp.net  Wed Feb 21 18:32:35 2001
From: dwig at advancedmp.net (Don Dwiggins)
Date: 21 Feb 2001 09:32:35 -0800
Subject: iterators (was: python-dev summary)
In-Reply-To: Kevin Digweed's message of "Sun, 18 Feb 2001 19:26:19 +0000"
References: <mailman.982256267.9885.python-list@python.org>
	<3A8CAC77.91555B93@seebelow.org>
	<96ipv7$hug$2@nntp9.atl.mindspring.net>
	<kud7cjuf4b.fsf@lasipalatsi.fi> <3A9021DB.D18A19D4@dial.pipex.com>
Message-ID: <un1bgx7po.fsf@advancedmp.net>

> I saw a post recently where someone
> did "(None, None, None, ThingIWant, OtherThingIWant, None) =
> funcreturning6_tuple();". I confess to doing something similar when I
> first started Python. 
...
> OTOH, if the other variations on the 'for' syntax were to be allowed, I
> think that "for (,,, ThingIWant, OtherthingIWant,,) =
> funcreturning6_tuple();" should also be allowed. (BTW, I'm being
> sarcastic).

Actually, you can already do it Prolog-style:
  (_, _, _, ThingIWant, OtherthingIWant, _) = funcreturning6_tuple()

-- 
Don Dwiggins                    "Solvitur Ambulando"
Advanced MP Technology
dwig at advancedmp.net




From qrczak at knm.org.pl  Thu Feb 22 17:08:12 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Feb 2001 16:08:12 GMT
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net> <kud7cjuf4b.fsf@lasipalatsi.fi> <3A9021DB.D18A19D4@dial.pipex.com> <mailman.982776809.17537.python-list@python.org>
Message-ID: <slrn99aebc.gsa.qrczak@qrnik.zagroda>

21 Feb 2001 09:32:35 -0800, Don Dwiggins <dwig at advancedmp.net> pisze:

> Actually, you can already do it Prolog-style:
>   (_, _, _, ThingIWant, OtherthingIWant, _) = funcreturning6_tuple()

Except that the object bound to _ will be retained until the function
returns or _ is rebound.

Unfortunately the interpreter is probably not allowed to optimize
it out by realizing that this variable is never referenced, because
somebody might want to rely on the fact that the destructor is not
called too early. I wish it was allowed to optimize it out. Or is
it already? Relying on the moment of calling a destructor is evil.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From skip at mojam.com  Fri Feb 16 19:35:37 2001
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 16 Feb 2001 12:35:37 -0600 (CST)
Subject: iterators (was: python-dev summary)
In-Reply-To: <96ipv7$hug$2@nntp9.atl.mindspring.net>
References: <mailman.982256267.9885.python-list@python.org>
	<3A8CAC77.91555B93@seebelow.org>
	<96ipv7$hug$2@nntp9.atl.mindspring.net>
Message-ID: <14989.29433.578965.545111@beluga.mojam.com>

    Andrew> I see similar problems with ">>" and now with "k:v" iterators.
    Andrew> These all mean there will be multiple ways to do the same task.

The ">>" syntax can actually be thought of as reducing the number of
(standard) ways of printing.  Previously, you could use the print statement
for straightforward output (but only to sys.stdout), but if you wanted to
write to anything besides sys.stdout, you generally were forced to use
file's write method.  Now you have your choice: straightforward output or
finegrained control of output.  If your scientist friends are comfortable
with print, there's no need to introduce them to the more complex use of
write methods and the % formatting operator.  

No comment on the k:v iterator.  I don't believe that's a done deal.

-- 
Skip Montanaro (skip at mojam.com)
Support Mojam & Musi-Cal: http://www.musi-cal.com/sponsor.shtml
(847)971-7098



From dalke at acm.org  Fri Feb 16 09:54:16 2001
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 16 Feb 2001 01:54:16 -0700
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org>
Message-ID: <96ipv9$hug$3@nntp9.atl.mindspring.net>

Here's the rationale for PEP 234

>  1. It provides an extensible iterator interface.

I notice it isn't as powerful as C++ iterators.  For
example, there's no way to specify a bidirectional iterator.
I also don't see how mutations in the container are
handled, but I don't know how C++ does that either.

I also note that there are different types of iterators
which may be applied to an object.  For example, a file
can be iterated a line at a time or a record at a time.
To support this currently, I write a wrapper object which
handles __getitem__ as appropriate.

What's wrong with that approach?  For example,

  import iterate
  for key, val in iterate.items(obj):
      pass

where iterate.items is a function which does essentially
the same thing as mp_iteritems.  Off the top of my head
(without much thought and likely to be error prone) it
could look similar to:

iterate.py:

  class IterateItems:
      def __init__(self, obj, next_key):
          self.__obj = obj
          self.__next_key = next_key
      def __getitem__(self, i):
          k = self.__next_key[i]
          return k, self.__obj[k]

  def items(obj):
      if hasattr(obj, "__iteritems__"):
          return obj.__iteritems__()
      elif hasattr(obj, "__iter__"):
          return IterateItems(obj, obj.__iter__())
      elif hasattr(obj, "__items__"):
          return obj.__items__()
      elif hasattr(obj, "__getitem__"):
          return IterateItems(obj, obj.keys())
      ...

Similar constructs for
  iterate.keys(obj)
  iterate.values(obj)
  ...

My point is that this iterator scheme does not and cannot
handle every situation so there will have to be ways to
map the desired solution to the iterator scheme.  Why not
expand on that and use wrappers/explicit factory fuctions
for everything instead of using new syntax.

>  2. It resolves the endless "i indexing sequence" debate.

None of my clients have ever brought this up as a concern.
For that matter, I'm not sure what you mean by it.
If you had said "i indexing dictionary" I might understand.

There are endless "tab vs. spaces vs. {}" debates.  Does
that mean something needs to be done with Python to settle
those debates.  Meaning, I don't view this as a useful
argument.

Maybe it really means that you are using this phrase as
a shorthand description of a problem and I don't know how
to expand it properly.

>  3. It allows performance enhancements to dictionary iteration.

My alternate suggestion supports this as well, but without
syntax changes.

>  4. It allows one to provide an interface for just iteration
      without pretending to provide random access to elements.

Hmm.  I've had success in throwing an exception when the
iterator was used in a non-forward manner.   I tracked
a "self.__n" with the current position, and enforced
that "i == self.__n" in "__getitem__(self, i)".  No one
has ever complained about it.

But I do like the idea.  Consider one place where I had
a forward iterator through a database (provided by the
underlying C API, which didn't allow a random access
iterator.)  I did the check against self.__n as above,
but it wasn't obvious in the code that only forward
iteration was allowed.  Instead, if there is a iterate
module with a "forward" factory function, then it can
work like:

iterate.py:

  class ForwardIterator:
      def __init__(self, obj):
          self.__obj = obj
          self.__n = 0
      def __getitem__(self, i):
          assert self.__n == i, "only forward iteration is allowed"
          self.__n += 1
          return self.__obj[i]

  def forward(obj):
      if hasattr(obj, "__iter__"):
          return obj.__iter__()
      elif hasattr(obj, "__getitem__"):
          return ForwardIterator(obj)
      ...

and my code would be similar to:

  class ForwardDatabaseIterator:
      def __init__(self, stream):
          self.stream = stream
      def __getitem__(self, i):
          # guaranteed to be called in increasing order, so I
          # don't need to check for random access
          handle = daylight.dt_next(self.stream)
          if h == 0:
              daylight.dt_dealloc(self.stream)
              return Record(handle)


  class Database:
      def __iter__(self):
         stream = daylight.dt_stream(self.handle, daylight.TYP_RECORD)
         return ForwardDatabaseIterator(stream)

  import iterate
  db = open_database("user at host", "password")
  for record in iterate.forward(db):
      pass

>  5. It is backward-compatible with all existing user-defined
>     classes and extension objects that emulate sequences and
>     mappings, even mappings that only implement a subset of
>     {__getitem__, keys, values, items}.

Same for my alternate suggestion because it uses the same
techniques for backwards compatibility that you use.

Note that I am *not* proposing that this module wrapper
should be used.  I would rather keep things as they are
(see my comments on this in c.l.py).  Instead, read it as
a push that if this functionality is added to Python that
it be done without adding new syntax, and a proposal
suggestion how that might be done.

For Ka-Ping Yee.  I'm not on python-dev and it's hard
to search the pipermail archives, so I don't know if I caught
everything on this discussion.  In
http://mail.python.org/pipermail/python-dev/2001-February/012677.html
you say
> > Sorry, Ping, I didn't know you have a PEP for iterators already.
>
> I posted it on this very boutique (i mean, mailing list) a week ago
> and messages have been going back and forth on its thread since then.

I can't seem to find those mails.  I don't see mention of
"234" or "iterator" in any of the titles listed in the indicies,
and pipermail doesn't have any way to do a text search.  My
apologies for not following every link to find that thread.

That said, M-A Lemburg in
http://mail.python.org/pipermail/python-dev/2001-February/012679.html
said
> I'd suggest to implement generic iterators which implements your
> suggestions and put them into the builins or a special iterator
> module...
>
> from iterators import xitems, xkeys, xvalues
>
> for key, value in xitems(dict):
> for key in xkeys(dict):
> for value in xvalues(dict):

which is essentially what I said.  Your reply to his message,
in http://mail.python.org/pipermail/python-dev/2001-February/012682.html
doesn't comment on his suggestion and the PEP doesn't make
any comment on it.

(Oh, I downloaded all of Feb's mail to date.  I can't find any
other message which includes reference to "from iter".  I also
found the thread "Sets: elt in dict, lst.include" and other
threads, but as a flat file it's hard to follow what's going
on.  My apologies for not rewriting this email based on my
new knowledge - I don't think it is affected much by what I've
read.)

Finally, as a minor note, I almost never embed def's in def's
so examples like

            def __iter__(self):
                def iter(self=self):
                    line = self.file.readline()
                    if line: return line
                    else: raise IndexError
                return iter

take some effort for me to follow.  I tend to create new
classes instead, since I find them less confusing than
using default arguments as a way to pass state around.

  Sincerely,

                    Andrew Dalke
                    dalke at acm.org





From gzeljko at sezampro.yu  Fri Feb 16 14:35:37 2001
From: gzeljko at sezampro.yu (gzeljko)
Date: Fri, 16 Feb 2001 14:35:37 +0100
Subject: PEP-0234
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org>
Message-ID: <96jacs$qg6$1@neptun.beotel.net>

Perhaps I'm missing something obvious, but:

If you hev construct:

                    for ident1 in ident2:

I'm really surprised that people wish express intended
behaviour of ident2 messing with ident1. This is unnatural,
unflexible and dead-lock for language evolution (IMHO).

!?-ly-y'rs-gzeljko










From dwig at advancedmp.net  Fri Feb 16 19:45:20 2001
From: dwig at advancedmp.net (Don Dwiggins)
Date: 16 Feb 2001 10:45:20 -0800
Subject: About iterators as a new language feature
In-Reply-To: Grant Griffin's message of "Thu, 15 Feb 2001 22:28:39 -0600"
References: <mailman.982256267.9885.python-list@python.org>
	<3A8CAC77.91555B93@seebelow.org>
Message-ID: <uhf1ucvr3.fsf_-_@advancedmp.net>

Grant Griffin writes:
>> Discussion went round and round for a while and moved
>> on to more general iteration constructs, prompting Ka-Ping Yee to
>> write a PEP entitled "iterators":
>> 
>> <http://python.sourceforge.net/peps/pep-0234.html>
>> 
>> Please comment!

> I went through this one, and I really like it: it seems both very
> "useful" and very "beautiful".  What more could it possibly need?

> However...

> I regret Python's increasing loss of status as "executable
> pseudo-code".  This proposal, as well as 2.0's "list comprehensions",
> seem to be making Python harder for the uninitiated to read.  (To be
> fair, though, as a "moderately-initiated" Pythoneer, I've recently
> warmed up to list comprehensions.)

...

> I think Python is at a crossroads where it is changing from "executable
> pseudo-code" into something increasingly...well..."complex".  If
> "growth" in a language means "new features", I guess complexity is
> inevitable.  In all honesty, the techie in me is all in favor of that. 
> But my wiser part realizes that the techie part needs to be kept in
> check.  (Or, as the great Archie Bunker said, "Stifle yourself!")

> So, give me this great new feature of iterators--if you can't manage to
> save me from myself.

I'd like to suggest an alternative approach to adding features to the
language for this kind of thing: providing "iterator-like" methods in the
"collection classes".  I'd planned to work up a message on this theme,
describing how Smalltalk uses the Block (think lambda) as a basic language
element, and builds around it "execution mechanisms" that would be language
elements in other languages, and suggesting that something like that be
considered for Python.  For example, the Smalltalk language doesn't have any
syntax for "if-else", "for", or "while" -- the equivalents are built out of
classes and methods defined in the "library", and use Blocks as an essential
enabler.  In particular, all iteration is done using methods of the
Collection hierarchy.  There's a handful of standard methods, but an
enterprising programmer can easily create new ones for special purposes.

The direction I'm looking in is to do something similar for Python.  I'm
just getting started with the language, so there are probably some issues I
can't see yet.  However, I've seen that there are some Smalltalk speakers in
this forum, so I think there could be a productive discussion on this topic.

I guess the basic theme is, wherever possible, keep the language simple but
powerful, and build new structures and mechanisms in the language, rather
than added onto it.

-- 
Don Dwiggins                    "Solvitur Ambulando"
Advanced MP Technology
dwig at advancedmp.net




From jpet at eskimo.com  Sat Feb 17 07:11:12 2001
From: jpet at eskimo.com (Jeff Petkau)
Date: Sat, 17 Feb 2001 06:11:12 GMT
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <mailman.982313491.5504.python-list@python.org>
Message-ID: <4Coj6.3622$1T1.356977@paloalto-snr1.gtei.net>

Steve Purcell <stephen_purcell at yahoo.com> wrote in message
news:mailman.982313491.5504.python-list at python.org...
> This new proposal sounds 'neat', but just think about plugging the new
> iterator syntax into list comprehensions like the following (stolen from
the
> PEP):
>
> >>> nums = [1,2,3,4]
> >>> fruit = ["Apples", "Peaches", "Pears", "Bananas"]
> >>> print [(i,f) for i in nums for f in fruit if f[0] == "P" if i%2 == 1]
> [(1, 'Peaches'), (1, 'Pears'), (3, 'Peaches'), (3, 'Pears')]
>
> Sprinkle a few ':' iterator magic symbols in there, and it doesn't look
like
> Python any more. Of course, if the intention is to win over the Perl
> community...
>
> -Steve

I totally agree with this. I hope future language additions
take the form of builtin functions like zip() instead of new
syntax.

Functions are more in line with the 'explicit is better than
implicit' argument--does 'for x in dict' iterate over keys,
values, or items? What better way could we specify this than
keys(), values(), and items()?

For iterating through lists with indices, we could just toss
something like this into __builtins__:

    def indexed(x):
        """returns a list of (index,value) pairs
        for the mapping or sequence x."""

        if hasattr(x,'items'):
            return x.items()
        else:
            return zip(range(len(x)),x)

    for i,v in indexed([1,2,3]):
        print 'elem at',i,'is',v

The only argument I've seen for new syntax is efficiency--
dict.keys() et al can be expensive for big dictionaries.
But this is solved very nicely the way xrange() and
xreadlines() did it, with lazy iterators. Maybe Guido thinks
they're a hack, but they're a simple, efficient, consistent,
easy to use solution to a recurring problem.  We could use
more hacks like that!

Admittedly the xfoo() naming convention is ugly, but that's
the nature of backwards compatibility. New iterators could
be lazy by default, and if you want an actual list, just use
list(x).

Um, what topic did I start with again?

Oh, yeah. And function names are much easier to look up in the
documentation.

--Jeff





From hinsen at cnrs-orleans.fr  Fri Feb 16 15:17:48 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 16 Feb 2001 15:17:48 +0100
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <m33ddeempf.fsf@chinon.cnrs-orleans.fr>

Michael Hudson <mwh21 at cam.ac.uk> writes:

>  This is a summary of traffic on the python-dev mailing list between
>  Feb 1 and Feb 14 2001.  It is intended to inform the wider Python

Great news, and thanks for your efforts!

>  Skip Montanaro raised some concerns about Python's accelerated
>  release schedule, and it was pointed out that the default Python for
>  both debian unstable and Redhat 7.1 beta was still 1.5.2.  Have
>  *you* upgraded to Python 2.0?  If not, why not?

No. I keep a 2.0 installed on my personal account for compatibility
testing, but updating Python on all the machines I use at work is
a major effort (all extension modules need to be recompiled), so I do
this only when there's an improvement that is significant to my work.
This will be the case for 2.1 (e.g. rich comparisons), but it wasn't
in 2.0.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From qrczak at knm.org.pl  Sun Feb 18 22:41:55 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 18 Feb 2001 21:41:55 GMT
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <slrn990gd3.tc5.qrczak@qrnik.zagroda>

Thu, 15 Feb 2001 16:56:54 +0000 (GMT), Michael Hudson <mwh21 at cam.ac.uk> pisze:

>  Discussion went round and round for a while and moved
>  on to more general iteration constructs, prompting Ka-Ping Yee to
>  write a PEP entitled "iterators":
> 
>   <http://python.sourceforge.net/peps/pep-0234.html>

What about mp_iterkeys trying to use __iteritems__ after __iterkeys__ is
not available, before failing back to __iter__? Similarly __itervalues__.
It would allow a mapping to implement only __iteritems__, obtaining correct
iteration over keys and values for free.

What about introducing a "black hole" or "wildcard pattern" for
assignments?
    e0, e1, _, _ = some_4_tuple
    for k:_ in mapping:
    for _:v in mapping:
If there is code using _ as a variable name, maybe some different name.
But it's _ in some other languages (ML, Haskell, Prolog, Clean).

After these two things increasing the complexity :-( let me propose
a simplification. Have _ as above, but don't have mp_iterkeys nor
mp_itervalues. The mapping iteration protocol uses only mp_iteritems
and __iteritems__ / __iter__. Iteration over keys or values is
expressed as iteration over items, while throwing half of each pair
into a black hole.

For mp_iteritems implemented in terms of __iter__ (i.e. sequences
treating as mappings from integers) and for builtin types the
interpreter can generate as efficient code for k:_ and _:v iteration
as if there were separate iterkey and itervalue protocols. For user
classes it's a bit worse (an unnecessary 2-tuple is created), but
the simplification of concepts certainly justifies it.

So finally I propose adding black holes, and having only the iteritems
protocol and syntax. iterkeys and itervalues are just instances of
these concepts put together, not separate concepts.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From cjc26 at nospam.cornell.edu  Mon Feb 19 19:12:44 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Mon, 19 Feb 2001 18:12:44 GMT
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org> <slrn990gd3.tc5.qrczak@qrnik.zagroda>
Message-ID: <wmdk6.96422$v.7354987@typhoon.nyroc.rr.com>

* Marcin 'Qrczak' Kowalczyk <qrczak at knm.org.pl> menulis:
| 
| What about introducing a "black hole" or "wildcard pattern" for
| assignments?
|     e0, e1, _, _ = some_4_tuple
|     for k:_ in mapping:
|     for _:v in mapping:
| If there is code using _ as a variable name, maybe some different name.
| But it's _ in some other languages (ML, Haskell, Prolog, Clean).

You can already do stuff like

    e0, e1, _, _ = (10, 20, 40, 30)
    print e0  => 10
    print e1  => 20
    print _   => 30

You can use whatever variable name you want for the dummy variable.

As for dict iteration, I don't like the idea of inventing new syntax for
it; I would prefer having dict.items() return a lazy list, or maybe add
a xitems() method to return a lazy list.  (Actually, if there were some
way to signal to a method that it's being called in an iteration
context, then we wouldn't need all these x-methods...)


-- 
Cliff Crawford                 http://www.people.cornell.edu/pages/cjc26/
"huh?  Are you all there?  You missing some key parts up there?  Becoming  
a English major or an minor?   Do you need a annual report?  Are you slow
today, eh?  Did someone sever you parietal lobe?" - Mark


From eferro at inicia.es  Thu Feb 15 18:07:02 2001
From: eferro at inicia.es (Eduardo Ferro)
Date: Thu, 15 Feb 2001 18:07:02 +0100
Subject: ioctl problem
Message-ID: <96gva8$nnd$1@diana.bcn.ttd.net>

Hi!

Have anybody some example of the use of ioctl method when this ioctl
return some data in a estructure.
I use icotl to control the cdrom, and the ioctls without parameters works
well, but i don`t know how to use the ioctls that return me data

Thanks in advance :)


From donn at u.washington.edu  Thu Feb 15 19:04:56 2001
From: donn at u.washington.edu (Donn Cave)
Date: 15 Feb 2001 18:04:56 GMT
Subject: ioctl problem
References: <96gva8$nnd$1@diana.bcn.ttd.net>
Message-ID: <96h5o8$u8e$1@nntp6.u.washington.edu>

Quoth "Eduardo Ferro" <eferro at inicia.es>:

| Have anybody some example of the use of ioctl method when this ioctl
| return some data in a estructure.
| I use icotl to control the cdrom, and the ioctls without parameters works
| well, but i don`t know how to use the ioctls that return me data

  >>> import fcntl
  >>> print fcntl.ioctl.__doc__
  ioctl(fd, opt, [arg])

  Perform the requested operation on file descriptor fd.  The operation
  is defined by op and is operating system dependent.  Typically these
  codes can be retrieved from the library module IOCTL.  The argument arg
  is optional, and defaults to 0; it may be an int or a string. If arg is
  given as a string, the return value of ioctl is a string of that length,
  containing the resulting value put in the arg buffer by the operating system.
  The length of the arg string is not allowed to exceed 1024 bytes. If the arg
  given is an integer or if none is specified, the result value is an integer
  corresponding to the return value of the ioctl call in the C code.

That means you must lay your data out as a string of bytes.  Pass any
string of that length to ioctl(), then use the "struct" module's byte
unpacking functions to convert the return string to the data type you
want to work with.

  >>> import fcntl
  >>> n = fcntl.ioctl(0, 1074295912, 'XXXXXXXX')
  >>> n
  '\030\000P\000\360\000 \003'
  >>> import struct
  >>> struct.unpack('hhhh', n)
  (24, 80, 240, 800)

That's the TIOCGWINSZ ioctl (on NetBSD 1.5.)

	Donn Cave, donn at u.washington.edu


From eferro at inicia.es  Fri Feb 16 08:22:57 2001
From: eferro at inicia.es (Eduardo Ferro)
Date: Fri, 16 Feb 2001 08:22:57 +0100
Subject: ioctl problem
References: <96gva8$nnd$1@diana.bcn.ttd.net> <96h5o8$u8e$1@nntp6.u.washington.edu>
Message-ID: <96iiof$3dg$1@diana.bcn.ttd.net>

> 
>   Perform the requested operation on file descriptor fd.  The operation
>   is defined by op and is operating system dependent.  Typically these
>   codes can be retrieved from the library module IOCTL.  The argument
>   arg is optional, and defaults to 0; it may be an int or a string. If
>   arg is given as a string, the return value of ioctl is a string of
>   that length, containing the resulting value put in the arg buffer by
>   the operating system. The length of the arg string is not allowed to
>   exceed 1024 bytes. If the arg given is an integer or if none is
>   specified, the result value is an integer corresponding to the return
>   value of the ioctl call in the C code.

I' ll try it, i think that data will return at the "arg" argument... at
the lib doc it don`t say anything about it... but i forget to check the
doc string :)

Thanks


From adolfo.benin at riskmap.it  Thu Feb 15 19:55:33 2001
From: adolfo.benin at riskmap.it (Adolfo Benin)
Date: Thu, 15 Feb 2001 19:55:33 +0100
Subject: COM server info
In-Reply-To: <E14TOS4-00065t-00@mail.python.org>
Message-ID: <5.0.2.1.0.20010215193232.00a63eb0@mail1.cs.interbusiness.it>

Hi,
	I've developed a COM object in Python, then I regitered it as a DCOM on my 
machine.
Every time a client makes a request to the server DCOM, the Python 
interpreter is launched and the COM is created. But when the client ends, 
the COM server also ends.
So every call to the server creates a new python interpreter instance in 
memory!

Is there a way to keep the COM server alive using always the same COM 
object instance?
I have already performed this task succesfully using CORBA (single servant 
CORBA object), but I need to expose the server as a COM object too.

Have you got any suggestion?

Thanks,
Adolfo.




From aleaxit at yahoo.com  Thu Feb 15 23:23:14 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 15 Feb 2001 23:23:14 +0100
Subject: COM server info
References: <mailman.982263496.7413.python-list@python.org>
Message-ID: <96hkta0274s@news2.newsguy.com>

"Adolfo Benin" <adolfo.benin at riskmap.it> wrote in message
news:mailman.982263496.7413.python-list at python.org...
> Hi,
> I've developed a COM object in Python, then I regitered it as a DCOM on my
> machine.
> Every time a client makes a request to the server DCOM, the Python
> interpreter is launched and the COM is created. But when the client ends,
> the COM server also ends.
> So every call to the server creates a new python interpreter instance in
> memory!
>
> Is there a way to keep the COM server alive using always the same COM
> object instance?
> I have already performed this task succesfully using CORBA (single servant
> CORBA object), but I need to expose the server as a COM object too.
>
> Have you got any suggestion?

There are several approaches.  My suggestion is to eschew the
'singleton' design pattern: let however many COM objects the
clients wish be created, each with its own natural identity; they
can delegate everything to a single 'internal' COM object they
in turn create (which can, e.g., RegisterActiveObject itself with
ACTIVEOBJECT_STRONG to keep the process alive).  Using NT
Servers is also good in a similar approach.


Alex





From sales at buystainlessonline.com  Thu Feb 15 21:19:39 2001
From: sales at buystainlessonline.com (sales at buystainlessonline.com)
Date: Thu, 15 Feb 2001 15:19:39 -0500
Subject: Great Sites!! Three Rivers Demolition Photos!
Message-ID: <0o814w5n6i6y.2i0irr76vk@mackdads.com>

Visit www.Buyscraponline.com for a detailed pitorial of 
last sundays explosion.

Also, if you purchase or distribute Stainless Steel and other Nickel Alloys,
you should visit www.buystainlessonline.com

We also build and host web sites. We are very competitive.

Click Here to be REMOVED FROM THIS MAILING LIST!
http://www.buyscraponline.com/bluemelon/maillist.php

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010215/de11dd2b/attachment.htm>

From tjg at exceptionalminds.com  Thu Feb 15 21:30:08 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Thu, 15 Feb 2001 12:30:08 -0800
Subject: vim and python
Message-ID: <20010215123008.A9349@trufflehunter.avalongroup.net>

Since this is more python related than vim related I'll post it here. I have
recently had contact with a vim missionary (came to my door one Saturday
morning and started telling me that I needed to be saved from emacs). 

Since I had been feeling a bit frustrated with emacs, and since I do suffer
from tendonitis in one of my wrists, I thought I'd give it a shot.
Surprisingly, I quite like it, and it is certainly easier on my bad wrist.

I've been muddling through for a couple of days now, and discovered that
there is quite nice builtin python support. I played with it and got it sort
of working, but have run into what look like environmental problems.

I start vim from the directory where my source files are located, and
attempt to do a :pyf %, and I started getting import errors. Not a problem,
I simply added my source directory to the end of sys.path and tried again.
Now the import works correctly, but I am getting more errors and I can't
explain them.

Traceback (innermost last):
  File "<string>", line 1, in ?
  File "wlautoconf.py", line 20, in ?
    class acOptions(wlOptions):
NameError: wlOptions
Press RETURN or enter command to continue

wlOptions is a class that is imported from a file in my source directory.

So, does anyone have any tips on setting up a nice python/vim environment?

Thanks.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  30 days     1 hours ago<<



From kdahlhaus at yahoo.com  Sun Feb 18 17:17:48 2001
From: kdahlhaus at yahoo.com (Kevin Dahlhausen)
Date: Sun, 18 Feb 2001 16:17:48 GMT
Subject: vim and python
References: <mailman.982268898.28556.python-list@python.org>
Message-ID: <MASj6.2250$ws2.139240@ord-read.news.verio.net>

'Pyf' is attempting to source the file using the embedded Python
interpreter in Vim.  I don't know offhand if it contains different import
hooks or if it plays with the PYTHONPATH.   I think of it as more for
scripting Vim with Python that debugging Python code, although in theory
what you're doing should work fine.    The first thing to try is running
that python file from outside of Vim and see what happens.   

There was a short series of articles on Python / Vim here a few weeks
ago.  When google puts the archives back up, you might check for 'Vim and
Python Efficiently'


By the way, I commend your choice of editor and language!  :)



In article <mailman.982268898.28556.python-list at python.org>, "Timothy
Grant" <tjg at exceptionalminds.com> wrote:

> Since this is more python related than vim related I'll post it here. I
> have recently had contact with a vim missionary (came to my door one
> Saturday morning and started telling me that I needed to be saved from
> emacs). 
> 
> Since I had been feeling a bit frustrated with emacs, and since I do
> suffer from tendonitis in one of my wrists, I thought I'd give it a
> shot. Surprisingly, I quite like it, and it is certainly easier on my
> bad wrist.
> 
> I've been muddling through for a couple of days now, and discovered that
> there is quite nice builtin python support. I played with it and got it
> sort of working, but have run into what look like environmental
> problems.
> 
> I start vim from the directory where my source files are located, and
> attempt to do a :pyf %, and I started getting import errors. Not a
> problem, I simply added my source directory to the end of sys.path and
> tried again. Now the import works correctly, but I am getting more
> errors and I can't explain them.
> 
> Traceback (innermost last):
>   File "<string>", line 1, in ? File "wlautoconf.py", line 20, in ?
>     class acOptions(wlOptions):
> NameError: wlOptions Press RETURN or enter command to continue
> 
> wlOptions is a class that is imported from a file in my source
> directory.
> 
> So, does anyone have any tips on setting up a nice python/vim
> environment?
> 
> Thanks.
>


From tjg at exceptionalminds.com  Sun Feb 18 22:45:17 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Sun, 18 Feb 2001 13:45:17 -0800
Subject: vim and python
In-Reply-To: <MASj6.2250$ws2.139240@ord-read.news.verio.net>; from kdahlhaus@yahoo.com on Sun, Feb 18, 2001 at 04:17:48PM +0000
References: <mailman.982268898.28556.python-list@python.org> <MASj6.2250$ws2.139240@ord-read.news.verio.net>
Message-ID: <20010218134517.E862@trufflehunter.avalongroup.net>

On Sun, Feb 18, 2001 at 04:17:48PM +0000, Kevin Dahlhausen wrote:
> 'Pyf' is attempting to source the file using the embedded Python
> interpreter in Vim.  I don't know offhand if it contains different import
> hooks or if it plays with the PYTHONPATH.   I think of it as more for
> scripting Vim with Python that debugging Python code, although in theory
> what you're doing should work fine.    The first thing to try is running
> that python file from outside of Vim and see what happens.   
> 
It works fine outside Vim, hence the consternation.

> There was a short series of articles on Python / Vim here a few weeks
> ago.  When google puts the archives back up, you might check for 'Vim and
> Python Efficiently'
> 
If it was just a few weeks ago, I probably have it archived somewhere, I'll
go digging.

> 
> By the way, I commend your choice of editor and language!  :)
> 
Thanks, it is definitely improving the the tendonitis problem, but
productivity has been slipping while I learn the new mode of 
operation.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  33 days  1:54 hours ago<<



From mertz at gnosis.cx  Thu Feb 15 21:33:58 2001
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Thu, 15 Feb 2001 15:33:58 -0500
Subject: Detecting RegEx across Python versions
Message-ID: <20Dj6kKkXsbG092yn@bellatlantic.net>                

I have a module that wishes to match files by either "glob"
[fnmatch] or regex.  In a certain function, I hope to just check
a variable ('pattern') to see if it is a string or a compiled
regex, and match accordingly.  Unfortunately, Python 2 seems to
handle this matter differently than Python 1.5.2 (in a way that
does not seem necessary for the improved sre engine).

I have done the following, but would welcome something more
elegant

    [...]
    # kludge to detect a regular expression across python versions
    elif sys.version[0]=='1' and isinstance(pattern, re.RegexObject):
        if pattern.match(name):
            files.append(fname)
    elif sys.version[0]=='2' and type(pattern)==type(re.compile('')):
        if pattern.match(name):
            files.append(fname)
    elif type(pattern) is StringType:
        if fnmatch.fnmatch(name, pattern):
            files.append(fname)

Yours, Lulu...

--
---[ to our friends at TLAs (spread the word) ]--------------------------
Echelon North Korea Nazi cracking spy smuggle Columbia fissionable Stego
White Water strategic Clinton Delta Force militia TEMPEST Libya Mossad
---[ Postmodern Enterprises <mertz at gnosis.cx> ]--------------------------




From fredrik at pythonware.com  Fri Feb 16 11:00:54 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 16 Feb 2001 10:00:54 GMT
Subject: Detecting RegEx across Python versions
References: <mailman.982269918.610.python-list@python.org>
Message-ID: <qT6j6.17383$AH6.2382842@newsc.telia.net>

> I have done the following, but would welcome something more
> elegant
>
>     [...]
>     # kludge to detect a regular expression across python versions
>     elif sys.version[0]=='1' and isinstance(pattern, re.RegexObject):
>         if pattern.match(name):
>             files.append(fname)
>     elif sys.version[0]=='2' and type(pattern)==type(re.compile('')):
>         if pattern.match(name):
>             files.append(fname)
>     elif type(pattern) is StringType:
>         if fnmatch.fnmatch(name, pattern):
>             files.append(fname)

if isinstance(pattern, type("")):
    if fnmatch.fnmatch(name, pattern):
        files.append(fname)
else:
    if pattern.match(name):
        files.append(fname)

or:

try:
    if pattern.match(name):
        files.append(fname)
except AttributeError:
    if fnmatch.fnmatch(name, pattern):
        files.append(fname)

or:

if isinstance(pattern, type("")):
    pattern = re.compile(fnmatch.translate(pattern))

...

if pattern.match(name):
    ...

(insert os.path.normcase where necessary)

Cheers /F




From whisper at oz.net  Thu Feb 15 22:26:36 2001
From: whisper at oz.net (Dave LeBlanc)
Date: 15 Feb 2001 21:26:36 GMT
Subject: ZELP! I can't Zope!!
Message-ID: <96hhic$cqn$0@216.39.170.247>

I'll bet there's some nifty zope forum where I can get help and i'll
be jiggered if I can figure out which of the zope mailing lists are
appropriate for newbie zope help/questions. I figured there would be
some zopers (zopeheads?) lurking around c.l.p and figured i'd see if I
could get a little info on the following problem:

My most immediate problem is that I can't get the zope tutorial to
install by following the directions given for doing so. After
following the instructions to go to the "add products page" (which I'm
guessing is the "Products" page with the "add product" button on it,
when I try to run the tutorial via the help menu (left frame tree
menu), I still get "Can't find tutorial examples - please install
tutorial before continuing". 

Perhaps I didn't enter the correct "ID" when doing the "add product"
bit, but there is no information about what an ID is nor if there is a
specific value that needs to be entered for it. (Note to Zope central:
your documentation could be clearer here guys.)

This is Zope 2.3.0 "binary windows install" with Python 1.5.2 on an NT
4.0sp6 box.

TIA for any help!

Dave LeBlanc


From rerla at channing.harvard.edu  Fri Feb 16 20:32:20 2001
From: rerla at channing.harvard.edu (Ross Lazarus)
Date: Fri, 16 Feb 2001 14:32:20 -0500
Subject: ZELP! I can't Zope!!
References: <96hhic$cqn$0@216.39.170.247>
Message-ID: <3A8D8044.2C2342ED@channing.harvard.edu>

It's been so long that I can't remember exactly what you need to do to
get the tutorial working, but I know that there are lots of places to
look.

Start at http://www.zope.org of course. There's a site search which
might be useful. Not to mention the zope book at
http://www.zope.org/Members/michel/ZB/

I read the zope lists at egroups.com, but they're also archived at the
zope site above. 
zope at zope.org is the place to send questions like this. 
There's also zope-dev at zope.org for developer type questions.

Good luck

Dave LeBlanc wrote:
> 
> I'll bet there's some nifty zope forum where I can get help and i'll
> be jiggered if I can figure out which of the zope mailing lists are
> appropriate for newbie zope help/questions. I figured there would be
> some zopers (zopeheads?) lurking around c.l.p and figured i'd see if I
> could get a little info on the following problem:


From schnoerr at mailzone.com  Thu Feb 15 22:40:58 2001
From: schnoerr at mailzone.com (Claudius =?iso-8859-1?Q?Schn=F6rr?=)
Date: Thu, 15 Feb 2001 22:40:58 +0100
Subject: Q: Can methods and functions be overloaded?
Message-ID: <3A8C4CEA.B418B56E@mailzone.com>

Hello,

I would appreciate to write methods for different counts of
arguments, say
f(self, a) and f( self, b, c, d ) or so where b and a have to be handled
differently.

Is this possible? I think no because the second definition of f
overwrites the first one.

Any hints are welcome.
Please email a copy of your answer directly to 
    Claudius.Schnoerr at ddg.de

Thank you in advance,

Claudius



-- 
----------------------------------------------------------------
Dr.-Ing. Claudius Schn"orr        Tel: ++49-(0)211-5591570
Sigmaringenstr. 2                 Fax: ++49-(0)89-2443-45595
40547 D"usseldorf / Germany       e-mail: schnoerr at mailzone.com
----------------------------------------------------------------


From cfelling at iae.nl  Fri Feb 16 01:59:30 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 16 Feb 2001 01:59:30 +0100
Subject: Q: Can methods and functions be overloaded?
References: <3A8C4CEA.B418B56E@mailzone.com>
Message-ID: <96hu1i$cr4$1@animus.fel.iae.nl>

Claudius Schn?rr <schnoerr at mailzone.com> wrote:
> Hello,

> I would appreciate to write methods for different counts of
> arguments, say
> f(self, a) and f( self, b, c, d ) or so where b and a have to be handled
> differently.

> Is this possible? I think no because the second definition of f
> overwrites the first one.

Head on:)
but head up, Python is very flexible, so build your own overloader

class Overloaded:
    __overload = {}

    def f(self, *args):
        apply(self.__overload[len(args)], (self,) + args)

    def f1(self, a):
        print a
    __overload[1] = f1

    def f2(self, a, b):
        print a, 'and', b
    __overload[2] = f2


u = Overloaded()
u.f('spam')
u.f('spam', 'spam')

-- 
groetjes, carel


From sholden at holdenweb.com  Fri Feb 16 02:38:59 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 15 Feb 2001 20:38:59 -0500
Subject: Can methods and functions be overloaded?
References: <3A8C4CEA.B418B56E@mailzone.com>
Message-ID: <PA%i6.14755$GA5.232956@e420r-atl1.usenetserver.com>

"Claudius Schn?rr" <schnoerr at mailzone.com> wrote in message
news:3A8C4CEA.B418B56E at mailzone.com...
> Hello,
>
> I would appreciate to write methods for different counts of
> arguments, say
> f(self, a) and f( self, b, c, d ) or so where b and a have to be handled
> differently.
>
> Is this possible? I think no because the second definition of f
> overwrites the first one.
>
> Any hints are welcome.
> Please email a copy of your answer directly to
>     Claudius.Schnoerr at ddg.de
>
> Thank you in advance,
>
> Claudius
>
Well, you can use a notation which passes all positional arguments in a
list.

Without testing, something like:

    def meth1(self, *args):
        if len(args) == 1:
            # code one-argument case
        else:
            # code more-than-one-argument case

Another option might be to provide keyword default arguments for the
additional arguments, as in:

    def meth2(self, a, b=None, c=None):
        if b == c == None:
            # code one-argument case
        else:
            # code more-than-one-argument case

Does this help, or did you alredy know about it? I'm afraid Python doesn't
allow the same kind of method overloading by signature that, say, Java and
C++ do.

regards
 Steve




From cfelling at iae.nl  Fri Feb 16 16:24:36 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 16 Feb 2001 16:24:36 +0100
Subject: Can methods and functions be overloaded?
References: <3A8C4CEA.B418B56E@mailzone.com> <PA%i6.14755$GA5.232956@e420r-atl1.usenetserver.com>
Message-ID: <96jgnk$33i$1@animus.fel.iae.nl>

Steve Holden <sholden at holdenweb.com> wrote:
...
> Does this help, or did you alredy know about it? I'm afraid Python doesn't
> allow the same kind of method overloading by signature that, say, Java and
> C++ do.

Right, but you can build your own:)

class Overload:
    def __init__(self):
        self.fun = {}

    def __call__(self, *args):
        try:
            fun = self.fun[self.signature(*args)]
        except KeyError:
            raise TypeError, "invalid overloading signature"
        return fun(*args)

    def signature(self, *args):
        '''smart up if you need to play with class hierarchies'''
        return tuple([type(x) for x in args])

    def overload(self, fun, *args):
        self.fun[self.signature(*args)] = fun


class Overloaded:
    def __init__(self):
        f = self.f = Overload()
        f.overload(self.f_None)
        f.overload(self.f_num, 1)
        f.overload(self.f_num, 1.0)
        f.overload(self.f_string, "1")
        f.overload(self.f_more, 'spam', 1)

    def f_None(self):
        return None

    def f_num(self, n):
        return n * 3

    def f_string(self, s):
        return 'string is %s' % s

    def f_more(self, s, n):
        return s * n

o = Overloaded()

print o.f()
print o.f(10)
print o.f(5.4)
print o.f("spam", 9)
print o.f(9, "spam")


this yields:

None
30
16.2
spamspamspamspamspamspamspamspamspam
Traceback (most recent call last):
  File "t.py", line 48, in ?
    print o.f(9, "spam")
  File "t.py", line 9, in __call__
    raise TypeError, "invalid overloading signature"
TypeError: invalid overloading signature

-- 
groetjes, carel


From info at black-karma.de  Thu Feb 15 22:42:52 2001
From: info at black-karma.de (Harvest T. Moon)
Date: Thu, 15 Feb 2001 22:42:52 +0100
Subject: [newbie] easy tutorial on multithreading?
Message-ID: <96hifo$lq0fh$1@ID-22517.news.dfncis.de>

i'm looking for a rather simple instruction to multi-threading with python,
i.e. starting a function as seperate thread etc.

thnx i.a.
Harvest T. Moon
-- a life less ordinary!!




From aahz at panix.com  Fri Feb 16 00:02:18 2001
From: aahz at panix.com (Aahz Maruch)
Date: 15 Feb 2001 15:02:18 -0800
Subject: [newbie] easy tutorial on multithreading?
References: <96hifo$lq0fh$1@ID-22517.news.dfncis.de>
Message-ID: <96hn5q$do0$1@panix6.panix.com>

In article <96hifo$lq0fh$1 at ID-22517.news.dfncis.de>,
Harvest T. Moon <info at black-karma.de> wrote:
>
>i'm looking for a rather simple instruction to multi-threading with python,
>i.e. starting a function as seperate thread etc.

This is rather sketchy, but take a look at
http://starship.python.net/crew/aahz

There'll be something better up in a couple of weeks (my slides for
IPC9).
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From grisha at verio.net  Thu Feb 15 22:55:03 2001
From: grisha at verio.net (Gregory (Grisha) Trubetskoy)
Date: Thu, 15 Feb 2001 16:55:03 -0500
Subject: += as declaration wish
Message-ID: <Pine.BSF.4.32.0102151652290.71736-100000@localhost>

Am I missing something obvious when I think that it'd be nice if you could
do:

x += 3

and if x didn't exist, += would simply act like = rather than raise an
exception.

Grisha



From jafo at tummy.com  Thu Feb 15 23:54:49 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Thu, 15 Feb 2001 15:54:49 -0700
Subject: += as declaration wish
In-Reply-To: <Pine.BSF.4.32.0102151652290.71736-100000@localhost>; from grisha@verio.net on Thu, Feb 15, 2001 at 04:55:03PM -0500
References: <Pine.BSF.4.32.0102151652290.71736-100000@localhost>
Message-ID: <20010215155449.E23577@tummy.com>

On Thu, Feb 15, 2001 at 04:55:03PM -0500, Gregory (Grisha) Trubetskoy wrote:
>Am I missing something obvious when I think that it'd be nice if you could
>do:
>
>x += 3

What does x default to if it doesn't exist?  None?  0? ""?

If it worked, I'd still explicitly set the LHS first anyway, I'm fairly sure.

Sean
-- 
 Linux:  When you need to run like a greased weasel.
                 -- Sean Reifschneider, 1998
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From max at alcyone.com  Fri Feb 16 03:46:58 2001
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 15 Feb 2001 18:46:58 -0800
Subject: += as declaration wish
References: <Pine.BSF.4.32.0102151652290.71736-100000@localhost>
Message-ID: <3A8C94A2.438AAA3E@alcyone.com>

"Gregory (Grisha) Trubetskoy" wrote:

> Am I missing something obvious when I think that it'd be nice if you
> could
> do:
> 
> x += 3
> 
> and if x didn't exist, += would simply act like = rather than raise an
> exception.

Avoiding such magic like this is what makes Python a well-designed
language.  If you want x to default to something, you can set it to that
yourself before you start using +=.  (Besides, what should it default to
for ints? strings? lists? user-defined types?)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Love is the most subtle form of self-interest.
\__/ Holbrook Jackson
    Polly Wanna Cracka? / http://www.pollywannacracka.com/
 The Internet resource for interracial relationships.


From tim.one at home.com  Fri Feb 16 04:39:57 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 15 Feb 2001 22:39:57 -0500
Subject: += as declaration wish
In-Reply-To: <3A8C94A2.438AAA3E@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIELHIPAA.tim.one@home.com>

[Gregory (Grisha) Trubetskoy]
> Am I missing something obvious when I think that it'd be nice if you
> could do:
>
> x += 3
>
> and if x didn't exist, += would simply act like = rather than raise an
> exception.

Only that it's semantically confused, so Guido will never buy it.

[Erik Max Francis]
> Avoiding such magic like this is what makes Python a well-designed
> language.  If you want x to default to something, you can set it to that
> yourself before you start using +=.  (Besides, what should it default to
> for ints? strings? lists? user-defined types?)

That's not a problem.  Gregory didn't ask for a default:  he wants

   LHS op= RHS

to act like

try:
    LHS
except NameError:
    LHS = RHS
else:
    LHS op= RHS

No defaults are involved.  So it's not an intractable idea.  By analogy, we
could also make f(x) act like

try:
    x
except NameError:
    f()
else:
    f(x)

and so on and so on.  Then

    x += f(y, z)

gets *really* cute <wink>.

if-a-thing's-a-rotten-idea-in-the-large-it's-probably-not-a-great-
    idea-in-the-small-ly y'rs  - tim




From bbollenbach at homenospam.com  Fri Feb 16 05:21:48 2001
From: bbollenbach at homenospam.com (Brad Bollenbach)
Date: Fri, 16 Feb 2001 04:21:48 GMT
Subject: += as declaration wish
References: <Pine.BSF.4.32.0102151652290.71736-100000@localhost>
Message-ID: <wV1j6.1506$TU6.282659@news1.rdc1.mb.home.com>

The reason this isn't, and shouldn't be allowed is simple. From the Python
Documentation:

"An augmented assignment evaluates the target (which, unlike normal
assignment statements, cannot be an unpacking) and the expression list,
performs the binary operation specific to the type of assignment on the two
operands, and assigns the result to the original target."

In other words, when you do x += 1, it will evaluate x, and perform (x + 1)
before assigning back to x, which is an error in Python if x doesn't exist.

I'm curious: in what situations does this do more to help the look of your
code than not? "Working around" this should never cause any problems.

"Gregory (Grisha) Trubetskoy" <grisha at verio.net> wrote in message
news:Pine.BSF.4.32.0102151652290.71736-100000 at localhost...
>
> Am I missing something obvious when I think that it'd be nice if you could
> do:
>
> x += 3
>
> and if x didn't exist, += would simply act like = rather than raise an
> exception.
>
> Grisha
>




From ssthapa at barnacle-bill.cs.uchicago.edu  Fri Feb 16 19:49:19 2001
From: ssthapa at barnacle-bill.cs.uchicago.edu (Suchandra S Thapa)
Date: Fri, 16 Feb 2001 18:49:19 GMT
Subject: += as declaration wish
References: <mailman.982294882.3768.python-list@python.org>
Message-ID: <slrn98qthh.1ri.ssthapa@barnacle-bill.cs.uchicago.edu>

	I agree with you.  I think having  x += y create a new
x would create more problems than it solves.  Currently typos in the 
variable name would raise an error that is easily caught and fixed, having
a new variable silently created just seems like a bad idea. I personally would
love it if python went the other way and allowed an optional mode that 
required variables to be declared before they were used. 



From Jerry.Spicklemire at IFLYATA.COM  Thu Feb 15 23:09:20 2001
From: Jerry.Spicklemire at IFLYATA.COM (Spicklemire, Jerry)
Date: Thu, 15 Feb 2001 17:09:20 -0500
Subject: Upgrade? Why not?
Message-ID: <977A39E65CFCD3119ABF00D0B741D849E0FEF2@innt-73.ata.com>

Mark Hudson asked?

"Have you upgraded to Python 2.0?"

no (he stammered in mortification)

"If not, why not?"

Because I'm not a sysadmin, just a lowly coder. 
It was hard enough to get Python installed and 
compiled in the first place. Even though I do 
stand up and hollar for a Zope upgrade when it 
seems absolutely essential, if Python 1.5.2 is 
good enough for the standard Zope distribution, 
then as far as I can tell, "it ain't broke."

With that out of the way, let me assure all 
that from my perspective the new developments 
sound fabulous, wonderful, and seemingly have 
the unimaginable effect of making Python even 
more perfect. I was never very good at math or 
physics, so I just tuck these little mysteries 
into the same corner as time slowing to a crawl 
at the center of a black hole. Even though I 
can't imagine how it's possible, it is such a 
kick to watch it happen. Have you folks ever 
thought of doing a time-lapse photo thingy on 
the Python development process?

Later, 
Jerry S.




From jafo-pythonlist at tummy.com  Thu Feb 15 23:52:37 2001
From: jafo-pythonlist at tummy.com (Sean Reifschneider)
Date: Thu, 15 Feb 2001 15:52:37 -0700
Subject: Upgrade? Why not?
In-Reply-To: <977A39E65CFCD3119ABF00D0B741D849E0FEF2@innt-73.ata.com>; from Jerry.Spicklemire@IFLYATA.COM on Thu, Feb 15, 2001 at 05:09:20PM -0500
References: <977A39E65CFCD3119ABF00D0B741D849E0FEF2@innt-73.ata.com>
Message-ID: <20010215155237.D23577@tummy.com>

On Thu, Feb 15, 2001 at 05:09:20PM -0500, Spicklemire, Jerry wrote:
>"If not, why not?"

I write code that runs not only on my box, but on many of the tummy.com
boxes, the machines at our co-location facility, and our client's machines.
If I take advantage of things in 2.0 it requires me upgrading probably
at least half of those machines.

There's nothing I really *NEED*.  About the only thing I really may want is
SSL sockets for my network backup system, and I can make that just bypass the
appropriate code when running on a platform without.  So, I'll just keep
happily using 1.5.2 and wait until 2.0 is more prevalent before I start
using it.

My laptop?  Yeah, it has 1.5.2, 2.0 and 2.1a2 installed...  Mostly for my
Python presentations...

Sean
-- 
 Money is the root of all evil!  Man needs roots...
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From stephen_purcell at yahoo.com  Fri Feb 16 09:35:38 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 16 Feb 2001 09:35:38 +0100
Subject: Upgrade? Not for some time...
In-Reply-To: <977A39E65CFCD3119ABF00D0B741D849E0FEF2@innt-73.ata.com>; from Jerry.Spicklemire@IFLYATA.COM on Thu, Feb 15, 2001 at 05:09:20PM -0500
References: <977A39E65CFCD3119ABF00D0B741D849E0FEF2@innt-73.ata.com>
Message-ID: <20010216093538.E7411@freedom.puma-ag.com>

Spicklemire, Jerry wrote:
> Mark Hudson asked?
> "Have you upgraded to Python 2.0?"

I'm still using 1.5.2, with no plans to upgrade. My reasons:

 1. Python 1.5.2 and all its juicy C extensions (PIL, PyGreSQL etc) are
    officially supported with binary packages on Linux and elsewhere. Python
    2.0 will probably not make its way into the core of any big Linux distros
    for quite a few months. No source installs of Python + extensions for me.

 2. Until I feel like only cavemen are still using 1.5.2, I'll avoid using
    2.0 in case I end up writing funky code with list comprehensions that
    nobody can use.

 3. With the current rate of releases and core python development, I'll
    save the effort of upgrading until 2.1 (or whatever) becomes the '1.5.2'
    of the future!

Makes me think of Java's JDK 1.3 -- why bother? I haven't needed most of
JDK 1.2 yet, and at least *everyone* has that version. For that matter, some
people are still starting new projects using JDK 1.1, and Java 1.0 is the
only serious choice for applets.

I hope Python doesn't go the same way.

-Caveman Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From tanzer at slinky.cs.nyu.edu  Tue Feb 20 22:40:43 2001
From: tanzer at slinky.cs.nyu.edu (David Tanzer)
Date: 20 Feb 2001 16:40:43 -0500
Subject: Upgrade? Not for some time...
References: <977A39E65CFCD3119ABF00D0B741D849E0FEF2@innt-73.ata.com> <mailman.982312520.2099.python-list@python.org>
Message-ID: <0mwvaldodg.fsf@slinky.cs.nyu.edu>

Steve Purcell <stephen_purcell at yahoo.com> writes:

> Spicklemire, Jerry wrote:
>  2. Until I feel like only cavemen are still using 1.5.2, I'll avoid using
>     2.0 in case I end up writing funky code with list comprehensions that
>     nobody can use.
> 
I myself plan to put the list comprehensions to good use.




From dan at eevolved.com  Fri Feb 16 00:10:51 2001
From: dan at eevolved.com (Dan Parisien)
Date: Thu, 15 Feb 2001 23:10:51 GMT
Subject: segfault. which module to blame?
Message-ID: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com>

I am using shelve and a lot of threads to write to a dbm db. I implemented 
my own row-level locking with dictionaries and counters and when I run a 
test script that simulates extreme conditions (1500 threads, 10 writes 
each) it segfaults 1/100 times it runs.

That kinda sucks, don't you think? I'm wondering if someone who has 
experience could point me to which module is at fault: threading? or 
shelve+dbm? or neither? :) Or maybe some technique on how to crawl through 
the core file (I'm on mandrake gnu/linux 7.2 python 2.0.1)

I'm not expecting a correct answer, just a point in the right directions

Thank you :)
Dan



From dsh8290 at rit.edu  Fri Feb 16 02:01:54 2001
From: dsh8290 at rit.edu (D-Man)
Date: Thu, 15 Feb 2001 20:01:54 -0500
Subject: segfault. which module to blame?
In-Reply-To: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com>; from dan@eevolved.com on Thu, Feb 15, 2001 at 11:10:51PM +0000
References: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com>
Message-ID: <20010215200154.C2150@harmony.cs.rit.edu>

$ gdb `which python` core
<startup info from gdb>
gdb> backtrace
<backtrace from gdb>

Basically  gdb <executable image> <corefile>  will load the
executable's symbols (assuming it has some) and the core file.  It
will warn you if the exe has no symbols, or if the core came from a
different exe.  Backtrace will give you a trace similar to when a
python exception isn't caught (except it is for C, and thus much lower
level detail).

DDD (from gnu.org) is a nice (except that is uses motif) gui frontend
for gdb that will make browsing symbols, etc much easier (and you
won't need to learn yet another command line, yet).

HTH,
-D

On Thu, Feb 15, 2001 at 11:10:51PM +0000, Dan Parisien wrote:
| I am using shelve and a lot of threads to write to a dbm db. I implemented 
| my own row-level locking with dictionaries and counters and when I run a 
| test script that simulates extreme conditions (1500 threads, 10 writes 
| each) it segfaults 1/100 times it runs.
| 
| That kinda sucks, don't you think? I'm wondering if someone who has 
| experience could point me to which module is at fault: threading? or 
| shelve+dbm? or neither? :) Or maybe some technique on how to crawl through 
| the core file (I'm on mandrake gnu/linux 7.2 python 2.0.1)
| 
| I'm not expecting a correct answer, just a point in the right directions
| 
| Thank you :)
| Dan
| 



From dan at eevolved.com  Fri Feb 16 02:55:23 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 01:55:23 GMT
Subject: segfault. which module to blame? (entire traceback)
References: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com> <mailman.982285338.32590.python-list@python.org>
Message-ID: <fM%i6.239159$f36.9599321@news20.bellglobal.com>

D-Man wrote:

> 
> $ gdb `which python` core
> <startup info from gdb>
> gdb> backtrace
> <backtrace from gdb>

Cool.
It seems to crash in the garbage collecting routines. Maybe someone with 
more experience than me can help. Since I do a lot of shelve stuff (no 
threads in this test) there that explains cPickle. Anyways, maybe this is a 
known bug... Here's the entire traceback.

#0  0x809139d in visit_decref (op=0x869bd94, data=0x0) at ./gcmodule.c:153
#1  0x807ed67 in dict_traverse (op=0x86da2c4, visit=0x8091390 
<visit_decref>, arg=0x0)
    at dictobject.c:1146
#2  0x80913d6 in subtract_refs (containers=0x80c7818) at ./gcmodule.c:167
#3  0x80917f7 in collect (young=0x80c7818, old=0x80c7818) at 
./gcmodule.c:379
#4  0x80919e7 in collect_generations () at ./gcmodule.c:467
#5  0x8091acd in _PyGC_Insert (op=0x86e7dcc) at ./gcmodule.c:507
#6  0x8085537 in PyTuple_New (size=2) at tupleobject.c:87
#7  0x40182da7 in put2 (self=0x86da700, ob=0x82381a0) at ./cPickle.c:724
#8  0x40182d0a in put (self=0x86da700, ob=0x82381a0) at ./cPickle.c:699
#9  0x40183947 in save_string (self=0x86da700, args=0x82381a0, doput=1) at 
./cPickle.c:1141
#10 0x40184c0b in save (self=0x86da700, args=0x82381a0, pers_save=0) at 
./cPickle.c:1786
#11 0x40183fcf in save_dict (self=0x86da700, args=0x816e32c) at 
./cPickle.c:1402
#12 0x40184c91 in save (self=0x86da700, args=0x816e32c, pers_save=0) at 
./cPickle.c:1814
#13 0x40185015 in dump (self=0x86da700, args=0x816e32c) at ./cPickle.c:1930
#14 0x40185425 in Pickler_dump (self=0x86da700, args=0x810f52c) at 
./cPickle.c:2093
#15 0x805835e in call_builtin (func=0x8131cf8, arg=0x810f52c, kw=0x0) at 
ceval.c:2650
#16 0x805826b in PyEval_CallObjectWithKeywords (func=0x8131cf8, 
arg=0x810f52c, kw=0x0)
    at ceval.c:2618
#17 0x8057294 in eval_code2 (co=0x810ecd8, globals=0x8110254, locals=0x0, 
args=0x813d7a0,
    argcount=3, kws=0x0, kwcount=0, defs=0x0, defcount=0, owner=0x811f5ac) 
at ceval.c:1951
#18 0x80586be in call_function (func=0x81185a4, arg=0x86c1aec, kw=0x0) at 
ceval.c:2772
#19 0x805825d in PyEval_CallObjectWithKeywords (func=0x813aa34, 
arg=0x86c1aec, kw=0x0)
    at ceval.c:2616
#20 0x807286e in instance_ass_subscript (inst=0x812042c, key=0x814f108, 
value=0x816e32c)
    at classobject.c:925
#21 0x806da29 in PyObject_SetItem (o=0x812042c, key=0x814f108, 
value=0x816e32c) at abstract.c:117
#22 0x8055ead in eval_code2 (co=0x810aea0, globals=0x810b02c, locals=0x0, 
args=0x8109da8,
    argcount=1, kws=0x8109dac, kwcount=0, defs=0x0, defcount=0, 
owner=0x811df74) at ceval.c:1208
#23 0x8056fa5 in eval_code2 (co=0x810b1c8, globals=0x810b02c, locals=0x0, 
args=0x811d2b0,
    argcount=3, kws=0x0, kwcount=0, defs=0x0, defcount=0, owner=0x811df74) 
at ceval.c:1850
#24 0x80586be in call_function (func=0x810f6d4, arg=0x86b5fbc, kw=0x0) at 
ceval.c:2772
#25 0x805825d in PyEval_CallObjectWithKeywords (func=0x80d067c, 
arg=0x86b5fbc, kw=0x0)
---Type <return> to continue, or q <return> to quit---
    at ceval.c:2616
#26 0x807286e in instance_ass_subscript (inst=0x813d35c, key=0x86e9f20, 
value=0x814ee84)
    at classobject.c:925
#27 0x806da29 in PyObject_SetItem (o=0x813d35c, key=0x86e9f20, 
value=0x814ee84) at abstract.c:117
#28 0x8055ead in eval_code2 (co=0x810cb90, globals=0x810b02c, locals=0x0, 
args=0x80dda5c,
    argcount=3, kws=0x80dda68, kwcount=0, defs=0x0, defcount=0, 
owner=0x813bf5c) at ceval.c:1208
#29 0x8056fa5 in eval_code2 (co=0x810c0e8, globals=0x80d284c, 
locals=0x80d284c, args=0x0,
    argcount=0, kws=0x0, kwcount=0, defs=0x0, defcount=0, owner=0x0) at 
ceval.c:1850
#30 0x8054c14 in PyEval_EvalCode (co=0x810c0e8, globals=0x80d284c, 
locals=0x80d284c)
    at ceval.c:319
#31 0x8068a21 in run_node (n=0x80db910, filename=0xbffff9d5 "huge.py", 
globals=0x80d284c,
    locals=0x80d284c) at pythonrun.c:886
#32 0x80689d3 in run_err_node (n=0x80db910, filename=0xbffff9d5 "huge.py", 
globals=0x80d284c,
    locals=0x80d284c) at pythonrun.c:874
#33 0x80689a7 in PyRun_FileEx (fp=0x80ce828, filename=0xbffff9d5 "huge.py", 
start=257,
    globals=0x80d284c, locals=0x80d284c, closeit=1) at pythonrun.c:866
#34 0x8067fd4 in PyRun_SimpleFileEx (fp=0x80ce828, filename=0xbffff9d5 
"huge.py", closeit=1)
    at pythonrun.c:579
#35 0x8067bdf in PyRun_AnyFileEx (fp=0x80ce828, filename=0xbffff9d5 
"huge.py", closeit=1)
    at pythonrun.c:459
#36 0x8051b3d in Py_Main (argc=2, argv=0xbffff864) at main.c:289
#37 0x80515fa in main (argc=2, argv=0xbffff864) at python.c:10
#38 0x400b1cbe in __libc_start_main () from /lib/libc.so.6       



From jeremy at alum.mit.edu  Fri Feb 16 02:13:10 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Thu, 15 Feb 2001 20:13:10 -0500 (EST)
Subject: segfault. which module to blame? (entire traceback)
In-Reply-To: <fM%i6.239159$f36.9599321@news20.bellglobal.com>
References: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com>
	<mailman.982285338.32590.python-list@python.org>
	<fM%i6.239159$f36.9599321@news20.bellglobal.com>
Message-ID: <14988.32422.118120.352796@w221.z064000254.bwi-md.dsl.cnc.net>

Are there any C extensions being used with this process?  My first
guess would be bad reference counting in a third-party extension.
Just in case that guess is wrong, what are you pickling?

Jeremy



From dan at eevolved.com  Fri Feb 16 18:21:30 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 17:21:30 GMT
Subject: segfault. which module to blame? (entire traceback)
References: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com>	<mailman.982285338.32590.python-list@python.org>	<fM%i6.239159$f36.9599321@news20.bellglobal.com> <mailman.982340600.13592.python-list@python.org>
Message-ID: <ukdj6.263124$JT5.9137465@news20.bellglobal.com>

Jeremy Hylton wrote:

> Are there any C extensions being used with this process?  My first
> guess would be bad reference counting in a third-party extension.
> Just in case that guess is wrong, what are you pickling?
> 
> Jeremy
> 
Unfortunately everything I use came with BeOpen Python 2.0.1's rpm on 
python.org.

I write normal python dictionaries and strings to a shelve db. I also use 
locks (threading.Lock) to implement row-level locking in shelve. I've had 
segfaults before while coding normal python scripts. I think it was the 
dictionaries because that's pretty much all I used. Could it be bad 
reference counting in those? I would have to read the code, but could there 
be a problem if two keys hashed to the same value in the dictionary and 
that one is deleted before the other or something similar?

thanks Jeremy,
Dan


From dsh8290 at rit.edu  Fri Feb 16 19:43:34 2001
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 16 Feb 2001 13:43:34 -0500
Subject: segfault. which module to blame? (entire traceback)
In-Reply-To: <ukdj6.263124$JT5.9137465@news20.bellglobal.com>; from dan@eevolved.com on Fri, Feb 16, 2001 at 05:21:30PM +0000
References: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com> <mailman.982285338.32590.python-list@python.org> <fM%i6.239159$f36.9599321@news20.bellglobal.com> <mailman.982340600.13592.python-list@python.org> <ukdj6.263124$JT5.9137465@news20.bellglobal.com>
Message-ID: <20010216134334.B3862@harmony.cs.rit.edu>

On Fri, Feb 16, 2001 at 05:21:30PM +0000, Dan Parisien wrote:
[snip]
| segfaults before while coding normal python scripts. I think it was the 
| dictionaries because that's pretty much all I used. Could it be bad 
| reference counting in those? I would have to read the code, but could there 
| be a problem if two keys hashed to the same value in the dictionary and 
| that one is deleted before the other or something similar?

I'll find out by making a class that always hashes to the same value :

class Hashable :
...     def __hash__( self ) :
...             return 5
...
>>> h1 = Hashable()
>>> h2 = Hashable()
>>> id( h1 )
8185684
>>> id( h2 )
8182548
>>> d = { }
>>> d[ h1 ] = 1
>>> d[ h2 ] = 2
>>> print d
{<__main__.Hashable instance at 007CE754>: 1, <__main__.Hashable
instance at 007CDB14>: 2}
>>>
>>> del d[ h1 ]
>>> print d
{<__main__.Hashable instance at 007CDB14>: 2}
>>> del d[ h2 ]
>>>


It didn't crash for me (Win2k, Python 2.0) when I had 2 keys with the
same hash value.

-D




From tim.one at home.com  Fri Feb 16 20:31:27 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 14:31:27 -0500
Subject: segfault. which module to blame? (entire traceback)
In-Reply-To: <ukdj6.263124$JT5.9137465@news20.bellglobal.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEOCIPAA.tim.one@home.com>

[Dan Parisien]
> I write normal python dictionaries and strings to a shelve db. I also use
> locks (threading.Lock) to implement row-level locking in shelve. I've had
> segfaults before while coding normal python scripts.

That's a bug, then.  Did you report it?  I don't see segfaults while coding
normal Python scripts, so don't expect it to get fixed by magic.  There
don't appear to be any relevant bug reports from others open on SourceForge
either.

> I think it was the dictionaries because that's pretty much all I used.
> Could it be bad reference counting in those? I would have to read the
> code, but could there be a problem if two keys hashed to the same value
> in the dictionary and that one is deleted before the other or something
> similar?

While it's not impossible that dicts are screwing up, it's extremely
unlikely, because the dict code is very heavily used (don't forget that
Python uses dicts under the covers to manage your namespaces and sys.modules
and ... there is no Python program-- not even the one-liner "pass" --that
doesn't use dicts!).  So if dicts had refcount problems, it's almost certain
we'd be deluged with bug reports.  Because of the Birthday Paradox,
collisions aren't rare, so that code is heavily used too.

Try to come up with a minimal failing case, then open a bug on SourceForge.




From dan at eevolved.com  Fri Feb 16 20:54:13 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 19:54:13 GMT
Subject: segfault. which module to blame? (entire traceback)
References: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com> <mailman.982285338.32590.python-list@python.org> <fM%i6.239159$f36.9599321@news20.bellglobal.com> <mailman.982340600.13592.python-list@python.org> <ukdj6.263124$JT5.9137465@news20.bellglobal.com> <mailman.982349059.17107.python-list@python.org>
Message-ID: <Fzfj6.241885$f36.9683324@news20.bellglobal.com>

D-Man wrote:

> On Fri, Feb 16, 2001 at 05:21:30PM +0000, Dan Parisien wrote:
> [snip]
> | segfaults before while coding normal python scripts. I think it was the
> | dictionaries because that's pretty much all I used. Could it be bad
> | reference counting in those? I would have to read the code, but could
> | there be a problem if two keys hashed to the same value in the
> | dictionary and that one is deleted before the other or something
> | similar?
> 
> I'll find out by making a class that always hashes to the same value :
> 
> class Hashable :
> ...     def __hash__( self ) :
> ...             return 5
> ...
> >>> h1 = Hashable()
> >>> h2 = Hashable()
> >>> id( h1 )
> 8185684
> >>> id( h2 )
> 8182548
> >>> d = { }
> >>> d[ h1 ] = 1
> >>> d[ h2 ] = 2
> >>> print d
> {<__main__.Hashable instance at 007CE754>: 1, <__main__.Hashable
> instance at 007CDB14>: 2}
> >>>
> >>> del d[ h1 ]
> >>> print d
> {<__main__.Hashable instance at 007CDB14>: 2}
> >>> del d[ h2 ]
> >>>
> 
> 
> It didn't crash for me (Win2k, Python 2.0) when I had 2 keys with the
> same hash value.
> 
> -D

I didn't think so :)

Just a hypothesis. One that shouldn't have made it out to the newsgroup...

Anyways, the two crashes are not related (although they are both segfaults, 
the similarity ends there)

I'm almost certain the problem is with dbm.

Dan



From boswell at caltech.edu  Fri Feb 16 00:47:07 2001
From: boswell at caltech.edu (Dustin Boswell)
Date: Thu, 15 Feb 2001 15:47:07 -0800
Subject: how to compile https support into Python2.0
Message-ID: <3A8C6A7B.42C809C4@caltech.edu>

I originally posted a question asking how to get Python to support
urls like https://secure-server.com so you could use urllib.urlopen()
transparently.

After a bunch of searching on the web, here's what I did (on Redhat 6.2)

to get the Python Distribution to compile correctly with SSL...

(If it doesn't work, or there's an easier/better way, please correct
me.)

OPENSSL
This is a package that you need to have installed before you
can build the Python source distribution (in the SSL way).

1) go to http://www.openssl.org/source/
 and download openssl-0.9.6.tar.gz
2) gunzip openssl-0.9.6.tar.gz
3) tar -xvf openssl-0.9.6.tar (make sure you are in the
directory you want to put the source in; this can be in
any temporary place, since the build will automatically
place it in the right spot later)
4) cd openssl-0.9.6
5) read the INSTALL file and...
6) ./config ; make ; make test ; make install
7) if that didn't work, I can't help you

This should have installed files into a /usr/local/ssl
directory (along with various others)


PYTHON
Here's where and how to build a Python interpreter..

1) Go to http://www.python.org/2.0/#download
   and download BeOpen-Python-2.0.tar.gz
2) gunzip BeOpen-Python-2.0.tar.gz
3) tar -xvf BeOpen-Python-2.0.tar ( same deal as step 3 above )
4) goto
http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/python/dist/src/?cvsroot=python#dirlist

        and make sure you have fresh copies of:
        - Modules/socketmodule.c
        - Lib/urllib.py
        - Lib/httplib.py
        - (possilby others, although I didn't, and haven't (yet) run
        into problems
5) in Python-2.0/Modules/ do:
     cp Setup.in Setup
6) emacs (or, gulp, vi if you prefer) Setup
7) uncomment the four lines so that it looks like:
# Socket module compiled with SSL support; you must edit the SSL
variable:
SSL=/usr/local/ssl
_socket socketmodule.c \
        -DUSE_SSL -I$(SSL)/include -I$(SSL)/include/openssl \
        -L$(SSL)/lib -lssl -lcrypto

8) in the top directory ( Python-2.0/ ) do ...
        ./configure ; make ; make test (took a while) ; make install


Dustin Boswell
boswell at caltech.edu



From p.g at figu.no  Fri Feb 16 01:04:34 2001
From: p.g at figu.no (Per Gummedal)
Date: Fri, 16 Feb 2001 01:04:34 +0100
Subject: [ann]: Python 2.0 for AS/400, beta 3
Message-ID: <194-1908137407.20010216010434@figu.no>

A new update. http://home.no.net/pgummeda

Changes:
 - The File400 method, read, has been replaced with the following:
   readf, readl, readp, readn, readrrn, readeq, readne, readpe
   Some new methods: posf, posl, posb, posa
   getRecord now returns a Record400 object.
   Use _<fieldname> to access fields in a File or Record.
   ( print f._CUSNO )
   blkrcd=Y is now default when file is opened for reading.
   Some bugs has been fixed.
   
 - New module os400 has the following functions.
   Program - Returns a Program object. Example of use:
             pgm = os400.Program('GETPRICE','*LIBL',(('c',10),('d',5,1)))
             pgm.call('TEST1',0))
             price = pgm[1]

   rtvdtaara - Retrieve data area

   chgdtaara - Change data area

   
Per





From syver at NOSPAMcyberwatcher.com  Fri Feb 16 02:00:14 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 16 Feb 2001 02:00:14 +0100
Subject: Python emacs mode (newbie)
Message-ID: <96hu7d$4pq$1@troll.powertech.no>

Is there anybody who has any idea on how to implement a module loading
function in GNU Emacs like the one Pythonwin has?

The basic idea is that instead of writing a file name you write a module
name, like ex:
httplib or win32com.client.dynamic and the editor will load the specified
module.

I guess the implementation would work around searching the python path, but
I am total newbie in elisp (and lisp itself for that matter) and could use
some pointers as to how to search the filesystem from elisp, get the
registry keys or environment variables for the pythonpath, and how to load a
file into the current window.




From crmartin at indra.com  Fri Feb 16 02:14:20 2001
From: crmartin at indra.com (Charles R Martin)
Date: 15 Feb 2001 18:14:20 -0700
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no>
Message-ID: <m3ae7nzaxf.fsf@localhost.localdomain>

I don't know Python at all, but what you're describing sounds like
require/provide. 

See C-h f require RET
    C-h f provide RET


"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:

> Is there anybody who has any idea on how to implement a module loading
> function in GNU Emacs like the one Pythonwin has?
> 
> The basic idea is that instead of writing a file name you write a module
> name, like ex:
> httplib or win32com.client.dynamic and the editor will load the specified
> module.
> 
> I guess the implementation would work around searching the python path, but
> I am total newbie in elisp (and lisp itself for that matter) and could use
> some pointers as to how to search the filesystem from elisp, get the
> registry keys or environment variables for the pythonpath, and how to load a
> file into the current window.
> 
> 

-- 
Whatever you can do or dream you can do, begin it.  Boldness has genius, power
and magic in it.  Begin it now."  --Goethe
------------------------------------------------------------------------------
				 Charlie Martin, Broomfield, CO USA 40 N 105 W


From evgenyr at cs.washington.edu  Fri Feb 16 09:18:05 2001
From: evgenyr at cs.washington.edu (Evgeny Roubinchtein)
Date: Fri, 16 Feb 2001 08:18:05 GMT
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no>
Message-ID: <863ddf59e4.fsf@ruby.localdomain>

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:

> Is there anybody who has any idea on how to implement a module loading
> function in GNU Emacs like the one Pythonwin has?
> 
> The basic idea is that instead of writing a file name you write a module
> name, like ex:
> httplib or win32com.client.dynamic and the editor will load the specified
> module.


How do you want this to work?  So you want a function like
(my-find-python-module "foo"), and up comes the window containing the
window containing the module foo. 

A slightly different approach would be to require that the user have
an inferior python process running, and ask that process to locate the
module, then Emacs just loads it...  (i.e., Python, rather than Emacs
is doing the path searching). 

If you haven't been exposed to lisp at all, you'll probably want to
start with the Emacs lisp introduction, which can be found at:

http://www.gnu.org/manual/emacs-lisp-intro/emacs-lisp-intro.html

You will then want to use the Emacs Lisp Reference Manual:

http://www.gnu.org/manual/elisp-manual-20-2.5/elisp.html

Happy hacking! 

-- 
Evgeny

CPE: Create Parity Error


From Kai.Grossjohann at CS.Uni-Dortmund.DE  Fri Feb 16 13:12:07 2001
From: Kai.Grossjohann at CS.Uni-Dortmund.DE (Kai =?iso-8859-1?q?Gro=DFjohann?=)
Date: 16 Feb 2001 13:12:07 +0100
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no>
Message-ID: <vafwvaqltd4.fsf@lucy.cs.uni-dortmund.de>

On Fri, 16 Feb 2001, Syver Enstad wrote:

> The basic idea is that instead of writing a file name you write a
> module name, like ex: httplib or win32com.client.dynamic and the
> editor will load the specified module.

You mean you want something like `C-x C-f', but with module names, not
file names?

There is a library, ffap.el, which provides a framework for this.  But
I couldn't see any Python-specific support in it.  However, this could
presumably fairly easily be added.  See the variable ffap-alist for
details.  After adding Python support, you can hit `M-x
find-file-at-point RET' in the *.py buffer and you can open the right
file. 

kai
-- 
Be indiscrete.  Do it continuously.


From syver at NOSPAMcyberwatcher.com  Sat Feb 17 03:05:17 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Sat, 17 Feb 2001 03:05:17 +0100
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no> <vafwvaqltd4.fsf@lucy.cs.uni-dortmund.de>
Message-ID: <96kmdc$47q$1@troll.powertech.no>

"Kai Gro?johann" <Kai.Grossjohann at CS.Uni-Dortmund.DE> wrote in message
news:vafwvaqltd4.fsf at lucy.cs.uni-dortmund.de...
> On Fri, 16 Feb 2001, Syver Enstad wrote:
> You mean you want something like `C-x C-f', but with module names, not
> file names?

Yes!
>
> There is a library, ffap.el, which provides a framework for this.  But
> I couldn't see any Python-specific support in it.  However, this could
> presumably fairly easily be added.  See the variable ffap-alist for
> details.  After adding Python support, you can hit `M-x
> find-file-at-point RET' in the *.py buffer and you can open the right
> file.
You mean to open modules by parsing import statements in a file?
Ex.
module.py

import sys
import MyClass

......
And then you could place the cursor over let's say MyClass and press some
key combination to load the py file for the module. Is that what you mean?

This would be cool as an added bonus, but it's the find-file like
functionality that is crucial.




From Kai.Grossjohann at CS.Uni-Dortmund.DE  Sat Feb 17 17:53:08 2001
From: Kai.Grossjohann at CS.Uni-Dortmund.DE (Kai =?iso-8859-1?q?Gro=DFjohann?=)
Date: 17 Feb 2001 17:53:08 +0100
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no> <vafwvaqltd4.fsf@lucy.cs.uni-dortmund.de> <96kmdc$47q$1@troll.powertech.no>
Message-ID: <vafofw1i74b.fsf@lucy.cs.uni-dortmund.de>

On Sat, 17 Feb 2001, Syver Enstad wrote:

> You mean to open modules by parsing import statements in a file?
> Ex.
> module.py
> 
> import sys
> import MyClass
> 
> ......  And then you could place the cursor over let's say MyClass
> and press some key combination to load the py file for the
> module. Is that what you mean?
> 
> This would be cool as an added bonus, but it's the find-file like
> functionality that is crucial.

Oh, I'm sorry.  With M-x find-file-at-point RET, you could type it on
the `import MyClass' statement above and then you'll get a default
filename prompt of /some/path/here/MyClass.py.

But ffap.el does not provide for a facility where you can enter a
module name and then it opens the right file.

Hm.

But it would be trivial to do: C-x b foo RET, M-x python-mode RET,
type in the module name, type M-x find-file-at-point RET, voil?!

kai
-- 
Be indiscrete.  Do it continuously.


From syvere at cyberwatcher.com  Sun Feb 18 22:32:35 2001
From: syvere at cyberwatcher.com (syvere at cyberwatcher.com)
Date: 18 Feb 2001 22:32:35 +0100
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no> <vafwvaqltd4.fsf@lucy.cs.uni-dortmund.de> <96kmdc$47q$1@troll.powertech.no> <vafofw1i74b.fsf@lucy.cs.uni-dortmund.de>
Message-ID: <ulmr3smmk.fsf@TECH.i-did-not-set--mail-host-address--so-shoot-me>

Kai.Grossjohann at CS.Uni-Dortmund.DE (Kai =?iso-8859-1?q?Gro=DFjohann?=) writes:




> > Oh, I'm sorry.  With M-x find-file-at-point RET, you could type it
> on the `import MyClass' statement above and then you'll get a
> default filename prompt of /some/path/here/MyClass.py.

You mean this is implemented, or do you mean I'll have to implement this
to support ffap? 

I am not with you here, what are you trying to say.
> But ffap.el does not provide for a facility where you can enter a
> module name and then it opens the right file.
> 
> Hm.
> 
> But it would be trivial to do: C-x b foo RET, M-x pytho
> n-mode RET, type in the module name, type M-x find-file-at-point
> RET, voil?!  I am not that familiar with Emacs, so... You the
> current buffer to display the foo file, activate python-mode, type a
> python module name in the buffer, you use the ffap function
> find-file-at-point?

Are suggesting that one could in someway override the implementation of
find-file-at-point for python mode so that it also works on modulens?

By the way the python module loading seems to work fine on my machine
now. But I am trying to get pdb to work from Emacs and I don't seem to
have any luck with it. PS. I am running on Win2000 with NT Emacs. Any
help on getting pdb up and running would be very much appreciated.


From Kai.Grossjohann at CS.Uni-Dortmund.DE  Sun Feb 18 23:01:47 2001
From: Kai.Grossjohann at CS.Uni-Dortmund.DE (Kai =?iso-8859-1?q?Gro=DFjohann?=)
Date: 18 Feb 2001 23:01:47 +0100
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no> <vafwvaqltd4.fsf@lucy.cs.uni-dortmund.de> <96kmdc$47q$1@troll.powertech.no> <vafofw1i74b.fsf@lucy.cs.uni-dortmund.de> <ulmr3smmk.fsf@TECH.i-did-not-set--mail-host-address--so-shoot-me>
Message-ID: <vaf8zn3y7jo.fsf@lucy.cs.uni-dortmund.de>

On 18 Feb 2001, syvere at cyberwatcher.com wrote:

> Kai.Grossjohann at CS.Uni-Dortmund.DE (Kai
> =?iso-8859-1?q?Gro=DFjohann?=) writes:
> 
>> Oh, I'm sorry.  With M-x find-file-at-point RET, you could type it
>> on the `import MyClass' statement above and then you'll get a
>> default filename prompt of /some/path/here/MyClass.py.
> 
> You mean this is implemented, or do you mean I'll have to implement
> this to support ffap?
> 
> I am not with you here, what are you trying to say.

By adding an entry to ffap-alist, you can implement the behavior I
describe.  ffap.el does not currently support Python.

But by adding an entry into ffap-alist, you can NOT implement the
behavior that you were asking for: being able to enter a module name
in the minibuffer and have Emacs open the right *.py file.

Hm.  I think it should be possible to also get the behavior you want,
with just a little Lisp hacking.

kai
-- 
Be indiscrete.  Do it continuously.


From thomas.heller at ion-tof.com  Mon Feb 19 13:39:53 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Mon, 19 Feb 2001 13:39:53 +0100
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no> <vafwvaqltd4.fsf@lucy.cs.uni-dortmund.de> <96kmdc$47q$1@troll.powertech.no> <vafofw1i74b.fsf@lucy.cs.uni-dortmund.de> <ulmr3smmk.fsf@TECH.i-did-not-set--mail-host-address--so-shoot-me>
Message-ID: <96r46q$mngak$1@ID-59885.news.dfncis.de>

The following lets you enter a (possibly dotted)
python module name in the minibuffer after
entering M-x find-pymodule RET.
Comments welcome, I'm new to elisp.
This is for the .emacs file:

(defun find-pymodule ()
  (interactive)
  (let ((keyword (symbol-near-point)))
    (let ((text (shell-command-to-string
   (format "cmd.exe /c find-pymodule.py %s"
    (read-from-minibuffer "module? " keyword)))))
      (find-file text))))

and this is somewhere on the path, in find-pymodule.py:

import imp, sys, string

def find(parts, path=None):
    top = parts[0]
    file, pathname, stuff = imp.find_module(top, path)
    if not parts[1:]:
        return pathname
    mod = imp.load_module(top, file, pathname, stuff)
    return find(parts[1:], mod.__path__)

if __name__ == '__main__':
    try:
        sys.stdout.write(find(string.split(sys.argv[1], '.')))
    except ImportError:
        pass

I prefer to do the actual searching in python, not elisp

Thomas




From bryan at eevolved.com  Fri Feb 16 02:03:49 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Fri, 16 Feb 2001 01:03:49 GMT
Subject: Zlib header removal
Message-ID: <V%_i6.239101$f36.9591081@news20.bellglobal.com>

Hi,

I'm implementing an encrypted network protocol in which it is clear that 
both parties will be using zlib as the compression mechanism.  It would be 
beneficial to remove the zlib compression header from the the compressed 
plaintext prior to encryption, to avoid possible known plaintext attacks.  
Is there a way to do this? The zlib module doesn't explain the header well. 
Can I just chop off a specific number of bytes from the beginning? If so 
how many?

Thanks.
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"I want to know God's thoughts; the rest are details."-- Einstein



From rcameszREMOVETHIS at dds.removethistoo.nl  Sun Feb 18 02:23:34 2001
From: rcameszREMOVETHIS at dds.removethistoo.nl (Robert Amesz)
Date: Sun, 18 Feb 2001 01:23:34 GMT
Subject: Zlib header removal
References: <V%_i6.239101$f36.9591081@news20.bellglobal.com>
Message-ID: <Xns904C187BF506Frcamesz@127.0.0.1>

Bryan Mongeau wrote:

>Hi,
>
>I'm implementing an encrypted network protocol in which it is clear
>that both parties will be using zlib as the compression mechanism. 
>It would be beneficial to remove the zlib compression header from
>the the compressed plaintext prior to encryption, to avoid possible
>known plaintext attacks.  Is there a way to do this? The zlib module
>doesn't explain the header well. Can I just chop off a specific
>number of bytes from the beginning? If so how many?


AFAIK, Zlib follows the format in RFC1950 exactly, so it's not 
difficult to answer that question. Looking at the specification (see 
below) it would seem that you could chop off the first two bytes if, 
and only if,  you always compress the data using the exact same 
settings. 

Two bytes, that's not really a lot of known plaintext, but every little 
bit might be a liability, I guess.


HTH, Robert Amesz

---
>From RFC1950:
   2.2. Data format

      A zlib stream has the following structure:

           0   1
         +---+---+
         |CMF|FLG|   (more-->)
         +---+---+

      (if FLG.FDICT set)

           0   1   2   3
         +---+---+---+---+
         |     DICTID    |   (more-->)
         +---+---+---+---+

         +=====================+---+---+---+---+
         |...compressed data...|    ADLER32    |
         +=====================+---+---+---+---+

      Any data which may appear after ADLER32 are not part of the zlib
      stream.

      CMF (Compression Method and flags)
         This byte is divided into a 4-bit compression method and a 4-
         bit information field depending on the compression method.

            bits 0 to 3  CM     Compression method
            bits 4 to 7  CINFO  Compression info

      CM (Compression method)
         This identifies the compression method used in the file. CM 
         = 8 denotes the "deflate" compression method with a window 
         size up to 32K.  This is the method used by gzip and PNG 
         (see references [1] and [2] in Chapter 3, below, for the 
         reference documents).  CM = 15 is reserved.  It might be 
         used in a future version of this specification to indicate
         the presence of an extra field before the compressed data. 

      CINFO (Compression info)
         For CM = 8, CINFO is the base-2 logarithm of the LZ77 window 
         size, minus eight (CINFO=7 indicates a 32K window size).
         Values of CINFO above 7 are not allowed in this version of
         the specification.  CINFO is not defined in this
         specification for CM not equal to 8. 

      FLG (FLaGs)
         This flag byte is divided as follows:

            bits 0 to 4  FCHECK  (check bits for CMF and FLG)
            bit  5       FDICT   (preset dictionary)
            bits 6 to 7  FLEVEL  (compression level)

         The FCHECK value must be such that CMF and FLG, when viewed as
         a 16-bit unsigned integer stored in MSB order (CMF*256 + FLG),
         is a multiple of 31.

      FDICT (Preset dictionary)
         If FDICT is set, a DICT dictionary identifier is present 
         immediately after the FLG byte. The dictionary is a sequence
         of bytes which are initially fed to the compressor without 
         producing any compressed output. DICT is the Adler-32
         checksum of this sequence of bytes (see the definition of
         ADLER32 below).  The decompressor can use this identifier to
         determine which dictionary has been used by the compressor.

      FLEVEL (Compression level)
         These flags are available for use by specific compression
         methods.  The "deflate" method (CM = 8) sets these flags as
         follows:

            0 - compressor used fastest algorithm
            1 - compressor used fast algorithm
            2 - compressor used default algorithm
            3 - compressor used maximum compression, slowest algorithm

         The information in FLEVEL is not needed for decompression; it
         is there to indicate if recompression might be worthwhile.

      compressed data
         For compression method 8, the compressed data is stored in the
         deflate compressed data format as described in the document
         "DEFLATE Compressed Data Format Specification" by L. Peter
         Deutsch. (See reference [3] in Chapter 3, below)

         Other compressed data formats are not specified in this
         version of the zlib specification. 

      ADLER32 (Adler-32 checksum)
         This contains a checksum value of the uncompressed data 
         (excluding any dictionary data) computed according to
         Adler-32 algorithm. This algorithm is a 32-bit extension and
         improvement of the Fletcher algorithm, used in the ITU-T
         X.224 / ISO 8073 standard. See references [4] and [5] in
         Chapter 3, below) 

         Adler-32 is composed of two sums accumulated per byte: s1 is 
         the sum of all bytes, s2 is the sum of all s1 values. Both
         sums are done modulo 65521. s1 is initialized to 1, s2 to
         zero.  The Adler-32 checksum is stored as s2*65536 + s1 in
         most- significant-byte first (network) order.


From erictodd at texas.net  Fri Feb 16 06:51:10 2001
From: erictodd at texas.net (Eric Myers)
Date: Fri, 16 Feb 2001 05:51:10 GMT
Subject: Need help w/ wxStaticBitmap (novice)
Message-ID: <id3j6.6955$5O3.698176@news5.aus1.giganews.com>

Hello.  Using wxDesigner, I've created a small dialog (dialer_wdr.py) whose 
job is to change as it catches certain events.  The contents of the dialog 
are as follows:
-A button, whose Id is ID_BUTTON
-A static text box whose Id is ID_TEXT
-A static bitmap whose Id is ID_STATICBITMAP, and whose initial bitmap is 
MyBitmapsFunc(0)

The event-catching and dialog-changing is handled by my implementation 
file,  which is named dialer.py.  When the event referred to is detected, 
dialer.py executes the following:

self.FindWindowById(ID_TEXT).SetLabel("new text")
--and the static text successfully gets changed to "new text"

self.FindWindowById(ID_BUTTON).SetLabel("new caption")
--and the button's caption successfully gets changed

self.FindWindowById(ID_STATICBITMAP).SetBitmap(MyBitmapsFunc(1))
--and I get an error that says "AttributeError: 'wxWindowPtr' instance has 
no attribute 'SetBitmap'"

I don't understand why this technique works in the first two situations but 
gives me problems with the bitmap thing.  If for some reason I can't 
accomplish what I'm trying to do in this way, can somebody please point me 
in the right direction?

                            Thanks



From mfletch at tpresence.com  Fri Feb 16 11:49:35 2001
From: mfletch at tpresence.com (Mike Fletcher)
Date: Fri, 16 Feb 2001 05:49:35 -0500
Subject: Need help w/ wxStaticBitmap (novice)
Message-ID: <CC59868A0A76D311B1F50090277B321C1E14B3@VRTMAIL>

The problem is likely that find window is defined as:

	wxWindow* FindWindow(long id)

So you're getting back a wxWindow pointer/object instead of a
wxStaticBitmap.  You need to cast it to a wxStaticBitmap (yes, it's ugly to
have to cast things in Python):

	wxPyTypeCast(
		self.FindWindowById(
			ID_STATICBITMAP
		),
		'wxStaticBitmap',
	).SetBitmap(MyBitmapsFunc(1))

Haven't tested that, but it sounds right.  No idea, incidentally, why you're
able to get the other two to work, seems to me they'd require a cast too.

HTH,
Mike

-----Original Message-----
From: Eric Myers [mailto:erictodd at texas.net]
Sent: Friday, February 16, 2001 12:51 AM
To: python-list at python.org
Subject: Need help w/ wxStaticBitmap (novice)


Hello.  Using wxDesigner, I've created a small dialog (dialer_wdr.py) whose 
job is to change as it catches certain events.  The contents of the dialog 
are as follows:
-A button, whose Id is ID_BUTTON
-A static text box whose Id is ID_TEXT
-A static bitmap whose Id is ID_STATICBITMAP, and whose initial bitmap is 
MyBitmapsFunc(0)

The event-catching and dialog-changing is handled by my implementation 
file,  which is named dialer.py.  When the event referred to is detected, 
dialer.py executes the following:

self.FindWindowById(ID_TEXT).SetLabel("new text")
--and the static text successfully gets changed to "new text"

self.FindWindowById(ID_BUTTON).SetLabel("new caption")
--and the button's caption successfully gets changed

self.FindWindowById(ID_STATICBITMAP).SetBitmap(MyBitmapsFunc(1))
--and I get an error that says "AttributeError: 'wxWindowPtr' instance has 
no attribute 'SetBitmap'"

I don't understand why this technique works in the first two situations but 
gives me problems with the bitmap thing.  If for some reason I can't 
accomplish what I'm trying to do in this way, can somebody please point me 
in the right direction?

                            Thanks

-- 
http://mail.python.org/mailman/listinfo/python-list



From erictodd at texas.net  Fri Feb 16 15:47:00 2001
From: erictodd at texas.net (erictodd at texas.net)
Date: Fri, 16 Feb 2001 14:47:00 GMT
Subject: Need help w/ wxStaticBitmap (novice)
References: <mailman.982320618.1389.python-list@python.org>
Message-ID: <E3bj6.256124$Df2.15641981@news3.aus1.giganews.com>

In article <mailman.982320618.1389.python-list at python.org>, Mike  
Fletcher <mfletch at tpresence.com> wrote:
> The problem is likely that find window is defined as:
> 
> 	wxWindow* FindWindow(long id)
> 
> So you're getting back a wxWindow pointer/object instead of a
> wxStaticBitmap.  You need to cast it to a wxStaticBitmap (yes, it's ugly to
> have to cast things in Python):
> 
> 	wxPyTypeCast(
> 		self.FindWindowById(
> 			ID_STATICBITMAP
> 		),
> 		'wxStaticBitmap',
> 	).SetBitmap(MyBitmapsFunc(1))
> 
> Haven't tested that, but it sounds right.  No idea, incidentally, why you're
> able to get the other two to work, seems to me they'd require a cast too.
> 
> HTH,
> Mike
> 

Yes, that worked straight out of the box.  I thank you very much--that was 
an incredibly prompt reply!

-Eric


From dan at eevolved.com  Fri Feb 16 08:54:01 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 07:54:01 GMT
Subject: shelve and the best db possible
Message-ID: <t05j6.259674$JT5.9093034@news20.bellglobal.com>

I've been using shelve in dbm mode for a while now and Im' having problems. 
Either there are random segfaults or writes take an exponentially longer 
time to do the bigger the db gets (100 rows -> 1 second, 1000 rows -> 20 
seconds, 10000 rows -> 600 seconds then crash)

I really love shelve, but dbm doesn't seem to scale very well :( Could 
anyone point me to a good db with a shelve interface? I've tried to get 
bsddb3 to go, but the test scripts all die :(

Thanks,
Dan


From janne at oops.nnets.fi  Fri Feb 16 12:59:26 2001
From: janne at oops.nnets.fi (Janne Sinkkonen)
Date: 16 Feb 2001 13:59:26 +0200
Subject: shelve and the best db possible
References: <t05j6.259674$JT5.9093034@news20.bellglobal.com>
Message-ID: <m3ae7m2601.fsf@kinos.nnets.fi>

Dan Parisien <dan at eevolved.com> writes:

> I really love shelve, but dbm doesn't seem to scale very well :( Could 
> anyone point me to a good db with a shelve interface? I've tried to get 
> bsddb3 to go, but the test scripts all die :(

I have used an old bsddb 2.x interface from TCS with a largish data
base (btree, hundreds of MB's, hundreds of thousands of updates)
without any problems. Don't know whether it's still available somewhere.

The module name is bsddb (or TCS.bsddb)

-- 
Janne


From robin at alldunn.com  Sat Feb 17 01:42:55 2001
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 16 Feb 2001 16:42:55 -0800
Subject: shelve and the best db possible
References: <t05j6.259674$JT5.9093034@news20.bellglobal.com> <m3ae7m2601.fsf@kinos.nnets.fi>
Message-ID: <4Pjj6.44$uHc.19005522@news.randori.com>

>
> I have used an old bsddb 2.x interface from TCS with a largish data
> base (btree, hundreds of MB's, hundreds of thousands of updates)
> without any problems. Don't know whether it's still available somewhere.
>

http://alldunn.com/python/index.html#DB


--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!





From tim.one at home.com  Fri Feb 16 09:05:10 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 03:05:10 -0500
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
In-Reply-To: <Pine.LNX.4.10.10102151650400.1491-100000@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEMBIPAA.tim.one@home.com>

[Michael Hudson]
> ...
>    * Imports on case-insensitive file systems *
>
>  There was quite some discussion about how to handle imports on a
>  case-insensitive file system (eg. on Windows).  I didn't follow the
>  details, but Tim Peters is on the case (sorry), so I'm confident it
>  will get sorted out.

You can be sure the whitespace will be consistent, anyway <wink>.

OK, this one sucks.  It should really have gotten a PEP, but it cropped up
too late in the release cycle and it can't be delayed (see below).

Here's the scoop:  file systems vary across platforms in whether or not they
preserve the case of filenames, and in whether or not the platform C library
file-opening functions do or don't insist on case-sensitive matches:


                        case-preserving   case-destroying
                     +-------------------+------------------+
    case-sensitive   | most Unix flavors | brrrrrrrrrr      |
                     +-------------------+------------------+
    case-insensitive | Windows           | some unfortunate |
                     | MacOSX HFS+       | network schemes  |
                     | Cygwin            |                  |
                     +-------------------+------------------+

In the upper left box, if you create "fiLe" it's stored as "fiLe", and only
open("fiLe") will open it (open("file") will not, nor will the 14 other
variations on that theme).

In the lower right box, if you create "fiLe", there's no telling what it's
stored as-- but most likely as "FILE" --and any of the 16 obvious variations
on open("FilE") will open it.

The lower left box is a mix:  creating "fiLe" stores "fiLe" in the platform
directory, but you don't have to match case when opening it; any of the 16
obvious variations on open("FILe") work.

NONE OF THAT IS CHANGING!  Python will continue to follow platform
conventions wrt whether case is preserved when creating a file, and wrt
whether open() requires a case-sensitive match.  In practice, you should
always code as if matches were case-sensitive, else your program won't be
portable.  But then you should also always open binary files with the "b"
flag, and you don't do that either <wink>.

What's proposed is to change the semantics of Python "import" statements,
and there *only* in the lower left box.

Support for MaxOSX HFS+, and for Cygwin, is new in 2.1, so nothing is
changing there.  What's changing is Windows behavior.  Here are the current
rules for import on Windows:

1. Despite that the filesystem is case-insensitive, Python insists on
   a case-sensitive match.  But not in the way the upper left box works:
   if you have two files, FiLe.py and file.py on sys.path, and do

       import file

   then if Python finds FiLe.py first, it raises a NameError.  It does
   *not* go on to find file.py; indeed, it's impossible to import any
   but the first case-insensitive match on sys.path, and then only if
   case matches exactly in the first case-insensitive match.

2. An ugly exception:  if the first case-insensitive match on sys.path
   is for a file whose name is entirely in upper case (FILE.PY or
   FILE.PYC or FILE.PYO), then the import silently grabs that, no matter
   what mixture of case was used in the import statement.  This is
   apparently to cater to miserable old filesystems that really fit in
   the lower right box.  But this exception is unique to Windows, for
   reasons that may or may not exist <frown>.

3. And another exception:  if the envar PYTHONCASEOK exists, Python
   silently grabs the first case-insensitive match of any kind.

So these Windows rules are pretty complicated, and neither match the Unix
rules nor provide semantics natural for the native filesystem.  That makes
them hard to explain to Unix *or* Windows users.  Nevertheless, they've
worked fine for years, and in isolation there's no compelling reason to
change them.

However, that was before the MacOSX HFS+ and Cygwin ports arrived.  They
also have case-preserving case-insensitive filesystems, but the people doing
the ports despised the Windows rules.  Indeed, a patch to make HFS+ act like
Unix for imports got past a reviewer and into the code base, which
incidentally made Cygwin also act like Unix (but this met the unbounded
approval of the Cygwin folks, so they sure didn't complain -- they had
patches of their own pending to do this, but the reviewer for those balked).

At a higher level, we want to keep Python consistent, and I in particular
want Python to do the same thing on *all* platforms with case-preserving
case-insensitive filesystems.  Guido too, but he's so sick of this argument
don't ask him to confirm that <0.9 wink>.

The proposed new semantics for the lower left box:

A. If the PYTHONCASEOK envar exists, same as before:  silently accept
   the first case-insensitive match of any kind; raise ImportError if
   none found.

B. Else search sys.path for the first case-sensitive match; raise
   ImportError if none found.

#B is the same rule as is used on Unix, so this will improve cross-platform
portability.  That's good.  #B is also the rule the Mac and Cygwin folks
want (and wanted enough to implement themselves, multiple times, which is a
powerful argument in PythonLand).  It can't cause any existing
non-exceptional Windows import to fail, because any existing non-exceptional
Windows import finds a case-sensitive match first in the path -- and it
still will.  An exceptional Windows import currently blows up with a
NameError or ImportError, in which latter case it still will, or in which
former case will continue searching, and either succeed or blow up with an
ImportError.

#A is needed to cater to case-destroying filesystems mounted on Windows, and
*may* also be used by people so enamored of "natural" Windows behavior that
they're willing to set an envar to get it.  That's their problem <wink>.  I
don't intend to implement #A for Unix too, but that's just because I'm not
clear on how I *could* do so efficiently (I'm not going to slow imports
under Unix just for theoretical purity).

The potential damage is here:  #2 (matching on ALLCAPS.PY) is proposed to be
dropped.  Case-destroying filesystems are a vanishing breed, and support for
them is ugly.  We're already supporting (and will continue to support)
PYTHONCASEOK for their benefit, but they don't deserve multiple hacks in
2001.

Flame at will.

or-flame-at-tim-your-choice-ly y'rs  - tim




From claird at starbase.neosoft.com  Fri Feb 16 14:54:12 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 16 Feb 2001 07:54:12 -0600
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
References: <mailman.982310784.27748.python-list@python.org>
Message-ID: <A907283AAE8B306C.9C2AADB86E4F5A84.1395E06F2108C534@lp.airnews.net>

In article <mailman.982310784.27748.python-list at python.org>,
Tim Peters <tim.one at home.com> wrote:
>[Michael Hudson]
>> ...
>>    * Imports on case-insensitive file systems *
>>
>>  There was quite some discussion about how to handle imports on a
>>  case-insensitive file system (eg. on Windows).  I didn't follow the
>>  details, but Tim Peters is on the case (sorry), so I'm confident it
>>  will get sorted out.
>
>You can be sure the whitespace will be consistent, anyway <wink>.
			.
			.
			.
>them is ugly.  We're already supporting (and will continue to support)
>PYTHONCASEOK for their benefit, but they don't deserve multiple hacks in
>2001.
>
>Flame at will.
>
>or-flame-at-tim-your-choice-ly y'rs  - tim
>
>

1.  Thanks.  Along with all the other benefits, I find
    this explanation FAR more entertaining than anything
    network television broadcasts (although nearly as
    tendentious as "The West Wing").
2.  I hope a few OS/400 and OpenVMS refugees convert and
    walk through the door soon.  *That* would make for a
    nice dose of fun.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From aahz at panix.com  Fri Feb 16 17:33:04 2001
From: aahz at panix.com (Aahz Maruch)
Date: 16 Feb 2001 08:33:04 -0800
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
References: <mailman.982310784.27748.python-list@python.org>
Message-ID: <96jko0$fje$1@panix2.panix.com>

In article <mailman.982310784.27748.python-list at python.org>,
Tim Peters <tim.one at home.com> wrote:
>
>Flame at will.
>
>or-flame-at-tim-your-choice-ly y'rs  - tim

Very clear, very cogent, very consistent.  Thanks for explaining all
that.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From tjreedy at udel.edu  Sat Feb 17 00:06:23 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 16 Feb 2001 18:06:23 -0500
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
References: <mailman.982310784.27748.python-list@python.org>
Message-ID: <96kbdd$j7r$1@news.udel.edu>

One of the hassles with Win9X is that Windows Explorer displays, I believe,
some files differently from how they are stored.  So I have sometimes had
trouble opening files.  (Did the ALLCAPS exception apply here too?)  As
long as os.listdir('.') displays filenames in a form acceptible to import
and open(), I should be ok.





From tim.one at home.com  Sat Feb 17 05:32:51 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 23:32:51 -0500
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
In-Reply-To: <96kbdd$j7r$1@news.udel.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEPJIPAA.tim.one@home.com>

[Terry Reedy]
> One of the hassles with Win9X is that Windows Explorer displays,
> I believe, some files differently from how they are stored.

This depends on the setting of Explorer's

    View -> Folder Options ... -> View -> Allow all uppercase names

option.  It that's on, file names are shown as stored.  But it's off by
default.  Then an ALLCAPS name is shown as Allcaps (i.e., first letter
uppercase, all the rest lowercase), while other names are shown as stored.

This has no effect on any Python file in the distribution, though, because
they all end with (lowercase) ".py", so they never meet the criteria "Allow
all uppercase names" is looking for.

> So I have sometimes had trouble opening files.

I don't understand that, as open() is case-insensitive on Windows.  Perhaps
you're thinking of Explorer's "Hide file extensions for known file types"
option?  *That* could give you trouble when opening files -- but isn't
relevant to Python imports.

> (Did the ALLCAPS exception apply here too?)

Always to import, never to anything else.  For each of the four file names

    ab
    Ab
    aB
    AB

on Windows, every one of the following works fine to open it:

    open("ab")
    open("Ab")
    open("aB")
    open("AB")

Python doesn't do anything special to make that happen -- that's just the
way the Windows system libraries work.

> As long as os.listdir('.') displays filenames in a form acceptible
> to import and open(), I should be ok.

You're forcing me to be careful here <0.9 wink>:  the names returned by
listdir('.') work fine for open() as-is on Windows, but also would if you
ran them through string.upper() first, or through string.lower() first, or
if you picked a case at random for each character in the name.

But the names will rarely work in an import statement (and don't today
either), because e.g.

    import random.py

will only work if you happen to have a *package* named random containing a
*module* named py (so is in a file named py.py).  In normal use, you have to
strip the extension off the filename before importing it.

Beyond that, yes, listdir() returns names using the case choices with which
they were created.  And Explorer options don't affect what listdir() does.




From qrczak at knm.org.pl  Mon Feb 19 21:59:33 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 19 Feb 2001 20:59:33 GMT
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
References: <mailman.982310784.27748.python-list@python.org> <96kbdd$j7r$1@news.udel.edu>
Message-ID: <slrn99329l.2fk.qrczak@qrnik.zagroda>

Fri, 16 Feb 2001 18:06:23 -0500, Terry Reedy <tjreedy at udel.edu> pisze:

> One of the hassles with Win9X is that Windows Explorer displays, I believe,
> some files differently from how they are stored.

And stores files differently than was asked to. For example when
you try to create a directory called Spam, it will create SPAM and
display it as Spam. And when asked to create SPAM, it will create it
but display Spam. It behaves correctly when the name has more than
8 characters or when a character outside ASCII is used.

This is Win95 - I don't know if they fixed this in later versions.

Worse: using the Explorer you can create a filename with some non-ASCII
characters (e.g. <<French quotes>> or bullet), which will be silently
converted to other characters, perhaps different in long and short
names (in these cases: '<<' and '>>' character pairs, and '\7' -
a control character which displays as bullet in DOS). Details depend
on the default codepages (the above is for CP-1250 and CP-852). This
can make the file inaccessible under either Windows or DOS or both
(because the converted character is sometimes not legal in a filename,
but the filename was checked for correctness before conversion). You
can't delete or rename such file from either Windows or DOS or both,
until you repair the filesystem. Scandisk is not always able to do it.

not-using-Windows-anymore-ly y'rs
-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From jeremy at alum.mit.edu  Mon Feb 19 21:18:03 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Mon, 19 Feb 2001 15:18:03 -0500 (EST)
Subject: [Python-Dev] Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
In-Reply-To: <LNBBLJKPBEHFEDALKOLCEEMBIPAA.tim.one@home.com>
References: <Pine.LNX.4.10.10102151650400.1491-100000@localhost.localdomain>
	<LNBBLJKPBEHFEDALKOLCEEMBIPAA.tim.one@home.com>
Message-ID: <14993.32635.85544.343209@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "TP" == Tim Peters <tim.one at home.com> writes:

  TP> [Michael Hudson]
  >> ...
  >> * Imports on case-insensitive file systems *
  >>
  >> There was quite some discussion about how to handle imports on a
  >> case-insensitive file system (eg. on Windows).  I didn't follow
  >> the details, but Tim Peters is on the case (sorry), so I'm
  >> confident it will get sorted out.

  TP> You can be sure the whitespace will be consistent, anyway
  TP> <wink>.

  TP> OK, this one sucks.  It should really have gotten a PEP, but it
  TP> cropped up too late in the release cycle and it can't be delayed
  TP> (see below).

It would be good to capture this in an informational PEP that just
describes what the policy is and why.  If nothing else, it could be a
copy of Tim's message immortalized with a PEP number.

Jeremy



From tjreedy at udel.edu  Tue Feb 20 06:39:08 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 20 Feb 2001 00:39:08 -0500
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
References: <mailman.982384399.26499.python-list@python.org>
Message-ID: <96sve1$69e$1@news.udel.edu>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.982384399.26499.python-list at python.org...
> [Terry Reedy]
> > One of the hassles with Win9X is that Windows Explorer displays,
> > I believe, some files differently from how they are stored.
>
> This depends on the setting of Explorer's
>
>     View -> Folder Options ... -> View -> Allow all uppercase names
>
> option.  It that's on, file names are shown as stored.  But it's off by
> default.  Then an ALLCAPS name is shown as Allcaps (i.e., first letter
> uppercase, all the rest lowercase), while other names are shown as
stored.

By coincidence, I discovered this 'feature' a couple of days ago, just
after posting this, in the process of turning on the display of hidden
files.

> > So I have sometimes had trouble opening files.
>
> I don't understand that, as open() is case-insensitive on Windows.

I do not believe that this has always been my experience, but the problem I
had was a couple of years ago, so I will wait until I reproduce the problem
before saying more.




From cyberian_bear at hotmail.com  Fri Feb 16 09:37:43 2001
From: cyberian_bear at hotmail.com (cyberian bear)
Date: Fri, 16 Feb 2001 03:37:43 -0500
Subject: equivalent of NULL or OM in Python
Message-ID: <2G5j6.1$CF4.251@typhoon.nyu.edu>

1.I'm doing a program in Python which perform +, -, *, / on two polynomials.
Their coefficient and exponentials are stored in tuples. So for example
3X^2+4X+4 would be stored like [(2,3),(1,4),(0,4)] but what if the term is
absent completely from one of the polynomials then i would have to check if
one of the tuuples is not present at all and not just zero. For  example
[(2,3),(0,4)]. In Pascal there is NULL in SETL there is OM but is there a
corresponding statement in Python. I've checked through several sources
including my textbook but didn't find the satisfactory answer. Maybe I
should just check if the second term of every tuple(i.e the coefficient) is
zero which means that the term is not present.
2. Also can anyone give me a clue how to iterate over all the coefficient
terms in the tuples in the first polynomials' tuples.
My guess is should be something like
for coefficient in polynomial1:
where polynomial is a list of tuples
but how do i tell it to iterate specifically over the second term in every
tuple and not over the whole tuples
cb




From hgg9140 at seanet.com  Fri Feb 16 14:53:19 2001
From: hgg9140 at seanet.com (Harry George)
Date: 16 Feb 2001 05:53:19 -0800
Subject: equivalent of NULL or OM in Python
References: <2G5j6.1$CF4.251@typhoon.nyu.edu>
Message-ID: <m3pugibupc.fsf@wilma.localdomain>

I think you are talking about "None".

For iterating over polynomials, I could put "m3na" (numerical
analysis, in Modula-3) on a web site for you.  It does polynomial
math.  For iterating over tuples in general, they are sequences so you
can do

   for term in poly:
      do_something_to_term


"cyberian bear" <cyberian_bear at hotmail.com> writes:

> 1.I'm doing a program in Python which perform +, -, *, / on two polynomials.
> Their coefficient and exponentials are stored in tuples. So for example
> 3X^2+4X+4 would be stored like [(2,3),(1,4),(0,4)] but what if the term is
> absent completely from one of the polynomials then i would have to check if
> one of the tuuples is not present at all and not just zero. For  example
> [(2,3),(0,4)]. In Pascal there is NULL in SETL there is OM but is there a
> corresponding statement in Python. I've checked through several sources
> including my textbook but didn't find the satisfactory answer. Maybe I
> should just check if the second term of every tuple(i.e the coefficient) is
> zero which means that the term is not present.
> 2. Also can anyone give me a clue how to iterate over all the coefficient
> terms in the tuples in the first polynomials' tuples.
> My guess is should be something like
> for coefficient in polynomial1:
> where polynomial is a list of tuples
> but how do i tell it to iterate specifically over the second term in every
> tuple and not over the whole tuples
> cb
> 
> 

-- 
Harry George
hgg9140 at seanet.com


From danielk at aracnet.com  Fri Feb 16 15:15:29 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Fri, 16 Feb 2001 06:15:29 -0800
Subject: equivalent of NULL or OM in Python
References: <2G5j6.1$CF4.251@typhoon.nyu.edu>
Message-ID: <dgcq8t4iahltqlcot0brik37iuf1h8hucq@4ax.com>

In this case, accessing an element outside of the range of values would produce
an IndexError exception. You can 'catch' this exception like this

poly = [(2,3),(0,4)]
try:
    thirdpart = poly[2]
except IndexError:
    print 'There are only 2 parts.'

In Python, when an object has not been initialized, it is set to 'None' (quotes
are for clarity), but I don't think this is useful in this situation.

For question 2:

for element in poly:
    print element[1] # prints the second part of each tuple 

Hope this helps,

Daniel Klein
Portland OR USA



On Fri, 16 Feb 2001 03:37:43 -0500, "cyberian bear" <cyberian_bear at hotmail.com>
wrote:

>1.I'm doing a program in Python which perform +, -, *, / on two polynomials.
>Their coefficient and exponentials are stored in tuples. So for example
>3X^2+4X+4 would be stored like [(2,3),(1,4),(0,4)] but what if the term is
>absent completely from one of the polynomials then i would have to check if
>one of the tuuples is not present at all and not just zero. For  example
>[(2,3),(0,4)]. In Pascal there is NULL in SETL there is OM but is there a
>corresponding statement in Python. I've checked through several sources
>including my textbook but didn't find the satisfactory answer. Maybe I
>should just check if the second term of every tuple(i.e the coefficient) is
>zero which means that the term is not present.
>2. Also can anyone give me a clue how to iterate over all the coefficient
>terms in the tuples in the first polynomials' tuples.
>My guess is should be something like
>for coefficient in polynomial1:
>where polynomial is a list of tuples
>but how do i tell it to iterate specifically over the second term in every
>tuple and not over the whole tuples
>cb
>



From tim.one at home.com  Fri Feb 16 22:51:13 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 16:51:13 -0500
Subject: equivalent of NULL or OM in Python
In-Reply-To: <dgcq8t4iahltqlcot0brik37iuf1h8hucq@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEOMIPAA.tim.one@home.com>

[Daniel Klein]
> ...
> In Python, when an object has not been initialized, it is set to
> 'None' (quotes are for clarity),

No.  An unbound variable is not bound to anything; that's why it's called
unbound <0.9 wink>.  Attempting to reference an uninitialized variable in
Python raises an exception (a subclass of NameError).




From sholden at holdenweb.com  Fri Feb 16 15:40:47 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 16 Feb 2001 09:40:47 -0500
Subject: equivalent of NULL or OM in Python
References: <2G5j6.1$CF4.251@typhoon.nyu.edu>
Message-ID: <L1bj6.16746$GA5.260574@e420r-atl1.usenetserver.com>

--
Tools, training and technology to help you meet your information needs



"cyberian bear" <cyberian_bear at hotmail.com> wrote in message
news:2G5j6.1$CF4.251 at typhoon.nyu.edu...
> 1.I'm doing a program in Python which perform +, -, *, / on two
polynomials.
> Their coefficient and exponentials are stored in tuples. So for example
> 3X^2+4X+4 would be stored like [(2,3),(1,4),(0,4)] but what if the term is
> absent completely from one of the polynomials then i would have to check
if
> one of the tuuples is not present at all and not just zero. For  example
> [(2,3),(0,4)]. In Pascal there is NULL in SETL there is OM but is there a
> corresponding statement in Python. I've checked through several sources
> including my textbook but didn't find the satisfactory answer. Maybe I
> should just check if the second term of every tuple(i.e the coefficient)
is
> zero which means that the term is not present.

Choice of data representation can be critical in algorithm design, and this
is quite a good example. However, I have not fully considered the symbolic
manipulations you want to perform, so I may be putting my foot in my mouth
here. That's how I know I have a size 10 mouth: the foot fits perfectly.

It would appear to be more regular to represent absent terms as a
coefficient of zero. Furthermore, if you reverse the order of the terms,
putting x^0 at the left, you can then use the index into the list as the
power. So your examples could be represented as follows:

3X^2+4X+4 : [4, 4, 3]
3x^2+4: [4, 0, 3]

Evaluation for a particular value of x then comes down to:

>>> poly = [4, 4, 3]
>>> r = 0
>>> x = 2
>>> for p in range(len(poly)):
...  r = r + (x^p)*poly[p]
...
>>> r
20

Polynomial addition would just add the equivalent coefficients (although you
would need to extend the shorter woth zeroes to make them equal lengths).
The rest is left as an exercise :-)

Would this representation be more tractable?

> 2. Also can anyone give me a clue how to iterate over all the coefficient
> terms in the tuples in the first polynomials' tuples.
> My guess is should be something like
> for coefficient in polynomial1:
> where polynomial is a list of tuples
> but how do i tell it to iterate specifically over the second term in every
> tuple and not over the whole tuples
> cb
>
>
Using your representation, what you want is something like:

>>> poly = [(2,3),(1,4),(0,4)]
>>> for i, c in poly:
...  print "Index:", i, "Coefficient:", c
...
Index: 2 Coefficient: 3
Index: 1 Coefficient: 4
Index: 0 Coefficient: 4

Obviously you can ignore the indexes if you want...

Does any of this help?

regards
 Steve




From sholden at holdenweb.com  Fri Feb 16 15:48:52 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 16 Feb 2001 09:48:52 -0500
Subject: equivalent of NULL or OM in Python
References: <2G5j6.1$CF4.251@typhoon.nyu.edu> <L1bj6.16746$GA5.260574@e420r-atl1.usenetserver.com>
Message-ID: <k9bj6.16770$GA5.260608@e420r-atl1.usenetserver.com>

"Steve Holden" <sholden at holdenweb.com> got too enthusiastic, and wrote
without checking:

> Evaluation for a particular value of x then comes down to:
>
> >>> poly = [4, 4, 3]
> >>> r = 0
> >>> x = 2
> >>> for p in range(len(poly)):
> ...  r = r + (x^p)*poly[p]
> ...
> >>> r
> 20

Woops, make that:

>>> for p in range(len(poly)):
...  r = r + (x**p)*poly[p]
...  print "p:", p, "r:", r
...
p: 0 r: 4
p: 1 r: 12
p: 2 r: 24

Got carried away and used the wrong power operator. Bitwise xor might have
been a little confusing here...

regards
 Steve



From digitig at cix.co.uk  Sun Feb 25 22:17:00 2001
From: digitig at cix.co.uk (Tim Rowe)
Date: Sun, 25 Feb 2001 21:17 +0000 (GMT Standard Time)
Subject: equivalent of NULL or OM in Python
References: <L1bj6.16746$GA5.260574@e420r-atl1.usenetserver.com>
Message-ID: <memo.20010225211730.34111A@digitig.compulink.co.uk>

In article <L1bj6.16746$GA5.260574 at e420r-atl1.usenetserver.com>, 
sholden at holdenweb.com (Steve Holden) wrote:

<snip>

> Choice of data representation can be critical in algorithm design, and 
> this
> is quite a good example. However, I have not fully considered the 
> symbolic
> manipulations you want to perform, so I may be putting my foot in my 
> mouth
> here. That's how I know I have a size 10 mouth: the foot fits perfectly.
> 
> It would appear to be more regular to represent absent terms as a
> coefficient of zero. Furthermore, if you reverse the order of the terms,
> putting x^0 at the left, you can then use the index into the list as the
> power. So your examples could be represented as follows:
> 
> 3X^2+4X+4 : [4, 4, 3]
> 3x^2+4: [4, 0, 3]

It depends what he wants to do. If he's doing things like CRC calculations 
the lists are likely to have a lot of zeroes. 

I think it would be worth considering numpy, as polynomial conversions are 
easily represented as array operations, and ISTR numpy is clued up on 
sparse arrays.


From DanielK at jBASE.com  Fri Feb 16 23:23:50 2001
From: DanielK at jBASE.com (Daniel Klein)
Date: Fri, 16 Feb 2001 14:23:50 -0800
Subject: equivalent of NULL or OM in Python
References: <mailman.982360339.1235.python-list@python.org>
Message-ID: <7Shj6.466$Vf.17201@typhoon.aracnet.com>

Thanks for setting me straight, Tim.

Dan

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.982360339.1235.python-list at python.org...
> [Daniel Klein]
> > ...
> > In Python, when an object has not been initialized, it is set to
> > 'None' (quotes are for clarity),
>
> No.  An unbound variable is not bound to anything; that's why it's called
> unbound <0.9 wink>.  Attempting to reference an uninitialized variable in
> Python raises an exception (a subclass of NameError).
>
>




From tim.one at home.com  Fri Feb 16 23:57:29 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 17:57:29 -0500
Subject: equivalent of NULL or OM in Python
In-Reply-To: <7Shj6.466$Vf.17201@typhoon.aracnet.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEPBIPAA.tim.one@home.com>

[Daniel Klein]
> Thanks for setting me straight, Tim.

And thank *you* for giving me the opportunity to excel <wink>.

honest-if-you-were-confused-10,000-were-too-ly y'rs  - tim




From greg at cosc.canterbury.ac.nz  Wed Feb 21 00:51:43 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Wed, 21 Feb 2001 12:51:43 +1300
Subject: equivalent of NULL or OM in Python
References: <2G5j6.1$CF4.251@typhoon.nyu.edu>
Message-ID: <3A93030F.6CD98D29@cosc.canterbury.ac.nz>

cyberian bear wrote:
> 
> Maybe I
> should just check if the second term of every tuple(i.e the coefficient) is
> zero which means that the term is not present.

It depends on how your algorithms work. If you find that,
e.g. it makes your multiplication algorithm awkward if
terms can be missed out, then always storing every term,
possibly with a zero coefficient, makes sense.

If you're doing that, you don't really need to store
the exponents, because they're implied by the position
in the list -- so all you need is a list of coefficients.
I would probably store it reversed, so that the coefficient
of x^i is in position i of the list, e.g.

  3x^2+2x+4 --> [4, 2, 3]

> how do i tell it to iterate specifically over the second term in every
> tuple and not over the whole tuples

Note that you can "unpack" a tuple by doing:

  exp, coeff = term

so one way to do what you want is

  for term in poly:
    exp, coeff = term
      #do something with coeff

or, more compactly,

  for exp, coeff in poly:
    #do something with coeff

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From wilsoncredit at email.com  Fri Feb 16 09:56:53 2001
From: wilsoncredit at email.com (wilsoncredit at email.com)
Date: Fri, 16 Feb 2001 08:56:53 GMT
Subject: ****SPECIAL REPORT****
Message-ID: <pX5j6.39556$Jp2.270876@news6.giganews.com>

--_NextPart_0000904A-000004CC-0388DEEA-5224
Content-Type: text/plain
Content-Transfer-Encoding: 7bit

Get a brand new, clean,
credit file in 15 days or less...
Get the credit you deserve...
For more info, send a email to
wilsoncredit at email.com
with "INFO" in the subject line...
--_NextPart_0000904A-000004CC-0388DEEA-5224--


From ransen_spam_me_not at nemo.it  Fri Feb 16 12:06:08 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Fri, 16 Feb 2001 11:06:08 GMT
Subject: sys.path and the Windows registry
Message-ID: <3a8d0450.162541@news.newsguy.com>

I'm just starting out with Python and have some simple
modules in a directoryb which is not on the sys.path.

I hoped that if I edited the registry and added my new
extra source file path I'd then see the new files in the
path browser (GUI Python 2).

I can't though, what am I missing?

TIA

Owen

--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From MarkH at ActiveState.com  Sat Feb 17 00:53:46 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 16 Feb 2001 23:53:46 GMT
Subject: sys.path and the Windows registry
References: <3a8d0450.162541@news.newsguy.com>
Message-ID: <3A8DBCAE.5080808@ActiveState.com>

Owen F. Ransen wrote:

> I'm just starting out with Python and have some simple
> modules in a directoryb which is not on the sys.path.
> 
> I hoped that if I edited the registry and added my new
> extra source file path I'd then see the new files in the
> path browser (GUI Python 2).
> 
> I can't though, what am I missing?

Create a new _subkey_ under the main PythonPath entry, calling it anything you like.  Add your new path as the default value for this new key.

The main PythonPath entry is reserved for the Python core, and it may choose to ignore it if it can find its own core.

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From ransen_spam_me_not at nemo.it  Sat Feb 17 18:42:02 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Sat, 17 Feb 2001 17:42:02 GMT
Subject: sys.path and the Windows registry
References: <3a8d0450.162541@news.newsguy.com> <3A8DBCAE.5080808@ActiveState.com>
Message-ID: <3a8e9737.2589893@news.newsguy.com>

Thanks Mark!

On Fri, 16 Feb 2001 23:53:46 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>Owen F. Ransen wrote:
>
>> I'm just starting out with Python and have some simple
>> modules in a directoryb which is not on the sys.path.
>> 
>> I hoped that if I edited the registry and added my new
>> extra source file path I'd then see the new files in the
>> path browser (GUI Python 2).
>> 
>> I can't though, what am I missing?
>
>Create a new _subkey_ under the main PythonPath entry, calling it anything you like.  Add your new path as the default value for this new key.
>
>The main PythonPath entry is reserved for the Python core, and it may choose to ignore it if it can find its own core.
>
>Mark.

--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From gavin at madzag.com  Fri Feb 16 13:11:26 2001
From: gavin at madzag.com (Gavin Tomlins)
Date: Fri, 16 Feb 2001 12:11:26 GMT
Subject: Interpreter Example - 4 score & 7 years ago
Message-ID: <ON8j6.29221$o85.141369@news-server.bigpond.net.au>

Greetings,

I'm currently pursuing references about using python as an 
interpreter/compiler.

A year + ago there was an example of writing an interpreter in python for 
chemical symbols/compounds. I was wondering if anyone had a reference or 
can recall the posting.

Regards

Gavin


From ivnowa at hvision.nl  Fri Feb 16 13:21:04 2001
From: ivnowa at hvision.nl (Hans Nowak)
Date: Fri, 16 Feb 2001 12:21:04 GMT
Subject: Interpreter Example - 4 score & 7 years ago
References: <ON8j6.29221$o85.141369@news-server.bigpond.net.au>
Message-ID: <3A8D193F.7727@hvision.nl>

Gavin Tomlins wrote:

> I'm currently pursuing references about using python as an
> interpreter/compiler.
> 
> A year + ago there was an example of writing an interpreter in python for
> chemical symbols/compounds. I was wondering if anyone had a reference or
> can recall the posting.

Is it this one?

  http://tor.dhs.org/~zephyrfalcon/snippets/source/63.py

Regards,

--Hans Nowak


From gavin at madzag.com  Fri Feb 16 13:39:13 2001
From: gavin at madzag.com (Gavin Tomlins)
Date: Fri, 16 Feb 2001 12:39:13 GMT
Subject: Interpreter Example - 4 score & 7 years ago
References: <ON8j6.29221$o85.141369@news-server.bigpond.net.au> <3A8D193F.7727@hvision.nl>
Message-ID: <Rb9j6.29296$o85.141216@news-server.bigpond.net.au>

>> A year + ago there was an example of writing an interpreter in python
>> for chemical symbols/compounds. I was wondering if anyone had a
>> reference or can recall the posting.
>
>Is it this one?
>  http://tor.dhs.org/~zephyrfalcon/snippets/source/63.py
>Regards,
>--Hans Nowak

The exact one I was looking for.

Thanks

Gavin


From matthew at manderson.co.uk  Fri Feb 16 13:37:59 2001
From: matthew at manderson.co.uk (Matthew Manderson)
Date: Fri, 16 Feb 2001 12:37:59 -0000
Subject: How upgrade Python 1.5 to 2.0 on Unix....?
Message-ID: <3a8d1ec0@news.telinco.net>

Hi all, I installed Python 2 on my raq3 and it works, however
it still defaults to Python1.5? how do I fix this?

The folders are:

usr/lib/python1.5
usr/lib/Python-2.0

Many thanks and regards, Matthew

____________________________

Matthew S Manderson ATT ATII
email: matthew at manderson.co.uk
___________________________




From gerhard.nospam at bigfoot.de  Fri Feb 16 20:43:56 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Fri, 16 Feb 2001 20:43:56 +0100
Subject: How upgrade Python 1.5 to 2.0 on Unix....?
References: <3a8d1ec0@news.telinco.net>
Message-ID: <3A8D82FC.6A2C1E63@bigfoot.de>

Matthew Manderson wrote:
> 
> Hi all, I installed Python 2 on my raq3 and it works, however
> it still defaults to Python1.5? how do I fix this?
> 
> The folders are:
> 
> usr/lib/python1.5
> usr/lib/Python-2.0

I guess then "python" command executes 1.5.2. The following commands
should fix this:

rm usr/bin/python
ln -s usr/bin/python2.0 usr/bin/python

Gerhard
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From marciann at tin.it  Fri Feb 16 13:40:17 2001
From: marciann at tin.it (Marco Iannaccone)
Date: Fri, 16 Feb 2001 12:40:17 GMT
Subject: Videogame scripting
Message-ID: <Rc9j6.17279$LM4.229776@twister1.tin.it>

Do you think Python is a good scripting language for videogames. I wanna use
it for both Windows and Linux. Is it easy enough to program it, and to use
it with C and C++? Is it fast?

--
Marco Iannaccone     marciann at tin.it
ICQ: 18748121           MetalCoder

"What is real?  How do you define real?  If you're talking about your
senses, what you feel, taste, smell, or see, then all you're talking about
are electrical signals interpreted by your brain."
                                                       Morpheus - The Matrix




From stephen_purcell at yahoo.com  Fri Feb 16 14:48:34 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 16 Feb 2001 14:48:34 +0100
Subject: Videogame scripting
In-Reply-To: <Rc9j6.17279$LM4.229776@twister1.tin.it>; from marciann@tin.it on Fri, Feb 16, 2001 at 12:40:17PM +0000
References: <Rc9j6.17279$LM4.229776@twister1.tin.it>
Message-ID: <20010216144834.B12097@freedom.puma-ag.com>

Marco Iannaccone wrote:
> Do you think Python is a good scripting language for videogames. I wanna use
> it for both Windows and Linux. Is it easy enough to program it, and to use
> it with C and C++? Is it fast?

Don't know the answers to all of those questions, but it seems like you
should check out PySDL:

          http://pysdl.sourceforge.net/

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From root at rainerdeyke.com  Fri Feb 16 19:34:53 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Fri, 16 Feb 2001 18:34:53 GMT
Subject: Videogame scripting
References: <Rc9j6.17279$LM4.229776@twister1.tin.it>
Message-ID: <hpej6.243861$ge4.84222425@news2.rdc2.tx.home.com>

"Marco Iannaccone" <marciann at tin.it> wrote in message
news:Rc9j6.17279$LM4.229776 at twister1.tin.it...
> Do you think Python is a good scripting language for videogames.

Yes.

> I wanna use
> it for both Windows and Linux. Is it easy enough to program it, and to use
> it with C and C++?

Yes.  Interfacing with C++ can be difficult if you use C++ exceptions, but
there are wrapper libraries that take care of that for you.

> Is it fast?

It is much slower than C/C++.  Depending on what you do with it, this may or
may not be significant.  Compared with other scripting languages, it's not
bad.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From aleaxit at yahoo.com  Fri Feb 16 21:02:18 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 16 Feb 2001 21:02:18 +0100
Subject: Videogame scripting
References: <Rc9j6.17279$LM4.229776@twister1.tin.it>
Message-ID: <96k28e21utj@news2.newsguy.com>

"Marco Iannaccone" <marciann at tin.it> wrote in message
news:Rc9j6.17279$LM4.229776 at twister1.tin.it...
> Do you think Python is a good scripting language for videogames. I wanna
use
> it for both Windows and Linux. Is it easy enough to program it, and to use
> it with C and C++? Is it fast?

No scripting language is really _fast_ (the FAST parts you'll be better
off keeping in C/C++ -- or you may want to move to high-order but
fully-compilable languages such as O'Caml).  But Python IS very easy,
powerful, and handy at the utmost, on both Linux and Windows; and
it interfaces well to C, and even better to C++.


Alex





From catunda at pobox.com  Fri Feb 16 14:06:19 2001
From: catunda at pobox.com (Marco Catunda)
Date: Fri, 16 Feb 2001 11:06:19 -0200
Subject: popen problem
Message-ID: <3A8D25CB.D1A1168C@pobox.com>

Hello,
	

	I've written a program using threads and popen. Basically, this program
	execute a pool of threads and each thread executes a command ping with 
	open function. Like bellow:

	outFile = os.popen( "ping localhost", "r" )
	lines = outFile.readlines()
	outFile.close()

	I have no idea why, at some cases, the outfile.close() function
generates an 
	IOError except. This except generating is totaly random. 

	Does anyone have an idea why it happens? And how can I correct it?
	Is popen thread safe?
	
 	Thanks for help.

Regards
-- Marco Catunda


From mix77 at usa.net  Fri Feb 16 14:24:40 2001
From: mix77 at usa.net (Mix)
Date: Fri, 16 Feb 2001 15:24:40 +0200
Subject: PIL (Python Image Library) RPM
Message-ID: <96ja30$5a3$1@ctb-nnrp2.saix.net>

Where can I find Python Image Library RPM (if there is one)?




From jafo at tummy.com  Sat Feb 17 05:29:31 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Fri, 16 Feb 2001 21:29:31 -0700
Subject: PIL (Python Image Library) RPM
In-Reply-To: <96ja30$5a3$1@ctb-nnrp2.saix.net>; from mix77@usa.net on Fri, Feb 16, 2001 at 03:24:40PM +0200
References: <96ja30$5a3$1@ctb-nnrp2.saix.net>
Message-ID: <20010216212931.D5348@tummy.com>

On Fri, Feb 16, 2001 at 03:24:40PM +0200, Mix wrote:
>Where can I find Python Image Library RPM (if there is one)?

That one I used to build Sketch is at:

ftp.tummy.com:/pub/tummy/RPMS/SRPMS/python-imaging-1.0-6.src.rpm

Sean
-- 
 Follow the dream, not the competition.
                 -- Nemesis Racing Team
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From gchiaramonte at yahoo.com  Fri Feb 16 14:41:39 2001
From: gchiaramonte at yahoo.com (Gene C)
Date: Fri, 16 Feb 2001 13:41:39 GMT
Subject: Python & Microsoft Script Control 
Message-ID: <n6aj6.848$JS2.124285@newsread2.prod.itd.earthlink.net>

Anyone know where I can find information on using Python with the Microsoft
Script Contol?

>From the Microsoft website:

Through the ActiveX Scripting Interface, the Microsoft Script Control allows
applications to support any scripting language. Microsoft Internet Explorer
4.0 delivers two scripting languages, including Visual Basic Scripting
Edition and JScriptT development software, the industry's leading
implementation of ECMAScript. Support for additional scripting languages,
such as Perl and Python, is available from third-party vendors.




From gchiaramonte at yahoo.com  Fri Feb 16 19:54:35 2001
From: gchiaramonte at yahoo.com (Gene)
Date: Fri, 16 Feb 2001 18:54:35 GMT
Subject: Python & Microsoft Script Control 
References: <n6aj6.848$JS2.124285@newsread2.prod.itd.earthlink.net>
Message-ID: <LHej6.1558$JS2.193636@newsread2.prod.itd.earthlink.net>

It's part of the ActivePython distribution.

"Gene C" <gchiaramonte at yahoo.com> wrote in message
news:n6aj6.848$JS2.124285 at newsread2.prod.itd.earthlink.net...
> Anyone know where I can find information on using Python with the
Microsoft
> Script Contol?
>
> From the Microsoft website:
>
> Through the ActiveX Scripting Interface, the Microsoft Script Control
allows
> applications to support any scripting language. Microsoft Internet
Explorer
> 4.0 delivers two scripting languages, including Visual Basic Scripting
> Edition and JScriptT development software, the industry's leading
> implementation of ECMAScript. Support for additional scripting languages,
> such as Perl and Python, is available from third-party vendors.
>
>
>




From aleaxit at yahoo.com  Fri Feb 16 21:00:25 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 16 Feb 2001 21:00:25 +0100
Subject: Python & Microsoft Script Control
References: <n6aj6.848$JS2.124285@newsread2.prod.itd.earthlink.net>
Message-ID: <96k28c11utj@news2.newsguy.com>

"Gene C" <gchiaramonte at yahoo.com> wrote in message
news:n6aj6.848$JS2.124285 at newsread2.prod.itd.earthlink.net...
> Anyone know where I can find information on using Python with the
Microsoft
> Script Contol?

I know of no location with specific info on that, apart from Microsoft's
scripting pages (not much on the scriptcontrol itself -- it's basically a
rehash of the helpfile that comes with the scriptcontrol).  Fortunately,
there isn't much difficulty involved, as long as you have win32all
installed with your Python, or are using the ActivePython distribution.

The MS help file is VB-centered, but that goes for 99% of the docs
of all Automation object models &c -- you need to be able to read
'just enough VB to get by', it's really pretty easy.

While I haven't yet managed to 'really' make Python a part of the
applications we develop at work, the ScriptControl IS helping -- with
the rest of ActiveScripting/Automation/COM, it makes it really trivial
to let an application be scripted through any scripting language at
all -- and Python really shines when thus placed on a level playing
field with VBScript &tc.

Only itch I'm aware of, so far, is some bug which makes the
Procedures collection of the ScriptControl unusable when the
Language property is set to Python -- I'm currently working
around that with a Python-specific kludge (if operations on
the Procedures collection fail with an 'index error', Python
specific workarounds [based on Python introspection] are
used), but anyway I gave Mark Hammond a tiny program which
easily reproduces the bug, and I trust he'll fix it eventually.


Alex





From MarkH at ActiveState.com  Sat Feb 17 01:00:40 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 17 Feb 2001 00:00:40 GMT
Subject: Python & Microsoft Script Control
References: <n6aj6.848$JS2.124285@newsread2.prod.itd.earthlink.net> <96k28c11utj@news2.newsguy.com>
Message-ID: <3A8DBE4C.9010200@ActiveState.com>

Alex Martelli wrote:


> While I haven't yet managed to 'really' make Python a part of the
> applications we develop at work, the ScriptControl IS helping -- with
> the rest of ActiveScripting/Automation/COM, it makes it really trivial
> to let an application be scripted through any scripting language at
> all -- and Python really shines when thus placed on a level playing
> field with VBScript &tc.

Have you checked out win32comext\axscript\test\testHost.py - this implements a first-class Active Scripting host, and allows you to define your ActiveScripting object module purely in Python.

Mark.



From aleaxit at yahoo.com  Sat Feb 17 08:38:33 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 17 Feb 2001 08:38:33 +0100
Subject: Python & Microsoft Script Control
References: <n6aj6.848$JS2.124285@newsread2.prod.itd.earthlink.net> <96k28c11utj@news2.newsguy.com> <3A8DBE4C.9010200@ActiveState.com>
Message-ID: <96l9sn0dp7@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A8DBE4C.9010200 at ActiveState.com...
> Alex Martelli wrote:
>
> > While I haven't yet managed to 'really' make Python a part of the
> > applications we develop at work, the ScriptControl IS helping -- with
> > the rest of ActiveScripting/Automation/COM, it makes it really trivial
> > to let an application be scripted through any scripting language at
> > all -- and Python really shines when thus placed on a level playing
> > field with VBScript &tc.
>
> Have you checked out win32comext\axscript\test\testHost.py - this
implements a first-class Active Scripting host, and allows you to define
your ActiveScripting object module purely in Python.

Haven't tried it, no -- looks like a neat hack indeed, but not applicable
to the issue of "how do I get Python into these `mainstream' Windows
applications", for which Microsoft ScriptControl offers a backdoor.


Alex





From Pierre.Blanchet at solsoft.fr  Fri Feb 16 15:13:36 2001
From: Pierre.Blanchet at solsoft.fr (Pierre Blanchet)
Date: Fri, 16 Feb 2001 15:13:36 +0100
Subject: PRoblem with UDP sockets
Message-ID: <E14Tldg-0000rD-00@ciguri.solsoft.fr>

	I'm trying to code a TFTP client with python 1.5.2 under 
Linux.

	I cannot find a way to send UDP packets shorter than 14 
bytes.

	By example, if i try to send (using sendto) the string ' 0 
4 0 1', i get on the wire '0 4 0 1 1 1 1 1 1 1 1 1 1 1 1' althougth 
sendto tell me it has send 4 bytes.

	Do i do something wrong ? Is there a bug in python in Linux 
?

	TIA,

	Pierre.

-- 
Pierre Blanchet		 				  Support Engineer
http://www.solsoft.com				Pierre.Blanchet at solsoft.fr
Tel.: +33 147 15 55 00                               Fax: +33 147 15 55 09





From sholden at holdenweb.com  Fri Feb 16 15:28:36 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 16 Feb 2001 09:28:36 -0500
Subject: PRoblem with UDP sockets
References: <mailman.982332924.15258.python-list@python.org>
Message-ID: <7Saj6.24117$n4.336846@e420r-atl2.usenetserver.com>

"Pierre Blanchet" <Pierre.Blanchet at solsoft.fr> wrote in message
news:mailman.982332924.15258.python-list at python.org...
> I'm trying to code a TFTP client with python 1.5.2 under
> Linux.
>
> I cannot find a way to send UDP packets shorter than 14
> bytes.
>
> By example, if i try to send (using sendto) the string ' 0
> 4 0 1', i get on the wire '0 4 0 1 1 1 1 1 1 1 1 1 1 1 1' althougth
> sendto tell me it has send 4 bytes.
>
> Do i do something wrong ? Is there a bug in python in Linux
> ?
>
> TIA,
>
> Pierre.
>
Have you checked the data length in the UPD header? Wondered whether this
might be more to do with the minimum Ethernet frame size...

regards
 Steve
--
Tools, training and technology to help you meet your information needs







From Pierre.Blanchet at solsoft.fr  Fri Feb 16 16:26:36 2001
From: Pierre.Blanchet at solsoft.fr (Pierre Blanchet)
Date: Fri, 16 Feb 2001 16:26:36 +0100
Subject: PRoblem with UDP sockets 
In-Reply-To: Your message of "Fri, 16 Feb 2001 09:28:36 EST."
             <7Saj6.24117$n4.336846@e420r-atl2.usenetserver.com> 
References: <mailman.982332924.15258.python-list@python.org>  <7Saj6.24117$n4.336846@e420r-atl2.usenetserver.com> 
Message-ID: <E14TmmK-00019F-00@ciguri.solsoft.fr>

On February 16 2001 at 9:28, 
	"Steve Holden" <sholden at holdenweb.com> wrote:
> "Pierre Blanchet" <Pierre.Blanchet at solsoft.fr> wrote in message
> news:mailman.982332924.15258.python-list at python.org...
> > I'm trying to code a TFTP client with python 1.5.2 under
> > Linux.
> >
> > I cannot find a way to send UDP packets shorter than 14
> > bytes.
> >
> > By example, if i try to send (using sendto) the string ' 0
> > 4 0 1', i get on the wire '0 4 0 1 1 1 1 1 1 1 1 1 1 1 1' althougth
> > sendto tell me it has send 4 bytes.
> >
> > Do i do something wrong ? Is there a bug in python in Linux
> > ?
> >
> > TIA,
> >
> > Pierre.
> >
> Have you checked the data length in the UPD header? Wondered whether this
> might be more to do with the minimum Ethernet frame size...
> 

	Nevermind. The bug was mine.

	Actually now my TFTPclient works :) But my packets are 
still longer than i need.

	But you were right, the minimum size of an ethernet packet 
is 60 ( as shown by ethereal) and there is a need for padding (in 
my case 14 bytes in the "trailer").

	Thanks Steve for your help,

	Pierre.

-- 
Pierre Blanchet		 				  Support Engineer
http://www.solsoft.com				Pierre.Blanchet at solsoft.fr
Tel.: +33 147 15 55 00                               Fax: +33 147 15 55 09





From sholden at holdenweb.com  Fri Feb 16 15:20:07 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 16 Feb 2001 09:20:07 -0500
Subject: Sping (was PIDDLE) not producing PDF file
Message-ID: <9Kaj6.24112$n4.336343@e420r-atl2.usenetserver.com>

This is a little program I used to produce digital camera thumbnail listings
under 1.5.2 which worked fine. I've loaded sping and PIL under 2.0 (where's
all the documentation gone?) and modified the code to accommodate the new
packaging structure. It runs, but seems to produce no PDF file. Can anyone
point out what I might be doing wrong?

regards
 Steve

from sping import *
from sping.PDF import PDFCanvas, Font
import PIL.Image, os.path, sys, glob

MAXLINE = 8
MAXCOL = 4
LINEHEIGHT = 100
COLWIDTH = 140
LEFTMARGIN = 45
TOPMARGIN = 45

fntLarge = Font(face="helvetica", size=16, bold=1)
fntMedium = Font(face="helvetica", size=12, bold=1)
fntSmall = Font(face="helvetica", size=8, bold=0)

imagename = []
for filepat in sys.argv[1:]:
    for infile in glob.glob(filepat):
        imagename.append(infile)

imagename.sort()
lineno = MAXLINE+1
colno = MAXCOL+1
pageno = 0

out = PDFCanvas(name = "C:\PhotoClips.pdf")

for i in range(len(imagename)):
    if colno >= MAXCOL:
        lineno = lineno+1
        colno = 0
    if lineno >= MAXLINE:
        lineno = 0
        colno = 0
        if pageno > 0: out.clear()
        pageno = pageno +1
    print imagename[i]; sys.stdout.flush()
    im = PIL.Image.open(imagename[i])
    imx = colno*COLWIDTH+LEFTMARGIN
    imy = lineno*LINEHEIGHT+TOPMARGIN
    out.drawImage(im.resize((220,146)), imx, imy, imx+109, imy+72)
    out.drawString(os.path.basename(imagename[i]),
        imx, imy+80, fntSmall)
    colno = colno +1

out.flush()




--
--
Tools, training and technology to help you meet your information needs






From sholden at holdenweb.com  Fri Feb 16 16:13:35 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 16 Feb 2001 10:13:35 -0500
Subject: Sping (was PIDDLE) not producing PDF file
References: <9Kaj6.24112$n4.336343@e420r-atl2.usenetserver.com>
Message-ID: <xwbj6.17244$GA5.260521@e420r-atl1.usenetserver.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:9Kaj6.24112$n4.336343 at e420r-atl2.usenetserver.com...
> This is a little program I used to produce digital camera thumbnail
listings
> under 1.5.2 which worked fine. I've loaded sping and PIL under 2.0
(where's
> all the documentation gone?) and modified the code to accommodate the new
> packaging structure. It runs, but seems to produce no PDF file. Can anyone
> point out what I might be doing wrong?
>
> regards
>  Steve
>
Seems to be "follow up my own postings" day today. Oh, well.

Christopher Lee was kind enough to pint out by e-mail that I omitted to
update my program with a call to the PDFCanvas' save() method, a change made
around six months ago.

So adding a final line of

out.save()

fixed everything.  Great software!

regards
 Steve





From Roy.Culley at switzerland.org  Fri Feb 16 16:45:04 2001
From: Roy.Culley at switzerland.org (Roy.Culley at switzerland.org)
Date: Fri, 16 Feb 2001 16:45:04 +0100
Subject: Creating a dictionary from log file records
Message-ID: <0uhj69.vpb.ln@gd2zzx>

I'm new to python and am trying to convert a perl script which analyses
firewall logs to python as a learning exercise.

The log files contain lines of multiple key / value pairs such as:

    interface qfe0 proto tcp src 1.1.1.1 dst 2.2.2.2 service smtp \
        s_port 44008 len 44 rule 7

Not all records are the same and the key / value pairs are not at
fixed positions. In perl, assuming the line is in $_, I can do:

    %Rec = split

Is there an equivalent simple way to do this with python? I've done
it by converting the data into a list and using a while loop to set
the dictionary entries. However, the log files have about 4 million
entries per day so I need something that is fast.

Any help / suggestions most appreciated.


From shaleh at valinux.com  Fri Feb 16 17:35:52 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 08:35:52 -0800 (PST)
Subject: Creating a dictionary from log file records
In-Reply-To: <0uhj69.vpb.ln@gd2zzx>
Message-ID: <XFMail.20010216083552.shaleh@valinux.com>

On 16-Feb-2001 Roy.Culley at switzerland.org wrote:
> I'm new to python and am trying to convert a perl script which analyses
> firewall logs to python as a learning exercise.
> 
> The log files contain lines of multiple key / value pairs such as:
> 
>     interface qfe0 proto tcp src 1.1.1.1 dst 2.2.2.2 service smtp \
>         s_port 44008 len 44 rule 7
> 
> Not all records are the same and the key / value pairs are not at
> fixed positions. In perl, assuming the line is in $_, I can do:
> 
>     %Rec = split
> 
> Is there an equivalent simple way to do this with python? I've done
> it by converting the data into a list and using a while loop to set
> the dictionary entries. However, the log files have about 4 million
> entries per day so I need something that is fast.
> 

so for every line in the file, you create a temporary perl hash.  What do you
do with the hash?  generating a new hash for every line can not be but so fast.



From Roy.Culley at switzerland.org  Fri Feb 16 19:11:18 2001
From: Roy.Culley at switzerland.org (Roy.Culley at switzerland.org)
Date: Fri, 16 Feb 2001 19:11:18 +0100
Subject: Creating a dictionary from log file records
References: <mailman.982341324.16674.python-list@python.org>
Message-ID: <6gqj69.qbc.ln@gd2zzx>

In article <mailman.982341324.16674.python-list at python.org>,
	"Sean 'Shaleh' Perry" <shaleh at valinux.com> wrote:
>
> On 16-Feb-2001 Roy.Culley at switzerland.org wrote:
>> I'm new to python and am trying to convert a perl script which analyses
>> firewall logs to python as a learning exercise.
>> 
>> The log files contain lines of multiple key / value pairs such as:
>> 
>>     interface qfe0 proto tcp src 1.1.1.1 dst 2.2.2.2 service smtp \
>>         s_port 44008 len 44 rule 7
>> 
>> Not all records are the same and the key / value pairs are not at
>> fixed positions. In perl, assuming the line is in $_, I can do:
>> 
>>     %Rec = split
>> 
>> Is there an equivalent simple way to do this with python? I've done
>> it by converting the data into a list and using a while loop to set
>> the dictionary entries. However, the log files have about 4 million
>> entries per day so I need something that is fast.
>> 
> 
> so for every line in the file, you create a temporary perl hash.  What do you
> do with the hash?  generating a new hash for every line can not be but so fast.

On a Sun E250 with a single 300MHz processor the perl script can
process a days log of 3.5 - 4 MB in a little over half an hour. From
the hash I make lots of other hashes which I then use to produce a log
report with many different types of summary data. For example,
services by IP source address, destination address (both accepted and
dropped connections), connections by src addr + dst addr + service,
etc. These are all in hashes. Many different combinations. This
enables us to see quickly scans (both IP addr and port), what services
are used most (helps order firewall rules), etc. It is quite a simple
perl script and most of the time is spent sorting the hashes and
producing the report. The initial data extraction into the hashes is
very fast.


From stevee at recombinant.demon.co.uk  Fri Feb 16 22:18:34 2001
From: stevee at recombinant.demon.co.uk (Stephen D Evans)
Date: Fri, 16 Feb 2001 21:18:34 +0000
Subject: Creating a dictionary from log file records
References: <0uhj69.vpb.ln@gd2zzx>
Message-ID: <3A8D9929.CD69DAA1@recombinant.demon.co.uk>

Roy,

For an alternative Python solution (without resorting to C) you could try
using compile() to create code objects for eval or exec rather than using a
while or for loop. On my Intel Pentium 200 machine the exec and eval
methods in the following code ran about 4% faster than the for loop. Please
note that I have only tested this on Python 2.0

####
filename = 'logfile.log'

# Old FORTRAN technique. Static allocation of list length.
# If an IndexError occurs, change the value and start again.
MAX_LEN = 20

# pre-create a list of code objects that return list 'L' as a dictionary
eval_code_object_list = [None] * MAX_LEN
exec_code_object_list = [None] * MAX_LEN
for i in range(0, MAX_LEN, 2): # assume list of even length
    text = ''
    for j in range(0, i, 2):
        text='%sL[%d]:L[%d],' % (text, j, j+1)
    text = '{%s}' % text

    eval_code_object_list[i] = compile(text, '', 'eval')

    text = 'dict=%s' % text
    exec_code_object_list[i] = compile(text, '', 'exec')


import time # for some stats

start_time = time.clock()

# for Python 1.5.2
print 'Python readlines():',
lines = open(filename, 'rt').readlines()

print time.clock() - start_time
start_time = time.clock()

# an alternative method for Python 2.0 (not as fast), will fail on Python
1.5.2
print 'Python read().splitlines():',
lines = open(filename, 'rt').read().splitlines()

print time.clock() - start_time

# Python 2.1 could use the 'for line in open(filename).xreadlines():'
method
# for large files

start_time = time.clock()
print 'for loop:',

## The list 'L' is converted to a dictionary 'dict' using a simple for
loop.
for line in lines:
    L = line.split()
    dict = {}
    for i in range(0, len(L), 2):
        dict[L[i]] = L[i+1]

print time.clock() - start_time
start_time = time.clock()
print 'eval:',

## The list 'L' is converted to a dictionary 'dict' using an index into the
list
## of code objects. Lookups and no iteration during the creation of 'dict'
for line in lines:
    L = line.split()
    dict = eval(eval_code_object_list[len(L)])

print time.clock() - start_time
start_time = time.clock()
print 'exec:',

## as above but use exec to produce dict
for line in lines:
    L = line.split()
    exec exec_code_object_list[len(L)]
    # dict is here !

print time.clock() - start_time

####

I normally do not bother with optimisations that give less than 10%
improvement.

There has been a request for a Python equivalent of the Perl split
mentioned
somewhere in the Python world...

Stephen D Evans

Roy.Culley at switzerland.org wrote:

> I'm new to python and am trying to convert a perl script which analyses
> firewall logs to python as a learning exercise.
>
> The log files contain lines of multiple key / value pairs such as:
>
>     interface qfe0 proto tcp src 1.1.1.1 dst 2.2.2.2 service smtp \
>         s_port 44008 len 44 rule 7
>
> Not all records are the same and the key / value pairs are not at
> fixed positions. In perl, assuming the line is in $_, I can do:
>
>     %Rec = split
>
> Is there an equivalent simple way to do this with python? I've done
> it by converting the data into a list and using a while loop to set
> the dictionary entries. However, the log files have about 4 million
> entries per day so I need something that is fast.
>
> Any help / suggestions most appreciated.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010216/48cbc034/attachment.htm>

From neelk at alum.mit.edu  Fri Feb 16 16:49:28 2001
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 16 Feb 2001 15:49:28 GMT
Subject: PEP-0234: Iterators
Message-ID: <slrn98qmpv.4b8.neelk@alum.mit.edu>

I have a couple of comments about the iterators proposal. The need for
improving the iteration protocol has been apparent for a really long
time, and I'm really glad to see that a PEP for it has appeared.
However, in its current form I think that what the PEP proposes can be
simplified a bit.

You can get a simplification if the separate protocols for sequence
and mappings were combined, and an iterator was an object with methods
"current_index", "current_value", and "next_state". The "current
index" of a sequence would be the position in the list, and the
"current index" of a mapping would the key that yields that value.

So then a loop like:

for elt in coll:
   # Stuff

would be equivalent to:

iterator = coll.__iter__()
while 1:
    elt = iterator.current_value()

    # Stuff

    try:
        iterator = iterator.next_state()
    except IndexError:
        break

The "key:value" syntax could then be the same for mappings and
sequences:

for key:value in coll:
    # Stuff

iterator = coll.__iter__()
while 1:
    key = iterator.current_index()
    val = iterator.current_value()

    # Stuff

    try: 
	iterator = iterator.next_state()
    except IndexError:
	break

And similar variations (":value", "key:") would follow logically.


Neel


From jeremy at alum.mit.edu  Fri Feb 16 17:57:09 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 16 Feb 2001 11:57:09 -0500 (EST)
Subject: PEP-0234: Iterators
In-Reply-To: <slrn98qmpv.4b8.neelk@alum.mit.edu>
References: <slrn98qmpv.4b8.neelk@alum.mit.edu>
Message-ID: <14989.23525.754713.423514@w221.z064000254.bwi-md.dsl.cnc.net>

Neel (and others who have responded),

You should also send a copy of your comments to the PEP author, in the
case of 234, Ka-Ping Yee <ping at lfw.org>.  There's no problem
discussing it on python-list, but the PEP author might miss them
without an explicit cc.

Jeremy



From olivier.deckmyn at mail.dotcom.fr  Fri Feb 16 17:05:37 2001
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Fri, 16 Feb 2001 17:05:37 +0100
Subject: PIL Fonts
Message-ID: <96jj4p$172n$1@feed.teaser.net>

Hi all !

Is there a way to use TrueType or PostScript font with ImageDraw in PIL ?
I know it exists a conversion tool to make PIL fonts from X fonts...But I
don't have any "nice" X font :(

Any idea ?
ps : is the python t1lib project discontinued ?


---
We are Micro$oft. You will be assimilated. Resistance is futile.





From chris.gonnerman at usa.net  Sat Feb 17 07:03:38 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 17 Feb 2001 00:03:38 -0600
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net>
Message-ID: <010a01c098a7$68082ae0$a100000a@local>

----- Original Message -----
From: "Olivier Deckmyn" <olivier.deckmyn at mail.dotcom.fr>
Subject: PIL Fonts
>...
> Is there a way to use TrueType or PostScript font with ImageDraw in PIL ?
> I know it exists a conversion tool to make PIL fonts from X fonts...But I
> don't have any "nice" X font :(
>...

I got so irritated with PIL over font support that I took over the gdmodule
project from Richard Jones.  If you do a lot of graphics with text on them,
in PNG or JPEG format, then get the GD library (and FreeType, and libjpeg
and libpng... hope you already have some installed) and finally build my
updated gdmodule.  You get graphics with TrueType font support!

BTW the core GD library doesn't do a real great job on JPEG files IMHO; I
must admit, I create things like button frames in the GIMP or on Windoze
using Paint Shop Pro, save them as PNG,  then fill in the text with Python
and the gdmodule, and finally use GIMP or ImageMagick to convert to GIF or
JPEG.  Ugly, but after the button frames are created I can fully script the
rest.

My gdmodule page is:

    http://newcenturycomputers.net/projects/gdmodule





From chris.gonnerman at usa.net  Sat Feb 17 17:03:22 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 17 Feb 2001 10:03:22 -0600
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net> <010a01c098a7$68082ae0$a100000a@local> <001701c098b7$e1f68300$0200000a@kashmir2k>
Message-ID: <000b01c098fb$2bcf6800$a100000a@local>

----- Original Message -----
From: "Olivier Deckmyn" <olivier.deckmyn at mail.dotcom.fr>
Subject: Re: PIL Fonts


> Thanx for your support !
>
> Unfortunaly, gd lacks for filters support :-(
>
> I wonder how I can use GD just for making TrueType rendering and insert
play
> with them from PIL...
>
> Thanx anyway !
> Olivier.

It's kludgy, but you can create your graphics in PIL, save them as PNG, load
them with GD,
add text, save with GD, open in PIL and revise more if needed...

I've been told there is a FreeType module to plug in to PIL, which would be
great, but I can't
find any examples for using it.  (To be fair, it has been a while since I
looked.)






From fredrik at pythonware.com  Sat Feb 17 19:51:21 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 17 Feb 2001 18:51:21 GMT
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net>
Message-ID: <JKzj6.17963$AH6.2439731@newsc.telia.net>

Olivier Deckmyn wrote:
> Is there a way to use TrueType or PostScript font with ImageDraw in PIL ?
> I know it exists a conversion tool to make PIL fonts from X fonts...But I
> don't have any "nice" X font :(

TrueType (freetype) support for Python/PIL is available here:

    http://starship.python.net/crew/kernr/Projects.html

and

    http://effbot.org/pil (windows only, at this time)

Cheers /F




From olivier.deckmyn at mail.dotcom.fr  Mon Feb 19 12:12:15 2001
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Mon, 19 Feb 2001 12:12:15 +0100
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net> <JKzj6.17963$AH6.2439731@newsc.telia.net>
Message-ID: <96qv2n$2nao$1@feed.teaser.net>

Is there any luck one can have the unix/sources version of _imagingft module
?

Anyway, thanx again Fred for your support !



"Fredrik Lundh" <fredrik at pythonware.com> a ?crit dans le message news:
JKzj6.17963$AH6.2439731 at newsc.telia.net...
> Olivier Deckmyn wrote:
> > Is there a way to use TrueType or PostScript font with ImageDraw in PIL
?
> > I know it exists a conversion tool to make PIL fonts from X fonts...But
I
> > don't have any "nice" X font :(
>
> TrueType (freetype) support for Python/PIL is available here:
>
>     http://starship.python.net/crew/kernr/Projects.html
>
> and
>
>     http://effbot.org/pil (windows only, at this time)
>
> Cheers /F
>
>




From fredrik at pythonware.com  Mon Feb 19 14:33:40 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Feb 2001 13:33:40 GMT
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net> <JKzj6.17963$AH6.2439731@newsc.telia.net> <96qv2n$2nao$1@feed.teaser.net>
Message-ID: <Ug9k6.18262$AH6.2517644@newsc.telia.net>

Olivier Deckmyn wrote:
> Is there any luck one can have the unix/sources version of
> _imagingft module?

sure, but not now -- it's a bit too messy for release, and
I don't have time to clean it up right now.

> Anyway, thanx again Fred for your support !

looks like someone hasn't read the PIL faq...

    http://www.pythonware.com/products/pil/faq.htm

> We are Micro$oft. You will be assimilated. Resistance is futile.

you mean "all your base are belong to us", don't you?

Cheers /F




From rtexier at elikya.com  Mon Feb 19 11:35:53 2001
From: rtexier at elikya.com (Romuald Texier)
Date: Mon, 19 Feb 2001 10:35:53 +0000
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net>
Message-ID: <96qpe1$81a$1@news.irisa.fr>

Olivier Deckmyn wrote:

> Hi all !
> 
> Is there a way to use TrueType or PostScript font with ImageDraw in PIL ?
> I know it exists a conversion tool to make PIL fonts from X fonts...But I
> don't have any "nice" X font :(
> 
> Any idea ?
> ps : is the python t1lib project discontinued ?

Why not using PyGimp ? It a Python binding for gimp : all the power of this 
graphics factory within your scripts ! (contrarily to Script-fu, Python has 
the control ; it is actually a wrapper to libgimp)

http://www.daa.com.au/~james/pygimp/

-- 
Romuald Texier


From olivier.deckmyn at mail.dotcom.fr  Mon Feb 19 12:17:06 2001
From: olivier.deckmyn at mail.dotcom.fr (Olivier Deckmyn)
Date: Mon, 19 Feb 2001 12:17:06 +0100
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net> <96qpe1$81a$1@news.irisa.fr>
Message-ID: <96qvbq$2nce$1@feed.teaser.net>

Doest this mean that a python script can delegate the making and render of
an Image to Gimp and get it back as a JPEG in a string for example ?


"Romuald Texier" <rtexier at elikya.com> a ?crit dans le message news:
96qpe1$81a$1 at news.irisa.fr...
> Olivier Deckmyn wrote:
>
> > Hi all !
> >
> > Is there a way to use TrueType or PostScript font with ImageDraw in PIL
?
> > I know it exists a conversion tool to make PIL fonts from X fonts...But
I
> > don't have any "nice" X font :(
> >
> > Any idea ?
> > ps : is the python t1lib project discontinued ?
>
> Why not using PyGimp ? It a Python binding for gimp : all the power of
this
> graphics factory within your scripts ! (contrarily to Script-fu, Python
has
> the control ; it is actually a wrapper to libgimp)
>
> http://www.daa.com.au/~james/pygimp/
>
> --
> Romuald Texier




From rtexier at elikya.com  Tue Feb 20 12:16:50 2001
From: rtexier at elikya.com (Romuald Texier)
Date: Tue, 20 Feb 2001 11:16:50 +0000
Subject: PIL Fonts
References: <96jj4p$172n$1@feed.teaser.net> <96qpe1$81a$1@news.irisa.fr> <96qvbq$2nce$1@feed.teaser.net>
Message-ID: <96tg6s$cip$1@news.irisa.fr>

Olivier Deckmyn wrote:

> Doest this mean that a python script can delegate the making and render of
> an Image to Gimp and get it back as a JPEG in a string for example ?

I do not know if you can get it back in a string. In a file, sure. If you 
want it in a string, that means you want to further process it, isn't it ? 
So why not  make all processing in Gimp and only get a file ? The problem 
is if you want to make dynamic images without disk usage, though.

  If you find a way, I would be interested in it ;o)

Sincerely,

-- 
Romuald Texier


From mnajmeddine at geo.census.gov  Fri Feb 16 17:51:15 2001
From: mnajmeddine at geo.census.gov (Mohamed Najmeddine)
Date: Fri, 16 Feb 2001 11:51:15 -0500
Subject: CGI/networking
Message-ID: <LNBBJMBHBAJKJGAJCGABEEIHCJAA.mnajmeddine@geo.census.gov>

Hello,
I'm trying to create a web interface. The users should be able to execute
the cgi scripts that will access their CDROM drive and check the files that
the CD contains and compared that list to a file installed on the WEBSERVER
(on a different machine UNIX). My problem is how to get the cgi script
access the hard drive of the PC users. If anyone could point me to the right
direction, I would appreciate it. (Do I need to use Sockets... and are
modules that would facilitate my job)
Thanks,
Med.




From glenfant at equod.com.nospam  Fri Feb 16 18:20:35 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Fri, 16 Feb 2001 18:20:35 +0100
Subject: CGI/networking
References: <mailman.982342354.21385.python-list@python.org>
Message-ID: <96jn5t$e7s$1@reader1.imaginet.fr>

Hopefully this is not possible for privacy/security reasons !

"Mohamed Najmeddine" <mnajmeddine at geo.census.gov> a ?crit dans le message
news: mailman.982342354.21385.python-list at python.org...
> Hello,
> I'm trying to create a web interface. The users should be able to execute
> the cgi scripts that will access their CDROM drive and check the files
that
> the CD contains and compared that list to a file installed on the
WEBSERVER
> (on a different machine UNIX). My problem is how to get the cgi script
> access the hard drive of the PC users. If anyone could point me to the
right
> direction, I would appreciate it. (Do I need to use Sockets... and are
> modules that would facilitate my job)
> Thanks,
> Med.
>
>



From sholden at holdenweb.com  Fri Feb 16 18:34:02 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 16 Feb 2001 12:34:02 -0500
Subject: CGI/networking
References: <mailman.982342354.21385.python-list@python.org>
Message-ID: <bAdj6.17801$GA5.265382@e420r-atl1.usenetserver.com>

"Mohamed Najmeddine" <mnajmeddine at geo.census.gov> wrote in message
news:mailman.982342354.21385.python-list at python.org...
> Hello,
> I'm trying to create a web interface. The users should be able to execute
> the cgi scripts that will access their CDROM drive and check the files
that
> the CD contains and compared that list to a file installed on the
WEBSERVER
> (on a different machine UNIX). My problem is how to get the cgi script
> access the hard drive of the PC users. If anyone could point me to the
right
> direction, I would appreciate it. (Do I need to use Sockets... and are
> modules that would facilitate my job)
> Thanks,
> Med.
>
>
This would depend on either:

a) Delivering active content (Javascript/VBScript/Java) to the client, and
having it execute, examine the filestore, and return results to the web
server, or

b) Having a network process on the client listening for connections from
your cgi agent, and again returning results.

In short, much effort has been expended to (try to) ensure that such
solutions are not available. Active content is normally prevented
frommanipulating the client filestore for security reasons. You could modify
your clients to allow such access in various ways, but all of them represent
fairly nasty security holes.

regards
 Steve




From bshapiro at funnygarbage.com  Fri Feb 16 18:04:27 2001
From: bshapiro at funnygarbage.com (Ben Shapiro)
Date: Fri, 16 Feb 2001 12:04:27 -0500
Subject: pywin and wxPython
Message-ID: <96jmi1$jst$1@bob.news.rcn.net>

Hey all...
Here's the predicament..
I have an OCX control that is responsible for playback of mp3 files that i
am trying to
use from python.  This control is wrapped using makepy.  Mark H pointed me
in the right
direction to use pywin to host the OCX..  only problem is that my app uses
wxPython as the
GUI framework.  From what i can tell you need to attach the activex control
to an MFC window for
it to load....  Any example of how to do this would be appreciated greatly!

My existing code is below:

        p = win32ui.CreateWnd()
        win32ui.EnableControlContainer()
        mp3player =
gencache.EnsureModule('{2DF09BC5-B050-11D2-81C1-00C0DFEAA961}', 0, 1, 0)
        self.olectl = activex.MakeControlInstance(mp3player.AMPL3DLite)
        self.olectl.CreateControl("", win32con.WS_TABSTOP , (7,43,500,300),
p , 131)  <---- fails here  / nothing seems to happen if i comment out this
line
        self.olectl.Open("c:\\code\\3LP\\pyfileunder\\music\\test.mp3")
        self.olectl.Play()




From bshapiro at funnygarbage.com  Fri Feb 16 18:09:08 2001
From: bshapiro at funnygarbage.com (Ben Shapiro)
Date: Fri, 16 Feb 2001 12:09:08 -0500
Subject: pywin and wxPython
References: <96jmi1$jst$1@bob.news.rcn.net>
Message-ID: <96jmqp$kvu$1@bob.news.rcn.net>

oops i forgot to mention that this activeX control has no GUI
so i really just need to load it without displaying anything


"Ben Shapiro" <bshapiro at funnygarbage.com> wrote in message
news:96jmi1$jst$1 at bob.news.rcn.net...
> Hey all...
> Here's the predicament..
> I have an OCX control that is responsible for playback of mp3 files that i
> am trying to
> use from python.  This control is wrapped using makepy.  Mark H pointed me
> in the right
> direction to use pywin to host the OCX..  only problem is that my app uses
> wxPython as the
> GUI framework.  From what i can tell you need to attach the activex
control
> to an MFC window for
> it to load....  Any example of how to do this would be appreciated
greatly!
>
> My existing code is below:
>
>         p = win32ui.CreateWnd()
>         win32ui.EnableControlContainer()
>         mp3player =
> gencache.EnsureModule('{2DF09BC5-B050-11D2-81C1-00C0DFEAA961}', 0, 1, 0)
>         self.olectl = activex.MakeControlInstance(mp3player.AMPL3DLite)
>         self.olectl.CreateControl("", win32con.WS_TABSTOP ,
(7,43,500,300),
> p , 131)  <---- fails here  / nothing seems to happen if i comment out
this
> line
>         self.olectl.Open("c:\\code\\3LP\\pyfileunder\\music\\test.mp3")
>         self.olectl.Play()
>
>




From SBrunning at trisystems.co.uk  Fri Feb 16 18:04:30 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 16 Feb 2001 17:04:30 -0000
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 200
	1-02-01 - 2001-02-15)
Message-ID: <31575A892FF6D1118F5800600846864D78B91D@intrepid>

> From:	claird at starbase.neosoft.com [SMTP:claird at starbase.neosoft.com]
> 2.  I hope a few OS/400 and OpenVMS refugees convert and
>     walk through the door soon.  *That* would make for a
>     nice dose of fun.
 
As far as OS/400 goes, check out Per Gummedal's fine work at
<http://home.no.net/pgummeda/>. Also, see
<http://groups.yahoo.com/group/JythonAS400>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From info at black-karma.de  Fri Feb 16 18:13:13 2001
From: info at black-karma.de (Harvest T. Moon)
Date: Fri, 16 Feb 2001 18:13:13 +0100
Subject: Q: Tkinter text widget
Message-ID: <96jn22$kqb59$1@ID-22517.news.dfncis.de>

how do i change the text color for specific lines in a text widget, so i can
have different lines with different colored text? (or at least some words
different colored/styled i.e. bold etc.)

Harvest T. Moon
-- a life less ordinary!!




From fredrik at pythonware.com  Sat Feb 17 10:52:40 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 17 Feb 2001 09:52:40 GMT
Subject: Tkinter text widget
References: <96jn22$kqb59$1@ID-22517.news.dfncis.de>
Message-ID: <IRrj6.14067$Qb7.2220744@newsb.telia.net>

"Harvest T. Moon" wrote:
> how do i change the text color for specific lines in a text widget, so i can
> have different lines with different colored text? (or at least some words
> different colored/styled i.e. bold etc.)

use tags:

    text.tag_config("highlight", foreground="red")
    text.insert(INSERT, "red text", "highlight") # insert coloured text
    text.tag_add("highlight", 5.0, 6.0) # highlight line 5

also see:

http://www.pythonware.com/library/tkinter/introduction/text.htm
=> "concepts"

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From aboufer at atlsci.com  Fri Feb 16 18:24:49 2001
From: aboufer at atlsci.com (aboufer at atlsci.com)
Date: Fri, 16 Feb 2001 12:24:49 -0500
Subject: Python threads ( termination )
Message-ID: <3A8D6261.7EE44EF0@atlsci.com>

Greetings,

    I am trying to write a multi-threaded app using python, and I am
trying to find
a way to stop/terminate a thread from within another.

Here is an example :

import threading

mySuccess = 0

class MyThread( threading.Thread) :
    def __init__( self, name='hello') :
        threading.Thread.__init__( self, name=name)

    def run(self) :
        global mySuccess
        import random

        for x in range(10) :
                threading._sleep( random() )
                print 'Hello ', x

        mySuccess = 1

myTest() :
    t1 = MyThread()
    t1.start()

    t1.join( 0.1 )

    if mySuccess == 1 :
        print 'Thread exited okay'
    else :
        print 'Thread did not respond !!'
        I need a way to kill thread t1 here <------------------------

if __name__ == '__main__' :
        myTest()


In this example, my function myTest calls join() on thread t1, with a
specified time_out. What I would like to
do is to terminate/stop thread t1 if it does not respond within timeout.



Thanks all



From dan at eevolved.com  Fri Feb 16 18:23:39 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 17:23:39 GMT
Subject: Python threads ( termination )
References: <3A8D6261.7EE44EF0@atlsci.com>
Message-ID: <vmdj6.263138$JT5.9137465@news20.bellglobal.com>

aboufer at atlsci.com wrote:

> Greetings,
> 
>     I am trying to write a multi-threaded app using python, and I am
> trying to find
> a way to stop/terminate a thread from within another.
> 
> Here is an example :
> 
> import threading
> 
> mySuccess = 0
> 
> class MyThread( threading.Thread) :
>     def __init__( self, name='hello') :
>         threading.Thread.__init__( self, name=name)
> 
>     def run(self) :
>         global mySuccess
>         import random
> 
>         for x in range(10) :
>                 threading._sleep( random() )
>                 print 'Hello ', x
> 
>         mySuccess = 1
> 
> myTest() :
>     t1 = MyThread()
>     t1.start()
> 
>     t1.join( 0.1 )
> 
>     if mySuccess == 1 :
>         print 'Thread exited okay'
>     else :
>         print 'Thread did not respond !!'
>         I need a way to kill thread t1 here <------------------------
> 
> if __name__ == '__main__' :
>         myTest()
> 
> 
> In this example, my function myTest calls join() on thread t1, with a
> specified time_out. What I would like to
> do is to terminate/stop thread t1 if it does not respond within timeout.
> 
> 
> 
> Thanks all
> 

You could use 2 threading.Event instances to communicate between threads 
and have the 'bad' thread terminate itself.

There is no real way to terminate a thread from another thread...

Dan



From jmarshal at mathworks.com  Fri Feb 16 18:52:07 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 16 Feb 2001 17:52:07 GMT
Subject: Python threads ( termination )
References: <3A8D6261.7EE44EF0@atlsci.com> <vmdj6.263138$JT5.9137465@news20.bellglobal.com>
Message-ID: <96jpc7$h5k$1@news.mathworks.com>

Dan Parisien <dan at eevolved.com> wrote:

> You could use 2 threading.Event instances to communicate between threads 
> and have the 'bad' thread terminate itself.

> There is no real way to terminate a thread from another thread...

> Dan

You can do this if it's the main thread you're trying to kill.  If you
send the process a signal, it's caught by the main thread.  In a
project I'm working on, we need to run arbitrary Python code which may
do Bad Things (like go into an infinite loop).  So we have the main
thread execute this untrusted code, and have a second thread
occasionally checking up on the main one to be sure progress is being
made.  If the main thread is spending too much time executing a chunk
of code, the second thread signals the process, raising an exception
in the main thread, which then moves on and does something productive.

Of course this is only useful if there's only one thread you're
interested in manipulating.


From hamish_lawson at yahoo.co.uk  Fri Feb 16 18:36:55 2001
From: hamish_lawson at yahoo.co.uk (=?iso-8859-1?q?Hamish=20Lawson?=)
Date: Fri, 16 Feb 2001 17:36:55 +0000 (GMT)
Subject: What became of BeOpen?
Message-ID: <20010216173655.20521.qmail@web204.mail.yahoo.com>

What actually became of BeOpen, PythonLab's former home? I understand
that they were in some difficulties, and, judging from the fact that
the beopen.com domain is up for sale, I'm guessing that they folded.

Hamish Lawson


____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From hamish_lawson at yahoo.co.uk  Fri Feb 16 18:38:37 2001
From: hamish_lawson at yahoo.co.uk (=?iso-8859-1?q?Hamish=20Lawson?=)
Date: Fri, 16 Feb 2001 17:38:37 +0000 (GMT)
Subject: What became of BeOpen?
Message-ID: <20010216173837.3547.qmail@web216.mail.yahoo.com>

What actually became of BeOpen, PythonLab's former home? I understand
that they were in some difficulties, and, judging from the fact that
the beopen.com domain is up for sale, I'm guessing that they folded.

Hamish Lawson


____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From owen at astroNOJNK.washington.edu.invalid  Fri Feb 16 18:41:05 2001
From: owen at astroNOJNK.washington.edu.invalid (Russell E. Owen)
Date: Fri, 16 Feb 2001 09:41:05 -0800
Subject: Discussion about PEP 234: iterators
Message-ID: <96jonj$i0e$1@nntp6.u.washington.edu>

This is in reaction to the new iterator proposal 
<http://python.sourceforge.net/peps/pep-0234.html>:

My own reaction is "great!" one misgiving: it does not address a very 
common need: to iterate over a dictionary in key-sorted order. Is there 
some variant syntax that could handle this case?

(Better yet, and going out on thin ice here, could this be made the 
default? I realize that's unlikely and may anger some people to even 
suggest it, but I would like to point out that some languages actually 
store dictionaries with their keys ordered, so it's at least technically 
feasible).

If this issue cannot be addressed, could we at least get "sort" to 
return the sorted collection, so one could do:

for key in dictionary.keys().sort():

-- Russell


From aleaxit at yahoo.com  Fri Feb 16 20:39:56 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 16 Feb 2001 20:39:56 +0100
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu>
Message-ID: <96k28b01utj@news2.newsguy.com>

"Russell E. Owen" <owen at astroNOJNK.washington.edu.invalid> wrote in message
news:96jonj$i0e$1 at nntp6.u.washington.edu...
> This is in reaction to the new iterator proposal
> <http://python.sourceforge.net/peps/pep-0234.html>:
>
> My own reaction is "great!" one misgiving: it does not address a very
> common need: to iterate over a dictionary in key-sorted order. Is there
> some variant syntax that could handle this case?
>
> (Better yet, and going out on thin ice here, could this be made the
> default? I realize that's unlikely and may anger some people to even
> suggest it, but I would like to point out that some languages actually
> store dictionaries with their keys ordered, so it's at least technically
> feasible).

One can choose to store mappings using sorted order on keys (that,
for example, is how C++'s standard std::map is specified to work),
but performance for all other tasks can suffer quite substantially
when compared to hash tables.  Python uses dictionaries so widely
that _any_ hit on dictionary performance would not be acceptable.

I would suggest leaving dictionaries alone and supplying another
instance of mapping, based on red-black trees or some variant
thereon, if the sorted-keys requirement is indeed so widespread.


> If this issue cannot be addressed, could we at least get "sort" to
> return the sorted collection, so one could do:
>
> for key in dictionary.keys().sort():

Why not

def sort(adisposablelist):
    adisposablelist.sort()
    return adisposablelist

and then

for key in sort(dictionary.keys()):


Is the syntax-sugar difference so big, after all?


Alex





From othello at javanet.com  Fri Feb 16 23:09:47 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Fri, 16 Feb 2001 17:09:47 -0500
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <96k28b01utj@news2.newsguy.com>
Message-ID: <3A8DA52A.254849DB@javanet.com>

I think dictionary iteration should match iterations on other objects/types
as closely as possible unless performance is impacted.

for elem in list                # does not automatically sort
for key in dict                # so this should not automatically sort

for key:value in dict        # proposed
for index:item in list        # so this should work

if elem in list                # this works
if key in dict                # so this should work

map( fun, list )            # this works
map( func, dict )        # so this should work with functions that take a key as
a parameter

filter( boolfun, list )        # gives a possibly smaller list
filter( boolfun, dict )        # so this should return a possibly smaller
dictionary


Thanks,

Raymond






From tim.hochberg at ieee.org  Fri Feb 16 23:39:16 2001
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Fri, 16 Feb 2001 22:39:16 GMT
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <96k28b01utj@news2.newsguy.com> <3A8DA52A.254849DB@javanet.com>
Message-ID: <o_hj6.18$8t2.15688@news1.rdc1.az.home.com>

"Raymond Hettinger" <othello at javanet.com> wrote in message
news:3A8DA52A.254849DB at javanet.com...
> I think dictionary iteration should match iterations on other
objects/types
> as closely as possible unless performance is impacted.

Ok then, let me put on my pedantic hat.
.
.
Oof, that's a tight fit.

> for elem in list                # does not automatically sort
> for key in dict                # so this should not automatically sort
>
> for key:value in dict        # proposed
> for index:item in list        # so this should work

Alright, looking at what you have so far, you've established a one to one
correspondence between dictionary keys and sequence indices as well as
between dictionary values and sequence values. That seems to make sense, but
this next bit is where we run into trouble -- consistency runs head on into
practicality and peoples expectations.

> if elem in list                # this works
> if key in dict                # so this should work

If one is going to go by the correspondence that you've set up above, then
really the dictionary equivalent should be:

if value in dict:

Unfortunately, people tend to think of a dictionary as containing keys more
than they think of it as containing definitions (values). So the form you
propose is better from that point of view. It's also probably more useful in
a practical sense as well. It is kind of distressing that it's not
consistent with the correspondence that's set up between keys<->indices and
values<->values though.

> map( fun, list )            # this works
> map( func, dict )        # so this should work with functions that take a
key as
> a parameter

This is not obvious to me. I would think that this would need a function
that takes values as parameters and would produce:

{key1 : func(val1), key2 : func(val2) ....}

I can't even figure out how this would work if func operated on keys. That
seems to be a bad sign for the obviousness of this construct.....

> filter( boolfun, list )        # gives a possibly smaller list
> filter( boolfun, dict )        # so this should return a possibly smaller
> dictionary

But how would it work? If were to guess at the meaning of something like
this I'd get:

dict = {}
for key, value in dict.keys():
   if boolfun(value):
      dict[key] = value

But being as we disagreed above, I'm not sure that we agree here.

I too would like to see everything as consistent as possible (but no
consistenter[sic]), but I'm not sure that agreement can be reached on that.

-tim






From joconnor at cybermesa.com  Fri Feb 16 23:56:38 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Fri, 16 Feb 2001 15:56:38 -0700
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <96k28b01utj@news2.newsguy.com> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com>
Message-ID: <3A8DB026.CE29126C@cybermesa.com>


Tim Hochberg wrote:

> "Raymond Hettinger" <othello at javanet.com> wrote in message
> news:3A8DA52A.254849DB at javanet.com...
> > I think dictionary iteration should match iterations on other
> objects/types
> > as closely as possible unless performance is impacted.
> ...

>
>
> > if elem in list                # this works
> > if key in dict                # so this should work
>
> If one is going to go by the correspondence that you've set up above, then
> really the dictionary equivalent should be:
>
> if value in dict:
>
> Unfortunately, people tend to think of a dictionary as containing keys more
> than they think of it as containing definitions (values). So the form you
> propose is better from that point of view. It's also probably more useful in
> a practical sense as well. It is kind of distressing that it's not
> consistent with the correspondence that's set up between keys<->indices and
> values<->values though.

Actually, given my Smalltalk background, I tend to see Dictionaries as
containing Associations (an Assocaiation is a key value pair)  Thus looping over
a Dictionary returns an association on each iteration.

To carry this to Python, I would like to see looping over Dictionaries return a
Tuple of the key,value pair.  Allowing either:

    for assoc in dictionary:
        print "key %s = value %s" % (assoc[0], assoc[1])

or

    for (key, value) in dictionary:
        print "key %s = value %s" % (key,value)

Take care,
Jay O'Connor
joconnor at cybermesa.com



From pingster at cs.berkeley.edu  Sun Feb 18 11:48:58 2001
From: pingster at cs.berkeley.edu (Ka-Ping Yee)
Date: Sun, 18 Feb 2001 02:48:58 -0800
Subject: Join python-iter@yahoogroups.com to discuss PEP 234
In-Reply-To: <3A8DB026.CE29126C@cybermesa.com>
References: <96jonj$i0e$1@nntp6.u.washington.edu> <96k28b01utj@news2.newsguy.com> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com> <3A8DB026.CE29126C@cybermesa.com>
Message-ID: <Pine.SOL.4.30.0102180246010.13768-100000@guaymas.CS.Berkeley.EDU>

Hello all,

I'd like to thank you for your thoughts and comments about
PEP 234 and invite you to join a discussion group on the
topic of iteration in Python.

Greg Wilson has graciously established python-iter at yahoogroups.com
as a place where we can focus a conversation on this issue.
If you're interested in talking about it, please join!

    http://groups.yahoo.com/group/python-iter

Thanks!


-- Ping




From n8gray at caltech.edu.is.my.email.address  Sun Feb 18 08:10:47 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Sat, 17 Feb 2001 23:10:47 -0800
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <96k28b01utj@news2.newsguy.com> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com>
Message-ID: <96nst0$pe9@gap.cco.caltech.edu>

Tim Hochberg wrote:
> [snip discussion of consistency of "if thing in dict:"]
>
> Unfortunately, people tend to think of a dictionary as containing keys
> more than they think of it as containing definitions (values). So the form
> you propose is better from that point of view. It's also probably more
> useful in a practical sense as well. It is kind of distressing that it's
> not consistent with the correspondence that's set up between
> keys<->indices and values<->values though.

Hmm... I tend to think of a dictionary as containing values more than keys. 
 The keys are just there so you can get to the values, right?  Others tend 
to think dicts contain (key, value) pairs.  Everybody has a different 
opinion on this one.  Rather than supplying an arbitrary default, which 
would lead to confusion, how about we let the programmer specify things 
explicitly.

How about we make it:
if key in dict.keys():
and
if value in dict.values():
and maybe even
if (key, value) in dict.items():

That way it's perfectly clear what the programmer intended.  I think 
there's a really good chance of Guido accepting this proposal, and I've 
already written a fully backwards-compatible patch.  Anybody up for a trip 
back in time?

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From aahz at panix.com  Sun Feb 18 17:00:00 2001
From: aahz at panix.com (Aahz Maruch)
Date: 18 Feb 2001 08:00:00 -0800
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com> <96nst0$pe9@gap.cco.caltech.edu>
Message-ID: <96ori0$mq$1@panix3.panix.com>

In article <96nst0$pe9 at gap.cco.caltech.edu>,
Nathaniel Gray  <n8gray at caltech.edu.is.my.email.address> wrote:
>
>Hmm... I tend to think of a dictionary as containing values more than keys. 
> The keys are just there so you can get to the values, right?  Others tend 
>to think dicts contain (key, value) pairs.  Everybody has a different 
>opinion on this one.  Rather than supplying an arbitrary default, which 
>would lead to confusion, how about we let the programmer specify things 
>explicitly.
>
>How about we make it:
>if key in dict.keys():
>and
>if value in dict.values():
>and maybe even
>if (key, value) in dict.items():

The problem with this is absolutely horrendous speed/memory performance
with any large dictionary.  Your code generates a list AND THEN DOES A
LINEAR SEARCH.  No, no, no, no, no, no, no.  We've already got

if dict.has_key(key):

and

if key in dict:

is a thin layer of sugar-coating designed to create greater usage
equivalence between lists and dicts.  I'm not particularly fond of it,
but it does have the advantage of not being forced to remember whether
has_key() has an underscore in it.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From jpet at eskimo.com  Mon Feb 19 04:32:41 2001
From: jpet at eskimo.com (Jeff Petkau)
Date: Mon, 19 Feb 2001 03:32:41 GMT
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com> <96nst0$pe9@gap.cco.caltech.edu> <96ori0$mq$1@panix3.panix.com>
Message-ID: <tt0k6.2546$%M1.357971@paloalto-snr1.gtei.net>

Aahz Maruch <aahz at panix.com> wrote in message
news:96ori0$mq$1 at panix3.panix.com...
> In article <96nst0$pe9 at gap.cco.caltech.edu>,
> Nathaniel Gray  <n8gray at caltech.edu.is.my.email.address> wrote:
> >
> >How about we make it:
> >if key in dict.keys():
> >if value in dict.values():
> >if (key, value) in dict.items():
>
> The problem with this is absolutely horrendous speed/memory performance
> with any large dictionary.  Your code generates a list AND THEN DOES A
> LINEAR SEARCH.  No, no, no, no, no, no, no.  We've already got

Instead of inventing new syntax for inefficent idioms, we
could just make the idioms more efficent, by having values(),
keys(), and items() return proxy objects with __contains__()
and sequential __getitem__() functions.

Alas, this isn't backwards compatible with old code that
expects a real list object.  Possible solutions, in descending
order of sanity:

- make new functions called xkeys(), xvalues(), etc. to
  return the proxies.  Pretty ugly, but consistent with
  xrange() and xreadlines().

- just live with it.

- make 'dict.keys' return the proxy object, which is
  callable.  Make proxy.__call__() return an actual
  list from the proxy so dict.keys() works as before.
  So cute I'm ashamed I even thought of it.

--Jeff Petkau (jpet at eskimo.com),
  currently specializing in poorly thought out solutions.





From hzhu at users.sourceforge.net  Tue Feb 20 04:23:22 2001
From: hzhu at users.sourceforge.net (Huaiyu Zhu)
Date: Tue, 20 Feb 2001 03:23:22 GMT
Subject: Best solution (Re: Discussion about PEP 234: iterators)
References: <96jonj$i0e$1@nntp6.u.washington.edu> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com> <96nst0$pe9@gap.cco.caltech.edu> <96ori0$mq$1@panix3.panix.com> <tt0k6.2546$%M1.357971@paloalto-snr1.gtei.net>
Message-ID: <slrn993opb.4f7.hzhu@rocket.knowledgetrack.com>

On Mon, 19 Feb 2001 03:32:41 GMT, Jeff Petkau <jpet at eskimo.com> wrote:

>- make 'dict.keys' return the proxy object, which is
>  callable.  Make proxy.__call__() return an actual
>  list from the proxy so dict.keys() works as before.
>  So cute I'm ashamed I even thought of it.

What a bright idea!  This is the best solution I've seen so far.  I'm
surprised no one has commented on this yet.  So let me spell it out to
generate more interest.

With this proposal, all of these work as expected

for k in dict.keys:
for k, v in dict.items:
for v in dict.values:

without creating extra list.  And all of these work as expected

if k in dict.keys:
if k, v in dict.items:
if v in dict.values:

Only the last need linear search, which is natural.

There will be no confusion over what 'in' means for dict.  There is no
clutter of code.  And "explicit is better than implicit".

Furthermore, all the old codes work as well, like

a = dict.keys()
v = dict.values()[3]
x, y = dict.items()[2]

Whatever the actual implementation would be, this seems to offer the
cleanest syntax space for iterators and has_item proxies and so on whenever
they become available.  

In the actual implementation, all iterators need to provide three magical
methods: __getitem__, __contains__ and __call__ with the usual sematic
restrictions:

__contains__ return true iff __getitem__ will get it at least once.
__call__ will return a list of objects returned by __getitem__.
__getitem__ raise IndexError at the end.


Huaiyu



From stephen_purcell at yahoo.com  Tue Feb 20 08:57:34 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 20 Feb 2001 08:57:34 +0100
Subject: Best solution (Re: Discussion about PEP 234: iterators)
In-Reply-To: <slrn993opb.4f7.hzhu@rocket.knowledgetrack.com>; from hzhu@users.sourceforge.net on Tue, Feb 20, 2001 at 03:23:22AM +0000
References: <96jonj$i0e$1@nntp6.u.washington.edu> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com> <96nst0$pe9@gap.cco.caltech.edu> <96ori0$mq$1@panix3.panix.com> <tt0k6.2546$%M1.357971@paloalto-snr1.gtei.net> <slrn993opb.4f7.hzhu@rocket.knowledgetrack.com>
Message-ID: <20010220085734.A13854@freedom.puma-ag.com>

Huaiyu Zhu wrote:
> 
> >- make 'dict.keys' return the proxy object, which is
> >  callable.  Make proxy.__call__() return an actual
> >  list from the proxy so dict.keys() works as before.
> >  So cute I'm ashamed I even thought of it.
> 
> What a bright idea!  This is the best solution I've seen so far.  I'm
> surprised no one has commented on this yet.  So let me spell it out to
> generate more interest.
> 
> With this proposal, all of these work as expected
> 
> for k in dict.keys:
> for k, v in dict.items:
> for v in dict.values:
> 
> without creating extra list.  And all of these work as expected
> 
> if k in dict.keys:
> if k, v in dict.items:
> if v in dict.values:


I love it. Where do I vote? :-)

This is *much* nicer looking than the peculiar new ':' syntax.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From s713221 at student.gu.edu.au  Tue Feb 20 12:02:36 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Tue, 20 Feb 2001 21:02:36 +1000
Subject: Best solution (Re: Discussion about PEP 234: iterators)
References: <96jonj$i0e$1@nntp6.u.washington.edu> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com> <96nst0$pe9@gap.cco.caltech.edu> <96ori0$mq$1@panix3.panix.com> <tt0k6.2546$%M1.357971@paloalto-snr1.gtei.net> <slrn993opb.4f7.hzhu@rocket.knowledgetrack.com>
Message-ID: <3A924ECC.6E0A1A@student.gu.edu.au>

Huaiyu Zhu wrote:
> 
> On Mon, 19 Feb 2001 03:32:41 GMT, Jeff Petkau <jpet at eskimo.com> wrote:
> 
> >- make 'dict.keys' return the proxy object, which is
> >  callable.  Make proxy.__call__() return an actual
> >  list from the proxy so dict.keys() works as before.
> >  So cute I'm ashamed I even thought of it.
> 
> What a bright idea!  This is the best solution I've seen so far.  I'm
> surprised no one has commented on this yet.  So let me spell it out to
> generate more interest.
> 
> With this proposal, all of these work as expected
> 
> for k in dict.keys:
> for k, v in dict.items:
> for v in dict.values:
> 
> without creating extra list.  And all of these work as expected
> 
> if k in dict.keys:
> if k, v in dict.items:
> if v in dict.values:
> 
> Only the last need linear search, which is natural.
> 
> There will be no confusion over what 'in' means for dict.  There is no
> clutter of code.  And "explicit is better than implicit".
> 
> Furthermore, all the old codes work as well, like
> 
> a = dict.keys()
> v = dict.values()[3]
> x, y = dict.items()[2]
> 
> Whatever the actual implementation would be, this seems to offer the
> cleanest syntax space for iterators and has_item proxies and so on whenever
> they become available.
> 
> In the actual implementation, all iterators need to provide three magical
> methods: __getitem__, __contains__ and __call__ with the usual sematic
> restrictions:
> 
> __contains__ return true iff __getitem__ will get it at least once.
> __call__ will return a list of objects returned by __getitem__.
> __getitem__ raise IndexError at the end.
> 
> Huaiyu

As long as there aren't any problems unforseen, this gets my vote as
well.

Joal Heagney/AncientHart


From paul.moore at uk.origin-it.com  Tue Feb 20 13:44:36 2001
From: paul.moore at uk.origin-it.com (Paul Moore)
Date: Tue, 20 Feb 2001 13:44:36 +0100
Subject: Best solution (Re: Discussion about PEP 234: iterators)
References: <96jonj$i0e$1@nntp6.u.washington.edu> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com> <96nst0$pe9@gap.cco.caltech.edu> <96ori0$mq$1@panix3.panix.com> <tt0k6.2546$%M1.357971@paloalto-snr1.gtei.net> <slrn993opb.4f7.hzhu@rocket.knowledgetrack.com> <mailman.982655845.6958.python-list@python.org>
Message-ID: <hGaSOhXrHdMq3n3Rc3cNSL1gHMp9@4ax.com>

On Tue, 20 Feb 2001 08:57:34 +0100, Steve Purcell
<stephen_purcell at yahoo.com> wrote:

>Huaiyu Zhu wrote:
[...]
>
>I love it. Where do I vote? :-)

You need to copy the proposal to the PEP author, Ka-Ping Yee
<ping at lfw.org>. He may be reading here, but best to be certain.

BTW, I like the look of this, too, but I suspect that there may be
subtleties which need addressing. The full PEP had a lot of issues
around ensuring backwards compatibility, dealing with user-defined
types and the like, which I don't follow well enough to comment
myself.

Paul.



From n8gray at caltech.edu.is.my.email.address  Sun Feb 18 09:29:38 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Sun, 18 Feb 2001 00:29:38 -0800
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <96k28b01utj@news2.newsguy.com> <3A8DA52A.254849DB@javanet.com>
Message-ID: <96o1h3$qc6@gap.cco.caltech.edu>

Raymond Hettinger wrote:

> I think dictionary iteration should match iterations on other
> objects/types as closely as possible unless performance is impacted.
> 
> for elem in list                # does not automatically sort
> for key in dict                # so this should not automatically sort

What about 
for value in dict          # Should this sort?

> for key:value in dict        # proposed
> for index:item in list        # so this should work
> 
> if elem in list                # this works
> if key in dict                # so this should work

Don't forget
if value in dict

> map( fun, list )            # this works
> map( func, dict )        # so this should work with functions that take a
> key as a parameter

And those that take a value as a parameter too, right?

Don't worry, I'm sure RMS can throw something together with soundex that'll 
be able to tell when you meant keys and when you meant values.

I know I'm being a stickler, but the default of "if thing in dict" is not 
at all clear to me, and it appears I'm not the only one.  Anecdotal 
evidence from Thomas Wouters (I believe) suggests that a significant number 
of programmers expect "thing" to be a value, not a key.  Others expect a 
(key, value) pair.  It just seems there's no clear default.  Oh, and some 
guy named Guido used to agree with me on this one too...

Unlike many of the other folks in opposition to "if key in dict", I'm not 
totally turned off by the "for key:value in dict" syntax and all of its 
variants.  The syntax is easy to learn and it's explicit, if shorthand, for 
the programmer's intent.  Why not use the same syntax for the "if key:value 
in dict" version?  That way nobody'll be happy!  ;^)

(except for me, but I won't complain)

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From tim.one at home.com  Sat Feb 17 06:19:49 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 17 Feb 2001 00:19:49 -0500
Subject: Discussion about PEP 234: iterators
In-Reply-To: <96jonj$i0e$1@nntp6.u.washington.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEPMIPAA.tim.one@home.com>

[Russell E. Owen]
> This is in reaction to the new iterator proposal
> <http://python.sourceforge.net/peps/pep-0234.html>:
>
> My own reaction is "great!" one misgiving: it does not address a very
> common need: to iterate over a dictionary in key-sorted order. Is there
> some variant syntax that could handle this case?

There's always *some* variant syntax that could be forced to work.  For
example,

   for key lambda dict:

<wink>.

> (Better yet, and going out on thin ice here, could this be made the
> default?

No.  dicts can grow very large, and forcing people to pay for a sort when
they don't want it would raise justified howls of protest.  If you want it
sorted, other replies have shown you easy ways to get it sorted.

> I realize that's unlikely and may anger some people to even suggest it,

Not at all.  I only get pissed off when I have to *reply* <wink>.

> but I would like to point out that some languages actually  store
> dictionaries with their keys ordered, so it's at least technically
> feasible).

In fact, the language Guido worked on before Python (ABC) implemented dicts
as balanced binary trees.  So they were always ordered.  He really disliked
it, because the truly common dict operations (insert and lookup) were much
slower than they are using Python's hash tables.  Note too that Python
(unlike ABC) doesn't define a total ordering over all values of all types,
so it's not currently possible "even in theory" to store all dicts in sorted
order ("sorted order" isn't currently well-defined in all cases).

you're-one-two-line-function-away-from-bliss-ly y'rs  - tim




From mertz at gnosis.cx  Fri Feb 16 18:53:13 2001
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Fri, 16 Feb 2001 12:53:13 -0500
Subject: Detecting RegEx across Python versions (fwd)
Message-ID: <JkWj6kKkXkJH092yn@bellatlantic.net>                

I asked earlier:
> I have done the following, but would welcome something more
> elegant
>
>     [...]
>     # kludge to detect a regular expression across python versions
>     elif sys.version[0]=='1' and isinstance(pattern, re.RegexObject):
>         if pattern.match(name):
>             files.append(fname)
>     elif sys.version[0]=='2' and type(pattern)==type(re.compile('')):
>         if pattern.match(name):
>             files.append(fname)
>     elif type(pattern) is StringType:
>         if fnmatch.fnmatch(name, pattern):
>             files.append(fname)

To which the /F-bot generously replied:
| if isinstance(pattern, type("")):
|     if fnmatch.fnmatch(name, pattern):
|         files.append(fname)
| else:
|     if pattern.match(name):
|         files.append(fname)
| ----
| try:
|     if pattern.match(name):
|         files.append(fname)
| except AttributeError:
|     if fnmatch.fnmatch(name, pattern):
|         files.append(fname)
| ----
| if isinstance(pattern, type("")):
|     pattern = re.compile(fnmatch.translate(pattern))
| if pattern.match(name):

Lundh's solutions are more compact, and probably per my request of more
elegant.  But I probably should have been explicit with the enthymeme:
I would like something that is both more elegant AND no less paranoid.

That is, I only want to glob a genuine string, and I only want to regex
a genuine re object.  Being the suspicious type--wholly untrusting of my
callers :-) (since they might be me)--I don't want to let this cause
uncaught exceptions on the wrong type for pattern.  For example (/F's
first example, roughly):

    >>> pattern = open('spam','w')
    >>> if isinstance(pattern, type('')):
    ...     print "Got String"
    ... else:
    ...     if pattern.match('spam'):
    ...         print "Got RegEx"
    ...
    Traceback (innermost last):
      File "<stdin>", line 4, in ?
    AttributeError: match

My own inelegant code would just pass things through without matching,
but also without raising errors.  /F's other versions will raise similar
uncaught errors.

I suppose I could wrap another 'try' around things, and catch those
non-string/non-regex things.  But I'd still really love to be able to
write a clean, version independent, expression of "X is a compiled regex
(and not any other thing)."

Yours, Lulu...



From william_wicker at spectratechnologies.com  Fri Feb 16 19:35:12 2001
From: william_wicker at spectratechnologies.com (William Wicker)
Date: Fri, 16 Feb 2001 18:35:12 -0000
Subject: Win32 COM: Passing NULL as a value
Message-ID: <904A72E79williamwickerspectra@207.126.101.100>

I am trying to work via the COM interface to an object that exposes a 
method something like this:

AnObject.DoSomething(doWhat, toWho)

where toWho is either a dispatch pointer, or NULL, if toWho is not 
significant.

When I try to do this with

obj = win32com.client.Dispatch("AnObject")
obj.DoSomething("kick", None)

or

obj.DoSomething("kick", 0)

I get type mismatch errors back from the COM object.

Is there a way to pass a NULL pointer via win32 COM?

P.S.

I'm trying this with the ActiveState Python 2.0 distribution (python win 
build 202)

Thanks!

    	William.


From MarkH at ActiveState.com  Sat Feb 17 00:58:24 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Fri, 16 Feb 2001 23:58:24 GMT
Subject: Win32 COM: Passing NULL as a value
References: <904A72E79williamwickerspectra@207.126.101.100>
Message-ID: <3A8DBDC4.4040705@ActiveState.com>

William Wicker wrote:


> I am trying to work via the COM interface to an object that exposes a 
> method something like this:
> 
> AnObject.DoSomething(doWhat, toWho)
> 
> where toWho is either a dispatch pointer, or NULL, if toWho is not 
> significant.
> 
> When I try to do this with
> 
> obj = win32com.client.Dispatch("AnObject")
> obj.DoSomething("kick", None)

None is used to pass a variant with VT_NULL.  It works for every object I have come across, so I am afraid I have no idea.

Mark.



From aleaxit at yahoo.com  Sat Feb 17 08:48:23 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 17 Feb 2001 08:48:23 +0100
Subject: Win32 COM: Passing NULL as a value
References: <904A72E79williamwickerspectra@207.126.101.100> <3A8DBDC4.4040705@ActiveState.com>
Message-ID: <96lag10eee@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A8DBDC4.4040705 at ActiveState.com...
> William Wicker wrote:
>
> > I am trying to work via the COM interface to an object that exposes a
> > method something like this:
> >
> > AnObject.DoSomething(doWhat, toWho)
> >
> > where toWho is either a dispatch pointer, or NULL, if toWho is not
> > significant.
> >
> > When I try to do this with
> >
> > obj = win32com.client.Dispatch("AnObject")
> > obj.DoSomething("kick", None)
>
> None is used to pass a variant with VT_NULL.  It works for every object I
have come across, so I am afraid I have no idea.

If the AnObject used canonical API's to handle its arguments
(VariantChangeType, directly or under the covers) it should
indeed work -- a VariantChangeType from VT_NULL to VT_DISPATCH
should succeed and give a VARIANT with pdispVal of 0.  So, I
guess from the symptoms described that AnObject is not being
'canonical' in its handling of arguments.

Is there a way to prepare a VARIANT directly in Python, with
VT_DISPATCH and pdispVal==0?  When met with such issues
in the past I used a little helper, an in-process COM server I
wrote that supplies a few objects for explicit fine-grained
issues of Automation ("prepare a variant with exactly this vt
tag-part and exactly this bitpattern in its value-union part"
being a good example).  But Pythoncom looks like it's powerful
enough to handle it directly if one digs deep enough, rather
than needing an auxiliary C++ coded library of objects...?


Alex





From william_wicker at spectratechnologies.com  Sat Feb 17 21:48:37 2001
From: william_wicker at spectratechnologies.com (William Wicker)
Date: Sat, 17 Feb 2001 20:48:37 -0000
Subject: Win32 COM: Passing NULL as a value
References: <904A72E79williamwickerspectra@207.126.101.100> <3A8DBDC4.4040705@ActiveState.com> <96lag10eee@news2.newsguy.com>
Message-ID: <904B97719williamwickerspectra@207.126.101.100>

aleaxit at yahoo.com (Alex Martelli) wrote in
<96lag10eee at news2.newsguy.com>: 

>"Mark Hammond" <MarkH at ActiveState.com> wrote in message
>news:3A8DBDC4.4040705 at ActiveState.com...
>> William Wicker wrote:
>>
>> > I am trying to work via the COM interface to an object that exposes
>> > a method something like this:
>> >
>> > AnObject.DoSomething(doWhat, toWho)
>> >
>> > where toWho is either a dispatch pointer, or NULL, if toWho is not
>> > significant.
>> >
>> > When I try to do this with
>> >
>> > obj = win32com.client.Dispatch("AnObject")
>> > obj.DoSomething("kick", None)
>>
>> None is used to pass a variant with VT_NULL.  It works for every
>> object I 
>have come across, so I am afraid I have no idea.
>
>If the AnObject used canonical API's to handle its arguments
>(VariantChangeType, directly or under the covers) it should
>indeed work -- a VariantChangeType from VT_NULL to VT_DISPATCH
>should succeed and give a VARIANT with pdispVal of 0.  So, I
>guess from the symptoms described that AnObject is not being
>'canonical' in its handling of arguments.
>

Urgh.

I wish I were more suprised by this answer. Thanks to you and Mark.

    	William.


From MarkH at ActiveState.com  Sun Feb 18 23:20:49 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 18 Feb 2001 22:20:49 GMT
Subject: Win32 COM: Passing NULL as a value
References: <904A72E79williamwickerspectra@207.126.101.100> <3A8DBDC4.4040705@ActiveState.com> <96lag10eee@news2.newsguy.com>
Message-ID: <3A9049E8.8050002@ActiveState.com>

Alex Martelli wrote:


> Is there a way to prepare a VARIANT directly in Python, with
> VT_DISPATCH and pdispVal==0?  

The only way is to call InvokeTypes() with the correct type tuples 
setup.  Nothing higher level has been created, but I would be willing to 
discuss how such a feature should be added...

Mark.



From bridgman at wyeth.gsfc.nasa.gov  Fri Feb 16 19:49:26 2001
From: bridgman at wyeth.gsfc.nasa.gov (Tom Bridgman)
Date: Fri, 16 Feb 2001 13:49:26 -0500
Subject: Converting an instance to a subclass?
Message-ID: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov>

Here's my problem (and we're currently still at Python v1.5.2 on IRIX,
if that matters).

I've defined a set of classes for manipulating some data, say class A.  

class A:
    def __init__(self,a,b,c)
        ...

and I have a number of methods that return me instances or lists of
instances of this class.  These are part of my general library of
classes for this project.

However, I need to write a utility which will define methods that are
only needed by the utility.  I really don't want these methods to be
'permanent' members of the class so I define them in a subclass as part
of the utility.

Class B(A):
   def Cleanup(self,x,y,z):
       ...

Is there a way I can cast the instances of class A into instances of
class B so I can use the additional methods?  I can't find anything
about it in "Programming Python" but then I'm not quite sure where to
look either.

For some reason I suspect I will be embarassed when I find out how
simple this might be. :^)

Thanks,
Tom
--
Dr. William T."Tom" Bridgman           Scientific Visualization Studio
Global Science & Technology, Inc.      NASA/Goddard Space Flight Center
Email: bridgman at wyeth.gsfc.nasa.gov    Code 935
Phone: 301-286-1346                    Greenbelt, MD 20771
FAX:   TBD                             http://svs.gsfc.nasa.gov/


From dsh8290 at rit.edu  Fri Feb 16 20:42:33 2001
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 16 Feb 2001 14:42:33 -0500
Subject: Converting an instance to a subclass?
In-Reply-To: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov>; from bridgman@wyeth.gsfc.nasa.gov on Fri, Feb 16, 2001 at 01:49:26PM -0500
References: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov>
Message-ID: <20010216144233.C4090@harmony.cs.rit.edu>

On Fri, Feb 16, 2001 at 01:49:26PM -0500, Tom Bridgman wrote:
[snip] 
| and I have a number of methods that return me instances or lists of
| instances of this class.  These are part of my general library of
| classes for this project.
| 
| However, I need to write a utility which will define methods that are
| only needed by the utility.  I really don't want these methods to be
| 'permanent' members of the class so I define them in a subclass as part
| of the utility.
| 
| Class B(A):
|    def Cleanup(self,x,y,z):
|        ...
| 

This is a good way to add functionality that isn't always needed --
you can use instances of 'A' when it's not needed, and instances of
'B' when it is.

| Is there a way I can cast the instances of class A into instances of
| class B so I can use the additional methods?  I can't find anything
| about it in "Programming Python" but then I'm not quite sure where to
| look either.

Since you use the word 'cast', I take it you have some C, C++ or Java
experience.  Those lanugages are statically typed, so if you declared
having one type, but really have another, you must explicitly tell the
compiler with a cast.  Python is dynamically typed, so if you really
have a "different" type (it can't really be different since you didn't
declare it in the first place, but) you have it and python is happy.

What I suspect is happening is that you are creating instances of A
when you really want B.

>>> class A : pass
...
>>> class B( A ) : pass
...
>>> obj1 = A()
>>> obj2 = B()
>>> obj1.__class__
<class __main__.A at 007A8934>
>>> obj2.__class__
<class __main__.B at 007AA9A4>
>>>

In this example, obj2 is an instance of class "B".  Since "B" is a
subclass of "A", obj2 is also an instance of class "A".  You can use
obj2 whereever an instance of A or B is expected.  On the other hand,
obj1 is an instance of class "A", and won't work where an instance of
class B is expected.  There is no way automatically to turn obj1 into
an instance of class B.  There are a couple of ways to work around
though.

1)  modify the factory functions to return instances of "B" instead.
    Since it is a subclass, all existing code will work.

2)  Make a conversion that will create a new instance of class "B" and
    init all data members based on an instance of class "A"

3)  Play some obscure tricks using python dynamism to add those
    functions to obj1 even though they don't belong there


#1 may be the easiest, but then you might as well just put the
functions in class "A" and only use them when you need to.

#2 could be done in B's __init__ function :

   class B( A ) :
       def __init__( self , instance_of_A ) :
           self.value = instance_of_A.value
           # ...

   Then when you want an instance of B, you can create it from the
   given instance of A.

#3 isn't good since it is obscure and will be difficult to maintain.

HTH,
-D




From bridgman at wyeth.gsfc.nasa.gov  Fri Feb 16 21:09:07 2001
From: bridgman at wyeth.gsfc.nasa.gov (Tom Bridgman)
Date: Fri, 16 Feb 2001 15:09:07 -0500
Subject: Converting an instance to a subclass?
References: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov> <mailman.982352599.31822.python-list@python.org>
Message-ID: <3A8D88E3.9FF4CA4B@wyeth.gsfc.nasa.gov>

I'm seriously considering your option #2.

These utility scripts are doing things which are really non-standard for
the process I'm developing.  They just exist solely for the purpose of
converting an old data format over to a new one, installing new
defaults, checking that support files are in the right locations, moving
them if they're not, etc.  The methods defined for this should never be
used once we switch over to the new system so I don't want them in my
standard class library to confuse some future maintainer of the code.

Thanks for your input,
Tom

D-Man wrote:
> 
> On Fri, Feb 16, 2001 at 01:49:26PM -0500, Tom Bridgman wrote:
> [snip]
> | and I have a number of methods that return me instances or lists of
> | instances of this class.  These are part of my general library of
> | classes for this project.
> |
> | However, I need to write a utility which will define methods that are
> | only needed by the utility.  I really don't want these methods to be
> | 'permanent' members of the class so I define them in a subclass as part
> | of the utility.
> |
> | Class B(A):
> |    def Cleanup(self,x,y,z):
> |        ...
> |
> 
> This is a good way to add functionality that isn't always needed --
> you can use instances of 'A' when it's not needed, and instances of
> 'B' when it is.
> 
> | Is there a way I can cast the instances of class A into instances of
> | class B so I can use the additional methods?  I can't find anything
> | about it in "Programming Python" but then I'm not quite sure where to
> | look either.
> 
> Since you use the word 'cast', I take it you have some C, C++ or Java
> experience.  Those lanugages are statically typed, so if you declared
> having one type, but really have another, you must explicitly tell the
> compiler with a cast.  Python is dynamically typed, so if you really
> have a "different" type (it can't really be different since you didn't
> declare it in the first place, but) you have it and python is happy.
> 
> What I suspect is happening is that you are creating instances of A
> when you really want B.
> 
> >>> class A : pass
> ...
> >>> class B( A ) : pass
> ...
> >>> obj1 = A()
> >>> obj2 = B()
> >>> obj1.__class__
> <class __main__.A at 007A8934>
> >>> obj2.__class__
> <class __main__.B at 007AA9A4>
> >>>
> 
> In this example, obj2 is an instance of class "B".  Since "B" is a
> subclass of "A", obj2 is also an instance of class "A".  You can use
> obj2 whereever an instance of A or B is expected.  On the other hand,
> obj1 is an instance of class "A", and won't work where an instance of
> class B is expected.  There is no way automatically to turn obj1 into
> an instance of class B.  There are a couple of ways to work around
> though.
> 
> 1)  modify the factory functions to return instances of "B" instead.
>     Since it is a subclass, all existing code will work.
> 
> 2)  Make a conversion that will create a new instance of class "B" and
>     init all data members based on an instance of class "A"
> 
> 3)  Play some obscure tricks using python dynamism to add those
>     functions to obj1 even though they don't belong there
> 
> #1 may be the easiest, but then you might as well just put the
> functions in class "A" and only use them when you need to.
> 
> #2 could be done in B's __init__ function :
> 
>    class B( A ) :
>        def __init__( self , instance_of_A ) :
>            self.value = instance_of_A.value
>            # ...
> 
>    Then when you want an instance of B, you can create it from the
>    given instance of A.
> 
> #3 isn't good since it is obscure and will be difficult to maintain.
> 
> HTH,
> -D

--
Dr. William T."Tom" Bridgman           Scientific Visualization Studio
Global Science & Technology, Inc.      NASA/Goddard Space Flight Center
Email: bridgman at wyeth.gsfc.nasa.gov    Code 935
Phone: 301-286-1346                    Greenbelt, MD 20771
FAX:   TBD                             http://svs.gsfc.nasa.gov/


From oconnor at bioreason.com  Fri Feb 16 21:48:34 2001
From: oconnor at bioreason.com (Jay O'Connor)
Date: Fri, 16 Feb 2001 13:48:34 -0700
Subject: Converting an instance to a subclass?
References: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov> <mailman.982352599.31822.python-list@python.org>
Message-ID: <3A8D9222.ED817082@bioreason.com>


D-Man wrote:

> #2 could be done in B's __init__ function :
> 
>    class B( A ) :
>        def __init__( self , instance_of_A ) :
>            self.value = instance_of_A.value
>            # ...
> 
>    Then when you want an instance of B, you can create it from the
>    given instance of A.
> 

For conveneience you could add a converter in A

class A:

	def asB (self):
		b = B(self)
		return b


and in B you could do

class B:
	def __init__(self, instanceOfA):
		self.__dict__ = instanceOfA.__dict__

The only problem with adding the covenience method to A is that it makes
A aware of it's subclasses hierarchy, which may not be desireable.

Take care,
Jay O'Connor
joconnor at cybermesa.com


From oconnor at bioreason.com  Fri Feb 16 21:43:26 2001
From: oconnor at bioreason.com (Jay O'Connor)
Date: Fri, 16 Feb 2001 13:43:26 -0700
Subject: Converting an instance to a subclass?
References: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov>
Message-ID: <3A8D90EE.47B1BA35@bioreason.com>


Tom Bridgman wrote:
...
 
> However, I need to write a utility which will define methods that are
> only needed by the utility.  I really don't want these methods to be
> 'permanent' members of the class so I define them in a subclass as part
> of the utility.
> 
> Class B(A):
>    def Cleanup(self,x,y,z):
>        ...
> 
> Is there a way I can cast the instances of class A into instances of
> class B so I can use the additional methods?  I can't find anything
> about it in "Programming Python" but then I'm not quite sure where to
> look either.


One technique you can use is to add new instance methods to the objects
you wish to treat differently

------

#!/usr/bin/python

import new

class MyClass():

	def __init__(self):
		x = 100



# Seperate function, not part of MyClass
# but expects 'self' and references variables created in
MyClass.__init__()
def extraFunction(self):
	print "x = " + `self.x`

object = MyClass()

object.extra = new.instancemethod (extraFunction, object, MyClass)

object.extra()
------

This will print "x = 100"


This only adds the new method as an instance method to that one object. 
It does not affect other instances of MyClass, or MyClass itself

So just declare your extra methods seperate from the class and add them
as needed to the appropriate objects

Take care,
Jay O'Connor
joconnor at cybermesa.com


From tjreedy at udel.edu  Fri Feb 16 23:46:58 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Fri, 16 Feb 2001 17:46:58 -0500
Subject: Converting an instance to a subclass?
References: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov>
Message-ID: <96ka96$iqd$1@news.udel.edu>

> Is there a way I can cast the instances of class A into instances of
> class B so I can use the additional methods

Is this what you want (with 1.5.2)?

>>> class a: pass
>>> class b(a): pass
>>> i = a()
>>> i.__class__
<class __main__.a at 7934f0>
>>> i.__class__ = b
>>> i.__class__
<class __main__.b at 7932b0>

> I can't find anything  about it in "Programming Python"

I believe that __class__ became writeable after that was written -- or it
may have been too esoteric (and dangerous) to include.

Terry J. Reedy





From bridgman at wyeth.gsfc.nasa.gov  Wed Feb 21 15:06:15 2001
From: bridgman at wyeth.gsfc.nasa.gov (Tom Bridgman)
Date: Wed, 21 Feb 2001 09:06:15 -0500
Subject: Converting an instance to a subclass?
References: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov> <96ka96$iqd$1@news.udel.edu>
Message-ID: <3A93CB57.795226AE@wyeth.gsfc.nasa.gov>

Terry et al.,

This is a really slick trick and looks like it does what I need.

Thanks to all who suggested it.  I was right.  I was *really* simple. 

Tom

Terry Reedy wrote:
> 
> > Is there a way I can cast the instances of class A into instances of
> > class B so I can use the additional methods
> 
> Is this what you want (with 1.5.2)?
> 
> >>> class a: pass
> >>> class b(a): pass
> >>> i = a()
> >>> i.__class__
> <class __main__.a at 7934f0>
> >>> i.__class__ = b
> >>> i.__class__
> <class __main__.b at 7932b0>
> 
> > I can't find anything  about it in "Programming Python"
> 
> I believe that __class__ became writeable after that was written -- or it
> may have been too esoteric (and dangerous) to include.
> 
> Terry J. Reedy

--
Dr. William T."Tom" Bridgman           Scientific Visualization Studio
Global Science & Technology, Inc.      NASA/Goddard Space Flight Center
Email: bridgman at wyeth.gsfc.nasa.gov    Code 935
Phone: 301-286-1346                    Greenbelt, MD 20771
FAX:   TBD                             http://svs.gsfc.nasa.gov/


From dan at eevolved.com  Fri Feb 16 21:05:45 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 20:05:45 GMT
Subject: TCS.bsddb
Message-ID: <tKfj6.263574$JT5.9155320@news20.bellglobal.com>

I have just downloaded, compiled and installed TCS.bsddb, robin dunn's 
bsddb1.2 for sleepycat software's berkley db version 2.7.7.

I am getting an error on a simple
---
from TCS.bsddb import dbShelve

db = dbShelve.open("test.db")
---
it raises the exception

Traceback (most recent call last):
  File "./shelve.py", line 5, in ?
    db = dbShelve.open("test.db")
  File "/usr/local/lib/python2.0/site-packages/TCS/bsddb/dbShelve.py", line 
165, in open
    d.open(filename, filetype, flags, mode)
  File "/usr/local/lib/python2.0/site-packages/TCS/bsddb/db.py", line 638, 
in open
    val = apply(dbc.DB_open,(self.this,arg0,)+args)
db.error: (22, 'Invalid argument') 

I checked, and the arguments are fine. Anybody have this error before?

Dan



From janne at oops.nnets.fi  Fri Feb 16 23:48:27 2001
From: janne at oops.nnets.fi (Janne Sinkkonen)
Date: 17 Feb 2001 00:48:27 +0200
Subject: TCS.bsddb
References: <tKfj6.263574$JT5.9155320@news20.bellglobal.com>
Message-ID: <m3hf1uz1l0.fsf@kinos.nnets.fi>

Dan Parisien <dan at eevolved.com> writes:

> I have just downloaded, compiled and installed TCS.bsddb, robin dunn's 
> bsddb1.2 for sleepycat software's berkley db version 2.7.7.
> 
> I am getting an error on a simple
> ---
> from TCS.bsddb import dbShelve
> 
> db = dbShelve.open("test.db")

Mine works, with Python 2.1a2 and with 1.5.x, but my bsddb is an older 2.x

-- 
Janne


From janne at oops.nnets.fi  Fri Feb 16 23:58:01 2001
From: janne at oops.nnets.fi (Janne Sinkkonen)
Date: 17 Feb 2001 00:58:01 +0200
Subject: TCS.bsddb
References: <tKfj6.263574$JT5.9155320@news20.bellglobal.com> <m3hf1uz1l0.fsf@kinos.nnets.fi>
Message-ID: <m3d7ciz152.fsf@kinos.nnets.fi>

Janne Sinkkonen <janne at oops.nnets.fi> writes:

> Mine works, with Python 2.1a2 and with 1.5.x, but my bsddb is an older 2.x

Sorry, I tried to say that the underlying Berkeley database is an
older 2.x version (2.1.3 probably). 

-- 
Janne


From dan at eevolved.com  Sat Feb 17 00:36:09 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 23:36:09 GMT
Subject: TCS.bsddb
References: <tKfj6.263574$JT5.9155320@news20.bellglobal.com> <m3hf1uz1l0.fsf@kinos.nnets.fi> <m3d7ciz152.fsf@kinos.nnets.fi>
Message-ID: <JPij6.137201$Pm2.2466565@news20.bellglobal.com>

Janne Sinkkonen wrote:

> Janne Sinkkonen <janne at oops.nnets.fi> writes:
> 
> > Mine works, with Python 2.1a2 and with 1.5.x, but my bsddb is an older
> > 2.x
> 
> Sorry, I tried to say that the underlying Berkeley database is an
> older 2.x version (2.1.3 probably).
> 
My problem is that I'm running Mandrake 7.2 and it fills /lib with 
incorrectly (I'm assuming) versioned libs. If I delete certain symbolic 
links my tests will work... Unfortunately as soon as I run ldconfig, my 
changes are wiped

Oh, and cvs depends on libdb.so.2 (one of the symbolic links I deleted). 
cvs segfaulted in the middle of a commit, now I'm locked out :)

Not a good day today...
Dan



From othello at javanet.com  Fri Feb 16 22:42:35 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Fri, 16 Feb 2001 16:42:35 -0500
Subject: Pipes in PythonWin
Message-ID: <3A8D9ECA.C0BB3C38@javanet.com>

Any idea how to get the pipe functions to work from within
PythonWin?

At the the command line, this script works:
   print os.popen('cal').read()

Within PythonWin, I get:
  WindowsError: [Errno 2] The system cannot find the file specified

Is there some enabling step or work around?

Thanks,

Raymond



From MarkH at ActiveState.com  Sat Feb 17 01:01:24 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 17 Feb 2001 00:01:24 GMT
Subject: Pipes in PythonWin
References: <3A8D9ECA.C0BB3C38@javanet.com>
Message-ID: <3A8DBE77.9070107@ActiveState.com>

Raymond Hettinger wrote:

> Any idea how to get the pipe functions to work from within
> PythonWin?
> 
> At the the command line, this script works:
>    print os.popen('cal').read()
> 
> Within PythonWin, I get:
>   WindowsError: [Errno 2] The system cannot find the file specified
> 
> Is there some enabling step or work around?

This is a bug.  To work around it copy "w9xpopen.exe" to the same 
directory as pythonwin.exe

Mark.
-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From othello at javanet.com  Mon Feb 19 01:30:48 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Sun, 18 Feb 2001 19:30:48 -0500
Subject: Pipes in PythonWin
References: <3A8D9ECA.C0BB3C38@javanet.com> <3A8DBE77.9070107@ActiveState.com>
Message-ID: <3A906938.15618A63@javanet.com>

Mark Hammond wrote:

> Raymond Hettinger wrote:
>
> > Any idea how to get the pipe functions to work from within
> > PythonWin?
> >
> > At the the command line, this script works:
> >    print os.popen('cal').read()
> >
> > Within PythonWin, I get:
> >   WindowsError: [Errno 2] The system cannot find the file specified
> >
> > Is there some enabling step or work around?
>
> This is a bug.  To work around it copy "w9xpopen.exe" to the same
> directory as pythonwin.exe
>
> Mark.

It works and I'm back in business!

Thanks,

Raymond



From lreilly at cs.strath.ac.uk  Fri Feb 16 23:32:00 2001
From: lreilly at cs.strath.ac.uk (Lee Reilly CS1997)
Date: Fri, 16 Feb 2001 22:32:00 +0000
Subject: coding standards - commenting / documentation
Message-ID: <3A8DAA60.D37A2C18@cs.strath.ac.uk>

Hi there,

Can anyone tell me if there is a standard template or recongnised
standard for documenting (especially commenting) python code. Or, if
anybody had a good, structured way of doing this maybe you'd like to
send me a sample?

Thanks very much in advance,

Lee


From danielk at aracnet.com  Sat Feb 17 04:26:27 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Fri, 16 Feb 2001 19:26:27 -0800
Subject: coding standards - commenting / documentation
References: <3A8DAA60.D37A2C18@cs.strath.ac.uk>
Message-ID: <sprr8t0ersljfa258o1r1mshqcu1cedlon@4ax.com>

On Fri, 16 Feb 2001 22:32:00 +0000, Lee Reilly CS1997 <lreilly at cs.strath.ac.uk>
wrote:
>Can anyone tell me if there is a standard template or recongnised
>standard for documenting (especially commenting) python code.

http://www.python.org/doc/essays/styleguide.html


From rwil at americasm01.nt.com  Fri Feb 16 23:32:12 2001
From: rwil at americasm01.nt.com (Willison, Rick [WDLN2:2X38:EXCH])
Date: Fri, 16 Feb 2001 17:32:12 -0500
Subject: Unix Python Installation Troubles
Message-ID: <3A8DAA6C.24E19E6D@americasm01.nt.com>


I've inherited a bunch of Python scripts, and I'm trying to install
the freeware on my Unix machine. It looks like _tkinter can't be found
during the
make. It thinks that I'm running an old tk/tcl, when in fact I have
tk-8.0 ...!

I've included the Setup file, and the spot where the make blows up.

Can you help??

...Rick
-------------- next part --------------
A non-text attachment was scrubbed...
Name: python_install.log
Type: application/x-unknown-content-type-txtfile
Size: 702 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010216/8b1a8cbf/attachment.bin>
-------------- next part --------------
#-*- makefile -*-
# The file Setup is used by the makesetup script to construct the files
# Makefile and config.c, from Makefile.pre and config.c.in,
# respectively.  The file Setup itself is initially copied from
# Setup.in; once it exists it will not be overwritten, so you can edit
# Setup to your heart's content.  Note that Makefile.pre is created
# from Makefile.pre.in by the toplevel configure script.

# (VPATH notes: Setup and Makefile.pre are in the build directory, as
# are Makefile and config.c; the *.in files are in the source
# directory.)

# Each line in this file describes one or more optional modules.
# Comment out lines to suppress modules.
# Lines have the following structure:
#
# <module> ... [<sourcefile> ...] [<cpparg> ...] [<library> ...]
#
# <sourcefile> is anything ending in .c (.C, .cc, .c++ are C++ files)
# <cpparg> is anything starting with -I, -D, -U or -C
# <library> is anything ending in .a or beginning with -l or -L
# <module> is anything else but should be a valid Python
# identifier (letters, digits, underscores, beginning with non-digit)
#
# (As the makesetup script changes, it may recognize some other
# arguments as well, e.g. *.so and *.sl as libraries.  See the big
# case statement in the makesetup script.)
#
# Lines can also have the form
#
# <name> = <value>
#
# which defines a Make variable definition inserted into Makefile.in
#
# Finally, if a line contains just the word "*shared*" (without the
# quotes but with the stars), then the following modules will not be
# included in the config.c file, nor in the list of objects to be
# added to the library archive, and their linker options won't be
# added to the linker options, but rules to create their .o files and
# their shared libraries will still be added to the Makefile, and
# their names will be collected in the Make variable SHAREDMODS.  This
# is used to build modules as shared libraries.  (They can be
# installed using "make sharedinstall", which is implied by the
# toplevel "make install" target.)  (For compatibility,
# *noconfig* has the same effect as *shared*.)
#
# In addition, *static* reverses this effect (negating a previous
# *shared* line).

# NOTE: As a standard policy, as many modules as can be supported by a
# platform should be present.  The distribution comes with all modules
# enabled that are supported by most platforms and don't require you
# to ftp sources from elsewhere.


# Some special rules to define PYTHONPATH.
# Edit the definitions below to indicate which options you are using.
# Don't add any whitespace or comments!

# Directories where library files get installed.
# DESTLIB is for Python modules; MACHDESTLIB for shared libraries.
DESTLIB=$(LIBDEST)
MACHDESTLIB=$(BINLIBDEST)

# NOTE: all the paths are now relative to the prefix that is computed
# at run time!

# Standard path -- don't edit.
# No leading colon since this is the first entry.
# Empty since this is now just the runtime prefix.
DESTPATH=

# Site specific path components -- should begin with : if non-empty
SITEPATH=

# Standard path components for test modules
TESTPATH=

# Path components for machine- or system-dependent modules and shared libraries
MACHDEPPATH=:plat-$(MACHDEP)

COREPYTHONPATH=$(DESTPATH)$(SITEPATH)$(TESTPATH)$(MACHDEPPATH)$(STDWINPATH)$(TKPATH)
PYTHONPATH=$(COREPYTHONPATH)


# The modules listed here can't be built as shared libraries for
# various reasons; therefore they are listed here instead of in the
# normal order.

# Some modules that are normally always on:

regex regexmodule.c regexpr.c	# Regular expressions, GNU Emacs style
pcre pcremodule.c pypcre.c	# Regular expressions, Perl style (for re.py)
posix posixmodule.c		# posix (UNIX) system calls
signal signalmodule.c		# signal(2)

# The SGI specific GL module:

GLHACK=-Dclear=__GLclear
#gl glmodule.c cgensupport.c -I$(srcdir) $(GLHACK) -lgl -lX11

# The thread module is now automatically enabled, see Setup.thread.

# Pure module.  Cannot be linked dynamically.
# -DWITH_QUANTIFY, -DWITH_PURIFY, or -DWITH_ALL_PURE
#WHICH_PURE_PRODUCTS=-DWITH_ALL_PURE
#PURE_INCLS=-I/usr/local/include
#PURE_STUBLIBS=-L/usr/local/lib -lpurify_stubs -lquantify_stubs
#pure puremodule.c $(WHICH_PURE_PRODUCTS) $(PURE_INCLS) $(PURE_STUBLIBS)

# Uncommenting the following line tells makesetup that all following
# modules are to be built as shared libraries (see above for more
# detail; also note that *static* reverses this effect):

#*shared*

# GNU readline.  Unlike previous Python incarnations, GNU readline is
# now incorporated in an optional module, configured in the Setup file
# instead of by a configure script switch.  You may have to insert a
# -L option pointing to the directory where libreadline.* lives,
# and you may have to change -ltermcap to -ltermlib or perhaps remove
# it, depending on your system -- see the GNU readline instructions.
# It's okay for this to be a shared library, too.

#readline readline.c -lreadline -ltermcap


# Modules that should always be present (non UNIX dependent):

array arraymodule.c	# array objects
cmath cmathmodule.c # -lm # complex math library functions
math mathmodule.c # -lm # math library functions, e.g. sin()
strop stropmodule.c	# fast string operations implemented in C
struct structmodule.c	# binary structure packing/unpacking
time timemodule.c # -lm # time operations and variables
operator operator.c	# operator.add() and similar goodies

#_locale _localemodule.c  # access to ISO C locale support


# Modules with some UNIX dependencies -- on by default:
# (If you have a really backward UNIX, select and socket may not be
# supported...)

fcntl fcntlmodule.c	# fcntl(2) and ioctl(2)
pwd pwdmodule.c		# pwd(3) 
grp grpmodule.c		# grp(3)
select selectmodule.c	# select(2); not on ancient System V
socket socketmodule.c	# socket(2); not on ancient System V
#_socket socketmodule.c	# socket(2); use this one for BeOS sockets
errno errnomodule.c	# posix (UNIX) errno values

# The crypt module is now disabled by default because it breaks builds
# on many systems (where -lcrypt is needed), e.g. Linux (I believe).
#crypt cryptmodule.c # -lcrypt	# crypt(3); needs -lcrypt on some systems


# Some more UNIX dependent modules -- off by default, since these
# are not supported by all UNIX systems:

#nis nismodule.c 	# Sun yellow pages -- not everywhere
#termios termios.c	# Steen Lumholt's termios module
#resource resource.c	# Jeremy Hylton's rlimit interface


# Multimedia modules -- off by default.
# These don't work for 64-bit platforms!!!
# These represent audio samples or images as strings:

#audioop audioop.c	# Operations on audio samples
#imageop imageop.c	# Operations on images
#rgbimg rgbimgmodule.c   # Read SGI RGB image files (but coded portably)


# The stdwin module provides a simple, portable (between X11 and Mac)
# windowing interface.  You need to ftp the STDWIN library, e.g. from
# ftp://ftp.cwi.nl/pub/stdwin.  (If you get it elsewhere, be sure to
# get version 1.0 or higher!)  The STDWIN variable must point to the
# STDWIN toplevel directory.

# Uncomment and edit as needed:
#STDWIN=/ufs/guido/src/stdwin

# Uncomment these lines:
#STDWINPATH=:lib-stdwin
#LIBTEXTEDIT=$(STDWIN)/$(MACHDEP)/Packs/textedit/libtextedit.a
#LIBX11STDWIN=$(STDWIN)/$(MACHDEP)/Ports/x11/libstdwin.a
#stdwin stdwinmodule.c -I$(STDWIN)/H $(LIBTEXTEDIT) $(LIBX11STDWIN) -lX11

# Use this instead of the last two lines above for alphanumeric stdwin:
#LIBALFASTDWIN=$(STDWIN)/$(MACHDEP)/Ports/alfa/libstdwin.a
#stdwin stdwinmodule.c -I$(STDWIN)/H $(LIBTEXTEDIT) $(LIBALFASTDWIN) -ltermcap


# The md5 module implements the RSA Data Security, Inc. MD5
# Message-Digest Algorithm, described in RFC 1321.  The necessary files
# md5c.c and md5.h are included here.

md5 md5module.c md5c.c


# The sha module implements the SHA checksum algorithm.
# (NIST's Secure Hash Algorithm.)
sha shamodule.c


# The mpz module interfaces to the GNU Multiple Precision library.
# You need to ftp the GNU MP library.  
# The GMP variable must point to the GMP source directory.
# This was originally written and tested against GMP 1.2 and 1.3.2.
# It has been modified by Rob Hooft to work with 2.0.2 as well, but I
# haven't tested it recently.

# A compatible MP library unencombered by the GPL also exists.  It was
# posted to comp.sources.misc in volume 40 and is widely available from
# FTP archive sites. One URL for it is:
# ftp://gatekeeper.dec.com/.b/usenet/comp.sources.misc/volume40/fgmp/part01.Z

#GMP=/ufs/guido/src/gmp
#mpz mpzmodule.c -I$(GMP) $(GMP)/libgmp.a


# SGI IRIX specific modules -- off by default.

# These module work on any SGI machine:

# *** gl must be enabled higher up in this file ***
#fm fmmodule.c $(GLHACK) -lfm -lgl		# Font Manager
#sgi sgimodule.c			# sgi.nap() and a few more

# This module requires the header file
# /usr/people/4Dgifts/iristools/include/izoom.h:
#imgfile imgfile.c -limage -lgutil -lgl -lm	# Image Processing Utilities


# These modules require the Multimedia Development Option (I think):

#al almodule.c -laudio			# Audio Library
#cd cdmodule.c -lcdaudio -lds -lmediad	# CD Audio Library
#cl clmodule.c -lcl -lawareaudio	# Compression Library
#sv svmodule.c yuvconvert.c -lsvideo -lXext -lX11	# Starter Video


# The FORMS library, by Mark Overmars, implements user interface
# components such as dialogs and buttons using SGI's GL and FM
# libraries.  You must ftp the FORMS library separately from
# ftp://ftp.cs.ruu.nl/pub/SGI/FORMS.  It was tested with FORMS 2.2a.
# NOTE: if you want to be able to use FORMS and curses simultaneously
# (or both link them statically into the same binary), you must
# compile all of FORMS with the cc option "-Dclear=__GLclear".

# The FORMS variable must point to the FORMS subdirectory of the forms
# toplevel directory:

#FORMS=/ufs/guido/src/forms/FORMS
#fl flmodule.c -I$(FORMS) $(GLHACK) $(FORMS)/libforms.a -lfm -lgl


# SunOS specific modules -- off by default:

#sunaudiodev sunaudiodev.c


# George Neville-Neil's timing module:

#timing timingmodule.c


# The _tkinter module.
#
# The TKPATH variable is always enabled, to save you the effort.
TKPATH=:lib-tk

# The command for _tkinter is long and site specific.  Please
# uncomment and/or edit those parts as indicated.  If you don't have a
# specific extension (e.g. Tix or BLT), leave the corresponding line
# commented out.  (Leave the trailing backslashes in!  If you
# experience strange errors, you may want to join all uncommented
# lines and remove the backslashes -- the backslash interpretation is
# done by the shell's "read" command and it may not be implemented on
# every system.

# *** Always uncomment this (leave the leading underscore in!):
_tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
# *** Uncomment and edit to reflect where your Tcl/Tk headers are:
	-I/opt/corp/contrib/tk-8.0/lib/sparc-solaris2.5.1/include \
# *** Uncomment and edit to reflect where your X11 header files are:
#	-I/usr/X11R6/include \
# *** Or uncomment this for Solaris:
	-I/usr/openwin/include \
# *** Uncomment and edit for Tix extension only:
#	-DWITH_TIX -ltix4.1.8.0 \
# *** Uncomment and edit for BLT extension only:
#	-DWITH_BLT -I/usr/local/blt/blt8.0-unoff/include -lBLT8.0 \
# *** Uncomment and edit for PIL (TkImaging) extension only:
#	-DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
# *** Uncomment and edit for TOGL extension only:
#	-DWITH_TOGL togl.c \
# *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
	-L/opt/corp/contrib/tk-8.0/lib/sparc-solaris2.5.1 \
# *** Uncomment and edit to reflect your Tcl/Tk versions:
	-ltk8.0 -ltcl8.0 \
# *** Uncomment and edit to reflect where your X11 libraries are:
#	-L/usr/X11R6/lib \
# *** Or uncomment this for Solaris:
	-L/usr/openwin/lib \
# *** Uncomment these for TOGL extension only:
#	-lGL -lGLU -lXext -lXmu \
# *** Uncomment for AIX:
#	-lld \
# *** Always uncomment this; X11 libraries to link with:
	-lX11

# Lance Ellinghaus's modules:

rotor rotormodule.c		# enigma-inspired encryption
#syslog syslogmodule.c		# syslog daemon interface


# Lance's curses module.  This requires the System V version of
# curses, sometimes known as ncurses (e.g. on Linux, link with
# -lncurses instead of -lcurses; on SunOS 4.1.3, insert -I/usr/5include
# -L/usr/5lib before -lcurses).

#curses cursesmodule.c -lcurses -ltermcap



# Tommy Burnette's 'new' module (creates new empty objects of certain kinds):

new newmodule.c


# Generic (SunOS / SVR4) dynamic loading module.
# This is not needed for dynamic loading of Python modules --
# it is a highly experimental and dangerous device for calling
# *arbitrary* C functions in *arbitrary* shared libraries:

#dl dlmodule.c


# Modules that provide persistent dictionary-like semantics.  You will
# probably want to arrange for at least one of them to be available on
# your machine, though none are defined by default because of library
# dependencies.  The Python module anydbm.py provides an
# implementation independent wrapper for these; dumbdbm.py provides
# similar functionality (but slower of course) implemented in Python.

# The standard Unix dbm module:

#dbm dbmmodule.c 	# dbm(3) may require -lndbm or similar

# Anthony Baxter's gdbm module.  GNU dbm(3) will require -lgdbm:

#gdbm gdbmmodule.c -I/usr/local/include -L/usr/local/lib -lgdbm


# Berkeley DB interface.
#
# This requires the Berkeley DB code, see
# ftp://ftp.cs.berkeley.edu/pub/4bsd/db.1.85.tar.gz
#
# Edit the variables DB and DBPORT to point to the db top directory
# and the subdirectory of PORT where you built it.
#
# (See http://www.jenkon-dev.com/~rd/python/ for an interface to
# BSD DB 2.1.0.)

#DB=/depot/sundry/src/berkeley-db/db.1.85
#DBPORT=$(DB)/PORT/irix.5.3
#bsddb bsddbmodule.c -I$(DBPORT)/include -I$(DBPORT) $(DBPORT)/libdb.a



# David Wayne Williams' soundex module (obsolete -- this will disappear!)
#soundex soundex.c

# Helper module for various ascii-encoders
binascii binascii.c

# Fred Drake's interface to the Python parser
parser parsermodule.c

# Digital Creations' cStringIO and cPickle
cStringIO cStringIO.c
cPickle cPickle.c


# Lee Busby's SIGFPE modules.
# The library to link fpectl with is platform specific.
# Choose *one* of the options below for fpectl:

# For SGI IRIX (tested on 5.3):
#fpectl fpectlmodule.c -lfpe

# For Solaris with SunPro compiler (tested on Solaris 2.5 with SunPro C 4.2):
# (Without the compiler you don't have -lsunmath.)
#fpectl fpectlmodule.c -R/opt/SUNWspro/lib -lsunmath -lm

# For other systems: see instructions in fpectlmodule.c.
#fpectl fpectlmodule.c ...

# Test module for fpectl.  No extra libraries needed.
#fpetest fpetestmodule.c

# Andrew Kuchling's zlib module.
# This require zlib 1.1.3 (or later).
# See http://www.cdrom.com/pub/infozip/zlib/
#zlib zlibmodule.c -I$(prefix)/include -L$(exec_prefix)/lib -lz


# Example -- included for reference only:
# xx xxmodule.c

From phlip_cpp at my-deja.com  Fri Feb 16 23:42:58 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 16 Feb 2001 22:42:58 GMT
Subject: string, split, sort, None, huh?
Message-ID: <96kadi$f2t@dispatch.concentric.net>

Thy Pon:

Gonna split a document by linefeeds into strings, then sort the strings. 
Child's play, huh?

        lines = split ( """Mary
had
a
little
lamb""", "\n" )

        print repr(lines)
        print repr(lines.sort())

Now the output:

        ['Mary', 'had', 'a', 'little', 'lamb']
        None

So where'd that 'None' come from? The Sorting Mini-HOWTO here said strings 
are just as easy as numbers:

        http://www.python.org/doc/howto/sorting/node2.html

Variations with 'cmp' and 'lambda x, y: cmp(x, y)' also get 'None'.

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  Proud victim of the dreaded boomerang effect  --




From glennj at pcard31e.ca.nortel.com  Fri Feb 16 23:55:41 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 16 Feb 2001 22:55:41 GMT
Subject: string, split, sort, None, huh?
References: <96kadi$f2t@dispatch.concentric.net>
Message-ID: <96kb5d$bb0$2@bcarh8ab.ca.nortel.com>

Phlip wrote:
>        print repr(lines)
>        print repr(lines.sort())
>
>Now the output:
>
>        ['Mary', 'had', 'a', 'little', 'lamb']
>        None

sort() sorts the list in place, and returns None.  Try:

lines = ['Mary', 'had', 'a', 'little', 'lamb']
print repr(lines)
lines.sort()
print repr(lines)


-- 
Glenn 


From phlip_cpp at my-deja.com  Sat Feb 17 00:37:44 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 16 Feb 2001 23:37:44 GMT
Subject: string, split, sort, None, huh?
References: <96kadi$f2t@dispatch.concentric.net> <96kb5d$bb0$2@bcarh8ab.ca.nortel.com>
Message-ID: <96kdk8$dtr@freepress.concentric.net>

Glenn W Jackman wrote:

> ...sort() sorts the list in place...

Jay O'Connor wrote:

> ...sort() sorts the sequence in in place...

Oh, man - I gotta lay off that crack pipe, huh?  @-}

Thanks, guys.

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  My opinions are those of your employer  --


From joconnor at cybermesa.com  Sat Feb 17 00:55:37 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Fri, 16 Feb 2001 16:55:37 -0700
Subject: string, split, sort, None, huh?
References: <96kadi$f2t@dispatch.concentric.net> <96kb5d$bb0$2@bcarh8ab.ca.nortel.com> <96kdk8$dtr@freepress.concentric.net>
Message-ID: <3A8DBDF8.FC5F2A35@cybermesa.com>


Phlip wrote:

> Glenn W Jackman wrote:
>
> > ...sort() sorts the list in place...
>
> Jay O'Connor wrote:
>
> > ...sort() sorts the sequence in in place...
>
> Oh, man - I gotta lay off that crack pipe, huh?  @-}
>
> Thanks, guys.
>

Happy to Help :)

Take care,
Jay O'Connor
joconnor at cybermesa.com




From joconnor at cybermesa.com  Sat Feb 17 00:01:35 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Fri, 16 Feb 2001 16:01:35 -0700
Subject: string, split, sort, None, huh?
References: <96kadi$f2t@dispatch.concentric.net>
Message-ID: <3A8DB14E.5B4D39FB@cybermesa.com>


Phlip wrote:

> Thy Pon:
>
> Gonna split a document by linefeeds into strings, then sort the strings.
> Child's play, huh?
>
>         lines = split ( """Mary
> had
> a
> little
> lamb""", "\n" )
>
>         print repr(lines)
>         print repr(lines.sort())
>
> Now the output:
>
>         ['Mary', 'had', 'a', 'little', 'lamb']
>         None
>
> So where'd that 'None' come from?

It came from the fact that  lines.sort() sorts the sequence in in place and
returns None, which is what your second repr() is getting

Try instead:

    print repr (lines)
    lines.sort()
    print repr(lines)

Take care,
Jay O'Connor
joconnor at cybermesa.com



From shaleh at valinux.com  Sat Feb 17 00:21:12 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 15:21:12 -0800 (PST)
Subject: string, split, sort, None, huh?
In-Reply-To: <96kadi$f2t@dispatch.concentric.net>
Message-ID: <XFMail.20010216152112.shaleh@valinux.com>

On 16-Feb-2001 Phlip wrote:
> Thy Pon:
> 
> Gonna split a document by linefeeds into strings, then sort the strings. 
> Child's play, huh?
> 
>         lines = split ( """Mary
> had
> a
> little
> lamb""", "\n" )
> 
>         print repr(lines)
>         print repr(lines.sort())
> 

list.sort() does not return a list, it sorts it in place.

>>> string = "Mary\nhad\na\nlittle\nlamb."
>>> print string
Mary
had
a
little
lamb.
>>> from string import split
>>> pieces = split(string, '\n')
>>> print pieces
['Mary', 'had', 'a', 'little', 'lamb.']
>>> pieces.sort()
>>> print pieces
['Mary', 'a', 'had', 'lamb.', 'little']




From glennj at pcard31e.ca.nortel.com  Fri Feb 16 23:50:08 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 16 Feb 2001 22:50:08 GMT
Subject: missing modules in BeOpen-Python-2.0?
Message-ID: <96kar0$bb0$1@bcarh8ab.ca.nortel.com>

I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
and was unable to import:  crypt, dbm, dl, nis

Why is this release missing these modules (that are documented in the
Global Module Index)?

-- 
Glenn 


From aahz at panix.com  Sat Feb 17 03:42:29 2001
From: aahz at panix.com (Aahz Maruch)
Date: 16 Feb 2001 18:42:29 -0800
Subject: missing modules in BeOpen-Python-2.0?
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com>
Message-ID: <96koel$fms$1@panix2.panix.com>

In article <96kar0$bb0$1 at bcarh8ab.ca.nortel.com>,
Glenn W Jackman  <glennj at nortelnetworks.com> wrote:
>
>I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
>and was unable to import:  crypt, dbm, dl, nis
>
>Why is this release missing these modules (that are documented in the
>Global Module Index)?

You need to compile them.  Check the README.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From glennj at pcard31e.ca.nortel.com  Sat Feb 17 23:39:16 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 17 Feb 2001 22:39:16 GMT
Subject: missing modules in BeOpen-Python-2.0?
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com> <96koel$fms$1@panix2.panix.com>
Message-ID: <96muik$grf$1@bcarh8ab.ca.nortel.com>

Aahz Maruch wrote:
>In article <96kar0$bb0$1 at bcarh8ab.ca.nortel.com>,
>Glenn W Jackman  <glennj at nortelnetworks.com> wrote:
>>
>>I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
>>and was unable to import:  crypt, dbm, dl, nis
>>
>>Why is this release missing these modules (that are documented in the
>>Global Module Index)?
>
>You need to compile them.  Check the README.

None of the READMEs say anything about it:
    $ rpm -ql BeOpen-Python | grep README
    /usr/doc/BeOpen-Python-2.0/README
    /usr/local/lib/python2.0/distutils/README
    /usr/local/lib/python2.0/site-packages/README
    /usr/local/lib/python2.0/test/README

I don't want to sound beligerant, but why should I have to compile
these modules?  This release includes many modules in
/usr/local/lib/python2.0/lib-dynload, including gdbm (but not dbm?)

-- 
Glenn 


From aahz at panix.com  Sun Feb 18 01:18:10 2001
From: aahz at panix.com (Aahz Maruch)
Date: 17 Feb 2001 16:18:10 -0800
Subject: missing modules in BeOpen-Python-2.0?
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com> <96koel$fms$1@panix2.panix.com> <96muik$grf$1@bcarh8ab.ca.nortel.com>
Message-ID: <96n4c2$o9q$1@panix3.panix.com>

In article <96muik$grf$1 at bcarh8ab.ca.nortel.com>,
Glenn W Jackman  <glennj at nortelnetworks.com> wrote:
>Aahz Maruch wrote:
>>In article <96kar0$bb0$1 at bcarh8ab.ca.nortel.com>,
>>Glenn W Jackman  <glennj at nortelnetworks.com> wrote:
>>>
>>>I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
>>>and was unable to import:  crypt, dbm, dl, nis
>>>
>>>Why is this release missing these modules (that are documented in the
>>>Global Module Index)?
>>
>>You need to compile them.  Check the README.
>
>None of the READMEs say anything about it:
>    $ rpm -ql BeOpen-Python | grep README
>    /usr/doc/BeOpen-Python-2.0/README
>    /usr/local/lib/python2.0/distutils/README
>    /usr/local/lib/python2.0/site-packages/README
>    /usr/local/lib/python2.0/test/README

I have to admit that I haven't compiled 2.0 myself, so I'm not sure
exactly where it's documented, but I know it's in there somewhere.

>I don't want to sound beligerant, but why should I have to compile
>these modules?  This release includes many modules in
>/usr/local/lib/python2.0/lib-dynload, including gdbm (but not dbm?)

IIRC, these modules are unavailable on Windows, so they're not compiled
by default.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From glennj at pcard31e.ca.nortel.com  Sun Feb 18 18:44:29 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 18 Feb 2001 17:44:29 GMT
Subject: missing modules in BeOpen-Python-2.0?
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com> <96koel$fms$1@panix2.panix.com> <96muik$grf$1@bcarh8ab.ca.nortel.com> <96n4c2$o9q$1@panix3.panix.com>
Message-ID: <96p1lt$qdl$1@bcarh8ab.ca.nortel.com>

Aahz Maruch wrote:
>>>>I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
>>>>and was unable to import:  crypt, dbm, dl, nis
>>>
>>>You need to compile them.  Check the README.
>>
>>I don't want to sound beligerant, but why should I have to compile
>>these modules?  This release includes many modules in
>>/usr/local/lib/python2.0/lib-dynload, including gdbm (but not dbm?)
>
>IIRC, these modules are unavailable on Windows, so they're not compiled
>by default.

According to the global module index for BeOpen-Python-2.0, there are
20 modules specific to "unix", 
    CGIHTTPServer, commands, crypt, dbm, dl, fcntl, gdbm, grp, nis,
    pipes, posix, posixfile, pwd, readline, resource, rlcompleter,
    syslog, TERMIOS, termios, tty
and only the 4 mentioned are missing.  Of the existing 16 in the
distribution, 8 of them exist in the lib-dynload directory.  
    fcntl, gdbm, grp, pwd, readline, resource, syslog, termios
So a bunch have come already compiled.

-- 
Glenn 


From jafo at tummy.com  Sun Feb 25 06:42:15 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Sat, 24 Feb 2001 22:42:15 -0700
Subject: New Python-2.0 SRPM.
In-Reply-To: <96kar0$bb0$1@bcarh8ab.ca.nortel.com>; from glennj@pcard31e.ca.nortel.com on Fri, Feb 16, 2001 at 10:50:08PM +0000
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com>
Message-ID: <20010224224215.A4706@tummy.com>

On Fri, Feb 16, 2001 at 10:50:08PM +0000, Glenn W Jackman wrote:
>I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
>and was unable to import:  crypt, dbm, dl, nis

I've released a new Python-2.0 SRPM, this one includes the above modules,
adds "-devel", "-tkinter", and "-doc" packages as well.

The SRPM is at ftp.tummy.com:/pub/tummy/RPMS/SRPMS/python-2.0-3tummy.src.rpm

You can build a binary RPM using "rpm --rebuild python-2.0-3tummy.src.rpm".
It will then place the binary RPMs under (by default)
"/usr/src/redhat/RPMS/i386".

Enjoy,
Sean
-- 
 "Who's the guy in the suit?"  "FBI."  "He looks a little to much like a
 lounge lizard for FBI."  "He's working under-cover."  -- Sean & Evelyn 2000
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From glennj at pcard31e.ca.nortel.com  Sun Feb 25 20:36:03 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 25 Feb 2001 19:36:03 GMT
Subject: New Python-2.0 SRPM.
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com> <mailman.983079786.28619.python-list@python.org>
Message-ID: <97bmr3$f2m$1@bcarh8ab.ca.nortel.com>

Sean Reifschneider wrote:
>On Fri, Feb 16, 2001 at 10:50:08PM +0000, Glenn W Jackman wrote:
>>I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
>>and was unable to import:  crypt, dbm, dl, nis
>
>I've released a new Python-2.0 SRPM, this one includes the above modules,
>adds "-devel", "-tkinter", and "-doc" packages as well.
>
>The SRPM is at ftp.tummy.com:/pub/tummy/RPMS/SRPMS/python-2.0-3tummy.src.rpm
>
>You can build a binary RPM using "rpm --rebuild python-2.0-3tummy.src.rpm".
>It will then place the binary RPMs under (by default)
>"/usr/src/redhat/RPMS/i386".

$ rpm --rebuild python-2.0-3tummy.src.rpm
Installing python-2.0-3tummy.src.rpm
only packages with major numbers <= 3 are supported by this version of RPM
$ rpm -qp python-2.0-3tummy.src.rpm
only packages with major numbers <= 3 are supported by this version of RPM
query of python-2.0-3tummy.src.rpm failed
$ rpm --version
RPM version 3.0.4


-- 
Glenn


From jafo at tummy.com  Mon Feb 26 09:13:43 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 01:13:43 -0700
Subject: New Python-2.0 SRPM.
In-Reply-To: <97bmr3$f2m$1@bcarh8ab.ca.nortel.com>; from glennj@pcard31e.ca.nortel.com on Sun, Feb 25, 2001 at 07:36:03PM +0000
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com> <mailman.983079786.28619.python-list@python.org> <97bmr3$f2m$1@bcarh8ab.ca.nortel.com>
Message-ID: <20010226011343.A3221@tummy.com>

On Sun, Feb 25, 2001 at 07:36:03PM +0000, Glenn W Jackman wrote:
>$ rpm --rebuild python-2.0-3tummy.src.rpm
>Installing python-2.0-3tummy.src.rpm
>only packages with major numbers <= 3 are supported by this version of RPM

Ahh, not following errata?  While you're there picking up the BIND
and other errata, you'll want to snag the latest RPM updates.  3.0.5
can deal with version 4 RPM files.  If you're using RedHat, you'll
find that at a mirror under "redhat/updates/<version>/<platform>".

The SRPMs were built on a RedHat 7.0-based box, but with the appropriate
errata applied you should be able to build them on a 6.x or lower machine
within reason.  If you run into problems, let me know -- I haven't
tested it on a pre-7 box.

Sean
-- 
 "Having computers is a lot like having kids, except it's not as much of a
 problem if they die while you're away."  -- Sean Reifschneider, 1997
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From glennj at pcard31e.ca.nortel.com  Mon Feb 26 18:55:07 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 26 Feb 2001 17:55:07 GMT
Subject: New Python-2.0 SRPM.
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com> <mailman.983079786.28619.python-list@python.org> <97bmr3$f2m$1@bcarh8ab.ca.nortel.com> <mailman.983175247.6037.python-list@python.org>
Message-ID: <97e59r$c3f$1@bcarh8ab.ca.nortel.com>

Sean Reifschneider wrote:
>The SRPMs were built on a RedHat 7.0-based box, but with the appropriate
>errata applied you should be able to build them on a 6.x or lower machine
>within reason.  If you run into problems, let me know -- I haven't
>tested it on a pre-7 box.

Thanks for your efforts, but I think I'll stick with python 1.5.2
until I have a more compelling reason to upgrade.

-- 
Glenn 


From phys137 at mailcity.com  Sat Feb 17 00:02:24 2001
From: phys137 at mailcity.com (charles loboz)
Date: Sat, 17 Feb 2001 09:02:24 +1000
Subject: python 3000 IDE? be warned:  contentious
Message-ID: <96k7t3$kr8$1@mail.pl.unisys.com>

1. It seems that there are several Python IDEs in which many smart people
invested plenty of effort. We have WinPython, IDLE (do we still have it ?
;-)), ActiveState is coming up with Komodo (build on Mozilla) and I probably
missed a few here - apologies. Plus we have commercial ones.

2. Competition is good, splitting efforts probably less good. And there's a
fine line between them.

3. My abbreviated personal view about technology directions: Win will rule
the desktop, not only because of it's current popularity - but because it
has component model.  Linux doesn't have a native object model - which
creates scalability (in size and functionality) problems for applications.
Java on Linux will provide the component model but only for the Java world,
not Linux.

4. Good and lasting IDE is about modularity and ease of extension. While I
applaud the efforts of ActiveState I'm not sure that using Mozilla gives
enough fuel for the long run.

5. www.netbeans.org - please, please, please read the whitepaper on that!

6. Essentially this is an open source IDE for Java - but can be used for
most anything else. It's highly modular - you can customize/replace editor
or file system etc (see 5). There is some fantastic stuff there as well as
plenty of good design. Jython folks obviously are already having a peek at
it (naturally) - but this could be used for general Python, too. And used on
both platforms - NT and UNIX.

7. I'm not sure how well Netbeans functionality can be blended with
winpython. but then, i suspect that winpython functionality will blend with
.NET anyway.





From neilh at scintilla.org  Fri Feb 16 23:33:20 2001
From: neilh at scintilla.org (Neil Hodgson)
Date: Fri, 16 Feb 2001 22:33:20 GMT
Subject: python 3000 IDE? be warned:  contentious
References: <96k7t3$kr8$1@mail.pl.unisys.com>
Message-ID: <QUhj6.29748$o85.145040@news-server.bigpond.net.au>

charles loboz:

> 2. Competition is good, splitting efforts probably less good. And there's
a
> fine line between them.
> ...
> 4. Good and lasting IDE is about modularity and ease of extension. While I
> applaud the efforts of ActiveState I'm not sure that using Mozilla gives
> enough fuel for the long run.

   Not all the effort is duplicated. Several IDEs use Scintilla as their
editing component. To share work, you have to find a bundle of functionality
that is widely applicable and refrain from dictating policies that will
conflict with particular uses.

   Other areas that may benefit from shared components are the debugging
interface between the IDE and the running program and interfaces to source
code control and project management. If the debugging interface used in
Komodo works well then it would be a good candidate for wide adoption.

   Neil





From aleaxit at yahoo.com  Sat Feb 17 09:03:30 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 17 Feb 2001 09:03:30 +0100
Subject: python 3000 IDE? be warned:  contentious
References: <96k7t3$kr8$1@mail.pl.unisys.com>
Message-ID: <96lbfu0frm@news2.newsguy.com>

"charles loboz" <phys137 at mailcity.com> wrote in message
news:96k7t3$kr8$1 at mail.pl.unisys.com...
    [snip]
> 3. My abbreviated personal view about technology directions: Win will rule
> the desktop, not only because of it's current popularity - but because it
> has component model.  Linux doesn't have a native object model - which

I share your enthusiasm for component-based technology, and consider
it the only current _technical_ plus of Win systems.  But, as for Linux, I
pin some hopes on XPCOM -- the COM clone technology of Mozilla.  It's
ok if it ain't "native" -- indeed, cross-platform seems preferable.  And
ActiveState has a Python XPCOM package -- I admit to not having tried
it yet (as one first needs to build Mozilla from sources, then add the
package) but what I see looks interesting indeed.  .NET is superior to
COM on most fronts (it should, as it was developed by roughly the same
folks with the advantage of years of practical experience!), but COM and
clones are good enough for most uses.

This has little to do with IDE's, per se -- component technology is
just as precious whatever kind of applications one is developing, be
they IDE's, accounting systems, or ray-tracers.  It's an issue of
whether each application is an island onto itself (or, at best, offers
some non-standardized ad hoc kludges to let other programs talk to
it), or rather, systematically, each and every application is designed
based on a collection of reusable components -- and is, implicitly, a
reusable component itself for potential higher-level applications.


Alex





From phys137 at mailcity.com  Sat Feb 17 23:09:26 2001
From: phys137 at mailcity.com (charles loboz)
Date: Sun, 18 Feb 2001 08:09:26 +1000
Subject: python 3000 IDE? be warned:  contentious
References: <96k7t3$kr8$1@mail.pl.unisys.com> <96lbfu0frm@news2.newsguy.com>
Message-ID: <96mp60$ar9$1@mail.pl.unisys.com>

1. I was unaware of XPCOM - so I stand corrected on that.

2. This and other articles in this thread  mention certain number of
alternatives to COM as an argument that IDE can be written in them. It
undoubtedly is technically possible but it doesn't solve neither the problem
of writing a fully functional IDE nor developing it further efficiently.

3. The first reason for it is the scope of work already available through
NetBeans and its community. NetBeans has a three year (at least) history as
a product prepared by a lot of people. Overall architecture, code etc.

4. The second reason is the support coming from the platform. Technology -
Swing seems to be the dominant GUI after Win, threading model, event model,
property editors and enormous number of external beans already available
(even though created for different purposes) - all of these much less
available for other mentioned platforms (excluding Win, naturally). Scale -
there are plenty of Java programmers out there and the number is growing
much faster than for any other platform. The difference in size of the crowd
is probably two-three _orders_ of _magnitude (or more).

5. Above arguments apply to things like wxWindows, XPCOM. As for Kylix -
looks nice but _proprietary_ (I think), so automatically excluded

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:96lbfu0frm at news2.newsguy.com...
> "charles loboz" <phys137 at mailcity.com> wrote in message
> news:96k7t3$kr8$1 at mail.pl.unisys.com...
>     [snip]
> > 3. My abbreviated personal view about technology directions: Win will
rule
> > the desktop, not only because of it's current popularity - but because
it
> > has component model.  Linux doesn't have a native object model - which
>
> I share your enthusiasm for component-based technology, and consider
> it the only current _technical_ plus of Win systems.  But, as for Linux, I
> pin some hopes on XPCOM -- the COM clone technology of Mozilla.  It's
> ok if it ain't "native" -- indeed, cross-platform seems preferable.  And
> ActiveState has a Python XPCOM package -- I admit to not having tried
> it yet (as one first needs to build Mozilla from sources, then add the
> package) but what I see looks interesting indeed.  .NET is superior to
> COM on most fronts (it should, as it was developed by roughly the same
> folks with the advantage of years of practical experience!), but COM and
> clones are good enough for most uses.
>
> This has little to do with IDE's, per se -- component technology is
> just as precious whatever kind of applications one is developing, be
> they IDE's, accounting systems, or ray-tracers.  It's an issue of
> whether each application is an island onto itself (or, at best, offers
> some non-standardized ad hoc kludges to let other programs talk to
> it), or rather, systematically, each and every application is designed
> based on a collection of reusable components -- and is, implicitly, a
> reusable component itself for potential higher-level applications.
>
>
> Alex
>
>
>




From olivier.deckmyn at mail.dotcom.fr  Sat Feb 17 09:16:05 2001
From: olivier.deckmyn at mail.dotcom.fr (Olivier DECKMYN)
Date: Sat, 17 Feb 2001 09:16:05 +0100
Subject: python 3000 IDE? be warned:  contentious
References: <96k7t3$kr8$1@mail.pl.unisys.com>
Message-ID: <96lbmb$8qk$1@feed.teaser.fr>

Did you take a look at http://boa-constructor.sourceforge.net/ ?

It might really be a good starting point, isn't it ?

"charles loboz" <phys137 at mailcity.com> a ?crit dans le message news:
96k7t3$kr8$1 at mail.pl.unisys.com...
> 1. It seems that there are several Python IDEs in which many smart people
> invested plenty of effort. We have WinPython, IDLE (do we still have it ?
> ;-)), ActiveState is coming up with Komodo (build on Mozilla) and I
probably
> missed a few here - apologies. Plus we have commercial ones.
>
> 2. Competition is good, splitting efforts probably less good. And there's
a
> fine line between them.
>
> 3. My abbreviated personal view about technology directions: Win will rule
> the desktop, not only because of it's current popularity - but because it
> has component model.  Linux doesn't have a native object model - which
> creates scalability (in size and functionality) problems for applications.
> Java on Linux will provide the component model but only for the Java
world,
> not Linux.
>
> 4. Good and lasting IDE is about modularity and ease of extension. While I
> applaud the efforts of ActiveState I'm not sure that using Mozilla gives
> enough fuel for the long run.
>
> 5. www.netbeans.org - please, please, please read the whitepaper on that!
>
> 6. Essentially this is an open source IDE for Java - but can be used for
> most anything else. It's highly modular - you can customize/replace editor
> or file system etc (see 5). There is some fantastic stuff there as well as
> plenty of good design. Jython folks obviously are already having a peek at
> it (naturally) - but this could be used for general Python, too. And used
on
> both platforms - NT and UNIX.
>
> 7. I'm not sure how well Netbeans functionality can be blended with
> winpython. but then, i suspect that winpython functionality will blend
with
> .NET anyway.
>
>
>




From muk_pr at yahoo.com  Sat Feb 17 19:23:58 2001
From: muk_pr at yahoo.com (Victor Black)
Date: Sat, 17 Feb 2001 20:23:58 +0200
Subject: python 3000 IDE? be warned:  contentious
References: <96k7t3$kr8$1@mail.pl.unisys.com>
Message-ID: <t8tgkpq7lp3hc7@corp.supernews.com>

<snip>
>3. My abbreviated personal view about technology directions: Win will rule
>the desktop, not only because of it's current popularity - but because it
<snip>

 Have you heard of kylix?
 http://www.borland.com/kylix/





From gustafl at algonet.se  Sat Feb 17 00:02:50 2001
From: gustafl at algonet.se (Gustaf Liljegren)
Date: 16 Feb 2001 23:02:50 GMT
Subject: socket.error
Message-ID: <Xns904B3C674D9gustaflalgonetse@194.213.69.148>

I'm using 1.5.2 in Windows 98.

When running the httplib module I get the following error when using a non-
existing host name:

Traceback (innermost last):
  File "httptest.py", line 28, in ?
    h = httplib.HTTP(domain)
  File "d:\python\Lib\httplib.py", line 51, in __init__
    if host: self.connect(host, port)
  File "d:\python\Lib\httplib.py", line 79, in connect
    self.sock.connect(host, port)
  File "<string>", line 1, in connect
socket.error: host not found

The code to produce this error is simply:

import httplib
h = httplib.HTTP('www.nonvalidhost.com')

Python doesn't seem to like when I write "except socket.error:", so I 
wonder how to handle it properly.

Regards,

Gustaf Liljegren


From dougfort at downright.com  Sat Feb 17 03:29:33 2001
From: dougfort at downright.com (Doug Fort)
Date: Fri, 16 Feb 2001 21:29:33 -0500
Subject: socket.error
References: <Xns904B3C674D9gustaflalgonetse@194.213.69.148>
Message-ID: <hmlj6.17356$Sl.757877@iad-read.news.verio.net>

Here's how we do it.  Note that you must 'import socket'.  The 'message'
object is usually a tuple.

            except socket.error, message:
                self._message = "Socket Error: %s" % (str(message))
--
Doug Fort
Senior Meat Manager
Downright Software LLC

"Gustaf Liljegren" <gustafl at algonet.se> wrote in message
news:Xns904B3C674D9gustaflalgonetse at 194.213.69.148...
> I'm using 1.5.2 in Windows 98.
>
> When running the httplib module I get the following error when using a
non-
> existing host name:
>
> Traceback (innermost last):
>   File "httptest.py", line 28, in ?
>     h = httplib.HTTP(domain)
>   File "d:\python\Lib\httplib.py", line 51, in __init__
>     if host: self.connect(host, port)
>   File "d:\python\Lib\httplib.py", line 79, in connect
>     self.sock.connect(host, port)
>   File "<string>", line 1, in connect
> socket.error: host not found
>
> The code to produce this error is simply:
>
> import httplib
> h = httplib.HTTP('www.nonvalidhost.com')
>
> Python doesn't seem to like when I write "except socket.error:", so I
> wonder how to handle it properly.
>
> Regards,
>
> Gustaf Liljegren




From dalke at acm.org  Sat Feb 17 00:53:54 2001
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 16 Feb 2001 16:53:54 -0700
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
References: <mailman.982316541.18203.python-list@python.org>
Message-ID: <96kemo$tum$1@slb5.atl.mindspring.net>

Tim wrote:
>I liked life better when language changes were slugged out on
>c.l.py.  There are soooooo many archives and semi-private
>mailing lists and utterly pvt emails now-- and all of them,
> like this msg, with useless Subject lines -- I
>just don't know where to look anymore either.  And I spend hours
>on this stuff every day of my miserable life.

*Sniff* I remember the good old days, when there were only
a couple of messages a day and I looked forward to reading
each one.  *Siff*, *sniff*  Now I just hit "kill thread" a
lot :)  Maybe I should start using another language as
obscure as Python was 5 years ago.

>he's been pushing to
>move discussions to venues like egroups for at least a year.

Here's a thought.  Why not set up an NNTP server for
python development?

Bear in mind that I know nothing about how complicated that
is, but here's what I'm thinking about.

Keep c.l.py around as the main list.  Have a new news server
for development only, with a single main group which is
python-dev.  Newsgroups can be moderated or unmoderated,
and I imagine it would be easy to automatically pass on
mail from specified people.

When a discussion thread gets to be too involved, create
a new newsgroup (which can be open or closed) and spin off
the discussions there.  Newsgroup creation can be done
very easily, if the alt.* groups are any indication.

If someone like me wants to follow the articles, I just
connect to the news server and see which groups are
available.  This lets me use all my local tools for reading,
threading an searching rather than depending on pipermail.
It can even be better since I only need to download the
archive once rather than (as I was yesterday) flipping back
and forth among different pages.  It also lets me read
attachements directly.

Archiving of newsgroups can be done by keeping all
posts and/or migrating old posts to a pipermail archive.
And of course people who want to use email can use
standard gateways between news and mail.

Why wouldn't this work?

                    Andrew
                    dalke at acm.org





From aahz at panix.com  Sat Feb 17 03:45:12 2001
From: aahz at panix.com (Aahz Maruch)
Date: 16 Feb 2001 18:45:12 -0800
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
References: <mailman.982316541.18203.python-list@python.org> <96kemo$tum$1@slb5.atl.mindspring.net>
Message-ID: <96kojo$g9u$1@panix2.panix.com>

In article <96kemo$tum$1 at slb5.atl.mindspring.net>,
Andrew Dalke <dalke at acm.org> wrote:
>
>Why wouldn't this work?

There's no intrinsic reason this wouldn't work.  News servers are one of
the most robust bits of software around, taking relatively little
maintenance.  The main problem is on the client side: most people aren't
familiar with using multiple news servers, and many newsreaders don't
make it easy to do so.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From dalke at acm.org  Sat Feb 17 06:44:49 2001
From: dalke at acm.org (Andrew Dalke)
Date: Fri, 16 Feb 2001 22:44:49 -0700
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
References: <mailman.982316541.18203.python-list@python.org> <96kemo$tum$1@slb5.atl.mindspring.net> <96kojo$g9u$1@panix2.panix.com>
Message-ID: <96l38v$oku$1@slb0.atl.mindspring.net>

Aahz Maruch wrote in message <96kojo$g9u$1 at panix2.panix.com>...
> The main problem is on the client side: most people aren't
>familiar with using multiple news servers, and many newsreaders don't
>make it easy to do so.

Although Netscape and IE both allow that.  All you need is
the "news:" URL.

I believe xrn doesn't support multiple servers, but it's been
a long time since I used it.  Looking at sourceforge, it
does appear that trn supports multiple groups, so there's
still the vt100 support! :)

I also forgot to add, Mozilla hosts their development via
their own news server (see http://www.mozilla.org/community.html )
and Freshmeat makes their announcements available as well
(see http://freshmeat.net/faq/view/18/).

                    Andrew
                    dalke at acm.org







From michael at stroeder.com  Sat Feb 17 10:39:08 2001
From: michael at stroeder.com (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Sat, 17 Feb 2001 10:39:08 +0100
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
References: <mailman.982316541.18203.python-list@python.org> <96kemo$tum$1@slb5.atl.mindspring.net> <96kojo$g9u$1@panix2.panix.com>
Message-ID: <3A8E46BC.73002E12@stroeder.com>

Aahz Maruch wrote:
> 
> In article <96kemo$tum$1 at slb5.atl.mindspring.net>,
> Andrew Dalke <dalke at acm.org> wrote:
> >
> >Why wouldn't this work?
> 
> The main problem is on the client side: most people aren't
> familiar with using multiple news servers, and many newsreaders don't
> make it easy to do so.

And many people are sitting behind firewalls and do only have access
to one general internal news server.

BTW: This is the same discussion like splitting
news:comp.lang.python. This has advantages and disadvantages.

Ciao, Michael.


From dalke at acm.org  Sat Feb 17 12:52:36 2001
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 17 Feb 2001 04:52:36 -0700
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
References: <mailman.982316541.18203.python-list@python.org> <96kemo$tum$1@slb5.atl.mindspring.net> <96kojo$g9u$1@panix2.panix.com> <3A8E46BC.73002E12@stroeder.com>
Message-ID: <96lom4$qle$1@slb6.atl.mindspring.net>

Michael Str?der wrote:
>And many people are sitting behind firewalls and do only have access
>to one general internal news server.

But I did mention that the newsgroup messages can still be
archived and made available via pipermail, so python-dev
access would be no worse than what there is now.

>BTW: This is the same discussion like splitting
>news:comp.lang.python. This has advantages and disadvantages.

Excepting there is already a proliferation of spun off
lists in python-dev (eg, to egroups).  So this is a suggestion
on how to pull them together in one place.

                    Andrew
                    dalke at acm.org





From tim_one at email.msn.com  Sat Feb 17 08:07:00 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 17 Feb 2001 02:07:00 -0500
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
In-Reply-To: <96kemo$tum$1@slb5.atl.mindspring.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEABJAAA.tim_one@email.msn.com>

[Andrew Dalke]
> ...
> Here's a thought.  Why not set up an NNTP server for
> python development?

Believe it or not, that's already being talked about, by people who know far
more than me about this kind of stuff.  It's attractive!  The problems are
"the usual":  finding a well-connected machine with spare cycles and bytes
that won't be taken away, not stepping on anyone's political toes, and
conjuring up a wizard with the spare bandwidth to bring it into existence
and to care for it when it's sick forever after.  Note that over the last 3
months, even the Python mailing lists have been sick more often than not --
and we're still not sure that they're fully recovered, because nobody has
had the time to figure out *why* the last batch of gross problems just
seemed to go away by magic.  You really trust us to manage an NNTP-based
approach <0.9 wink>?

there's-a-reason-we-love-sourceforge-ly y'rs  - tim




From thomas at xs4all.net  Sat Feb 17 12:05:58 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 17 Feb 2001 12:05:58 +0100
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
In-Reply-To: <LNBBLJKPBEHFEDALKOLCOEABJAAA.tim_one@email.msn.com>; from tim_one@email.msn.com on Sat, Feb 17, 2001 at 02:07:00AM -0500
References: <96kemo$tum$1@slb5.atl.mindspring.net> <LNBBLJKPBEHFEDALKOLCOEABJAAA.tim_one@email.msn.com>
Message-ID: <20010217120558.Q4924@xs4all.nl>

On Sat, Feb 17, 2001 at 02:07:00AM -0500, Tim Peters wrote:

> > Here's a thought.  Why not set up an NNTP server for
> > python development?

> Believe it or not, that's already being talked about, by people who know far
> more than me about this kind of stuff.  It's attractive!  The problems are
> "the usual":  finding a well-connected machine with spare cycles and bytes
> that won't be taken away, not stepping on anyone's political toes, and
> conjuring up a wizard with the spare bandwidth to bring it into existence
> and to care for it when it's sick forever after.  Note that over the last 3
> months, even the Python mailing lists have been sick more often than not --
> and we're still not sure that they're fully recovered, because nobody has
> had the time to figure out *why* the last batch of gross problems just
> seemed to go away by magic.  You really trust us to manage an NNTP-based
> approach <0.9 wink>?

Pfiffle. Regardles of whether the idea is workable or not, these *technical*
issues are non-issues. I (and the company I work for, XS4ALL) have offered
sponsorship in the form of machines and maintenance multiple times. (In
fact, I keep offering it whenever I mail Barry, when the lists seem sickish
again :-) The original offer was for webservers/mailservers, but there is no
reason to make it (or include into it) NNTP servers. And we have ample
bandwidth and experience in running NNTP servers.

I don't trust *you* to manage a UNIX system, Tim, and I doubt I trust Guido
to touch one of my systems<wink>, but I certainly trust me and my colleagues
;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From aahz at panix.com  Sat Feb 17 16:24:31 2001
From: aahz at panix.com (Aahz Maruch)
Date: 17 Feb 2001 07:24:31 -0800
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
References: <96kemo$tum$1@slb5.atl.mindspring.net> <LNBBLJKPBEHFEDALKOLCOEABJAAA.tim_one@email.msn.com> <mailman.982408039.17122.python-list@python.org>
Message-ID: <96m53f$mb6$1@panix3.panix.com>

In article <mailman.982408039.17122.python-list at python.org>,
Thomas Wouters  <thomas at xs4all.net> wrote:
>
>Pfiffle. Regardles of whether the idea is workable or not, these *technical*
>issues are non-issues. I (and the company I work for, XS4ALL) have offered
>sponsorship in the form of machines and maintenance multiple times. (In
>fact, I keep offering it whenever I mail Barry, when the lists seem sickish
>again :-) The original offer was for webservers/mailservers, but there is no
>reason to make it (or include into it) NNTP servers. And we have ample
>bandwidth and experience in running NNTP servers.
>
>I don't trust *you* to manage a UNIX system, Tim, and I doubt I trust Guido
>to touch one of my systems<wink>, but I certainly trust me and my colleagues

Let me back this up: xs4all has one of the best reputations in the ISP
business.  I don't have any personal experience, but several people I
trust rank xs4all right up with panix.  I think this would likely be one
of the best things that could happen to the Python community, getting
the core hardware/software under professional management.

The only disadvantage I can think of is that xs4all isn't on the
continental US, but that means very little these days.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From thomas at xs4all.net  Fri Feb 23 17:17:03 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 23 Feb 2001 17:17:03 +0100
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
In-Reply-To: <96m53f$mb6$1@panix3.panix.com>; from aahz@panix.com on Sat, Feb 17, 2001 at 07:24:31AM -0800
References: <96kemo$tum$1@slb5.atl.mindspring.net> <LNBBLJKPBEHFEDALKOLCOEABJAAA.tim_one@email.msn.com> <mailman.982408039.17122.python-list@python.org> <96m53f$mb6$1@panix3.panix.com>
Message-ID: <20010223171702.B16781@xs4all.nl>

On Sat, Feb 17, 2001 at 07:24:31AM -0800, Aahz Maruch wrote:

[ Thomas Wouters suggests running an NNTP/list server for the Python
  community ]

> Let me back this up [ .. ]

Since I've seen some positive reactions to this idea, I'm wondering how many
more I could get by explicitly asking for it :) I'm willing to try and get
my employee to cough up the machines, the place to locate them in a system
room and the connectivity to the 'net (and boy, that's some connectivity:)
but I'll only do that if it's actually considered useful. I'm pretty much
out of touch with 'the Python community' as an entity, so I'll have to ask
straight up:

- Do we need another way to get a list/forum for Pythonic discussions,
  regardless of the exact medium ? I'm unsure what real alternatives there
  are, but the initial look of groups.yahoo.com ("Give me your personal
  data!") scares the old cyberhacker in me enough to protest against it. And
  those of you who lived in the Netherlands during XS4ALL's ad campaign and
  subsequent court cases, last year, should know why :)

  I know about lists.sourceforge.net, but I doubt they want to bother
  setting up lists for all Python projects. (Or maybe they do. I don't
  know.) The same goes about starship.python.net, though I'm sure they are a
  bit more willing to host python discussions :)

- For howmany projects/discussions, roughly ? Is it really worth setting up
  a seperate machine/network for it ?

- What kind of forum would be desirable ? I was thinking newsgroups (with or
  without an 'upfeed') gated to Mailman lists, but some webforum is possible
  too. 

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From emile at fenx.com  Sat Feb 24 03:24:45 2001
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 23 Feb 2001 18:24:45 -0800
Subject: use a news server? (was Re: python-dev summary, Jan. 16-31)
References: <96kemo$tum$1@slb5.atl.mindspring.net> <LNBBLJKPBEHFEDALKOLCOEABJAAA.tim_one@email.msn.com> <mailman.982408039.17122.python-list@python.org> <96m53f$mb6$1@panix3.panix.com> <mailman.982945084.3594.python-list@python.org>
Message-ID: <97763e$o6qu3$1@ID-11957.news.dfncis.de>

Thomas,

As long as you're asking...  ;-)

One thing I've occasionally thought of trying to do is set
up a gateway for *all* of the varied python related SIG's,
mailing lists and newsgroups through to a private back-end
news server.  I dropped out of most of my subscriptions last
year when there were a lot of mail related problems, but I
miss being able to follow the interesting topics as they
come across.  Certainly the volume of mail on the python and
Zope lists alone keep me from re-subscribing, but I'm quite
happy browsing the python list on news.cis.dfn.de, and would
enjoy having news group access to browse and follow the
other lists.  Perhaps an upfeed on the legitimate
newsgroups, but for the mailing list only groups, an e-mail
shot off to the list would suffice (assuming they accept
posts from non-subscribers... otherwise just subscribe).

Good luck with this.  If you're counting votes, let me know.
I'm more than willing to vote again.  ;-)

See you,

--

Emile van Sebille
emile at fenx.com
-------------------


"Thomas Wouters" <thomas at xs4all.net> wrote in message
news:mailman.982945084.3594.python-list at python.org...
> On Sat, Feb 17, 2001 at 07:24:31AM -0800, Aahz Maruch
wrote:
>
> [ Thomas Wouters suggests running an NNTP/list server for
the Python
>   community ]
>
> > Let me back this up [ .. ]
>
> Since I've seen some positive reactions to this idea, I'm
wondering how many
> more I could get by explicitly asking for it :) I'm
willing to try and get
> my employee to cough up the machines, the place to locate
them in a system
> room and the connectivity to the 'net (and boy, that's
some connectivity:)
> but I'll only do that if it's actually considered useful.
I'm pretty much
> out of touch with 'the Python community' as an entity, so
I'll have to ask
> straight up:
>
> - Do we need another way to get a list/forum for Pythonic
discussions,
>   regardless of the exact medium ? I'm unsure what real
alternatives there
>   are, but the initial look of groups.yahoo.com ("Give me
your personal
>   data!") scares the old cyberhacker in me enough to
protest against it. And
>   those of you who lived in the Netherlands during
XS4ALL's ad campaign and
>   subsequent court cases, last year, should know why :)
>
>   I know about lists.sourceforge.net, but I doubt they
want to bother
>   setting up lists for all Python projects. (Or maybe they
do. I don't
>   know.) The same goes about starship.python.net, though
I'm sure they are a
>   bit more willing to host python discussions :)
>
> - For howmany projects/discussions, roughly ? Is it really
worth setting up
>   a seperate machine/network for it ?
>
> - What kind of forum would be desirable ? I was thinking
newsgroups (with or
>   without an 'upfeed') gated to Mailman lists, but some
webforum is possible
>   too.
>
> --
> Thomas Wouters <thomas at xs4all.net>
>
> Hi! I'm a .signature virus! copy me into your .signature
file to help me spread!
>




From shaleh at valinux.com  Sat Feb 17 01:04:10 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 16:04:10 -0800 (PST)
Subject: ok, so how do I return a sorted list without doing it in place
Message-ID: <XFMail.20010216160410.shaleh@valinux.com>

So, several of us informed Philip that list.sort() is in place and returns
nothing.  How do i accomplish:

for key in dict.keys().sort():
  print "%s -> %s" % (key, dict[key])

???

I know I can do:

list = dict.keys()
list.sort()
.....

but it is just plain annoying.



From tjg at exceptionalminds.com  Sat Feb 17 01:20:24 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Fri, 16 Feb 2001 16:20:24 -0800
Subject: ok, so how do I return a sorted list without doing it in place
In-Reply-To: <XFMail.20010216160410.shaleh@valinux.com>; from shaleh@valinux.com on Fri, Feb 16, 2001 at 04:04:10PM -0800
References: <XFMail.20010216160410.shaleh@valinux.com>
Message-ID: <20010216162024.F19267@trufflehunter.avalongroup.net>

On Fri, Feb 16, 2001 at 04:04:10PM -0800, Sean 'Shaleh' Perry wrote:
> I know I can do:
> 
> list = dict.keys()
> list.sort()
> .....
> 
> but it is just plain annoying.

I agree. This has bitten me on the butt more than once. It also seems
somewhat unPythonic. Usually I expect what happens in Python, but that 
sort process always strikes me as unexpected.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  31 days  4:30 hours ago<<



From phlip_cpp at my-deja.com  Sat Feb 17 01:21:59 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 17 Feb 2001 00:21:59 GMT
Subject: ok, so how do I return a sorted list without doing it in place
References: <mailman.982368260.960.python-list@python.org>
Message-ID: <96kg77$88c@dispatch.concentric.net>

Proclaimed Sean 'Shaleh' Perry from the mountaintops:

> So, several of us informed Philip that list.sort() is in place and returns
> nothing.  How do i accomplish:
> 
> for key in dict.keys().sort():
>   print "%s -> %s" % (key, dict[key])
> 
> ???
> 
> I know I can do:
> 
> list = dict.keys()
> list.sort()
> .....
> 
> but it is just plain annoying.

I don't know why a more useful return value couldn't have been supplied, 
for call chaining like this at least, but Mr. Manners reminds the Gentle 
Poster that one should not cram too much stuff on one command line. Leave 
that for Perl or C++ coders.

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  My opinions are those of your employer  --


From kalle at gnupung.net  Sat Feb 17 02:01:22 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Sat, 17 Feb 2001 02:01:22 +0100
Subject: ok, so how do I return a sorted list without doing it in place
In-Reply-To: <96kg77$88c@dispatch.concentric.net>; from phlip_cpp@my-deja.com on Sat, Feb 17, 2001 at 12:21:59AM +0000
References: <mailman.982368260.960.python-list@python.org> <96kg77$88c@dispatch.concentric.net>
Message-ID: <20010217020122.B4141@apone.network.loc>

Sez Phlip:
> Proclaimed Sean 'Shaleh' Perry from the mountaintops:
[list.sort() returns None]
> 
> I don't know why a more useful return value couldn't have been supplied, 

See http://www.python.org/doc/current/lib/typesseq-mutable.html, footnote 5,
for the reasoning of the designers:

"""
The sort() and reverse() methods modify the list in place for economy of
space when sorting or reversing a large list. They don't return the sorted
or reversed list to remind you of this side effect.
"""

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010217/1c4b69c0/attachment.pgp>

From r2d2 at mao.acc.umu.se  Mon Feb 19 09:14:35 2001
From: r2d2 at mao.acc.umu.se (Niklas Frykholm)
Date: Mon, 19 Feb 2001 08:14:35 +0000 (UTC)
Subject: ok, so how do I return a sorted list without doing it in place
References: <mailman.982368260.960.python-list@python.org> <96kg77$88c@dispatch.concentric.net>
Message-ID: <slrn991lfb.u76.r2d2@mao.acc.umu.se>

In article <96kg77$88c at dispatch.concentric.net>, Phlip wrote:
>Proclaimed Sean 'Shaleh' Perry from the mountaintops:
>
>> So, several of us informed Philip that list.sort() is in place and returns
>> nothing.  How do i accomplish:
>> 
>> for key in dict.keys().sort():
>>   print "%s -> %s" % (key, dict[key])
>> 
>> ???
>> 
>> I know I can do:
>> 
>> list = dict.keys()
>> list.sort()
>> .....
>> 
>> but it is just plain annoying.
>
>I don't know why a more useful return value couldn't have been supplied, 
>for call chaining like this at least, but Mr. Manners reminds the Gentle 
>Poster that one should not cram too much stuff on one command line. Leave 
>that for Perl or C++ coders.

I really like the way ruby has solved this, by providing two functions:
sort and sort!, where sort! sorts the list in place and sort returns a
sorted copy. The exclamation mark is used to denote a method that
changes the object in-place. With this convention you do not have to
remember whether a function works in-place or on a copy.

Many other similar functions also have two variants, for example there is
a map function and a map! function.

// Niklas


From kalle at gnupung.net  Sat Feb 17 01:56:17 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Sat, 17 Feb 2001 01:56:17 +0100
Subject: ok, so how do I return a sorted list without doing it in place
In-Reply-To: <XFMail.20010216160410.shaleh@valinux.com>; from shaleh@valinux.com on Fri, Feb 16, 2001 at 04:04:10PM -0800
References: <XFMail.20010216160410.shaleh@valinux.com>
Message-ID: <20010217015617.A4141@apone.network.loc>

Sez Sean 'Shaleh' Perry:
> So, several of us informed Philip that list.sort() is in place and returns
> nothing.  How do i accomplish:
> 
> for key in dict.keys().sort():
>   print "%s -> %s" % (key, dict[key])
> 
> ???

# Courtesy of Alex Martelli, a few dozen posts earlier or so.

def sort(l):
    l.sort()
    return l

for key in sort(dict.keys()):
    print "%s -> %s" % (key, dict[key])

That's the best you can get, methinks.

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010217/cb5416ac/attachment.pgp>

From aahz at panix.com  Sat Feb 17 03:48:52 2001
From: aahz at panix.com (Aahz Maruch)
Date: 16 Feb 2001 18:48:52 -0800
Subject: ok, so how do I return a sorted list without doing it in place
References: <XFMail.20010216160410.shaleh@valinux.com> <mailman.982369099.4358.python-list@python.org>
Message-ID: <96koqk$gv6$1@panix2.panix.com>

In article <mailman.982369099.4358.python-list at python.org>,
Timothy Grant  <tjg at exceptionalminds.com> wrote:
>On Fri, Feb 16, 2001 at 04:04:10PM -0800, Sean 'Shaleh' Perry wrote:
>> I know I can do:
>> 
>> list = dict.keys()
>> list.sort()
>> .....
>> 
>> but it is just plain annoying.
>
>I agree. This has bitten me on the butt more than once. It also seems
>somewhat unPythonic. Usually I expect what happens in Python, but that 
>sort process always strikes me as unexpected.

It's unexpected in some respects, but doing it this way causes the
fewest surprises overall.  At least you rarely get a downstream bug
this way.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From grante at visi.com  Sat Feb 17 17:21:32 2001
From: grante at visi.com (Grant Edwards)
Date: Sat, 17 Feb 2001 16:21:32 GMT
Subject: ok, so how do I return a sorted list without doing it in place
References: <XFMail.20010216160410.shaleh@valinux.com> <mailman.982369099.4358.python-list@python.org> <96koqk$gv6$1@panix2.panix.com>
Message-ID: <slrn98t98c.g7f.grante@isis.visi.com>

In article <96koqk$gv6$1 at panix2.panix.com>, Aahz Maruch wrote:

>>> I know I can do:
>>> 
>>> list = dict.keys()
>>> list.sort()
>>> .....
>>> 
>>> but it is just plain annoying.
>>
>>I agree. This has bitten me on the butt more than once. It also seems
>>somewhat unPythonic. Usually I expect what happens in Python, but that 
>>sort process always strikes me as unexpected.

On many occasions, I would find it quite handy if there was a sequence
method that returned a sorted shallow copy:

 for k in dict.keys().sorted():
   whatever(k) 

>It's unexpected in some respects, but doing it this way causes the
>fewest surprises overall.  At least you rarely get a downstream bug
>this way.

In my mine the used of the adjective "sorted" as a method name conveys the
fact that it doesn't modify the list in place they way the imperative "sort"
does.  YMMV.

-- 
Grant Edwards                   grante             Yow!  World War III? No
                                  at               thanks!
                               visi.com            


From grante at visi.com  Sun Feb 18 01:00:47 2001
From: grante at visi.com (Grant Edwards)
Date: Sun, 18 Feb 2001 00:00:47 GMT
Subject: ok, so how do I return a sorted list without doing it in place
References: <XFMail.20010216160410.shaleh@valinux.com> <mailman.982369099.4358.python-list@python.org> <96koqk$gv6$1@panix2.panix.com> <slrn98t98c.g7f.grante@isis.visi.com>
Message-ID: <slrn98u45f.6e0.grante@isis.visi.com>

In article <slrn98t98c.g7f.grante at isis.visi.com>, Grant Edwards wrote:

> for k in dict.keys().sorted():
>   whatever(k) 
[...]
>In my mine the used of the adjective "sorted"

Jeez, I _swear_ that's not what I typed...

-- 
Grant Edwards                   grante             Yow!  Yow!! "Janitor
                                  at               trapped in sewer uses ESP
                               visi.com            to find decayed burger"!!


From walterm at cmat.edu.uy  Sat Feb 17 02:52:18 2001
From: walterm at cmat.edu.uy (Walter Moreira)
Date: Fri, 16 Feb 2001 22:52:18 -0300
Subject: 'in' operator
Message-ID: <20010216225218.A13190@cmat.edu.uy>

Hello.

Why the following test raise an error?

  >>> '' in 'yY'
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
  TypeError: 'in <string>' requires character as left operand

I would expect it to give 0, false. With the actual behavior, one must write

  resp = raw_input('Yes or no: ')
  if resp and resp in 'yY':
      ....

to consider the case when the user just press `return', instead of

  if resp in 'yY':

Is this a planned behavior?
Regards:
Walter

--
--------------
Walter Moreira  <> Centro de Matematica  <> Universidad de la Republica
email: walterm at cmat.edu.uy <> HomePage: http://www.cmat.edu.uy/~walterm
                 +-----------------------------------------------------
        /OD\_    |   "Agent Smith: Never send a human to do a machine's
 O o . |_o_o_)   |   job."                       -- From ``The Matrix''
	       --+--



From tim.one at home.com  Sat Feb 17 04:50:19 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 16 Feb 2001 22:50:19 -0500
Subject: 'in' operator
In-Reply-To: <20010216225218.A13190@cmat.edu.uy>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEPJIPAA.tim.one@home.com>

[Walter Moreira]
> Why the following test raise an error?
>
>   >>> '' in 'yY'
>   Traceback (most recent call last):
>     File "<stdin>", line 1, in ?
>   TypeError: 'in <string>' requires character as left operand

   elt in sequence

tests whether elt is a member of sequence.  The elements of strings are
characters.  Testing whether anything other than a character is an element
of a string is simply a TypeError:  it's a highly suspicious question
because only a character *could* be an element of a string.  And an empty
string is not a character.

> I would expect it to give 0, false. With the actual behavior, one
> must write
>
>   resp = raw_input('Yes or no: ')
>   if resp and resp in 'yY':
>       ....
> to consider the case when the user just press `return',

But fails to consider the case where the user enters more than one character
before pressing 'return' (try it:  it still blows up).  You're missing this
one:

    if resp in ("y", "Y"):

*Now* you're asking whether a string is in a sequence of strings, which
always makes sense and is much closer to your intent.  Even better:

    if resp.lower()[:1] == "y":

After all, if you're asking the user "Yes or no", it's impolite to take
"Yes" as meaning "no" <0.3 wink>.

> Is this a planned behavior?

Yes.  Python is trying to guide you toward writing code you won't regret.
TypeErrors are signs that you're doing something so questionable that Python
would rather refuse to do anything than guess your intent.

   x = 6 + "42"

"642" or 48?  Python refuses to guess -- so that's a TypeError too.

it's-for-your-own-good<wink>-ly y'rs  - tim




From root at rainerdeyke.com  Sat Feb 17 08:50:07 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 17 Feb 2001 07:50:07 GMT
Subject: 'in' operator
References: <mailman.982374799.24358.python-list@python.org>
Message-ID: <P2qj6.245980$ge4.85515321@news2.rdc2.tx.home.com>

"Walter Moreira" <walterm at cmat.edu.uy> wrote in message
news:mailman.982374799.24358.python-list at python.org...
> Why the following test raise an error?
>
>   >>> '' in 'yY'
>   Traceback (most recent call last):
>     File "<stdin>", line 1, in ?
>   TypeError: 'in <string>' requires character as left operand
>
> I would expect it to give 0, false.

I would expect it to give 1, since "" is a substring of "yY".  Given this
basic misunderstanding of what '"" in "yY"' is supposed to do, raising an
exception seems a reasonable compromise.  If you want to check if something
is either "y" or "Y", use 'resp in ("y", "Y")'.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From pearu at cens.ioc.ee  Sat Feb 17 10:21:05 2001
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Sat, 17 Feb 2001 11:21:05 +0200
Subject: 'in' operator
In-Reply-To: <P2qj6.245980$ge4.85515321@news2.rdc2.tx.home.com>
References: <mailman.982374799.24358.python-list@python.org> <P2qj6.245980$ge4.85515321@news2.rdc2.tx.home.com>
Message-ID: <Pine.LNX.4.21.0102171108010.7063-100000@kev.ioc.ee>

On Sat, 17 Feb 2001, Rainer Deyke wrote:

> "Walter Moreira" <walterm at cmat.edu.uy> wrote in message
> news:mailman.982374799.24358.python-list at python.org...
> > Why the following test raise an error?
> >
> >   >>> '' in 'yY'
> >   Traceback (most recent call last):
> >     File "<stdin>", line 1, in ?
> >   TypeError: 'in <string>' requires character as left operand
> >
> > I would expect it to give 0, false.
> 
> I would expect it to give 1, since "" is a substring of "yY".  Given this

Me too, and because of the analogy:
`an empty set is a subset of every set'
`an empty string is a substring of every string'

Pearu



From jepler at inetnebr.com  Sat Feb 17 17:35:28 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Sat, 17 Feb 2001 16:35:28 GMT
Subject: 'in' operator
References: <mailman.982374799.24358.python-list@python.org> <P2qj6.245980$ge4.85515321@news2.rdc2.tx.home.com> <Pine.LNX.4.21.0102171108010.7063-100000@kev.ioc.ee>
Message-ID: <slrn98ta2f.9uo.jepler@potty.housenet>

On Sat, 17 Feb 2001 11:21:05 +0200, Pearu Peterson
 <pearu at cens.ioc.ee> wrote:
>Me too, and because of the analogy:
>`an empty set is a subset of every set'
>`an empty string is a substring of every string'

But "in" doesn't test for a subset.  For instance,
	[1, 2] in [1, 2, 3, 4]
is 0.

However, I think that
	"" in "yY"
and
	0 in "yY"
should return 0, because for all items in "yY", none of them are
"" or 0.  (This is compared to raising an error)

Jeff


From tim.one at home.com  Sat Feb 17 19:34:59 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 17 Feb 2001 13:34:59 -0500
Subject: 'in' operator
In-Reply-To: <slrn98ta2f.9uo.jepler@potty.housenet>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEAOJAAA.tim.one@home.com>

[Jeff Epler]
> ...
> However, I think that
> 	"" in "yY"
> and
> 	0 in "yY"
> should return 0, because for all items in "yY", none of them are
> "" or 0.  (This is compared to raising an error)

As explained earlier, that's exactly why Python raises TypeError for

    x in y

when type(x) is such that it can't possibly succeed:  you think the first
example should return 0, the original questioner thought 1.  Python refuses
to guess which one of you wrote it <wink>.

say-what-you-mean-and-there's-no-problem-ly y'rs  - tim




From moshez at zadka.site.co.il  Sun Feb 18 00:37:58 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 18 Feb 2001 01:37:58 +0200 (IST)
Subject: 'in' operator
In-Reply-To: <slrn98ta2f.9uo.jepler@potty.housenet>
References: <slrn98ta2f.9uo.jepler@potty.housenet>, <mailman.982374799.24358.python-list@python.org> <P2qj6.245980$ge4.85515321@news2.rdc2.tx.home.com> <Pine.LNX.4.21.0102171108010.7063-100000@kev.ioc.ee>
Message-ID: <20010217233758.F0ED0A840@darjeeling.zadka.site.co.il>

On Sat, 17 Feb 2001 16:35:28 GMT, jepler at inetnebr.com (Jeff Epler) wrote:

> However, I think that
> 	"" in "yY"
> and
> 	0 in "yY"
> should return 0, because for all items in "yY", none of them are
> "" or 0.  (This is compared to raising an error)

I thought so too, but when I offered to change, Guido didn't agree --
and he was probably was right.

Practicality beats purity -- and most times people do a ``0 in "yY"'' it's
an error, so Python signals it. For example, the recent poster, who was
lucky to have Python raise an error, rather then silently return 0.
-- 
"I'll be ex-DPL soon anyway so I'm        |LUKE: Is Perl better than Python?
looking for someplace else to grab power."|YODA: No...no... no. Quicker,
   -- Wichert Akkerman (on debian-private)|      easier, more seductive.
For public key, finger moshez at debian.org  |http://www.{python,debian,gnu}.org



From tjreedy at udel.edu  Sun Feb 18 09:21:08 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 18 Feb 2001 03:21:08 -0500
Subject: 'in' operator
References: <mailman.982374799.24358.python-list@python.org>
Message-ID: <96o070$bt6$1@news.udel.edu>

"Walter Moreira" <walterm at cmat.edu.uy> wrote in message
news:mailman.982374799.24358.python-list at python.org...
> Why the following test raise an error?
>
>   >>> '' in 'yY'
>   Traceback (most recent call last):
>     File "<stdin>", line 1, in ?
>   TypeError: 'in <string>' requires character as left operand

Precisely for the reason stated.  A character is a string of length 1.  A
null string (of length 0) is not a character.  Neither are strings of
length >= 2 [for which one must use find() or match()].

> I would expect it to give 0, false.

The empty set is contained in every set as a subset.  The empty string is a
substring of every string.  So, if defined, I would expect 1 = true.

> With the actual behavior, one must write

>   resp = raw_input('Yes or no: ')
>   if resp and resp in 'yY':
...

This is a good example of what 'and' is meant for.  Note that if '' in
string were defined as true, you would still have to write the above,
presuming that you did not want to interpret it as yes.

> Is this a planned behavior?

Based on the error message, yes.

Terry J. Reedy





From jmarshal at mathworks.com  Tue Feb 20 15:47:12 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 20 Feb 2001 14:47:12 GMT
Subject: 'in' operator
References: <mailman.982374799.24358.python-list@python.org> <96o070$bt6$1@news.udel.edu>
Message-ID: <96u01g$b38$1@news.mathworks.com>

Terry Reedy <tjreedy at udel.edu> wrote:

> "Walter Moreira" <walterm at cmat.edu.uy> wrote in message
> news:mailman.982374799.24358.python-list at python.org...
>> Why the following test raise an error?
>>
>>   >>> '' in 'yY'
>>   Traceback (most recent call last):
>>     File "<stdin>", line 1, in ?
>>   TypeError: 'in <string>' requires character as left operand

> Precisely for the reason stated.  A character is a string of length 1.  A
> null string (of length 0) is not a character.  Neither are strings of
> length >= 2 [for which one must use find() or match()].

Wouldn't it be more natural for this to be a ValueError?  There is no
character type in Python -- '', 'y', and 'yY' are all strings.


From shaleh at valinux.com  Sat Feb 17 05:12:23 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 20:12:23 -0800
Subject: why the string function / module change
Message-ID: <20010216201223.A14948@valinux.com>

Ok, so I read the what's new.  I understand that for Unicode support the
string module needed help.  No problem.  By why introduce a new syntax that
makes 2.0 incompatible with 1.5?

foo = 'this is a string'
pieces = foo.split()

split(foo) worked just as well.  I understand there were problems making the
string module do this, but still.  Could they have not been _split() or
something to dissuade their use?



From shaleh at valinux.com  Sat Feb 17 06:20:02 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 21:20:02 -0800
Subject: why the string function / module change
In-Reply-To: <20010216201223.A14948@valinux.com>; from shaleh@valinux.com on Fri, Feb 16, 2001 at 08:12:23PM -0800
References: <20010216201223.A14948@valinux.com>
Message-ID: <20010216212002.A14970@valinux.com>

On Fri, Feb 16, 2001 at 08:12:23PM -0800, Sean 'Shaleh' Perry wrote:
> 
> split(foo) worked just as well.  I understand there were problems making the
> string module do this, but still.  Could they have not been _split() or
> something to dissuade their use?
> 

example of recent code I have seen:

keys[keyname] = ": ".join(s.split(": ")[1:])

foo.join(sequence) # that logic is just odd



From db3l at fitlinxx.com  Wed Feb 21 07:43:28 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 21 Feb 2001 01:43:28 -0500
Subject: why the string function / module change
References: <mailman.982383141.21677.python-list@python.org>
Message-ID: <u4rxocz8v.fsf@ctwd0143.fitlinxx.com>

Sean 'Shaleh' Perry <shaleh at valinux.com> writes:

> Ok, so I read the what's new.  I understand that for Unicode support the
> string module needed help.  No problem.  By why introduce a new syntax that
> makes 2.0 incompatible with 1.5?

The syntax change wasn't to support Unicode, it was to make string
methods actual methods of string objects, and thus more consistent
with other Python objects.

If you think about, the string module was the odd man out.  After all,
if you want to read or write to a file object, you use
<fileobj>.read() and <fileobj>.write().  You don't use
filemodule.read(<fileobj>) and filemodule.write(<fileobj>).

If you are appending to a list object you use <listobj>.append(), and
not listmodule.append(<listobj>).  Similarly with dictionaries, and so on...

The string module - I expect largely for legacy reasons - existed as a
module with functions that took string objects as parameters.  But
that's very different than methods right on the string objects, which
are much more consistent with other objects in Python.

> split(foo) worked just as well.  I understand there were problems making the
> string module do this, but still.  Could they have not been _split() or
> something to dissuade their use?

I'm not aware of any problems - and you can still use
"string.split(foo)" rather than "foo.split()" if you like (internally
the string module functions are just reflected through the methods on
the objects they are given).   Although its possible that the string
module may eventually be deprecated, it's not likely to be for quite a
while, if ever.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From scarblac at pino.selwerd.nl  Wed Feb 21 23:32:46 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 21 Feb 2001 22:32:46 GMT
Subject: why the string function / module change
References: <mailman.982383141.21677.python-list@python.org> <u4rxocz8v.fsf@ctwd0143.fitlinxx.com>
Message-ID: <slrn998gd8.18q.scarblac@pino.selwerd.nl>

David Bolen <db3l at fitlinxx.com> wrote in comp.lang.python:
> If you think about, the string module was the odd man out.  After all,
> if you want to read or write to a file object, you use
> <fileobj>.read() and <fileobj>.write().  You don't use
> filemodule.read(<fileobj>) and filemodule.write(<fileobj>).
> 
> If you are appending to a list object you use <listobj>.append(), and
> not listmodule.append(<listobj>).  Similarly with dictionaries, and so on...

But lists and dictionaries are mutable, and files are as well, in a way.
Strings are not. Nor are tuples and integers, and those don't have methods
either. So strings were hardly an odd one out.

I like the string methods (maybe " ".join() is taking it a bit far) but I
think this reason is a bit weak.

-- 
Remco Gerlich


From qrczak at knm.org.pl  Thu Feb 22 17:02:16 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Feb 2001 16:02:16 GMT
Subject: why the string function / module change
References: <mailman.982383141.21677.python-list@python.org> <u4rxocz8v.fsf@ctwd0143.fitlinxx.com> <slrn998gd8.18q.scarblac@pino.selwerd.nl>
Message-ID: <slrn99ae08.d33.qrczak@qrnik.zagroda>

21 Feb 2001 22:32:46 GMT, Remco Gerlich <scarblac at pino.selwerd.nl> pisze:

> > If you are appending to a list object you use <listobj>.append(), and
> > not listmodule.append(<listobj>).  Similarly with dictionaries, and so on...
> 
> But lists and dictionaries are mutable, and files are as well, in a way.
> Strings are not.

Why should it matter?

> Nor are tuples and integers, and those don't have methods either.

I see no fundamental reason why they should not, except that Python
already expressed them as builtin syntax or builtin functions which
recognize builtin types (+, ==, len).

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From man_at_moon at hotmail.com  Sat Feb 17 08:51:01 2001
From: man_at_moon at hotmail.com (David A.)
Date: Sat, 17 Feb 2001 07:51:01 -0000
Subject: Yet Another Newbie Question (YANQ)
Message-ID: <Xlsj6.95$DC1.647178@newsserver.ip.pt>

Hi to everyone,

    I am playing around with python and I have 1 quetions where I hope
someone can help

I have this problem:

>>> a='a'
>>> prt='print'
>>> t=prt+' '+a
>>> t
'print a'

now when I try to:
>>> eval(t)
Traceback (innermost last):
  File "<pyshell#18>", line 1, in ?
    eval(t)
  File "<string>", line 1
    print a
        ^
SyntaxError: invalid syntax

can someone explain me what I am doing wrong! Why can't I use eval(print)?
Any help is welcome!

David Asfaha
--
"Quidquid latine dictum sit, altum viditur."
Whatever is said in Latin sounds profound.










From phd at phd.pp.ru  Mon Feb 19 12:31:08 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 19 Feb 2001 14:31:08 +0300 (MSK)
Subject: Yet Another Newbie Question (YANQ)
In-Reply-To: <Xlsj6.95$DC1.647178@newsserver.ip.pt>
Message-ID: <Pine.LNX.4.30.0102191430481.24446-100000@fep132.fep.ru>

On Sat, 17 Feb 2001, David A. wrote:
> >>> a='a'
> >>> prt='print'
> >>> t=prt+' '+a
> >>> t
> 'print a'
>
> now when I try to:
> >>> eval(t)
> Traceback (innermost last):
>   File "<pyshell#18>", line 1, in ?
>     eval(t)
>   File "<string>", line 1
>     print a
>         ^
> SyntaxError: invalid syntax

   exec t

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From jhauser at ifm.uni-kiel.de  Mon Feb 19 12:39:06 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 19 Feb 2001 12:39:06 +0100
Subject: Yet Another Newbie Question (YANQ)
References: <Xlsj6.95$DC1.647178@newsserver.ip.pt>
Message-ID: <878zn2yk9x.fsf@lisboa.ifm.uni-kiel.de>

You should use exec() instead of eval().

HTH,
__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From ecastro at cicei.ulpgc.es  Tue Feb 20 18:17:07 2001
From: ecastro at cicei.ulpgc.es (Enrique)
Date: 20 Feb 2001 16:17:07 -0100
Subject: Yet Another Newbie Question (YANQ)
References: <Xlsj6.95$DC1.647178@newsserver.ip.pt>
Message-ID: <3A929760.11515D22@cicei.ulpgc.es>


"David A." wrote:

> Hi to everyone,
>
>     I am playing around with python and I have 1 quetions where I hope
> someone can help
>
> I have this problem:
>
> >>> a='a'
> >>> prt='print'
> >>> t=prt+' '+a
> >>> t
> 'print a'
>
> now when I try to:
> >>> eval(t)
> Traceback (innermost last):
>   File "<pyshell#18>", line 1, in ?
>     eval(t)
>   File "<string>", line 1
>     print a
>         ^
> SyntaxError: invalid syntax
>
> can someone explain me what I am doing wrong! Why can't I use eval(print)?
> Any help is welcome!

I thinks it is a semantic confusion:
eval() evaluates an _expression_    (like i*(a+b-5))
exec() executes an _statement_      (print a,            del b,    import
moduleX)

Enrique




From SBrunning at trisystems.co.uk  Wed Feb 21 10:09:51 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 21 Feb 2001 09:09:51 -0000
Subject: Yet Another Newbie Question (YANQ)
Message-ID: <31575A892FF6D1118F5800600846864D78B934@intrepid>

> From:	David A. [SMTP:man_at_moon at hotmail.com]
> I have this problem:
> 
> >>> a='a'
> >>> prt='print'
> >>> t=prt+' '+a
> >>> t
> 'print a'
> 
> now when I try to:
> >>> eval(t)
> Traceback (innermost last):
>   File "<pyshell#18>", line 1, in ?
>     eval(t)
>   File "<string>", line 1
>     print a
>         ^
> SyntaxError: invalid syntax
> 
> can someone explain me what I am doing wrong! Why can't I use eval(print)?
> Any help is welcome!
 
eval evaluates a Python expression - 'print a' isn't an expression. Use eval
like this:

>>> spam = '1 + 4'
>>> eggs = eval(spam)
>>> print eggs
5

What *you* want is 'exec'. try this:

>>> a = 'a'
>>> prt='print'
>>> t=prt+' '+a
>>> t
'print a'
>>> exec(t)
a

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk

On two occasions I have been asked (by members of Parliament!), 'Pray, Mr.
Babbage, if you put into the machine wrong figures, will the right answers
come out?' I am not able rightly to apprehend the kind of confusion of ideas
that could provoke such a question. - Charles Babbage




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From dante at oz.net  Sat Feb 17 09:34:00 2001
From: dante at oz.net (Michael Esveldt)
Date: 17 Feb 2001 08:34:00 GMT
Subject: ZWiki Web problem
Message-ID: <96ld1o$9a0$0@216.39.162.234>

I'm encountering the following problem in my ZWiki Web. I'm new to Zope 
version 2.3.0 with python 1.5.2. Any attempt to View a Wiki page 
produces the error below. I've tried reinstalling ZWiki from scratch, 
reimporting ZWiki Web and creating a brand new ZWiki Web.

Zope has encountered an error while publishing this resource.

Error Type: KeyError
Error Value: %00

<troubleshooting snipped>

Traceback (innermost last):
  File /home/zope/lib/python/ZPublisher/Publish.py, line 222, in 
publish_module
  File /home/zope/lib/python/ZPublisher/Publish.py, line 187, in publish
  File /home/zope/lib/python/Zope/__init__.py, line 221, in 
zpublisher_exception_hook
    (Object: Traversable)
  File /home/zope/lib/python/ZPublisher/Publish.py, line 171, in publish
  File /home/zope/lib/python/ZPublisher/mapply.py, line 160, in mapply
    (Object: FrontPage)
  File /home/zope/lib/python/ZPublisher/Publish.py, line 112, in 
call_object
    (Object: FrontPage)
  File /home/zope/lib/python/Products/ZWiki/ZWikiPage.py, line 210, in 
__call__
    (Object: FrontPage)
  File /home/zope/lib/python/OFS/DTMLMethod.py, line 189, in __call__
    (Object: standard_wiki_footer)
  File /home/zope/lib/python/DocumentTemplate/DT_String.py, line 538, in 
__call__
    (Object: standard_wiki_footer)
  File /home/zope/lib/python/DocumentTemplate/DT_Util.py, line 334, in 
eval
    (Object: 
bobobase_modification_time().toZone(zwiki_timezone).rfc822())
    (Info: bobobase_modification_time)
  File <string>, line 0, in ?
  File /home/zope/lib/python/DateTime/DateTime.py, line 1111, in toZone
KeyError: (see above)
-- 
dante at oz.net - Michael Esveldt - #FightThePower on Openprojects


From donod at home.com  Sat Feb 17 10:32:10 2001
From: donod at home.com (Don O'Donnell)
Date: Sat, 17 Feb 2001 09:32:10 GMT
Subject: Thoughts on List Methods and Return Values
Message-ID: <3A8E44E3.3E25C7B3@home.com>

A recent posting (once again) discussed the fact that the list sort
method doesn't return a value.  The reason being, as pointed out by
Kalle:

"""
The sort() and reverse() methods modify the list in place for economy of
space when sorting or reversing a large list. They don't return the
sorted
or reversed list to remind you of this side effect.
"""

I always considered this a fairly legitimate reason and learned to live
with it.

Till today, when I needed to take a sequence of alternate paths supplied
as a function parameter, and insert the current directory as item [0],
to come up with a list of search paths.  Since I wanted to allow for
tuple as well as list input, I wrote:

	searchList = list(altSeq).insert(0, currPath)

And then, in order to eliminate possible empty strings, this became:

	searchList = filter(None, list(altSeq).insert(0, currPath))

After a few minutes of debugging, I realized why this doesn't work. 
Obviously, sort() and reverse() are not the only list methods that don't
return a value, insert() doesn't either.  So, I ended up with the
following:

	searchList = list(altSeq)
	searchList.insert(0, currPath)
	searchList = filter(None, searchList)

(I won't argue the fact that this may be more readable, but that's not
the issue.)

Of the 9 list methods listed on page 22 of Beazley's book, here's how
they break down:

Methods that return None:
	append, extend, insert, remove, reverse, sort

Methods that return a value:
	count, index, pop

Which leads me to the thought that if *all* the methods which now return
None were to modify the list in place as well as return the modified
list, we would soon get used to this behaviour and achieve some coding
economy by being able to chain operations.

In other words, the above argument against sort() and reverse()
returning their result seems to be because they would then behave
differently from the other list methods, leading us to be surprised by
this behaviour.  What if all the list methods returned a value?  Then
there would no longer be that element of surprise.  

I'm just throwing this out for discussion, knowing that it has no chance
of being implemented since it would break a lot of code.

Hmmm... after looking back at my initial incorrect (one-liner) code,
it's obvious that there's another problem -- the original input sequence
(altSeq) is modified even though it doesn't look like it is, and I might
use altSeq later on thinking it still contained it's original values. 
Whereas, in the correct 3-liner, it's obvious what's being modified and
what's not.  This can be easily remedied, however, by using the same
name for both input and result sequences:

	altSeq = filter(None, list(altSeq).insert(0, currPath))

But, of course, there no guarantee that this will always be done.

... So, maybe it's not such a great idea after all.  But after doing all
this typing I'm sure as hell not going to scrap this post  ;o)

talking-myself-out-of-it-ly y'rs

/Don


From fredrik at pythonware.com  Sat Feb 17 11:31:28 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 17 Feb 2001 10:31:28 GMT
Subject: Thoughts on List Methods and Return Values
References: <3A8E44E3.3E25C7B3@home.com>
Message-ID: <4qsj6.14072$Qb7.2221741@newsb.telia.net>

Don O'Donnell wrote:
> What if all the list methods returned a value?  Then
> there would no longer be that element of surprise.

    for item in mylist.reverse():
        print item
    for item in mylist.reverse():
        print item

Enough said /F




From aleaxit at yahoo.com  Sat Feb 17 19:02:12 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 17 Feb 2001 19:02:12 +0100
Subject: Thoughts on List Methods and Return Values
References: <3A8E44E3.3E25C7B3@home.com>
Message-ID: <96mea501uot@news2.newsguy.com>

"Don O'Donnell" <donod at home.com> wrote in message
news:3A8E44E3.3E25C7B3 at home.com...
    [snip]
> Methods that return None:
> append, extend, insert, remove, reverse, sort
    [snip]
> Which leads me to the thought that if *all* the methods which now return
> None were to modify the list in place as well as return the modified
> list, we would soon get used to this behaviour and achieve some coding
> economy by being able to chain operations.
    [snip]
> I'm just throwing this out for discussion, knowing that it has no chance
> of being implemented since it would break a lot of code.

Quite apart from any evaluation of this proposal, I fail to see what 'lot of
code' could possibly be broken by this change.  Is there a 'lot of code'
_relying_ on the None return, e.g.:
    foo = alist.sort()
as a shorthand for
    alist.sort()
    foo = None
...???  I sure hope not, and what code IS there that relies on this (if any)
somewhat looks like code that *deserves* to break...:-).


Alex





From tim.one at home.com  Sat Feb 17 20:42:49 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 17 Feb 2001 14:42:49 -0500
Subject: Thoughts on List Methods and Return Values
In-Reply-To: <96mea501uot@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEBCJAAA.tim.one@home.com>

[Don O'Donnell, about list methods]
> Methods that return None:
>    append, extend, insert, remove, reverse, sort
> ...
> Which leads me to the thought that if *all* the methods which now return
> None were to modify the list in place as well as return the modified
> list, we would soon get used to this behaviour and achieve some coding
> economy by being able to chain operations.
> ...

[Alex Martelli]
> Quite apart from any evaluation of this proposal, I fail to see
> what 'lot of code' could possibly be broken by this change.  Is
> there a 'lot of code' _relying_ on the None return, e.g.:
> ...

In a roundabout way, yes:

>>> x = [random.random() for i in xrange(1000000)]
>>> x.sort()

Oops!  Guess what that does to your interactive session if x.sort() starts
to return x.  People using my doctest.py would find that tons of their
docstring examples "suddenly don't work" because of that too.

BTW, it's not just lists, it's all mutate-in-place methods of all builtin
types that return None.

dict.update()-etc-ly y'rs  - tim




From aleaxit at yahoo.com  Sat Feb 17 23:06:25 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 17 Feb 2001 23:06:25 +0100
Subject: Thoughts on List Methods and Return Values
References: <mailman.982439002.3211.python-list@python.org>
Message-ID: <96msk502hoo@news2.newsguy.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.982439002.3211.python-list at python.org...
    [snip]
> > Which leads me to the thought that if *all* the methods which now return
> > None were to modify the list in place as well as return the modified
> > list, we would soon get used to this behaviour and achieve some coding
    [snip]
> [Alex Martelli]
> > Quite apart from any evaluation of this proposal, I fail to see
> > what 'lot of code' could possibly be broken by this change.  Is
> > there a 'lot of code' _relying_ on the None return, e.g.:
    [snip]
> In a roundabout way, yes:
>
> >>> x = [random.random() for i in xrange(1000000)]
> >>> x.sort()
>
> Oops!  Guess what that does to your interactive session if x.sort() starts

That's not "lots of code" that's going to be broken -- it's a currently
OK interactive-interpreter idiom that becomes non-ok (to be replaced
with, e.g., _=x.sort(), or whatever).  [Actually, it becomes OK again
if one uses the new interactive-session hook to prettyprint stuff returned
by interactive-session expressions, rather than just printing its repr, of
course; in fact, with a suitable prettyprint, it becomes a _plus_, since
one can see what's happening step by step as a list gets modified -- I
would argue that's a larger plus, e.g. for beginners' self-instruction,
than being able to chain calls...!-)].

> to return x.  People using my doctest.py would find that tons of their
> docstring examples "suddenly don't work" because of that too.

That's more serious, and I wish I had thought of that since I'm quite
an aficionado of your doctest.py.  The fixes I can think of would be
rather kludgey!

> BTW, it's not just lists, it's all mutate-in-place methods of all builtin
> types that return None.

Well, ALMOST all -- dict.setdefault 'may' mutate-in-place and may
not return None.  But that's a rather special case, of course.


Alex





From db3l at fitlinxx.com  Wed Feb 21 07:57:33 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 21 Feb 2001 01:57:33 -0500
Subject: Thoughts on List Methods and Return Values
References: <3A8E44E3.3E25C7B3@home.com>
Message-ID: <uzofgbk0y.fsf@ctwd0143.fitlinxx.com>

Don O'Donnell <donod at home.com> writes:

(...)

> After a few minutes of debugging, I realized why this doesn't work. 
> Obviously, sort() and reverse() are not the only list methods that don't
> return a value, insert() doesn't either.  So, I ended up with the
> following:
> 
> 	searchList = list(altSeq)
> 	searchList.insert(0, currPath)
> 	searchList = filter(None, searchList)
> 
> (I won't argue the fact that this may be more readable, but that's not
> the issue.)

(...)

> Which leads me to the thought that if *all* the methods which now return
> None were to modify the list in place as well as return the modified
> list, we would soon get used to this behaviour and achieve some coding
> economy by being able to chain operations.

Ah, so the point is trying to get coding economy in an arguably less
readable form?  Personally, I'd stick with giving readability (and
thus maintainability) the best chance at the expense of a little more
code text.

And as you note later,

> Hmmm... after looking back at my initial incorrect (one-liner) code,
> it's obvious that there's another problem -- the original input sequence
> (altSeq) is modified even though it doesn't look like it is, and I might
> use altSeq later on thinking it still contained it's original values. 
> Whereas, in the correct 3-liner, it's obvious what's being modified and
> what's not.  This can be easily remedied, however, by using the same
> name for both input and result sequences:

so if code just doesn't happen to have a convention for always
assigning to the same name as the modified list, it may be easy to
skip over while maintaining it.  I'd rather that Python itself gave a
little nudge in the right direction and ensured that the odds were in
the code's favor.

I've definitely noticed over time that when I tend to get a touch
annoyed at something in Python (and I've definitely had that feeling
at times with sort(), and the non-assignment in expressions, and so
on), that I'm sort of getting annoyed that I can't write a more
complicated piece of code that "feels" more elegant or compact, but in
reality is probably just a touch more obfuscated.  And in the end, the
"simpler" feeling code that's built out of more statements really is
just that - simpler to read, simpler to understand and simpler to
maintain.  While still behaving properly.  Amazing :-)

> ... So, maybe it's not such a great idea after all.  But after doing all
> this typing I'm sure as hell not going to scrap this post  ;o)

:-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From ransen_spam_me_not at nemo.it  Sat Feb 17 11:39:07 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Sat, 17 Feb 2001 10:39:07 GMT
Subject: Embedding Python into a VC6 app?
Message-ID: <3a8e543e.90577@news.newsguy.com>

Is there somewhere around a step by step guide to
embedding Python into a VC6 Windows application?


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From paul at fxtech.com  Sat Feb 17 17:45:30 2001
From: paul at fxtech.com (paul at fxtech.com)
Date: Sat, 17 Feb 2001 16:45:30 -0000
Subject: Embedding Python into a VC6 app?
In-Reply-To: <3a8e543e.90577@news.newsguy.com>
Message-ID: <96m9ra+qehf@eGroups.com>

--- In python-list at y..., ransen_spam_me_not at n... (Owen F. Ransen) 
wrote:
> Is there somewhere around a step by step guide to
> embedding Python into a VC6 Windows application?

Not really any step-by-step ones that I have seen. I am in the 
process of doing this myself, so was thinking of starting a "HOWTO" 
for Python embedding. I am sure people would find this interesting, 
and I'd love to have people contribute ideas.

To kick it off, this is what I did:

1. Obviously - download and install Python.
I did this with Python 2.0, but there was not a debug .lib included, 
so I got the latest CVS and built my own 2.1 DLL and LIB, then 
downloaded and installed the 2.1 alpha release.

2. Add a file to your project to isolate your Python stuff. Set an 
include path on it to include the Python headers directory.

3. In your file, do this:
#include <Python.h>

If the file compiles and your app links, you have set your headers up 
correctly. Also note that on Windows including Python.h will 
automatically add the appropriate link library for you (Python21_d.lib 
or Python21.lib, depending on the debug mode).

4. Add some code to initialize the Python interpreter, then register 
your app's extension functions. One of things you will want to do is 
redirect prints from within scripts (to stdout and stderr) to your own 
functions, so you can display the output in a window. Here is how I 
did it:

void InitPython()
{
    ::Py_SetProgramName("YOUR PROGRAM NAME");
    ::Py_Initialize();

    ::Py_InitModule("mymodule", my_methods);

    ::PyRun_SimpleString(
        "import mymodule\n"
        "class my_stdout:\n"
        "    def write(self, string):\n"
        "        mymodule.stdout(string)\n"
        "\n"
        "class my_stderr:\n"
        "    def write(self, string):\n"
        "        mymodule.stderr(string)\n"
        "\n"
        "import sys\n"
        "sys.stdout = my_stdout()\n"
        "sys.stderr = my_stderr()\n"
        "sys.path.append(PATH TO YOUR SCRIPTS)\n"
        "\n"
    );
}

static PyObject *s_stdout(PyObject *self, PyObject *args)
{
    char *string;
    if (!PyArg_ParseTuple(args, "s", &string))
        return NULL;

    // DO SOMETHING WITH THE STRING!
    // NOTE - Python new-lines only have a '\n' - you will need
    // to convert to "\r\n" on Windows

    Py_INCREF(Py_None);
    return Py_None;
}

static PyObject *s_stderr(PyObject *self, PyObject *args)
{
    char *string;
    if (!PyArg_ParseTuple(args, "s", &string))
        return NULL;

    // DO SOMETHING WITH THE STRING!

    Py_INCREF(Py_None);
    return Py_None;
}

struct PyMethodDef my_methods[] = 
{
    { "stdout",	s_stdout, METH_VARARGS },
    { "stderr",	s_stderr, METH_VARARGS },
    { NULL,	NULL }
};

When you are all done, call ::Py_Finalize().
Now you can call the other script functions to run your own scripts 
(or lines of code). 

You might want to check out http://www.mcmillan-inc.com/embed.html for 
some ideas.

I hope this helps get you started.





From ransen_spam_me_not at nemo.it  Sat Feb 17 18:42:04 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Sat, 17 Feb 2001 17:42:04 GMT
Subject: Embedding Python into a VC6 app?
References: <3a8e543e.90577@news.newsguy.com>
Message-ID: <3a8e9b6f.3670682@news.newsguy.com>

Oops, I've just found out how to do it, just 
read the docs Owen!


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From ransen_spam_me_not at nemo.it  Sun Feb 18 08:29:14 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Sun, 18 Feb 2001 07:29:14 GMT
Subject: Embedding Python into a VC6 app?
References: <mailman.982428393.27150.python-list@python.org>
Message-ID: <3a8fbc27.1282219@news.newsguy.com>

On Sat, 17 Feb 2001 16:45:30 -0000, paul at fxtech.com wrote:

>--- In python-list at y..., ransen_spam_me_not at n... (Owen F. Ransen) 
>wrote:
Thanks a lot for the pointers, I have managed to get a small
example going, but your post is going to help me a lot 
--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From mrbun at bigfoot.com  Sat Feb 17 12:00:34 2001
From: mrbun at bigfoot.com (Mike Bunyan)
Date: Sat, 17 Feb 2001 11:00:34 +0000
Subject: UserName automated authentication
Message-ID: <kcis8t81qi3olrspss0c8ck33b64jvqjp5@4ax.com>

I am experimenting with Zope to provide an intranet for 100 or so
people. Looking at achieving an automated (no brainer) authentication
procedure which runs when users access zope through their browser. The
network is NetWare, but I am not allowed to use python NetWare
modules, the users have WinNT4sp5 and MSIE5.5. NT authentication is
not available.
The idea is to use a python script (or sniffer) to get UserName and
authenticate this against a permissions table in mysql or gadfly. The
user does not have to complete any login and cookies would persist for
session.

I am puzzled by the behaviour of each of the following calls:
A = os.system('echo %username%')
B = win32api.GetUserName()
C = os.environ['USERNAME']

My innocence suggests that each should provide the same information.
Called from DOS or python gui the results are as expected except that
A produces:
>>>FBASSET
0
and I can only get 0 as a result and not catch the username.

When any of the above are called through an external zope procedure:
A = 0
B = 'system'
C fails because 'username' does not exist.

The way I read this is that the client workstation does not allow or
reveal 'secure' information through the browser.

Any ideas how to solve this problem, including alternative approaches
are welcome, and some illumination on what is actually happening will
be appreciated.

Mike B
___________________
mrbun at bigfoot.com


From stephen_purcell at yahoo.com  Sat Feb 17 12:31:50 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sat, 17 Feb 2001 12:31:50 +0100
Subject: UserName automated authentication
In-Reply-To: <kcis8t81qi3olrspss0c8ck33b64jvqjp5@4ax.com>; from mrbun@bigfoot.com on Sat, Feb 17, 2001 at 11:00:34AM +0000
References: <kcis8t81qi3olrspss0c8ck33b64jvqjp5@4ax.com>
Message-ID: <20010217123150.A18069@freedom.puma-ag.com>

Mike Bunyan wrote:
> I am experimenting with Zope to provide an intranet for 100 or so
> people.
> [snip]
> The idea is to use a python script (or sniffer) to get UserName and
> authenticate this against a permissions table in mysql or gadfly. The
> user does not have to complete any login and cookies would persist for
> session.
> 
> I am puzzled by the behaviour of each of the following calls:
> A = os.system('echo %username%')
> B = win32api.GetUserName()
> C = os.environ['USERNAME']

You can't use this kind of code from within Zope, because anything executed
there is effectively being executed by the user who started the Zope server
('system' -- the NT service manager?), not the user browsing the server.

You'll have to get users to enter their username and password, as with any
other web service.

For that, I'm sure there are a wealth of authentication/user database products
available for Zope.

Best wishes,

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From zzizz_ at notmail.com  Sat Feb 17 14:10:59 2001
From: zzizz_ at notmail.com (zzzzz)
Date: Sun, 18 Feb 2001 00:10:59 +1100
Subject: Launch Dial Up Connection in Win9X
Message-ID: <q0us8t89bduraek2teijtm6kgbuq5jn58k@4ax.com>

What's the best way to launch a dial up connection in Win9X?




From r2d2 at mao.acc.umu.se  Mon Feb 19 09:05:38 2001
From: r2d2 at mao.acc.umu.se (Niklas Frykholm)
Date: Mon, 19 Feb 2001 08:05:38 +0000 (UTC)
Subject: Launch Dial Up Connection in Win9X
References: <q0us8t89bduraek2teijtm6kgbuq5jn58k@4ax.com>
Message-ID: <slrn991kui.u76.r2d2@mao.acc.umu.se>

In article <q0us8t89bduraek2teijtm6kgbuq5jn58k at 4ax.com>, zzzzz wrote:
>What's the best way to launch a dial up connection in Win9X?

I had this problem two days ago :). I ended up interfacing a freeware
program called "RASDial95". It has a simple command line interface
that lets you dial up a certain service, check the connection status
or hang-up.

A google search on RASDial95 will give you download links.

// Niklas


From zzizz_ at notmail.com  Tue Feb 20 03:37:56 2001
From: zzizz_ at notmail.com (zzzzz)
Date: Tue, 20 Feb 2001 13:37:56 +1100
Subject: Launch Dial Up Connection in Win9X
References: <q0us8t89bduraek2teijtm6kgbuq5jn58k@4ax.com> <slrn991kui.u76.r2d2@mao.acc.umu.se>
Message-ID: <ivl39tstag90lhgecfm9thn40o9n50rjr5@4ax.com>

Is there any reason why I can't use win32ras module from Pythonwin on
win 9X? Is it NT only?

On Mon, 19 Feb 2001 08:05:38 +0000 (UTC), r2d2 at mao.acc.umu.se (Niklas
Frykholm) wrote:

>In article <q0us8t89bduraek2teijtm6kgbuq5jn58k at 4ax.com>, zzzzz wrote:
>>What's the best way to launch a dial up connection in Win9X?
>
>I had this problem two days ago :). I ended up interfacing a freeware
>program called "RASDial95". It has a simple command line interface
>that lets you dial up a certain service, check the connection status
>or hang-up.
>
>A google search on RASDial95 will give you download links.
>
>// Niklas

Regards,

zzzzz.
---------------------


From victor at chaoszero.com  Sat Feb 17 14:16:39 2001
From: victor at chaoszero.com (victor at chaoszero.com)
Date: Sat, 17 Feb 2001 14:16:39 0100
Subject: Chaos Zero Web Site
Message-ID: <982415799.710@inwind.it>

Hi, I'm Victor from CHAOS ZERO.

Visit http://www.chaoszero.com and you will find a new Crossover-Alternative Rock Band

In MUSIC section you'll find some promo mp3 and a full-lenght mp3 version of
the new rearranged cover of "The Wild Boys" by Duran Duran

We are searching for a label, distribution and promotion.

I'm just self-promoting my band and i'm doing it manually!

IF YOU'LL LIKE CHAOS ZERO, PLEASE FOWARD THIS MESSAGE TO YOUR FRIENDS
AND SPREAD THE VOICE!

Thanx. Victor

HTTP://WWW.CHAOSZERO.COM

victor at chaoszero.com




From hwanjoyu at students.uiuc.edu  Sat Feb 17 16:15:22 2001
From: hwanjoyu at students.uiuc.edu (hwan-jo yu)
Date: Sat, 17 Feb 2001 09:15:22 -0600
Subject: Q: break multiple loop
Message-ID: <Pine.GSO.4.10.10102170913380.22713-100000@ux8.cso.uiuc.edu>

Hi,
It seems that there is now "goto" statement in the python.
How can we terminate multiple for or while loop ?
Thanks.



From danielk at aracnet.com  Sat Feb 17 17:35:10 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sat, 17 Feb 2001 08:35:10 -0800
Subject: Q: break multiple loop
References: <Pine.GSO.4.10.10102170913380.22713-100000@ux8.cso.uiuc.edu>
Message-ID: <qu9t8tkqhk0rquvqtj2duacgmi7bufjcc7@4ax.com>

On Sat, 17 Feb 2001 09:15:22 -0600, hwan-jo yu <hwanjoyu at students.uiuc.edu>
wrote:

>How can we terminate multiple for or while loop ?

Post some pseudo-code of what you are trying to do and we will try to help you
out.

Dan



From stephen_purcell at yahoo.com  Sat Feb 17 17:42:54 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sat, 17 Feb 2001 17:42:54 +0100
Subject: Q: break multiple loop
In-Reply-To: <Pine.GSO.4.10.10102170913380.22713-100000@ux8.cso.uiuc.edu>; from hwanjoyu@students.uiuc.edu on Sat, Feb 17, 2001 at 09:15:22AM -0600
References: <Pine.GSO.4.10.10102170913380.22713-100000@ux8.cso.uiuc.edu>
Message-ID: <20010217174254.A18953@freedom.puma-ag.com>

hwan-jo yu wrote:
> It seems that there is now "goto" statement in the python.

True!

> How can we terminate multiple for or while loop ?

If you need to jump right out of nested loops, this is usually a sign that
there is an easier way to arrange your loop: maybe the inner loop should be
a separate function or method? But if you insist use a flag variable:

   keepgoing = 1
   while loop_condition1 and keepgoing:
      while loop_condition2:
         if we_should_stop_now():
             keepgoing = 0
             break

or

   keepgoing = 1
   while loop_condition1:
      while loop_condition2:
         if we_should_stop_now():
             keepgoing = 0
             break
      if not keepgoing: break

Another trick is with an exception:

   try:
       while loop_condition1:
           while loop_condition_2:
               if we_should_stop_now():
                    raise "stopped"
   except "stopped":
       pass

But forget I said that -- it's nasty.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From erno-news at erno.iki.fi  Sat Feb 17 17:59:38 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 17 Feb 2001 18:59:38 +0200
Subject: Q: break multiple loop
References: <Pine.GSO.4.10.10102170913380.22713-100000@ux8.cso.uiuc.edu>
Message-ID: <kuu25ttfd1.fsf@lasipalatsi.fi>

In article
<Pine.GSO.4.10.10102170913380.22713-100000 at ux8.cso.uiuc.edu>, hwan-jo
yu <hwanjoyu at students.uiuc.edu> writes:

| Hi,
| It seems that there is now "goto" statement in the python.
| How can we terminate multiple for or while loop ?

you can raise an exception.

  -- erno


From phlip_cpp at my-deja.com  Sat Feb 17 18:20:25 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 17 Feb 2001 17:20:25 GMT
Subject: Q: break multiple loop
References: <Pine.GSO.4.10.10102170913380.22713-100000@ux8.cso.uiuc.edu>
Message-ID: <96mbsp$bf8@dispatch.concentric.net>

Proclaimed hwan-jo yu from the mountaintops:

> Hi,
> It seems that there is now "goto" statement in the python.
> How can we terminate multiple for or while loop ?

        return

The rule "no absurdly long functions" trumps the rule "goto is the root of 
all evil". But it would be hard for a language to enforce the former. Show 
some discipline!

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  This machine last rebooted during the Second Millenium  --


From root at rainerdeyke.com  Sat Feb 17 18:36:56 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Sat, 17 Feb 2001 17:36:56 GMT
Subject: break multiple loop
References: <Pine.GSO.4.10.10102170913380.22713-100000@ux8.cso.uiuc.edu>
Message-ID: <YEyj6.249516$ge4.86293290@news2.rdc2.tx.home.com>

"hwan-jo yu" <hwanjoyu at students.uiuc.edu> wrote in message
news:Pine.GSO.4.10.10102170913380.22713-100000 at ux8.cso.uiuc.edu...
> Hi,
> It seems that there is now "goto" statement in the python.
> How can we terminate multiple for or while loop ?
> Thanks.

Either 'return' or 'raise'.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From paul at fxtech.com  Sat Feb 17 16:58:07 2001
From: paul at fxtech.com (paul at fxtech.com)
Date: Sat, 17 Feb 2001 15:58:07 -0000
Subject: Python 2.1 - tokenizer fgets crash (Windows)
Message-ID: <96m72f+g05a@eGroups.com>

This is weird. I am using the 2.1 debug DLL/lib built from the CVS 
sources in an embedded app. I can init the interpreter and run some 
startup code with PyRun_SimpleString() (to redirect stdout/stderr, 
etc). But when I try to use PyRun_AnyFile() it crashes in the 
tokenizer at the first fgets() call. Stepping into the code, 
everything looks "good" up to that call. I suspect some kind of 
library/DLL settings mismatch. 

Anyone have any clues?





From tim.one at home.com  Sat Feb 17 19:26:40 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 17 Feb 2001 13:26:40 -0500
Subject: Python 2.1 - tokenizer fgets crash (Windows)
In-Reply-To: <96m72f+g05a@eGroups.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEANJAAA.tim.one@home.com>

[paul at fxtech.com]
> This is weird. I am using the 2.1 debug DLL/lib built from the CVS
> sources in an embedded app. I can init the interpreter and run some
> startup code with PyRun_SimpleString() (to redirect stdout/stderr,
> etc). But when I try to use PyRun_AnyFile() it crashes in the
> tokenizer at the first fgets() call. Stepping into the code,
> everything looks "good" up to that call. I suspect some kind of
> library/DLL settings mismatch.
>
> Anyone have any clues?

You do:  "crashes" could mean anything on Windows.  What does it mean to
you?  Is it dying inside MS's implementation of fgets(), in Python; is it a
NULL-pointer error, an illegal instruction; etc etc.  Try it under a debug
build, if you haven't already (single-stepping can be highly misleading if
you're doing that with the non-debug build).





From paul at fxtech.com  Sat Feb 17 19:55:21 2001
From: paul at fxtech.com (paul at fxtech.com)
Date: Sat, 17 Feb 2001 18:55:21 -0000
Subject: Python 2.1 - tokenizer fgets crash (Windows)
In-Reply-To: <LNBBLJKPBEHFEDALKOLCIEANJAAA.tim.one@home.com>
Message-ID: <96mhep+u7oa@eGroups.com>

> You do:  "crashes" could mean anything on Windows.  What does it 
mean to
> you?  Is it dying inside MS's implementation of fgets(), in Python; 
is it a
> NULL-pointer error, an illegal instruction; etc etc.  Try it under a 
debug
> build, if you haven't already (single-stepping can be highly 
misleading if
> you're doing that with the non-debug build).

Yes, I'm using the debug build (as I mentioned) and I have 
single-stepped down into the code. It's crashing in Microsoft's 
fgets() function. The FILE * looks correct - the only thing I can 
think of is I'm doing the fopen() in my code and it is getting used in 
the DLL. Loading files from within the DLL (using Import) work fine.

I've gotten around it for now - I'm now importing my scripts as 
modules rather than running them.





From tim.one at home.com  Sat Feb 17 20:32:01 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 17 Feb 2001 14:32:01 -0500
Subject: Python 2.1 - tokenizer fgets crash (Windows)
In-Reply-To: <96mhep+u7oa@eGroups.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEBCJAAA.tim.one@home.com>

[paul at fxtech.com]
> Yes, I'm using the debug build (as I mentioned) and I have
> single-stepped down into the code. It's crashing in Microsoft's
> fgets() function. The FILE * looks correct - the only thing I can
> think of is I'm doing the fopen() in my code and it is getting used in
> the DLL. Loading files from within the DLL (using Import) work fine.

Ya, sure sounds like you're mixing DLL modes.  Can only suggest you
triple-check the DLLs you link with.  Make sure your own code is using the
MS multithreaded DLLs; don't know how you set up your project, but it's
(too!) easy to blow that part under DevStudio; see PC/example_nt/readme.txt
for details.




From johan at essay.org  Sat Feb 17 17:27:13 2001
From: johan at essay.org (=?iso-8859-1?Q?Johan_Fredrik_=D8hman?=)
Date: Sat, 17 Feb 2001 17:27:13 +0100
Subject: shared fails but static works
Message-ID: <96m8oa$7ek$1@maud.ifi.uio.no>

Hi,  I'm experiencing a strange problem when writing a C++ Python modules.  I assume this is a quite difficult problem to answer.  So I have included most of the info I see as important for those who want to have a look.

   The module (included below) works perfectly when I build the module staticly.  But when I enable *shared* in the Setupfile I get the following errormessage in /var/log/messages when the so. module is "run" from the python source code.

Errormessages in /var/log/messages

Feb 17 15:32:50 blekkulf python: PAM unable to dlopen(/lib/security/pam_unix.so)
Feb 17 15:32:50 blekkulf python: PAM [dlerror: /lib/security/pam_unix.so: undefined symbol: pam_get_item]
Feb 17 15:32:50 blekkulf python: PAM adding faulty module: /lib/security/pam_unix.so
Feb 17 15:32:50 blekkulf python: PAM unable to dlopen(/lib/security/pam_cracklib.so)
Feb 17 15:32:50 blekkulf python: PAM [dlerror: /lib/security/pam_cracklib.so: undefined symbol: pam_get_item]
Feb 17 15:32:50 blekkulf python: PAM adding faulty module: /lib/security/pam_cracklib.so
Feb 17 15:32:50 blekkulf python: PAM unable to dlopen(/lib/security/pam_limits.so)
Feb 17 15:32:50 blekkulf python: PAM [dlerror: /lib/security/pam_limits.so: undefined symbol: pam_get_item]
Feb 17 15:32:50 blekkulf python: PAM adding faulty module: /lib/security/pam_limits.so

I understand that the dloader has I problem here,  but I don't know how to solve it.   I have also tried the program as a standalone C++ program with no problems at all.  So the PAM library should be intact.

If you want to try youselv, be my guest and try to compile the program below as a shared module !

Any ideas ?

Thanks

JFO


#########################


#include <stdlib.h>
#include <unistd.h>
#include <string.h>
#include <iostream.h>

#include <stdio.h>

//extern "C" {
#include <security/pam_appl.h>
#include <security/pam_misc.h>
#include <Python.h>
//}


/* Static variables used to communicate between the conversation function
 * and the server_login function
 */
static const char *PAM_password;

/* hackish PAM conversation function
 * Here we assume that echo off means password.
 */
static int hack_conv (
                      int num_msg,
                      const struct pam_message **msg,
                      struct pam_response **resp,
                      void *)
{
#define COPY_STRING(s) (s) ? strdup(s) : (char*)NULL
  //cout << "PAM_PASSWORD:" << PAM_password <<"\n";
  struct pam_response *reply 
    = (struct pam_response*)malloc(sizeof(struct pam_response) * num_msg);
  if (!reply) return PAM_CONV_ERR;

  for (int replies = 0; replies < num_msg; replies++) {
    switch (msg[replies]->msg_style) {
    case PAM_PROMPT_ECHO_OFF:
      reply[replies].resp_retcode = PAM_SUCCESS;
      //cout << "PAM_PASSWORD:" << PAM_password <<"\n";
      reply[replies].resp = COPY_STRING(PAM_password);
      /* PAM frees resp */
      break;
    case PAM_TEXT_INFO:
      /* fall through */
    case PAM_ERROR_MSG:
      /* ignore it, but pam still wants a NULL response... */
      reply[replies].resp_retcode = PAM_SUCCESS;
      reply[replies].resp = NULL;
      break;
    case PAM_PROMPT_ECHO_ON:
      /* fall through */
    default:
      /* Must be an error of some sort... */
      free (reply);
      return PAM_CONV_ERR;
    }
  }
  *resp = reply;
  return PAM_SUCCESS;
}

static struct pam_conv hack_conversation = {
  &hack_conv,
  NULL
};


static int pam_check_pair(const char *user, const char *pass)
{
  //cout << "debug: user:" << user << pass;
  int retval;
  pam_handle_t *pamh = NULL;
  
  PAM_password = pass;
  
  retval = pam_start ("system-auth", user, &hack_conversation, &pamh);
  
  if (retval == PAM_SUCCESS) {
    retval = pam_authenticate(pamh, 0);
    //cout << "Returnvalue1: " << retval <<"\n";
    if (retval == PAM_SUCCESS){
      retval = pam_acct_mgmt(pamh, 0);
      //cout << "Returnvalue2: " << retval <<"\n";
    }
  }
  
  if (pamh != NULL)
    pam_end(pamh, PAM_SUCCESS);

  if (retval != PAM_SUCCESS){
    //cout << "Failure\n";
    return 0;
  }
  else {
    //cout << "Success\n";
    return 1;
  }
}


static PyObject* verify_user( PyObject *self, PyObject *args)
{
  char *user;
  char *password;
  int sts;

  if (!PyArg_ParseTuple(args, "ss", &user, &password))
    return NULL;
  //cout << "user: " << user  << " password: " << password;
  sts = pam_check_pair(user, password);
  return Py_BuildValue("i", sts);
}

static PyMethodDef jfopamMethods[] = {
  {"verify_user",  verify_user, METH_VARARGS},
  {NULL,      NULL}        /* Sentinel */
};

extern "C" void initjfopam()
  {
    (void) Py_InitModule("jfopam", jfopamMethods);
  }
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010217/83537e03/attachment.htm>

From paul at fxtech.com  Sat Feb 17 17:31:03 2001
From: paul at fxtech.com (paul at fxtech.com)
Date: Sat, 17 Feb 2001 16:31:03 -0000
Subject: Restarting embedded Python + extensions
Message-ID: <96m907+mmbd@eGroups.com>

I have Python embedded in a C++ app using Python scripts as extension 
modules. When testing extension scripts I want to be able to restart 
the Python session from scratch and reload all the modules, so I was 
trying to call Py_Finalize() followed by Py_Initialize(). This causes 
a crash (at least for Windows, 2.1 alpha, debug) in the builtin setup 
step in Py_Initialize().

So I was thinking of using a PyThreadState and Py_NewInterpreter() to 
run my extension scrips, and when I want to restart my extensions just 
call Py_EndInterpreter() and Py_NewInterpreter and reload my extension 
modules under the new interpreter.

Does this sound like the right approach?





From muk_pr at yahoo.com  Sat Feb 17 17:56:22 2001
From: muk_pr at yahoo.com (Victor Black)
Date: Sat, 17 Feb 2001 18:56:22 +0200
Subject: Better pyclbr
Message-ID: <t8tc20kgte58e2@corp.supernews.com>

Hi,
Are there any pyclbr like modules that can gimme more properties of  the
code(members,methods of the classes,global functions,variables).
PS I know they are available by dir() or other methods but what I want is
that will parse the code itself.

Regards.





From doughellmann at bigfoot.com  Sat Feb 17 21:49:51 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Sat, 17 Feb 2001 20:49:51 GMT
Subject: Better pyclbr
References: <t8tc20kgte58e2@corp.supernews.com>
Message-ID: <PtBj6.69389$m5.3968295@news1.rdc1.ga.home.com>

The parseinfo.py module of HappyDoc is based on a module of the same name
from the examples of the source distribution for 1.5.2.  I've added some
extra features such as extracting the raised exceptions, etc.

Check it out at http://happydoc.sourceforge.net.

Doug

In article <t8tc20kgte58e2 at corp.supernews.com>, "Victor Black"
<muk_pr at yahoo.com> wrote:

> Hi, Are there any pyclbr like modules that can gimme more properties of 
> the code(members,methods of the classes,global functions,variables). PS
> I know they are available by dir() or other methods but what I want is
> that will parse the code itself.
> 
> Regards.
> 
> 
>


From DOUGS at oceanic.com  Sat Feb 17 19:18:52 2001
From: DOUGS at oceanic.com (Doug Stanfield)
Date: Sat, 17 Feb 2001 08:18:52 -1000
Subject: file access
Message-ID: <8457258D741DD411BD3D0050DA62365907A625@huina.oceanic.com>

[Ben de Luca wrote:]
> > when i write the data to a file whats the best way? I guess 
> there is no way
> > to sumb the dictonaries strait to disk? i will have to parse it?

[Fredrik Lundh suggested:]
 
> http://www.python.org/doc/current/lib/module-pickle.html
> http://www.python.org/doc/current/lib/module-marshal.html

I'd also suggest checking out shelve:

http://www.python.org/doc/current/lib/module-shelve.html



From gahanr at gofree.indigo.ie  Sat Feb 17 20:12:57 2001
From: gahanr at gofree.indigo.ie (Robert Gahan)
Date: Sat, 17 Feb 2001 19:12:57 -0000
Subject: Writing to a COM port (windows)
Message-ID: <AhAj6.2566$r4.2220@news.indigo.ie>

Hi all,

I'm reasonably new to Python and very much a novice as far as windows
programming is concerned.

I am looking to write to a device which is connected to my COM1 port and am
looking for an
appropriate python method. (effectively looking to write to the device
driver)

The "C" equivalent for doing this is:

 hCom = CreateFile(szPort, GENERIC_READ | GENERIC_WRITE,
  FILE_SHARE_READ, NULL, OPEN_EXISTING,
  FILE_ATTRIBUTE_NORMAL, NULL);

 // setup device buffers
 SetupComm(hCom, 4096, 4096 ) ;

Is there any way of doing this from Python ???.

import os
f = open("COM1", 'w')
f.write("at012345678")
print f
f.close()

Traceback (innermost last):
  File "C:/Python20/smpp1", line 2, in ?
    f = open("COM1", 'w')
IOError: [Errno 13] Permission denied: 'COM1'

Thanks again,
Rob.




From samschul at pacbell.net  Sat Feb 17 21:18:03 2001
From: samschul at pacbell.net (Sam Schulenburg)
Date: Sat, 17 Feb 2001 12:18:03 -0800
Subject: Writing to a COM port (windows)
References: <AhAj6.2566$r4.2220@news.indigo.ie>
Message-ID: <O3Bj6.34$lK3.55234@news.pacbell.net>

Get Mark Hammonds win32all
http://www.activestate.com/Products/ActivePython/win32all.html

Mark has a win32comport_demo.py in the win32\demos directory.

Sam Schulenburg

"Robert Gahan" <gahanr at gofree.indigo.ie> wrote in message
news:AhAj6.2566$r4.2220 at news.indigo.ie...
> Hi all,
>
> I'm reasonably new to Python and very much a novice as far as windows
> programming is concerned.
>
> I am looking to write to a device which is connected to my COM1 port and
am
> looking for an
> appropriate python method. (effectively looking to write to the device
> driver)
>
> The "C" equivalent for doing this is:
>
>  hCom = CreateFile(szPort, GENERIC_READ | GENERIC_WRITE,
>   FILE_SHARE_READ, NULL, OPEN_EXISTING,
>   FILE_ATTRIBUTE_NORMAL, NULL);
>
>  // setup device buffers
>  SetupComm(hCom, 4096, 4096 ) ;
>
> Is there any way of doing this from Python ???.
>
> import os
> f = open("COM1", 'w')
> f.write("at012345678")
> print f
> f.close()
>
> Traceback (innermost last):
>   File "C:/Python20/smpp1", line 2, in ?
>     f = open("COM1", 'w')
> IOError: [Errno 13] Permission denied: 'COM1'
>
> Thanks again,
> Rob.
>
>
>




From gst at sysfrog.org  Sat Feb 17 21:20:11 2001
From: gst at sysfrog.org (Guenther Starnberger)
Date: 17 Feb 2001 21:20:11 +0100
Subject: problem with python/gtk
Message-ID: <87elwxysck.fsf@nirvana.sysfrog.org>

hi,

i have a problem using gtk with python.

in a programm i open a gtkwindow foo with foo.show().

closing the window with foo.hide() and reopening it with foo.show()
works.

but when i close the window with the windowmanager i can't reopen
it. i only get an empty window and the following error message:

----->8------>8------>8---------->8------>8------>8------>8------
Gtk-CRITICAL **: file gtkstyle.c: line 522 (gtk_style_attach):
assertion `style != NULL' failed.

Gtk-CRITICAL **: file gtkstyle.c: line 1128
(gtk_style_set_background): assertion `style != NULL' failed.

Gtk-CRITICAL **: file gtkstyle.c: line 3532 (gtk_paint_flat_box):
assertion `style != NULL' failed.
----->8------>8------>8---------->8------>8------>8------>8------

i tried to make an handler for the signal "destroy" which calls
foo.hide(). the handler gets called, but the error is still the same.

any suggestions?

cu
/gst

-- 
sysfrog.org


From lsloan at bigfoot.com  Sat Feb 17 23:26:29 2001
From: lsloan at bigfoot.com (Lance E Sloan)
Date: Sat, 17 Feb 2001 17:26:29 -0500
Subject: date calculations ala Perl's Date::Manip?
Message-ID: <3A8EFA96.37DE948C@my-deja.com>

I've been using Perl for a while and I'm now learning
Python.  I'm looking for a Python module that provides
the same fancy date calculations as Perl's Date::Manip.
What would you recommend?

I would appreciate it if responses to this article are
CC'ed to lsloan00 at my-deja.com as well as posted to this
newsgroup.  My newsserver is unreliable and I'd hate to
miss a response.

Thank you in advance!


From jbauer at rubic.com  Sun Feb 18 00:46:44 2001
From: jbauer at rubic.com (Jeff Bauer)
Date: Sat, 17 Feb 2001 17:46:44 -0600
Subject: date calculations ala Perl's Date::Manip?
References: <3A8EFA96.37DE948C@my-deja.com>
Message-ID: <3A8F0D64.FD238632@rubic.com>

Lance E Sloan wrote:
> I've been using Perl for a while and I'm now learning
> Python.  I'm looking for a Python module that provides
> the same fancy date calculations as Perl's Date::Manip.
> What would you recommend?

Here's my Python NormalDate module that I use for date 
calculations:

  http://starship.python.net/crew/jbauer/normaldate/

Jeff Bauer
Rubicon Research



From aahz at panix.com  Sun Feb 18 01:14:45 2001
From: aahz at panix.com (Aahz Maruch)
Date: 17 Feb 2001 16:14:45 -0800
Subject: date calculations ala Perl's Date::Manip?
References: <3A8EFA96.37DE948C@my-deja.com>
Message-ID: <96n45l$o41$1@panix3.panix.com>

[p&e as requested]

In article <3A8EFA96.37DE948C at my-deja.com>,
Lance E Sloan  <lsloan at bigfoot.com> wrote:
>
>I've been using Perl for a while and I'm now learning Python.  I'm
>looking for a Python module that provides the same fancy date
>calculations as Perl's Date::Manip.  What would you recommend?

Don't know what features are in Date::Manip, but mxDateTime is the
canonical Python module.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From kalle at gnupung.net  Sun Feb 18 02:07:08 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Sun, 18 Feb 2001 02:07:08 +0100
Subject: date calculations ala Perl's Date::Manip?
In-Reply-To: <96n45l$o41$1@panix3.panix.com>; from aahz@panix.com on Sat, Feb 17, 2001 at 04:14:45PM -0800
References: <3A8EFA96.37DE948C@my-deja.com> <96n45l$o41$1@panix3.panix.com>
Message-ID: <20010218020708.A1537@apone.network.loc>

Sez Aahz Maruch:
> [p&e as requested]
> 
> In article <3A8EFA96.37DE948C at my-deja.com>,
> Lance E Sloan  <lsloan at bigfoot.com> wrote:
> >
> >I've been using Perl for a while and I'm now learning Python.  I'm
> >looking for a Python module that provides the same fancy date
> >calculations as Perl's Date::Manip.  What would you recommend?
> 
> Don't know what features are in Date::Manip, but mxDateTime is the
> canonical Python module.

And is found on http://www.lemburg.com/files/python/

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010218/b33ecf9d/attachment.pgp>

From stevedperkins at NOSPAM.hotmail.com  Sat Feb 17 23:46:47 2001
From: stevedperkins at NOSPAM.hotmail.com (Steve D. Perkins)
Date: Sat, 17 Feb 2001 17:46:47 -0500
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org>
Message-ID: <96mv3102l3d@news2.newsguy.com>

Untitled Document    ARGH!

    I wll never understand why spam still exists, the internet reaching the
level of development and maturity is has.  I can't imagine that there's
ANYONE left anymore dumb and/or internet-inexperienced to purchase anything
advertised through spam.  Even if there is, I don't understand why spammers
woud target newsgroups for Tcl, Python, and Perl... technologies most often
used by crusty old-timers LEAST likely of all to be an effective target
audience...




From mindfield at badgers.emulationnet.com  Sun Feb 18 04:27:19 2001
From: mindfield at badgers.emulationnet.com (Mindfield)
Date: Sun, 18 Feb 2001 03:27:19 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com>
Message-ID: <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com>

On Sat, 17 Feb 2001 17:46:47 -0500, "Steve D. Perkins"
<stevedperkins at NOSPAM.hotmail.com> drooled an embarassing wet spot in
his lap as he typed:

>Untitled Document    ARGH!
>
>    I wll never understand why spam still exists, the internet reaching the
>level of development and maturity is has.  I can't imagine that there's
>ANYONE left anymore dumb and/or internet-inexperienced to purchase anything
>advertised through spam.  Even if there is, I don't understand why spammers
>woud target newsgroups for Tcl, Python, and Perl... technologies most often
>used by crusty old-timers LEAST likely of all to be an effective target
>audience...

Blame spambots.  They retrieve all available newsgroups on the
spammer's ISP, then post their crap to every single one of them.
Fortunately many ISPs have gotten wise and have placed posting caps on
duplicate messages, but spambots have found ways to circumvent that --
by placing a random number after the subject line, by rotating the
subject line from a list of subject lines, by placing random ASCII
text at the end of each message, etc. so the server doesn't think it's
a duplicate message.  It's really quite a loathesome little business,
and I've never figured out why people are still stupid enough to think
that it works, but I guess they still abide by the old motto, "There's
no such thing as bad publicity."

Ah, but there is if you lose your ISP for violating their TOS/AUP.
And there are plenty of people -- me, for instance -- who are
perfectly happy to send a nice note to their ISP's abuse department.
:)

-- 

Almost there...

"Ignorance breeds confidence more often than does knowledge."  -Darwin
The Emulation Newbie FAQ: http://www.emulationnet.com/emufaq.html

Remove the badgers before E-Mailing.


From sheila at spamcop.net  Sun Feb 18 05:01:13 2001
From: sheila at spamcop.net (Sheila King)
Date: Sun, 18 Feb 2001 04:01:13 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com>
Message-ID: <r5iu8tsqccj65ote3hfgo6dvjphskn15l3@4ax.com>

On Sun, 18 Feb 2001 03:27:19 GMT, Mindfield
<mindfield at badgers.emulationnet.com> wrote in comp.lang.python in article
<21gu8tsoknv01eegt4jak2lo7180kijno5 at 4ax.com>:

:It's really quite a loathesome little business,
:and I've never figured out why people are still stupid enough to think
:that it works, but I guess they still abide by the old motto, "There's
:no such thing as bad publicity."

It costs them next to nothing, and they must end up selling some of their
product or generating some revenue from the spam. If they didn't, the spamming
would decrease, but instead it is on the increase. Some people are stupid
enough to buy from spammers.

:Ah, but there is if you lose your ISP for violating their TOS/AUP.
:And there are plenty of people -- me, for instance -- who are
:perfectly happy to send a nice note to their ISP's abuse department.

Me, too.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From hs at paradise.nirvananet  Mon Feb 19 02:11:17 2001
From: hs at paradise.nirvananet (Hartmann Schaffer)
Date: 18 Feb 2001 20:11:17 -0500
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com> <r5iu8tsqccj65ote3hfgo6dvjphskn15l3@4ax.com>
Message-ID: <slrn990n47.382.hs@paradise.nirvananet>

In article <r5iu8tsqccj65ote3hfgo6dvjphskn15l3 at 4ax.com>, Sheila King wrote:
>On Sun, 18 Feb 2001 03:27:19 GMT, Mindfield
><mindfield at badgers.emulationnet.com> wrote in comp.lang.python in article
><21gu8tsoknv01eegt4jak2lo7180kijno5 at 4ax.com>:
>
>:It's really quite a loathesome little business,
>:and I've never figured out why people are still stupid enough to think
>:that it works, but I guess they still abide by the old motto, "There's
>:no such thing as bad publicity."
>
>It costs them next to nothing, and they must end up selling some of their
>product or generating some revenue from the spam. If they didn't, the spamming
>would decrease, but instead it is on the increase. Some people are stupid
>enough to buy from spammers.

frankly, i doubt this.  most of those spams are so inept that i would be
surprised if it generates any sales.  the only thing that seems to generate
sales are those spam generators (to clueless idiots)

hs


From fellowsd at cs.man.ac.uk  Mon Feb 19 11:52:09 2001
From: fellowsd at cs.man.ac.uk (Donal K. Fellows)
Date: Mon, 19 Feb 2001 10:52:09 +0000
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com> <r5iu8tsqccj65ote3hfgo6dvjphskn15l3@4ax.com> <slrn990n47.382.hs@paradise.nirvananet>
Message-ID: <3A90FAD9.B54BCFF6@cs.man.ac.uk>

Hartmann Schaffer wrote:
> frankly, i doubt this.  most of those spams are so inept that i would be
> surprised if it generates any sales.  the only thing that seems to generate
> sales are those spam generators (to clueless idiots)

Every time I think I've finally discovered just how stupid some people can
be, I'm unpleasantly surprised by some new asininity...

Donal.
-- 
Donal K. Fellows    http://www.cs.man.ac.uk/~fellowsd/    fellowsd at cs.man.ac.uk
-- If somebody in a suit or a uniform can sit you down and have you believe
   every word he says, you've just surrendered your license to be a thinking
   human being and traded it in for a zombie suit. -- mhoye at prince.carleton.ca


From fredericbonnet at free.fr  Wed Feb 21 11:17:14 2001
From: fredericbonnet at free.fr (Frederic BONNET)
Date: Wed, 21 Feb 2001 10:17:14 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com> <r5iu8tsqccj65ote3hfgo6dvjphskn15l3@4ax.com> <slrn990n47.382.hs@paradise.nirvananet> <3A90FAD9.B54BCFF6@cs.man.ac.uk>
Message-ID: <3A939540.19FB5346@free.fr>


"Donal K. Fellows" wrote:
> 
> Hartmann Schaffer wrote:
> > frankly, i doubt this.  most of those spams are so inept that i would be
> > surprised if it generates any sales.  the only thing that seems to generate
> > sales are those spam generators (to clueless idiots)
> 
> Every time I think I've finally discovered just how stupid some people can
> be, I'm unpleasantly surprised by some new asininity...

I think Charles De Gaulle said something like that: "Human intelligence may
have a limit, but believe me, human stupididy hasn't".
-- 
Fr?d?ric BONNET                          fredericbonnet at free.fr
---------------------------------------------------------------
"Theory may inform, but Practice convinces"
                                                    George Bain



From fredericbonnet at free.fr  Wed Feb 21 11:19:28 2001
From: fredericbonnet at free.fr (Frederic BONNET)
Date: Wed, 21 Feb 2001 10:19:28 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com> <r5iu8tsqccj65ote3hfgo6dvjphskn15l3@4ax.com> <slrn990n47.382.hs@paradise.nirvananet> <3A90FAD9.B54BCFF6@cs.man.ac.uk> <3A939540.19FB5346@free.fr>
Message-ID: <3A939613.2A32E483@free.fr>

Oh, and this one, too:

General Massu: "One should eliminate all the idiots"
De Gaulle: "Vast program..."

-- 
Fr?d?ric BONNET                          fredericbonnet at free.fr
---------------------------------------------------------------
"Theory may inform, but Practice convinces"
                                                    George Bain


From tim.one at home.com  Sun Feb 18 06:08:07 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 18 Feb 2001 00:08:07 -0500
Subject: Linda, this is cool!
In-Reply-To: <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOECCJAAA.tim.one@home.com>

[Steve D. Perkins]
> I will never understand why spam still exists, the internet reaching the
> level of development and maturity is has.

The sophistication of the technology has nothing to do with the
sophistication of the users, and new users keep pouring in.  Remember the
first time you ever got an email?  Didn't you get a thrill out of it?  It
can take quite a while for people to catch on to scams.  For example, look
at all the people who use Python or Perl <wink>.

> I can't imagine that there's ANYONE left anymore dumb and/or
> internet-inexperienced to purchase anything advertised through spam.

You must be joking.  The vast bulk of the world's population has yet to see
their first piece of spam.

> Even if there is, I don't understand why spammers would target
> newsgroups for Tcl, Python, and Perl... technologies most often
> used by crusty old-timers LEAST likely of all to be an effective
> target audience...

Because the marginal cost of sending out one more piece of email spam is
essentially nothing.  The "junk mail" you get in your snail-mail box is
carefully targeted, because that costs real money to send out.  Even so, a
direct mail marketer is ecstatic if they get a 3% response rate.  Spam
emailers can make a profit on a response rate thousands of times smaller.
If one person in a million responds, cool -- send 1, send 50 million, it
costs them about the same except for the effort of gathering that many
addresses.  While it makes compelling economic sense for a snail-mail
spammer to spend money on weeding out hopeless addresses, it makes no
economic sense for an email spammer to weed out anything, provided the cost
of sending out another million is a close approximation to nothing.

Sometimes the obvious explanation is right:  spam persists simply because
"good" spammers take in more money than they shell out.  It's the American
way <wink>.

Speaking of which, who do I complain to?  I responded to an ad in my inbox
guaranteeing I could lose 10 inches in one day!  Turns out they weren't
lying -- but a 3-inch dick isn't as much fun as my old one ...




From chrisw at nipltd.com  Wed Feb 21 16:25:20 2001
From: chrisw at nipltd.com (Chris Withers)
Date: Wed, 21 Feb 2001 15:25:20 +0000
Subject: Spamcop, anyone? ;-)
References: <LNBBLJKPBEHFEDALKOLCOECCJAAA.tim.one@home.com>
Message-ID: <3A93DDE0.B81DCC01@nipltd.com>

Tim Peters wrote:
> 
> Speaking of which, who do I complain to?  I responded to an ad in my inbox
> guaranteeing I could lose 10 inches in one day!  Turns out they weren't
> lying -- but a 3-inch dick isn't as much fun as my old one ...

hahaha :-)

Seriously though, take a look at http://spamcop.net if you get a chance. it's
quite satisfying to use, if nothing else :-)

Mindyou, I wonder if its nto a sinister plot in the long run: you report spam,
the have your address... I hope not :-S

paranoidly,

Chris



From robin at alldunn.com  Thu Feb 22 00:01:30 2001
From: robin at alldunn.com (Robin Dunn)
Date: Wed, 21 Feb 2001 15:01:30 -0800
Subject: Spamcop, anyone? ;-)
References: <LNBBLJKPBEHFEDALKOLCOECCJAAA.tim.one@home.com> <mailman.982769153.31105.python-list@python.org>
Message-ID: <aOXk6.5$7X5.2425107@news.randori.com>

"Chris Withers" <chrisw at nipltd.com> wrote in message
news:mailman.982769153.31105.python-list at python.org...
> Tim Peters wrote:
> >
> > Speaking of which, who do I complain to?  I responded to an ad in my
inbox
> > guaranteeing I could lose 10 inches in one day!  Turns out they weren't
> > lying -- but a 3-inch dick isn't as much fun as my old one ...
>
> hahaha :-)
>
> Seriously though, take a look at http://spamcop.net if you get a chance.
it's
> quite satisfying to use, if nothing else :-)
>
> Mindyou, I wonder if its nto a sinister plot in the long run: you report
spam,
> the have your address... I hope not :-S
>


>From http://spamcop.net/anonsignup.shtml:

Warning: Some users have reported that the email addresses used here find
their way onto spam lists, so you should consider creating a separate
"throwaway" account for reporting spam. There are many free email providers
you can use.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!





From no at email.com  Sun Feb 18 16:54:30 2001
From: no at email.com (Frank Miller)
Date: Sun, 18 Feb 2001 15:54:30 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com>
Message-ID: <WeSj6.439784$U46.12965104@news1.sttls1.wa.home.com>

"Mindfield" <mindfield at badgers.emulationnet.com> wrote in message
news:21gu8tsoknv01eegt4jak2lo7180kijno5 at 4ax.com...
> On Sat, 17 Feb 2001 17:46:47 -0500, "Steve D. Perkins"
> <stevedperkins at NOSPAM.hotmail.com> drooled an embarassing wet spot in
> his lap as he typed:
>
[snip]
>
> Fortunately many ISPs have gotten wise and have placed posting caps on
> duplicate messages, but spambots have found ways to circumvent that --
> by placing a random number after the subject line, by rotating the
> subject line from a list of subject lines, by placing random ASCII
> text at the end of each message, etc.

I have always wondered why spam had those extra characters at the end of the
subject lines and messages.  Thanks!  You've answered one of my 10 most
sought after questions!

FrankM




From brondsem at my-deja.com  Sun Feb 18 05:17:35 2001
From: brondsem at my-deja.com (Dave Brondsema)
Date: Sun, 18 Feb 2001 04:17:35 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com>
Message-ID: <z1Ij6.5619$0u5.1057082@news1.rdc1.mi.home.com>

"Steve D. Perkins" <stevedperkins at NOSPAM.hotmail.com> wrote in message
news:96mv3102l3d at news2.newsguy.com...
> Untitled Document    ARGH!
>
>     I wll never understand why spam still exists, the internet reaching
the
> level of development and maturity is has.  I can't imagine that there's
> ANYONE left anymore dumb and/or internet-inexperienced to purchase
anything
> advertised through spam.  Even if there is, I don't understand why
spammers
> woud target newsgroups for Tcl, Python, and Perl... technologies most
often
> used by crusty old-timers LEAST likely of all to be an effective target
> audience...
>
>

there will always be dumb people




From real.email at signature.this.is.invalid  Mon Feb 19 19:25:01 2001
From: real.email at signature.this.is.invalid (Csaba Raduly)
Date: Mon, 19 Feb 2001 18:25:01 +0000 (UTC)
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <z1Ij6.5619$0u5.1057082@news1.rdc1.mi.home.com>
Message-ID: <Xns904DB98F8quuxi@194.203.134.135>

And so it came to pass that "Dave Brondsema" <brondsem at my-deja.com>
on 18 Feb 2001 wrote <z1Ij6.5619$0u5.1057082 at news1.rdc1.mi.home.com>: 

[snip woes of spam]
>
>there will always be dumb people
>
>

P.T.Barnum had this figured out quite some time ago :-)
-- 
Csaba Raduly, Software Developer (OS/2), Sophos Anti-Virus
mailto:csaba.raduly at sophos.com      http://www.sophos.com/
US Support +1 888 SOPHOS 9      UK Support +44 1235 559933
You are in a maze of twisted little minds, all different.


From flavell at mail.cern.ch  Mon Feb 19 19:51:12 2001
From: flavell at mail.cern.ch (Alan J. Flavell)
Date: Mon, 19 Feb 2001 19:51:12 +0100
Subject: Linda, this is cool!
In-Reply-To: <Xns904DB98F8quuxi@194.203.134.135>
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <z1Ij6.5619$0u5.1057082@news1.rdc1.mi.home.com> <Xns904DB98F8quuxi@194.203.134.135>
Message-ID: <Pine.LNX.4.30.0102191950020.2583-100000@lxplus003.cern.ch>

On Mon, 19 Feb 2001, Csaba Raduly wrote:

> P.T.Barnum had this figured out quite some time ago :-)

"This way to the Egress".


[OT, so f'ups redirected]



From cyberian_bear at hotmail.com  Sun Feb 18 00:37:08 2001
From: cyberian_bear at hotmail.com (cyberian bear)
Date: Sat, 17 Feb 2001 18:37:08 -0500
Subject: wildcard in tuples
Message-ID: <8XDj6.13$sR5.356@typhoon.nyu.edu>

lets say i have a tuple in one list and i need to check it against every
tuple in second list.  So (e,c) --and i want to find every tuple in second
list which has tuples who are (e, *). What is the proper syntax for doing
this




From max at alcyone.com  Sun Feb 18 02:03:57 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 17 Feb 2001 17:03:57 -0800
Subject: wildcard in tuples
References: <8XDj6.13$sR5.356@typhoon.nyu.edu>
Message-ID: <3A8F1F7D.6781AF14@alcyone.com>

cyberian bear wrote:

> lets say i have a tuple in one list and i need to check it against
> every
> tuple in second list.  So (e,c) --and i want to find every tuple in
> second
> list which has tuples who are (e, *). What is the proper syntax for
> doing
> this

I'm not sure exactly what you're asking.  If all you want is to select
all 2-tuples that, say, have 2 as the first element, then you can do it
easily enough with list comprehensions:

>>> l = [(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]
>>> [ x for x in l if x[0] == 2 ]
[(2, 3), (2, 4)]

This can obviously be trivially modified for matching on the second
list.

You could do something similar using None as a wildcard:

>>> def selectTuple(list, tuple):
...     return [ x for x in list if (x[0] == tuple[0] or tuple[0] is \
...              None) and (x[1] == tuple[1] or tuple[1] is None) ]
... 
>>> l = [(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]
>>> selectTuple(l, (2, 3))
[(2, 3)]
>>> selectTuple(l, (2, None)) # tuples beginning with 2
[(2, 3), (2, 4)]
>>> selectTuple(l, (None, 4)) # tuples ending with 4
[(2, 4), (4, 4)]
>>> selectTuple(l, (None, None)) # all tuples
[(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Society attacks early when the individual is helpless.
\__/ B.F. Skinner
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.


From cyberian_bear at hotmail.com  Sun Feb 18 03:50:47 2001
From: cyberian_bear at hotmail.com (cyberian bear)
Date: Sat, 17 Feb 2001 21:50:47 -0500
Subject: wildcard in tuples
References: <8XDj6.13$sR5.356@typhoon.nyu.edu> <3A8F1F7D.6781AF14@alcyone.com>
Message-ID: <IMGj6.15$sR5.157@typhoon.nyu.edu>

Yep, that looks like just what i needed. but is there a way to modify those
statements so that they can be used in 'if then ' statement. So it will go
something like that:
if (# of matched terms is greater than 0) :
    do stuff

thanx
"Erik Max Francis" <max at alcyone.com> wrote in message
news:3A8F1F7D.6781AF14 at alcyone.com...
> cyberian bear wrote:
>
> > lets say i have a tuple in one list and i need to check it against
> > every
> > tuple in second list.  So (e,c) --and i want to find every tuple in
> > second
> > list which has tuples who are (e, *). What is the proper syntax for
> > doing
> > this
>
> I'm not sure exactly what you're asking.  If all you want is to select
> all 2-tuples that, say, have 2 as the first element, then you can do it
> easily enough with list comprehensions:
>
> >>> l = [(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]
> >>> [ x for x in l if x[0] == 2 ]
> [(2, 3), (2, 4)]
>
> This can obviously be trivially modified for matching on the second
> list.
>
> You could do something similar using None as a wildcard:
>
> >>> def selectTuple(list, tuple):
> ...     return [ x for x in list if (x[0] == tuple[0] or tuple[0] is \
> ...              None) and (x[1] == tuple[1] or tuple[1] is None) ]
> ...
> >>> l = [(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]
> >>> selectTuple(l, (2, 3))
> [(2, 3)]
> >>> selectTuple(l, (2, None)) # tuples beginning with 2
> [(2, 3), (2, 4)]
> >>> selectTuple(l, (None, 4)) # tuples ending with 4
> [(2, 4), (4, 4)]
> >>> selectTuple(l, (None, None)) # all tuples
> [(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Society attacks early when the individual is helpless.
> \__/ B.F. Skinner
>     Physics reference / http://www.alcyone.com/max/reference/physics/
>  A physics reference.




From max at alcyone.com  Sun Feb 18 04:07:00 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 17 Feb 2001 19:07:00 -0800
Subject: wildcard in tuples
References: <8XDj6.13$sR5.356@typhoon.nyu.edu> <3A8F1F7D.6781AF14@alcyone.com> <IMGj6.15$sR5.157@typhoon.nyu.edu>
Message-ID: <3A8F3C54.E33222B0@alcyone.com>

cyberian bear wrote:

> Yep, that looks like just what i needed. but is there a way to modify
> those
> statements so that they can be used in 'if then ' statement. So it
> will go
> something like that:
> if (# of matched terms is greater than 0) :
>     do stuff

With the selectTuple function I wrote, which returns a list of matched
tuples, it would simply be:

    if selectTuple(list, matchingTuple):
        do stuff

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ In Heaven all the interesting people are missing.
\__/ Friedrich Nietzsche
    Computer science / http://www.alcyone.com/max/reference/compsci/
 A computer science reference.


From b.l.crawford at worldnet.att.net  Sun Feb 18 00:41:26 2001
From: b.l.crawford at worldnet.att.net (Bonnie Crawford)
Date: Sat, 17 Feb 2001 23:41:26 GMT
Subject: Kansas city Opportunity
Message-ID: <G_Dj6.11411$X61.812836@bgtnsc04-news.ops.worldnet.att.net>

Terrific opportunity for a Staff Analyst Programmer.  Position involves
development and maintenance of an electronic proxy voting product.  Product
is primarily a web site which allows shareholders to log in and complete a
proxy card on line.  Development is done in Python language in a Solaris
environment using a DB2 database.  nature of programming includes CGI
programming, database programming, data file processing and report
generation.
Requirements:
Strong knowledge of Internet infrastructure such as SMTP, HTTP and FTP,
Background in object oriented programming, Python development preferred,
will consider C++, Java, or Object pascal experience.  Strong background as
UNIX developer preferably Solaris and background with programming database
applications.
Please email to b.l.crawford at att.net is interested.  Thanks!





From cjw at sympatico.ca  Sun Feb 18 02:45:39 2001
From: cjw at sympatico.ca (Colin J. Williams)
Date: Sun, 18 Feb 2001 01:45:39 GMT
Subject: A suspected bug
Message-ID: <3A8F2968.9240BD8C@sympatico.ca>

It seems to me that the bit of code below should report
a type inconsistency. 
X-Mozilla-Status: 0009 Sep 20 2000, 12:29:43) [MSC 32 bit (Intel)] on win32.
Portions Copyright 1994-2000 Mark Hammond (MarkH at ActiveState.com) - see
'Help/About PythonWin' for further copyright information.
>>> print 2 > '1'
0
>>> print 1 > '2'
0
>>>

If we replace the comparison operator by an arithmetic operator,
we get:
>>> z= 1+'1'
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
TypeError: number coercion failed
>>>

Which is what I would have expected.

Colin W.



From aahz at panix.com  Sun Feb 18 03:32:57 2001
From: aahz at panix.com (Aahz Maruch)
Date: 17 Feb 2001 18:32:57 -0800
Subject: A suspected bug
References: <3A8F2968.9240BD8C@sympatico.ca>
Message-ID: <96nc8p$148$1@panix3.panix.com>

In article <3A8F2968.9240BD8C at sympatico.ca>,
Colin J. Williams <cjw at sympatico.ca> wrote:
>
>It seems to me that the bit of code below should report
>a type inconsistency. 
>X-Mozilla-Status: 0009 Sep 20 2000, 12:29:43) [MSC 32 bit (Intel)] on win32.
>Portions Copyright 1994-2000 Mark Hammond (MarkH at ActiveState.com) - see
>'Help/About PythonWin' for further copyright information.
>>>> print 2 > '1'
>0
>>>> print 1 > '2'
>0
>>>>

What should the following code do:

foo = [1, '2']
foo.sort()

Because Python lists can contain heterogeneous information, the
comparison operators *must* work between types.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From dalke at acm.org  Sun Feb 18 04:02:23 2001
From: dalke at acm.org (Andrew Dalke)
Date: Sat, 17 Feb 2001 20:02:23 -0700
Subject: A suspected bug
References: <3A8F2968.9240BD8C@sympatico.ca> <96nc8p$148$1@panix3.panix.com>
Message-ID: <96ne4j$6j5$1@slb0.atl.mindspring.net>

Aahz Maruch wrote:
>What should the following code do:
>
>foo = [1, '2']
>foo.sort()
>
>Because Python lists can contain heterogeneous information, the
>comparison operators *must* work between types.

That's what I used to think was the accepted Python view
of the world, but from the update info for Python 2.1

]- Complex numbers use rich comparisons to define == and != but raise
]  an exception for <, <=, > and >=.  Unfortunately, this also means
]  that cmp() of two complex numbers raises an exception when the two
]  numbers differ.  Since it is not mathematically meaningful to compare
]  complex numbers except for equality, I hope that this doesn't break
]  too much code.

so foo.sort() won't even work for all built-in homogenous data
types.

'Course, I would prefer that complex numbers had a sort, even
if not "mathmatical".  There are a couple of ways to do it,
but I don't want to get into that tangent.

                    Andrew
                    dalke at acm.org





From tim.one at home.com  Sun Feb 18 04:31:00 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 17 Feb 2001 22:31:00 -0500
Subject: A suspected bug
In-Reply-To: <3A8F2968.9240BD8C@sympatico.ca>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEBPJAAA.tim.one@home.com>

[Colin J. Williams]
> It seems to me that the bit of code below should report
> a type inconsistency.
> X-Mozilla-Status: 0009 Sep 20 2000, 12:29:43) [MSC 32 bit
> (Intel)] on win32.
> Portions Copyright 1994-2000 Mark Hammond (MarkH at ActiveState.com) - see
> 'Help/About PythonWin' for further copyright information.
> >>> print 2 > '1'
> 0
> >>> print 1 > '2'
> 0
> >>>

It's functioning as designed and documented, so you may wish to argue that
it's a design flaw, but it won't be considered "a bug".  As the Reference
Manual says:

    The operators <, >, ==, >=, <=, and != compare the values of
    two objects.  The objects need not have the same type.  If both
    are numbers, they are coverted to a common type.  Otherwise,
    objects of different types always compare unequal, and are
    ordered consistently but arbitrarily.

It so happens that objects of different non-numeric types are compared today
by the string names of their types, so any integer compares less than any
string today (because "int" < "string").

I'd agree that's of marginal value, but that's the way it's always been, so
it would break stuff if it changed.  For example, sometimes I have giant
lists of objects of all kinds of types, and can reliably sort the list to
bring all the objects of the same type next to each other.  It does have its
charms.

it-must-be-so-unpythonic-that-it's-pythonic-again-ly y'rs  - tim




From fredrik at pythonware.com  Sun Feb 18 12:12:57 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 11:12:57 GMT
Subject: A suspected bug
References: <mailman.982467140.8836.python-list@python.org>
Message-ID: <Z6Oj6.18070$AH6.2468100@newsc.telia.net>

"Tim Peters wrote:
> I'd agree that's of marginal value, but that's the way it's always been, so
> it would break stuff if it changed.  For example, sometimes I have giant
> lists of objects of all kinds of types, and can reliably sort the list to
> bring all the objects of the same type next to each other.  It does have its
> charms.

except that it doesn't always work:

Python 2.0 (#8, Jan 29 2001, 22:28:01) on win32
>>> a = [u"foo", "b?r"]
>>> a.sort()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
UnicodeError: ASCII decoding error: ordinal not in range(128)

not to mention:

Python 2.1a2 (#10, Feb 18 2001, 00:16:17) on win32
>>> a = [u"foo", "b?r"]
>>> a.sort()
>>> a
UnicodeError: ASCII decoding error: ordinal not in range(128)
>>> a
['b\x84r', u'foo']
>>> a.sort()
>>> a = 10
UnicodeError: ASCII decoding error: ordinal not in range(128)
>>> a.sort()
>>> # hey, what's going on here?
...
UnicodeError: ASCII decoding error: ordinal not in range(128)
>>> quit
'Use Ctrl-Z plus Return to exit.'
(reboot)

Cheers /F




From amk at mira.erols.com  Sun Feb 18 16:36:31 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 18 Feb 2001 15:36:31 GMT
Subject: A suspected bug
References: <mailman.982467140.8836.python-list@python.org> <Z6Oj6.18070$AH6.2468100@newsc.telia.net>
Message-ID: <slrn98vrqv.i0.amk@mira.erols.com>

On Sun, 18 Feb 2001 11:12:57 GMT, Fredrik Lundh <fredrik at pythonware.com> wrote:
>except that it doesn't always work:

And as Andrew Dalke pointed out:
>>> L=[1j, 2j]
>>> L.sort()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: cannot compare complex numbers using <, <=, >, >=
>>>

--amk



From aahz at panix.com  Sun Feb 18 17:11:49 2001
From: aahz at panix.com (Aahz Maruch)
Date: 18 Feb 2001 08:11:49 -0800
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982467140.8836.python-list@python.org> <Z6Oj6.18070$AH6.2468100@newsc.telia.net> <slrn98vrqv.i0.amk@mira.erols.com>
Message-ID: <96os85$1qe$1@panix3.panix.com>

In article <slrn98vrqv.i0.amk at mira.erols.com>,
A.M. Kuchling <akuchlin at mems-exchange.org> wrote:
>On Sun, 18 Feb 2001 11:12:57 GMT, Fredrik Lundh <fredrik at pythonware.com> wrote:
>
>>except that it doesn't always work:
>
>And as Andrew Dalke pointed out:
>>>> L=[1j, 2j]
>>>> L.sort()
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>TypeError: cannot compare complex numbers using <, <=, >, >=
>>>>

Is there any chance this can be treated as a bug and fixed for the
release of 2.1?  Alternatively, given that we're already breaking code
with the change in the way complex numbers are handled, should cmp() now
raise an exception *every* time the type/class differs?

I think the current situation is going to cause major problems trying to
explain Python, particularly on the Unicode side.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From tim.one at home.com  Sun Feb 18 18:23:50 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 18 Feb 2001 12:23:50 -0500
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
In-Reply-To: <96os85$1qe$1@panix3.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDCJAAA.tim.one@home.com>

[Andrew Dalke]
> >>> L=[1j, 2j]
> >>> L.sort()
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: cannot compare complex numbers using <, <=, >, >=
> >>>

[Aahz Maruch]
> Is there any chance this can be treated as a bug and fixed for the
> release of 2.1?

Sure, but it's a small chance.  The reason all comparisons were
non-exceptional before isn't that Guido wanted that, but because the
implementation was such that it was literally impossible for a comparison to
raise an exception "that got noticed".  As soon as the implementation was
revamped to make raising visible exceptions from comparisons possible, Guido
was eager to drop comparisons outside of {==, !=} for complex numbers (as
Andrew said, there are several *possible* ways it could be done, but not a
one of them wouldn't surprise a majority of users).

What would change his mind is bug reports from people this actually hurts
(in practice).  So far no alpha user has filed a bug report against this.
Note that nobody is trying to hide that it changed, either:  there's a very
clear warning about this change of behavior in the alpha NEWS file.

> Alternatively, given that we're already breaking code with the change
> in the way complex numbers are handled, should cmp() now raise an
> exception *every* time the type/class differs?

Certainly not every time -- there's no practical or theoretical problem in
asking, e.g., whether 42 < 666L.  Forbidding that would be silly.  In
senseless cases, perhaps.  I suspect complex is being used to test the
waters before daring to dive in head first.

> I think the current situation is going to cause major problems trying to
> explain Python, particularly on the Unicode side.

Don't understand the last part.  I doubt anyone is comparing complex numbers
to Unicode strings today <wink>.  The only errors raised when mixing strings
and Unicode strings in comparisons occur in ambiguous cases where users
don't specify their intended encodings and so Python refuses to guess which
of the world's bazillion and one encodings they had in mind.  That's a
feature, albeit that it irritates people who think Python *should* read
their minds.

if-it-could-it-would-ly y'rs  - tim




From ssthapa at classes.cs.uchicago.edu  Tue Feb 20 18:53:50 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Tue, 20 Feb 2001 17:53:50 GMT
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982467140.8836.python-list@python.org> <Z6Oj6.18070$AH6.2468100@newsc.telia.net> <slrn98vrqv.i0.amk@mira.erols.com> <96os85$1qe$1@panix3.panix.com>
Message-ID: <slrn993luj.24c.ssthapa@ntcs-ip45.uchicago.edu>

Aahz Maruch <aahz at panix.com> wrote:
>>>>> L=[1j, 2j]
>>>>> L.sort()
>>Traceback (most recent call last):
>>  File "<stdin>", line 1, in ?
>>TypeError: cannot compare complex numbers using <, <=, >, >=
>>>>>
>
>Is there any chance this can be treated as a bug and fixed for the
>release of 2.1?  Alternatively, given that we're already breaking code
>with the change in the way complex numbers are handled, should cmp() now
>raise an exception *every* time the type/class differs?

  If you are referring to the comparision of complex numbers, I'm not
sure how the ordering would be done.   For example consider 1 and 1j, 
they have the same magnitude so mathematically speaking there is no
way to consistently order them like the real numbers where magnitude
is a more useful measure. IIRC, for any given magnitude n there are an
infinite number of complex numbers with that magnitude, namely all n*e^(ix)
for real x.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From aahz at panix.com  Tue Feb 20 19:42:16 2001
From: aahz at panix.com (Aahz Maruch)
Date: 20 Feb 2001 10:42:16 -0800
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982467140.8836.python-list@python.org> <slrn98vrqv.i0.amk@mira.erols.com> <96os85$1qe$1@panix3.panix.com> <slrn993luj.24c.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <96udq8$mh$1@panix3.panix.com>

In article <slrn993luj.24c.ssthapa at ntcs-ip45.uchicago.edu>,
 <s-thapa-11 at NOSPAM.alumni.uchicago.edu> wrote:
>Aahz Maruch <aahz at panix.com> wrote:
>>
>>Is there any chance this can be treated as a bug and fixed for the
>>release of 2.1?  Alternatively, given that we're already breaking code
>>with the change in the way complex numbers are handled, should cmp() now
>>raise an exception *every* time the type/class differs?
>
>  If you are referring to the comparision of complex numbers, I'm not
>sure how the ordering would be done.   For example consider 1 and 1j, 
>they have the same magnitude so mathematically speaking there is no
>way to consistently order them like the real numbers where magnitude
>is a more useful measure. IIRC, for any given magnitude n there are an
>infinite number of complex numbers with that magnitude, namely all n*e^(ix)
>for real x.

True enough.  Question is, in the Real World [tm], should everyone who
calls list.sort() be forced to wrap in try/except on the chance that a
pair of complex numbers will sneak in?
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From tim.one at home.com  Tue Feb 20 22:54:26 2001
From: tim.one at home.com (Tim Peters)
Date: Tue, 20 Feb 2001 16:54:26 -0500
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
In-Reply-To: <96udq8$mh$1@panix3.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEMHJAAA.tim.one@home.com>

[ssthapa at ntcs-ip45.uchicago.edu, reaffirms the lack of a non-surprising
 ordering for complex numbers]

[Aahz]
> True enough.  Question is, in the Real World [tm], should everyone who
> calls list.sort() be forced to wrap in try/except on the chance that a
> pair of complex numbers will sneak in?

We lost that the instant we let user-defined __cmp__ raise visible
exceptions; list.sort() simply isn't "safe" anymore regardless of what
Python does with complex numbers, and we can't go back without again
ignoring exceptions raised by __cmp__.

You can still define your own comparison function to pass to list.sort(),
that catches exceptions and hides them, instead supplying whatever ordering
*you* feel like making up in exceptional cases.  Much slower, though, but
that's the tradeoff.

In practice, I think I'm going to be more bothered by tuple comparisons.
For example, I often use bisect.insort() to maintain short lists of tuples
of the form

    (priority_integer, some_object)

or

    (first_time_to_deal_with_it, some_object)

ordered by the first element.  If there's ever a tie on the first element,
the builtin lexicographic comparison goes on to compare the second elements.
I'm sure I'll start getting exceptions due to that.  Again, I'll have to
wrap some_object in a class with a __cmp__ that hides those exceptions.

afraid-it's-clearly-more-pythonic-not-to-suppress-exceptions-by-
    magic-ly y'rs  - tim




From grante at visi.com  Tue Feb 20 21:19:15 2001
From: grante at visi.com (Grant Edwards)
Date: Tue, 20 Feb 2001 20:19:15 GMT
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982467140.8836.python-list@python.org> <Z6Oj6.18070$AH6.2468100@newsc.telia.net> <slrn98vrqv.i0.amk@mira.erols.com> <96os85$1qe$1@panix3.panix.com> <slrn993luj.24c.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <7jAk6.603$S7.51755@ruti.visi.com>

In article <slrn993luj.24c.ssthapa at ntcs-ip45.uchicago.edu>, ssthapa at classes.cs.uchicago.edu wrote:
>Aahz Maruch <aahz at panix.com> wrote:
>>>>>> L=[1j, 2j]
>>>>>> L.sort()
>>>Traceback (most recent call last):
>>>  File "<stdin>", line 1, in ?
>>>TypeError: cannot compare complex numbers using <, <=, >, >=
>>>>>>
>>
>>Is there any chance this can be treated as a bug and fixed for
>>the release of 2.1?  Alternatively, given that we're already
>>breaking code with the change in the way complex numbers are
>>handled, should cmp() now raise an exception *every* time the
>>type/class differs?
>
>If you are referring to the comparision of complex numbers, I'm
>not sure how the ordering would be done.

Understood.  But, for all of the other pairs of objects where
nobody is sure how the ordering should done an arbitrary,
consistent order is chosen rather than an exception being
raised.

>For example consider 1 and 1j, they have the same magnitude so
>mathematically speaking there is no way to consistently order
>them like the real numbers where magnitude is a more useful
>measure. IIRC, for any given magnitude n there are an infinite
>number of complex numbers with that magnitude, namely all
>n*e^(ix) for real x.

One could argue that there is no consistent way to order 1.23
and "Winnie the Pooh", but it is done anyway.  Complex numbers
should behave like everything else.

cmp() should always return an ordering, or always raise an
exception if the comparison is not meaningful.  I don't
particularly care which, but it ought to be consistent in order
to avoid violating the "law of least astonishment."

[I've recently read that Unicode strings are also broken in
this respect -- IMO, that needs to be fixed also.]

-- 
Grant Edwards                   grante             Yow!  Yow! Am I cleansed
                                  at               yet?!
                               visi.com            


From mbel44 at dial.pipex.net  Wed Feb 21 14:00:20 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Wed, 21 Feb 2001 13:00:20 +0000
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982467140.8836.python-list@python.org> <Z6Oj6.18070$AH6.2468100@newsc.telia.net> <slrn98vrqv.i0.amk@mira.erols.com> <96os85$1qe$1@panix3.panix.com> <slrn993luj.24c.ssthapa@ntcs-ip45.uchicago.edu> <7jAk6.603$S7.51755@ruti.visi.com>
Message-ID: <8ee79t0sinerobr9slnp3ceem6alfqjqn9@4ax.com>

grante at visi.com (Grant Edwards) wrote:

>One could argue that there is no consistent way to order 1.23
>and "Winnie the Pooh", but it is done anyway.  Complex numbers
>should behave like everything else.

>cmp() should always return an ordering, or always raise an
>exception if the comparison is not meaningful.  I don't
>particularly care which, but it ought to be consistent in order
>to avoid violating the "law of least astonishment."

>[I've recently read that Unicode strings are also broken in
>this respect -- IMO, that needs to be fixed also.]

In c++ there are two ways to compare something: The operators (> =
etc) all do the obvious thing but are not defined for types where
there is no natural order, such as complex numbers.

Alternatively there is a template 'cmp', which is used for ordering
types even if they dont have a natural order, and is used by the
implementations of set of map templates (C++ maps are tree-based,
similar to pythons hash-based dictionaries).

Im beginning to think that a similar scheme might make sense for
python: It would be a clean way of resolving the current quirky
behaviour of comparing unicode and non-unicode strings.





Toby Dickenson
tdickenson at geminidataloggers.com


From hzhu at mars.localdomain  Wed Feb 21 09:02:11 2001
From: hzhu at mars.localdomain (Huaiyu Zhu)
Date: Wed, 21 Feb 2001 08:02:11 GMT
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982706124.20851.python-list@python.org>
Message-ID: <slrn996t3d.1a8.hzhu@mars.localdomain>

There are two quite different reasons to define order: one is to impose a
meaningful (partial) order, another is to be able to sort things in
(whatever) possible order.

The first reason can be satisfied by rich comparison, using the magic
methods like __lt__, __ge__, and so on.

Can we then use cmp to satisfy the second mativation, namely to put
everything in an order?  This can be done without conflict with the first,
if it is required that __cmp__ be a well order (complete order) compatible
with __lt__.  (Every partial order can be extended to a well order, etc).

Eg, for complex number a+bj, one possibility is to sort them according to
(a,b) and put them between a and a+eps where eps>0.  Another is for
cmp(a,a+bj)==0.  Either would do to avoid exception in sorting.

Now that we have rich comparison, is there any reason, other than
historical, to allow __cmp__ to raise exceptions?  What would happen if sort
just treat exception of cmp(a,b) as cmp(a,b)==0?  Will it go into an
infinite loop?  

Just some random thoughts.

Huaiyu

On Tue, 20 Feb 2001 16:54:26 -0500, Tim Peters <tim.one at home.com> wrote:
>[ssthapa at ntcs-ip45.uchicago.edu, reaffirms the lack of a non-surprising
> ordering for complex numbers]
>
>[Aahz]
>> True enough.  Question is, in the Real World [tm], should everyone who
>> calls list.sort() be forced to wrap in try/except on the chance that a
>> pair of complex numbers will sneak in?
>
>We lost that the instant we let user-defined __cmp__ raise visible
>exceptions; list.sort() simply isn't "safe" anymore regardless of what
>Python does with complex numbers, and we can't go back without again
>ignoring exceptions raised by __cmp__.
>
>You can still define your own comparison function to pass to list.sort(),
>that catches exceptions and hides them, instead supplying whatever ordering
>*you* feel like making up in exceptional cases.  Much slower, though, but
>that's the tradeoff.
>
>In practice, I think I'm going to be more bothered by tuple comparisons.
>For example, I often use bisect.insort() to maintain short lists of tuples
>of the form
>
>    (priority_integer, some_object)
>
>or
>
>    (first_time_to_deal_with_it, some_object)
>
>ordered by the first element.  If there's ever a tie on the first element,
>the builtin lexicographic comparison goes on to compare the second elements.
>I'm sure I'll start getting exceptions due to that.  Again, I'll have to
>wrap some_object in a class with a __cmp__ that hides those exceptions.
>
>afraid-it's-clearly-more-pythonic-not-to-suppress-exceptions-by-
>    magic-ly y'rs  - tim
>
>


-- 
Huaiyu Zhu   <hzhu at users.sourceforge.net>


From tim.one at home.com  Thu Feb 22 07:42:31 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 22 Feb 2001 01:42:31 -0500
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
In-Reply-To: <slrn996t3d.1a8.hzhu@mars.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCGECFJBAA.tim.one@home.com>

[Huaiyu Zhu]
> ...
> Now that we have rich comparison, is there any reason, other than
> historical, to allow __cmp__ to raise exceptions?  What would
> happen if sort just treat exception of cmp(a,b) as cmp(a,b)==0?
> Will it go into an infinite loop?

list.sort() only looks at __lt__, unless you pass your own comparison
function.  So __cmp__ is usually irrelevant here.  (BTW, there's nothing you
can do to send list.sort() into an infinite loop, no matter how whacky a cmp
function you supply -- return a random result each time, and it will still
end eventually.)

But even if list.sort() did still use cmp(), I don't see what's gained by
treating exceptions as equality, except to open the door to unbounded
surprises:  as you said earlier, a partial ordering can be extended to a
full ordering, but this treatment of cmp exceptions isn't good enough to do
that.  For example, consider

    x, y, z == [5, 3j, 1]

Under the "exception implies equality" gimmick, we would have

    cmp(x, y) == 0 == cmp(y, z)

because those both raise exceptions.  But in a total ordering that would
imply cmp(x, z) == 0 (5==1) too.  Ain't so.  In fact, that specific list
would be a fixed-point for list.sort() (based on details of its
implementation I happen to know).  That list reversed would also be a
fixed-point.  So it doesn't even manage to get objects of the same type next
to each other.

Because complex numbers are built in, Python could impose a total ordering
on complex numbers (and, indeed, it used to!).  That surprises people too.
Mixing 8-bit with Unicode strings in the absence of intended encoding
information about the former is wholly analogous, and not so esoteric that
people are willing to endure senseless orderings:

    x, y, z = ["\xbb", u"\u3456", "\x81"]

This is "just like" the mixed int and complex case:  both that list and its
reverse would be sort fixed-points, and for the same reasons.

list.sort() simply can't do something sensible in the presence of
exceptions.  So what's better:  Pass exceptions up to the caller?  Or
silently return a senseless result?  Note that I don't say "senseless" just
because it doesn't make sense <wink>, but also because the result is
*unpredictable* to the point of uselessness (as in the examples above).

Disclaimer:  If list.sort() were willing to embrace a quadratic-time
algorithm, it could deduce a total ordering consistent with whatever partial
ordering is present, provided that the latter is itself consistent (but note
that we have no a priori guarantee about that, and user-defined __cmp__s
often aren't consistent).

doesn't-sound-likely-to-me-ly y'rs  - tim




From me at nospam.net  Sat Feb 24 20:27:16 2001
From: me at nospam.net (Scottie)
Date: Sat, 24 Feb 2001 11:27:16 -0800
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982706124.20851.python-list@python.org>
Message-ID: <t9g2oumq5n4d69@corp.supernews.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.982706124.20851.python-list at python.org...
> .... I think I'm going to be more bothered by tuple comparisons.
> For example, I often use bisect.insort() to maintain short lists of tuples
> of the form
>     (priority_integer, some_object)
> or
>     (first_time_to_deal_with_it, some_object)
> ordered by the first element.  If there's ever a tie on the first element,
> the builtin lexicographic comparison goes on to compare the second ...

I am trying to put together a nice priority queue package.
With the new ops, I feel I should standardize my code to use a single
test, rather than willy-nilly use < and <= in various places.
What opinions do others have on whether I should use "<" or "<="?
I realize exceptions may get thrown or strange orders result from
using non-total-ordering relations (I view as a user problem).

So here's my question:
I'm building these priority queues on either < (>) or <= (>=).  If you
were to build only one of these two on a type which would it be?
Which of the two would you expect to be used in such a package?
How should I make this choice?


and-what-is-the-meaning-of-life-ly yours,
-Scott David Daniels
Scott.Daniels at Acm.Org





From tim.one at home.com  Sun Feb 25 06:15:58 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 25 Feb 2001 00:15:58 -0500
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
In-Reply-To: <t9g2oumq5n4d69@corp.supernews.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEMOJBAA.tim.one@home.com>

[Scottie]
> ...
> I am trying to put together a nice priority queue package.
> With the new ops, I feel I should standardize my code to use a single
> test, rather than willy-nilly use < and <= in various places.
> What opinions do others have on whether I should use "<" or "<="?
> I realize exceptions may get thrown or strange orders result from
> using non-total-ordering relations (I view as a user problem).
>
> So here's my question:
> I'm building these priority queues on either < (>) or <= (>=).  If you
> were to build only one of these two on a type which would it be?
> Which of the two would you expect to be used in such a package?
> How should I make this choice?

Python's list.sort() originally used a full-blown cmp(), but in
*anticipation* of rich comparisons I rewrote it some years ago to only test
whether cmp did or did not return a value less than 0.  In other words,
rewrote it to *conceptually* use only <.

Rich comparisons finally are (will be) in 2.1, and Guido changed the
list.sort() comparison routine again to finish that job:  if Py_LT alone
settles the issue, that's all we ever call now.  So I expect that will
become a de facto standard.  Note that the standard module bisect.py also
only uses < (and, btw, note that bisect.insort() (also bisect.insort_right()
in 2.1) is very handy for building priority queues today (but "small"
ones -- it works with a straight Python list, so inserts "near the front"
can get unboundedly expensive as the list grows)).

> and-what-is-the-meaning-of-life-ly yours,

asking-the-question-prevents-the-answer-from-being-heard-ly y'rs  - tim




From cjw at sympatico.ca  Sun Feb 18 18:50:31 2001
From: cjw at sympatico.ca (Colin J. Williams)
Date: Sun, 18 Feb 2001 17:50:31 GMT
Subject: A suspected bug
References: <mailman.982467140.8836.python-list@python.org>
Message-ID: <3A9009B7.171E1D31@sympatico.ca>

Tim points to a paragraph in the Language Reference.  I've added
the paragraph which follows it below.

Tim Peters wrote:

> [Colin J. Williams]
> > It seems to me that the bit of code below should report
> > a type inconsistency.
> > X-Mozilla-Status: 0009 Sep 20 2000, 12:29:43) [MSC 32 bit
> > (Intel)] on win32.
> > Portions Copyright 1994-2000 Mark Hammond (MarkH at ActiveState.com) - see
> > 'Help/About PythonWin' for further copyright information.
> > >>> print 2 > '1'
> > 0
> > >>> print 1 > '2'
> > 0
> > >>>
>
> It's functioning as designed and documented, so you may wish to argue that
> it's a design flaw, but it won't be considered "a bug".  As the Reference
> Manual says:
>
>     The operators <, >, ==, >=, <=, and != compare the values of
>     two objects.  The objects need not have the same type.  If both
>     are numbers, they are coverted to a common type.  Otherwise,
>     objects of different types always compare unequal, and are
>     ordered consistently but arbitrarily.
>
>> (This unusual definition of comparison was used to simplify the definition
>> of operations like sorting and the in and not in operators. In the future,
>> the comparison rules for objects of different types are likely to change.)
>
> The David Beazley book Python: Essential Reference has, on page 38:

"If the types differ, a coercion operation is performed to convert one
of the types to the other:"

This looks clear enough, but one has to delve into the list below to see
that a coercion is only performed in some cases.

>
>
> It so happens that objects of different non-numeric types are compared today
> by the string names of their types, so any integer compares less than any
> string today (because "int" < "string").
>
> I'd agree that's of marginal value, but that's the way it's always been, so
> it would break stuff if it changed.  For example, sometimes I have giant
> lists of objects of all kinds of types, and can reliably sort the list to
> bring all the objects of the same type next to each other.  It does have its
> charms.
>

Tim makes a case for maintaining cmp, to group objects of like type.
Perhaps consideration should be given to providing a new 'comp'
which behaves as cmp, but extends the range of coercible types.

>
> it-must-be-so-unpythonic-that-it's-pythonic-again-ly y'rs  - tim

Many thanks to Paul Foley and to all who responded.

Colin W.





From jbell at iinet.net.au  Sun Feb 18 05:04:28 2001
From: jbell at iinet.net.au (John Bell)
Date: Sun, 18 Feb 2001 12:04:28 +0800
Subject: Automating Form Submission
Message-ID: <3A8F49C5.674ED0E4@iinet.net.au>

Pythonistas,

I am an experienced Perl programmer just beginning to look at Python as I need a scripting langusge for a large project which involves a wide range of technologies.  I know from bitter experience that Perl quickly becomes unwieldly as project size and complexity increases.

I am impressed by Python, and in evaluating its suitability for the project at hand have found extensions to cover all areas of need but one: Automation of Submission of Web based Forms.

I am probably missing something, but cannot find support for escaping of responses (as per Perl's URI::Escape::uri_escape), support for Forms processing (GET and POST methods as per Perl's LWP::Simple and LWP::UserAgent modules respectively) or any support for capture and use of Cookies.

At first view Python seems very good in the Web Server department, but not so good in the Web Client area.  I think I must be missing something, as I would have expected all this functionality to be written as part of the Grail project.

Please tell me I've got it wrong!

--

Regards,
John Bell


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010218/167632f6/attachment.htm>

From gregj at pobox.com  Sun Feb 18 11:37:18 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Sun, 18 Feb 2001 10:37:18 GMT
Subject: Automating Form Submission
References: <3A8F49C5.674ED0E4@iinet.net.au>
Message-ID: <3a8fa220.190450383@news.potlnd1.or.home.com>

On Sun, 18 Feb 2001 12:04:28 +0800, John Bell <jbell at iinet.net.au>
wrote:

>Pythonistas,
>
>I am an experienced Perl programmer just beginning to look at Python
>as I need a scripting langusge for a large project which involves a wide
>range of technologies.  I know from bitter experience that Perl quickly
>becomes unwieldly as project size and complexity increases.
>
>I am impressed by Python, and in evaluating its suitability for the
>project at hand have found extensions to cover all areas of need
>but one: Automation of Submission of Web based Forms.
>
>I am probably missing something, but cannot find support for escaping
>of responses (as per Perl's URI::Escape::uri_escape), support for
>Forms processing (GET and POST methods as per Perl's LWP::Simple
>and LWP::UserAgent modules respectively) or any support for capture
>and use of Cookies.
>
>At first view Python seems very good in the Web Server department,
>but not so good in the Web Client area.  I think I must be missing something,
>as I would have expected all this functionality to be written as part of the 
>Grail project.
>
>Please tell me I've got it wrong!

You've got it wrong. The standard Python distribution includes modules
for quoting/unquoting (escaping) URLs (urllib), submitting, receiving,
and processing GET and POST requests (httplib), and for parsing
MIME-encoded data (mimetools) and RFC822-format headers (rfc822). Of
course there is a module implementing CGI (cgi). Everything you need
is there.

You will find the book "Python Essential Reference" by David Beazley
(published by New Riders) very useful.


Greg Jorgensen
Deschooling Society
Portland, Oregon USA


From shaleh at valinux.com  Sun Feb 18 05:47:29 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Sat, 17 Feb 2001 20:47:29 -0800
Subject: how to unimport something
Message-ID: <20010217204729.D15374@valinux.com>

Suppose I have a program which has logical segments.  How do I run each segment
and have modules loaded in that segment get unloaded when I reach the next
one?



From s713221 at student.gu.edu.au  Sun Feb 18 09:15:00 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Sun, 18 Feb 2001 18:15:00 +1000
Subject: how to unimport something
References: <mailman.982471638.25003.python-list@python.org>
Message-ID: <3A8F8484.9B9C1F64@student.gu.edu.au>

Sean 'Shaleh' Perry wrote:
> 
> Suppose I have a program which has logical segments.  How do I run each segment
> and have modules loaded in that segment get unloaded when I reach the next
> one?

The direct answer is

Import a module into python
>>>import modulename

Delete a module
>>>del modulename

Mind you, you may want to post to the group an example of what you're
doing. Importing modules can take some time to load in, they may be able
to suggest a more efficient and safer method than module swapping to do
what it is you're trying to do.

If you're trying to do what I think you're doing, which is have a series
of modules with similar named functions, and dynamically "plug/unplug"
them from the interpreter, a better way may be to:

>>>import modulea,moduleb,modulec (Import all your plugin modules at the start, drop the cost of module import into startup.)

>>>module = modulea
>>>module.function() (Which is actually modulea.function())
>>>module = moduleb
>>>module.function() (Which is actually moduleb.function())
>>>module = modulec
>>>module.function() (Which is actually modulec.function())
Unless you had the same functions and classes in each module, and they
could accept each others argument patterns, there's the risk of your
script failing because a piece of calling code called module.function()
expecting modulea, but getting moduleb. Messy

If you were planning to add different functionalities to the interpreter
at the same time and remove that functionality later, you would
seriously risk writing code that calls modulea functionality, only to
crash your script because another piece of code wiped out modulea
earlier.

Finally, if you're thinking to only have the modules you need loaded
into the interpreter at any time, this may save some memory, but will
cost you in module loading time. Usually not worth it unless you're
writing scripts which eg. run for days on a resource-scarce server and
it only swaps modules a couple of times a day, or are running the
scripts in a severely memory scarce environment.

Joal Heagney/AncientHart


From shaleh at valinux.com  Sun Feb 18 10:19:12 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Sun, 18 Feb 2001 01:19:12 -0800
Subject: how to unimport something
In-Reply-To: <3A8F8484.9B9C1F64@student.gu.edu.au>; from s713221@student.gu.edu.au on Sun, Feb 18, 2001 at 06:15:00PM +1000
References: <mailman.982471638.25003.python-list@python.org> <3A8F8484.9B9C1F64@student.gu.edu.au>
Message-ID: <20010218011912.B15456@valinux.com>

On Sun, Feb 18, 2001 at 06:15:00PM +1000, s713221 at student.gu.edu.au wrote:
> 
> Mind you, you may want to post to the group an example of what you're
> doing. Importing modules can take some time to load in, they may be able
> to suggest a more efficient and safer method than module swapping to do
> what it is you're trying to do.
>

application does this:

init
do step A (takes some time)
do step B (takes even longer)
do step C (not that long)

each step is decently independent of the previous.  I plan to have A, B and C
implemented as separate modules (perhaps several for each step) and load each
as I need them.

Not every run of the program will reach all three, but most will.

It is possible for each step to take as long as a day or two (it could be given
a thousand files to munge). 



From aleaxit at yahoo.com  Sun Feb 18 12:09:15 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 12:09:15 +0100
Subject: how to unimport something
References: <mailman.982471638.25003.python-list@python.org> <3A8F8484.9B9C1F64@student.gu.edu.au> <mailman.982488021.18367.python-list@python.org>
Message-ID: <96oakn0vmq@news2.newsguy.com>

"Sean 'Shaleh' Perry" <shaleh at valinux.com> wrote in message
news:mailman.982488021.18367.python-list at python.org...
    [snip]
> It is possible for each step to take as long as a day or two (it could be
given

Then a checkpoint/restart architecture, rather than the ability to unload
code modules (which seems to offer little benefit), would seem to be the
key requirement -- if current goes away after 30 hours, you don't want
to waste all the processing you've already done by having to restart
from scratch.  Each 'restart' will only load the modules it needs, so,
rather than 'unloading', once you have the checkpoint/restart part
perfected, you might want at crucial points to stop 'this' process and
start 'another' one from the checkpoint -- process start/stop times are
NOT going to matter, given your timescale is several hours or more,
even on a heavy-process platform such as Windows or VMS, much less
on Unix.  Suitable for cp/rs might be mmap'ped files, so that the
switchover may actually pay no substantial overhead -- but the key
issue, as always in checkpoint/restart, is to identify the minimal useful
amount of *state* to be saved/reloaded.


Alex





From fredrik at pythonware.com  Sun Feb 18 12:34:42 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 11:34:42 GMT
Subject: how to unimport something
References: <mailman.982471638.25003.python-list@python.org> <3A8F8484.9B9C1F64@student.gu.edu.au> <mailman.982488021.18367.python-list@python.org>
Message-ID: <mrOj6.18074$AH6.2468734@newsc.telia.net>

Sean 'Shaleh' Perry wrote:
> application does this:
>
> init
> do step A (takes some time)
> do step B (takes even longer)
> do step C (not that long)
>
> each step is decently independent of the previous.  I plan to have A, B and C
> implemented as separate modules (perhaps several for each step) and load each
> as I need them.

not sure what this has to do with "unimporting", so I'm
probably missing something here, but is there any reason
you cannot do this in the obvious way:

import init_module
init_module.run()

if need_step_A:
    import step_A
    step_A.run()

if need_step_B:
    import step_B
    step_B.run()

if need_step_C:
    import step_C
    step_C.run()

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From shaleh at valinux.com  Mon Feb 19 19:27:42 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Mon, 19 Feb 2001 10:27:42 -0800
Subject: how to unimport something
In-Reply-To: <mrOj6.18074$AH6.2468734@newsc.telia.net>; from fredrik@pythonware.com on Sun, Feb 18, 2001 at 11:34:42AM +0000
References: <mailman.982471638.25003.python-list@python.org> <3A8F8484.9B9C1F64@student.gu.edu.au> <mailman.982488021.18367.python-list@python.org> <mrOj6.18074$AH6.2468734@newsc.telia.net>
Message-ID: <20010219102742.A29164@valinux.com>

On Sun, Feb 18, 2001 at 11:34:42AM +0000, Fredrik Lundh wrote:
> 
> not sure what this has to do with "unimporting", so I'm
> probably missing something here, but is there any reason
> you cannot do this in the obvious way:
>

possibly, I am only at the design phase, was simply curious about ways to
approach this. 



From s713221 at student.gu.edu.au  Sun Feb 18 09:15:20 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Sun, 18 Feb 2001 18:15:20 +1000
Subject: how to unimport something
References: <mailman.982471638.25003.python-list@python.org>
Message-ID: <3A8F8498.34B00D23@student.gu.edu.au>

Sean 'Shaleh' Perry wrote:
> 
> Suppose I have a program which has logical segments.  How do I run each segment
> and have modules loaded in that segment get unloaded when I reach the next
> one?

The direct answer is

Import a module into python
>>>import modulename

Delete a module
>>>del modulename

Mind you, you may want to post to the group an example of what you're
doing. Importing modules can take some time to load in, they may be able
to suggest a more efficient and safer method than module swapping to do
what it is you're trying to do.

If you're trying to do what I think you're doing, which is have a series
of modules with similar named functions, and dynamically "plug/unplug"
them from the interpreter, a better way may be to:

>>>import modulea,moduleb,modulec (Import all your plugin modules at the start, drop the cost of module import into startup.)

>>>module = modulea
>>>module.function() (Which is actually modulea.function())
>>>module = moduleb
>>>module.function() (Which is actually moduleb.function())
>>>module = modulec
>>>module.function() (Which is actually modulec.function())
Unless you had the same functions and classes in each module, and they
could accept each others argument patterns, there's the risk of your
script failing because a piece of calling code called module.function()
expecting modulea, but getting moduleb. Messy

If you were planning to add different functionalities to the interpreter
at the same time and remove that functionality later, you would
seriously risk writing code that calls modulea functionality, only to
crash your script because another piece of code wiped out modulea
earlier.

Finally, if you're thinking to only have the modules you need loaded
into the interpreter at any time, this may save some memory, but will
cost you in module loading time. Usually not worth it unless you're
writing scripts which eg. run for days on a resource-scarce server and
it only swaps modules a couple of times a day, or are running the
scripts in a severely memory scarce environment.

Joal Heagney/AncientHart


From s713221 at student.gu.edu.au  Sun Feb 18 09:16:52 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Sun, 18 Feb 2001 18:16:52 +1000
Subject: how to unimport something
References: <mailman.982471638.25003.python-list@python.org> <3A8F8498.34B00D23@student.gu.edu.au>
Message-ID: <3A8F84F4.A5E9C092@student.gu.edu.au>

Sorry about the double post people. I'm having some sporadic fun with my
university's email server/proxy server/newsserver/dnsserver *sighs*

Joal Heagney/AncientHart


From stevena at neosynapse.net  Sun Feb 18 07:17:29 2001
From: stevena at neosynapse.net (Steven D. Arnold)
Date: Sun, 18 Feb 2001 01:17:29 -0500
Subject: ANN: Pure Python Encryption [UPDATE]
Message-ID: <5.0.2.1.2.20010218011156.03cacbe8@phear.dementian.com>

We've posted an update to the Neosynapse Encryption Tools' IDEA 
implementation.  There was a bug that caused incorrect encryption and 
decryption, even though the test passed.  If you downloaded the previous 
version, you should get the new copy.  Apologies for the inconvenience; 
enjoy!  The software may be downloaded from:

         http://www.neosynapse.net/free_encrypt.htm


--
Steven D. Arnold                   President, Neosynapse Consulting Group
stevena at neosynapse.net                          http://www.neosynapse.net




From c941520 at alinga.newcastle.edu.au  Sun Feb 18 07:19:38 2001
From: c941520 at alinga.newcastle.edu.au (Ben de Luca)
Date: Sun, 18 Feb 2001 17:19:38 +1100
Subject: a few linux/unix related questions
Message-ID: <3a8e191f$0$25513$7f31c96c@news01.syd.optusnet.com.au>

I was wondering how i catch various signals in nix python so i can exit my
programs gracefully?

also how might i start  another application then have two way communication
between it's shell and the python program that started it?




From shaleh at valinux.com  Sat Feb 17 08:07:52 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 23:07:52 -0800
Subject: a few linux/unix related questions
In-Reply-To: <3a8e191f$0$25513$7f31c96c@news01.syd.optusnet.com.au>; from c941520@alinga.newcastle.edu.au on Sun, Feb 18, 2001 at 05:19:38PM +1100
References: <3a8e191f$0$25513$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <20010216230752.A14988@valinux.com>

On Sun, Feb 18, 2001 at 05:19:38PM +1100, Ben de Luca wrote:
> I was wondering how i catch various signals in nix python so i can exit my
> programs gracefully?
>

import signal
 
> also how might i start  another application then have two way communication
> between it's shell and the python program that started it?
>

os.popen() # just like in C

DESCRIPTION
       The  popen()  function opens a process by creating a pipe,
       forking, and invoking the shell.  Since a pipe is by defi-
       nition  unidirectional, the type argument may specify only
       reading or writing, not both; the resulting stream is cor-
       respondingly read-only or write-only.
 



From donn at oz.net  Sat Feb 17 08:29:08 2001
From: donn at oz.net (Donn Cave)
Date: 17 Feb 2001 07:29:08 GMT
Subject: a few linux/unix related questions
References: <3a8e191f$0$25513$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <96l984$8cj$0@216.39.151.169>

Quoth "Ben de Luca" <c941520 at alinga.newcastle.edu.au>:
| I was wondering how i catch various signals in nix python so i can exit my
| programs gracefully?
|
| also how might i start  another application then have two way communication
| between it's shell and the python program that started it?

Well, you could hardly have picked two thornier problems for Python
programming on UNIX.

The answer to the first one is "import signal", and see documention
of that module.  But bear in mind that signals are weird with Python.
The atomic unit of Python execution can span a C external function
call, and that can take an indefinitely long time.  So delivery of
signals at the Python level can be delayed.

The second one is either "import popen2", and suffer the inevitable
shortcomings of UNIX pipes when used in this way (as opposed to normal
use in one way batch data flow), or "import pty" and see if you can
get that to work (will have better luck if you're using Python 2.0.)
(This is not a Python-specific problem, by the way, same in C.)

	Donn Cave, donn at oz.net


From grante at visi.com  Sat Feb 17 17:27:07 2001
From: grante at visi.com (Grant Edwards)
Date: Sat, 17 Feb 2001 16:27:07 GMT
Subject: a few linux/unix related questions
References: <3a8e191f$0$25513$7f31c96c@news01.syd.optusnet.com.au> <mailman.982393640.28440.python-list@python.org>
Message-ID: <slrn98t9ir.g7f.grante@isis.visi.com>

In article <mailman.982393640.28440.python-list at python.org>, Sean 'Shaleh' Perry wrote:

>> also how might i start  another application then have two way communication
                                                         ^^^^^^^
>> between it's shell and the python program that started it?
>>
>
>os.popen() # just like in C
>
>DESCRIPTION
>       The  popen()  function opens a process by creating a pipe,
>       forking, and invoking the shell.  Since a pipe is by defi-
>       nition  unidirectional, the type argument may specify only
                ^^^^^^^^^^^^^^
>       reading or writing, not both; the resulting stream is cor-
>       respondingly read-only or write-only.

For bidirectional communications, use popen[234]

-- 
Grant Edwards                   grante             Yow!  I just remembered
                                  at               something about a TOAD!
                               visi.com            


From phrxy at csv.warwick.ac.uk  Sun Feb 18 08:09:59 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Sun, 18 Feb 2001 07:09:59 +0000
Subject: Unit testing data; What to test
Message-ID: <Pine.SOL.4.30.0102160309090.7729-100000@mimosa.csv.warwick.ac.uk>

Having decided testing was a Good Thing and that I ought to do it, I've
started to write tests, using PyUnit.

The first question is straightforward: do people have a standard, simple
way of handling data for tests, or do you just tend to stick most of it in
the test classes?  KISS I suppose.  But if the software is going to change
a lot, isn't it a good idea to separate the tests from their input and
expected result data?  Of course with big data -- I have some mathematical
functions that need to be checked for example -- you're obviously not
going to dump it directly the test code: I'm wondering more about data of
the order of tens of objects (objects in the Python sense).

In fact, do unit tests often end up having to be rewritten as code is
refactored?  Presumably yes, given the (low) level that unit tests are
written at.

The second (third) one is vague: What / How should one test?  Discuss.

I realise these questions are not strictly python-specific, but (let's
see, how do I justify this) it seems most of what is out there on the web
& USENET is either inappropriately formal, large-scale and
business-orientated for me (comp.software.testing and its FAQ, for
example), or merely describes a testing framework.  A few scattered XP
articles are the only exceptions I've found.

I'm sure there must be good and bad ways to test -- for example, I read
somewhere (can't find it now) that you should aim to end up so that each
bug generates, on (mode) average, one test failure, or at least a small
number.  The justification for this was that lots of tests failing as a
result of a single bug are difficult to deal with.  It seems to me that
this goal is a) impossible to achieve and b) pointless, since if multiple
test failures really are due to a single bug, they will all go away when
you fix it, just as compile-time errors often do.  No?


John



From aleaxit at yahoo.com  Sun Feb 18 09:16:32 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 09:16:32 +0100
Subject: Unit testing data; What to test
References: <Pine.SOL.4.30.0102160309090.7729-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <96o0bj0mib@news2.newsguy.com>

"John J. Lee" <phrxy at csv.warwick.ac.uk> wrote in message
news:Pine.SOL.4.30.0102160309090.7729-100000 at mimosa.csv.warwick.ac.uk...
    [snip]
> The first question is straightforward: do people have a standard, simple
> way of handling data for tests, or do you just tend to stick most of it in
> the test classes?  KISS I suppose.  But if the software is going to change
> a lot, isn't it a good idea to separate the tests from their input and
> expected result data?  Of course with big data -- I have some mathematical
> functions that need to be checked for example -- you're obviously not
> going to dump it directly the test code: I'm wondering more about data of
> the order of tens of objects (objects in the Python sense).

If you keep the (stimuli, expected_responses) sets separate from the
module being tested, you gain much the same benefits (and pay much
the same costs) as by keeping documentation separate from code for
other kinds of docs (think of tests as a form of executable docs...!).

Keeping some tests/docs together with the code simplifies distribution
issues: anybody who has your code also has this minimal set of tests
(and other docs).  However, in some cases, the total amount of tests
and other docs can easily swap the amount of code -- and in this case
keeping them together can easily be thought of as "too costly". The
code can become practically unreadable if there are an order of magnitude
more lines of docs (including tests) than lines of actual source in a .py.

My favourite approach (and I do wish I was more systematical in
practising what I preach:-) is to try and strike a balance: keep with
the code (docstrings, comments, special-purpose test functions) a
reasonable minimal amount of tests (& other docs) -- indicatively,
roughly the same order of magnitude as the source code itself;
but _also_ have a separate set of docs (& tests) for more extensive
needs.  The ideal line of separation would be, is this stuff going to
be needed only by users who are also going to read or change the
sources, or is it going to be more generally useful.  Docs & tests that
only work at the *interface* of the module, without concern for its
*internals*, may allow many users to treat the module as a "black
box", only reading/running/enriching the separate docs-and-tests.


> In fact, do unit tests often end up having to be rewritten as code is
> refactored?  Presumably yes, given the (low) level that unit tests are
> written at.

This is another good guideline: docs and tests best kept with the
source are those who'll likely need changing anyway when the code
is refactored.  The unit of reuse is the unit of release: when you
typically have internals changes that leave alone the interface of
the module, then that interface might usefully be documented and
tested "outside" the source code file -- you can imagine releasing
enhanced sources with identical "externals" docs/tests, or richer
tests/docs that require no re-release of the code itself.


> The second (third) one is vague: What / How should one test?  Discuss.
>
> I realise these questions are not strictly python-specific, but (let's

So is much of what gets discussed here, and, as a born rambler, I
love the ethos of this group, which welcomes such 'somewhat OT'
discussions!-)

> see, how do I justify this) it seems most of what is out there on the web
> & USENET is either inappropriately formal, large-scale and
> business-orientated for me (comp.software.testing and its FAQ, for
> example), or merely describes a testing framework.  A few scattered XP
> articles are the only exceptions I've found.

If XP fits your needs, you could definitely do worse than adopt it
wholesale!  Yes, much of what gets discussed about software
development deals with large-scale SW (in testing and elsewhere) --
that's because problems grow non-linearly with SW scale... when
you release an application that's about 1,000 SLOC, it does not
really matter much if your process and approach are jumbled; at
10,000 SLOC, it's a problem; at 100,000, an utter nightmare, so
you HAVE to be precise in defining your process then (or, call it
100, 1000, 10,000 FP -- but it's really about SLOC more than it
is about FP, which is why higher-level languages help so much).

Differently from what XP specifies, I think tests should be in two
categories (and, similarly, so should deliverables be, rather than
the just-one-deliverable which so characterizes XP -- that is a
tad too X for my conservative self, even though I buy into 70+%
of XP's tenets!).  Which, again, leads us to the internal/external
tests-and-docs split.  External tests and docs (possibly, in large
scale devt, on several scales: module aka unit, subsystem, whole
system) deal with externals/interfaces (not just GUI's &c -- I'm
talking about, e.g, module interfaces to other software; _of course_
'engine' and 'presentation' SHOULD almost invariably be separate
components, but that's another plane of 'split').  Internal tests
and docs deal with internals -- the kind of thing that needs to be
tweaked at each refactoring.

That's not the same dividing plane as the classic unit vs system
test distinction -- it's slanted differently, and perks up again at
each granularity level in a large-enough project (minimal granule
being the module -- perhaps, at Python level, package -- not the
single function or class, since functions and classes inside one
module _are_ typically coupled too strongly to test/release/doc
independently... there are exceptions, modules that are not very
cohesive but rather collections of somewhat unrelated stuff for
basically 'packaging' purposes, but they should be the exception
rather than the rule).


> I'm sure there must be good and bad ways to test -- for example, I read
> somewhere (can't find it now) that you should aim to end up so that each
> bug generates, on (mode) average, one test failure, or at least a small
> number.  The justification for this was that lots of tests failing as a
> result of a single bug are difficult to deal with.  It seems to me that
> this goal is a) impossible to achieve and b) pointless, since if multiple
> test failures really are due to a single bug, they will all go away when
> you fix it, just as compile-time errors often do.  No?

True.  However, if tests are designed in terms of a _sequence_, it IS
often possible to arrange for the most fundamental tests to be run
*first*, ensuring minimal workability of some lower-level, call it
'infrastructural', kind of objects, so that dependent (higher level)
parts can be tested _assuming_ the lower-level stuff works.  This
is more of a consideration for 'internals' kinds of tests, IMHO.


Alex





From shaleh at valinux.com  Sun Feb 18 10:15:53 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Sun, 18 Feb 2001 01:15:53 -0800
Subject: Unit testing data; What to test
In-Reply-To: <Pine.SOL.4.30.0102160309090.7729-100000@mimosa.csv.warwick.ac.uk>; from phrxy@csv.warwick.ac.uk on Sun, Feb 18, 2001 at 07:09:59AM +0000
References: <Pine.SOL.4.30.0102160309090.7729-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <20010218011553.A15456@valinux.com>

On Sun, Feb 18, 2001 at 07:09:59AM +0000, John J. Lee wrote:
> 
> Having decided testing was a Good Thing and that I ought to do it, I've
> started to write tests, using PyUnit.
> 
> The second (third) one is vague: What / How should one test?  Discuss.
> 
> I'm sure there must be good and bad ways to test -- for example, I read
> somewhere (can't find it now) that you should aim to end up so that each
> bug generates, on (mode) average, one test failure, or at least a small
> number.  The justification for this was that lots of tests failing as a
> result of a single bug are difficult to deal with.  It seems to me that
> this goal is a) impossible to achieve and b) pointless, since if multiple
> test failures really are due to a single bug, they will all go away when
> you fix it, just as compile-time errors often do.  No?
>

How I commonly see it done (and how I do it):

common cases
edge cases -- perhaps -1 could be valid in some cases and not others
              or maybe you have a buffer which is only soo big
bad cases -- but -2 never is

You want to see if your code fails or gives expected output.  The other big
thing is every time you find a bug, you fix it and then add a test case that
triggers the bug.  So if you run the buggy version you get an error and the
new version is clean.  This way, if you accidently remove the fix or cause
a similar bug elsewhere, your tests catch it.



From stephen_purcell at yahoo.com  Sun Feb 18 11:17:07 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sun, 18 Feb 2001 11:17:07 +0100
Subject: Unit testing data; What to test
In-Reply-To: <Pine.SOL.4.30.0102160309090.7729-100000@mimosa.csv.warwick.ac.uk>; from phrxy@csv.warwick.ac.uk on Sun, Feb 18, 2001 at 07:09:59AM +0000
References: <Pine.SOL.4.30.0102160309090.7729-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <20010218111707.E19924@freedom.puma-ag.com>

John J. Lee wrote:
> [snip] But if the software is going to change
> a lot, isn't it a good idea to separate the tests from their input and
> expected result data?

Do whatever is clearest, but try to arrange the tests such that their data
doesn't change often. High-maintenance test cases are either bad test cases
or a sign that the tested code is being changed for no good reason.

> In fact, do unit tests often end up having to be rewritten as code is
> refactored?  Presumably yes, given the (low) level that unit tests are
> written at.

Refactoring should normally not break tests -- it means "now the code is
working and passing the tests, let's remove cruft".

XP zealots might (don't quote me) go as far as saying that any change that
breaks the tests is a feature change, not a refactoring.

Refactoring of tests is a separate activity.

> The second (third) one is vague: What / How should one test?  Discuss.

Test *everything*, and do it thoroughly!

If you want answers implying less work, there are probably some guidelines
somewhere...

Personally, I like the XP testing approach:

  1. Identify a requirement that some code has to do X
  2. Write a test where we try to do X, and check that it worked
  3. Keep adding or changing code until the test passes, and no other test
     breaks.

Finding a bug is equivalent to identifying a requirement that some code should
not do Y -- write a test, and code until it passes.

A complete set of tests thus written forms a catalogue of what the code is
guaranteed to do correctly; use of the tested code in any other manner is
not guaranteed to work.

Some people make the mistake of wasting a lot of testing effort on checking
that their code does something "sensible" if misused. There's no time-efficient
way to catch mistakes made by people who didn't read the manual.

Think of the tests as a combination of 'programmers manual' and 'example code'.

> I'm sure there must be good and bad ways to test

True enough. The bad way is to *not* test.  :-)

Really, the point is that time spent wondering or reading about how best to
test would be better spent simply writing as many tests as possible.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From gherman at darwin.in-berlin.de  Wed Feb 21 11:10:19 2001
From: gherman at darwin.in-berlin.de (Dinu Gherman)
Date: Wed, 21 Feb 2001 11:10:19 +0100
Subject: Unit testing data; What to test
References: <Pine.SOL.4.30.0102160309090.7729-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <3A93940B.E09817A3@darwin.in-berlin.de>

"John J. Lee" wrote:
> 
> Having decided testing was a Good Thing and that I ought to do it, I've
> started to write tests, using PyUnit.
> 
> The first question is straightforward: do people have a standard, simple
> way of handling data for tests, or do you just tend to stick most of it in
> the test classes?  KISS I suppose.  But if the software is going to change
> a lot, isn't it a good idea to separate the tests from their input and
> expected result data?  Of course with big data -- I have some mathematical
> functions that need to be checked for example -- you're obviously not
> going to dump it directly the test code: I'm wondering more about data of
> the order of tens of objects (objects in the Python sense).
> 
> In fact, do unit tests often end up having to be rewritten as code is
> refactored?  Presumably yes, given the (low) level that unit tests are
> written at.
> 
> The second (third) one is vague: What / How should one test?  Discuss.
> 
> I realise these questions are not strictly python-specific, but (let's
> see, how do I justify this) it seems most of what is out there on the web
> & USENET is either inappropriately formal, large-scale and
> business-orientated for me (comp.software.testing and its FAQ, for
> example), or merely describes a testing framework.  A few scattered XP
> articles are the only exceptions I've found.
> 
> I'm sure there must be good and bad ways to test -- for example, I read
> somewhere (can't find it now) that you should aim to end up so that each
> bug generates, on (mode) average, one test failure, or at least a small
> number.  The justification for this was that lots of tests failing as a
> result of a single bug are difficult to deal with.  It seems to me that
> this goal is a) impossible to achieve and b) pointless, since if multiple
> test failures really are due to a single bug, they will all go away when
> you fix it, just as compile-time errors often do.  No?
> 
> John

I suggest reading some of the written material about XP
that you can find by searching your favourite online book
dealer for "extreme programming". I get five hits for the
series that was started with Kent Beck with Addison-Wesley,
two of which are not yet published... All of the first 
three that I know are *thin* books and easy to read!

In general I'd say only that much: knowbody knows better
*what* to test than people writing the code (after the use
cases / user stories have been determined)! So, if there 
is a lack of knowledge about what to test (I'm not sure this
is the case here, though, but this is the subject line ;-) 
there must be some deeper issue on the common understanding 
of what the system is expected to do.

As for organizing the tests: see the books I mentioned, but
don't expect a detailed process description. Everything that 
works for you will be fine! If something won't: adapt, shake 
and reiterate! ;-)

Regards,

Dinu

-- 
Dinu C. Gherman
ReportLab Consultant - http://www.reportlab.com
................................................................
"The only possible values [for quality] are 'excellent' and 'in-
sanely excellent', depending on whether lives are at stake or 
not. Otherwise you don't enjoy your work, you don't work well, 
and the project goes down the drain." 
                    (Kent Beck, "Extreme Programming Explained")


From phrxy at csv.warwick.ac.uk  Wed Feb 28 18:41:01 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Wed, 28 Feb 2001 17:41:01 +0000
Subject: Unit testing data; What to test
Message-ID: <Pine.SOL.4.30.0102251735030.6152-100000@mimosa.csv.warwick.ac.uk>

Sorry about the formatting etc. of this post -- I had to cut and paste
from Google.

> "John J. Lee" <phrxy at csv.warwick.ac.uk> wrote in message
> news:Pine.SOL.4.30.0102160309090.7729-100000 at mimosa.csv.warwick.ac.uk...
[...]
> If you keep the (stimuli, expected_responses) sets separate from the
> module being tested, you gain much the same benefits (and pay much
> the same costs) as by keeping documentation separate from code for
> other kinds of docs (think of tests as a form of executable docs...!).
>
> Keeping some tests/docs together with the code simplifies distribution
> issues: anybody who has your code also has this minimal set of tests
> (and other docs).

Well, that's not a problem, is it (unless the tests really are huge)?
They can be in the distribution without being mixed in with the code.

> However, in some cases, the total amount of tests
> and other docs can easily swap the amount of code -- and in this case
> keeping them together can easily be thought of as "too costly". The
> code can become practically unreadable if there are an order of
magnitude
> more lines of docs (including tests) than lines of actual source in a
.py.

Agreed.  My tests are separate from the code -- it hadn't occurred to me
to keep them in the same file.  It was the separation (or not) of test
data and test code I was wondering about, though.

> My favourite approach (and I do wish I was more systematical in
> practising what I preach:-) is to try and strike a balance: keep with
> the code (docstrings, comments, special-purpose test functions) a
> reasonable minimal amount of tests (& other docs) -- indicatively,
> roughly the same order of magnitude as the source code itself;
> but _also_ have a separate set of docs (& tests) for more extensive
> needs.
[...]

This is all good advice, but it doesn't actually answer my question.  ;-)
Anyway, I think I will try having some tests in the code, sounds like a
great idea considering the tests-as-docs aspect.

> > The second (third) one is vague: What / How should one test?  Discuss.
[...]
> > see, how do I justify this) it seems most of what is out there on the web
> > & USENET is either inappropriately formal, large-scale and
[...]
> > A few scattered XP
> > articles are the only exceptions I've found.
[...]
> If XP fits your needs, you could definitely do worse than adopt it
> wholesale!  Yes, much of what gets discussed about software
> development deals with large-scale SW (in testing and elsewhere) --
> that's because problems grow non-linearly with SW scale... when
> you release an application that's about 1,000 SLOC, it does not
> really matter much if your process and approach are jumbled; at

Let me assure you it _is_ possible to achieve an unmaintainable mess in
~1000 SLOC.  I have personally achieved this in, let's see (C-x C-f
munge.pl RET) ... 832 SLOC.  This is one of many things that Perl makes
easy (this is completely unfair to Perl of course -- what really makes it
easy is starting out thinking 'this is going to be a 100 line script' and
then extending by cut and paste -- I've learned my lesson, honest!).

> 10,000 SLOC, it's a problem; at 100,000, an utter nightmare, so
> you HAVE to be precise in defining your process then (or, call it

Yeah, of course, I understand that.  But there is still a place for lots
of testing for smaller efforts, minus the heavyweight formal process.

> 100, 1000, 10,000 FP -- but it's really about SLOC more than it
> is about FP, which is why higher-level languages help so much).
>
> Differently from what XP specifies, I think tests should be in two
> categories (and, similarly, so should deliverables be, rather than
> the just-one-deliverable which so characterizes XP -- that is a
> tad too X for my conservative self, even though I buy into 70+%
> of XP's tenets!).

Which two categories?  External and internal?  How is this different from
the XP unit test versus acceptance tests division?  Is your point just
that the external / internal division applies on more levels than just
final user / everything else -- which is what you seem to be talking about
below?

here's a snip explaining what XP people mean by acceptance tests for
anybody that hasn't read about it:

http://www.extremeprogramming.org/rules/functionaltests.html
----------

      Acceptance tests are created from user
stories. During an iteration the user stories selected
during the iteration planning meeting will be
translated into acceptance tests. The customer
specifies scenarios to test when a user story has
been correctly implemented. A story can have one
or many acceptance tests, what ever it takes to
ensure the functionality works.

      Acceptance tests are black box system tests.
Each acceptance test represents some expected
result from the system. Customers are responsible
for verifying the correctness of the acceptance tests
and reviewing test scores to decide which failed
tests are of highest priority. Acceptance tests are
also used as regression tests prior to a production
release.

----------

> Which, again, leads us to the internal/external
> tests-and-docs split.  External tests and docs (possibly, in large
> scale devt, on several scales: module aka unit, subsystem, whole
> system) deal with externals/interfaces (not just GUI's &c -- I'm
> talking about, e.g, module interfaces to other software; _of course_
> 'engine' and 'presentation' SHOULD almost invariably be separate
> components, but that's another plane of 'split').  Internal tests
> and docs deal with internals -- the kind of thing that needs to be
> tweaked at each refactoring.
>
> That's not the same dividing plane as the classic unit vs system
> test distinction -- it's slanted differently, and perks up again at
> each granularity level in a large-enough project (minimal granule
> being the module -- perhaps, at Python level, package -- not the
> single function or class, since functions and classes inside one
> module _are_ typically coupled too strongly to test/release/doc
> independently... there are exceptions, modules that are not very
> cohesive but rather collections of somewhat unrelated stuff for
> basically 'packaging' purposes, but they should be the exception
> rather than the rule).
[...]


John



From eatonalive%qwestinternet.net at pop3.qwestinternet.net  Sun Feb 18 09:00:11 2001
From: eatonalive%qwestinternet.net at pop3.qwestinternet.net (Jim Eaton)
Date: 18 Feb 2001 08:00:11 GMT
Subject: What to do after Python?
Message-ID: <3A8F81BC.108488CA@pop3.qwestinternet.net>

I am learning Python as a first language, and I have been wondering what
is a good language to learn after Python?  I'm thinking of either going
into C or Java but I'm not sure which one because I've heard many
arguments either way.  Will going into Java first be any detriment to
learning C later?  Any suggestions would be greatly appreciated.

Thank you in advance.


Josh
cooljage at hotmail.com



From s713221 at student.gu.edu.au  Sun Feb 18 09:22:05 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Sun, 18 Feb 2001 18:22:05 +1000
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <3A8F862D.4BE238F9@student.gu.edu.au>

Jim Eaton wrote:
> 
> I am learning Python as a first language, and I have been wondering what
> is a good language to learn after Python?  I'm thinking of either going
> into C or Java but I'm not sure which one because I've heard many
> arguments either way.  Will going into Java first be any detriment to
> learning C later?  Any suggestions would be greatly appreciated.
> 
> Thank you in advance.
> 
> Josh
> cooljage at hotmail.com

>From a very python-predudiced view, if you're using CPython, the C
languages will let you write C extensions for python. If you're using
JPython, then Java may be more useful.

Joal Heagney/AncientHart


From aleaxit at yahoo.com  Sun Feb 18 09:25:35 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 09:25:35 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <96o0su0mv6@news2.newsguy.com>

"Jim Eaton" <eatonalive%qwestinternet.net at pop3.qwestinternet.net> wrote in
message news:3A8F81BC.108488CA at pop3.qwestinternet.net...
> I am learning Python as a first language, and I have been wondering what
> is a good language to learn after Python?

There are several possibilities, but I would suggest SQL as the second
language.  It's incredibly useful AND when used well it urges you to
think in very different ways from a sequential imperative language
(such as Python, C, Java, ...) -- thus at the same time broadening
your horizons AND providing you with LOTS of real-world returns.

Similarly, for a THIRD language I would suggest some structured markup
language such as XML (with XPath, XSLT...) or HTML (4.0, with its DOM).

> I'm thinking of either going
> into C or Java but I'm not sure which one because I've heard many
> arguments either way.  Will going into Java first be any detriment to
> learning C later?  Any suggestions would be greatly appreciated.

A lower-level language such as Java or C might be a good 4th language.
Java is far simpler (as a language: the _libraries_ are extremely rich,
but that's another issue) and thus a better 4th step than C (the latter,
or C++, might come later).

Python (in the flavor of Jython, when you go for Java) will help you
hugely in each of these further study endeavours, BTW.


Alex





From max at alcyone.com  Sun Feb 18 11:13:42 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 02:13:42 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96o0su0mv6@news2.newsguy.com>
Message-ID: <3A8FA056.F7E767B4@alcyone.com>

Alex Martelli wrote:

> There are several possibilities, but I would suggest SQL as the second
> language.  It's incredibly useful AND when used well it urges you to
> think in very different ways from a sequential imperative language
> (such as Python, C, Java, ...) -- thus at the same time broadening
> your horizons AND providing you with LOTS of real-world returns.
> 
> Similarly, for a THIRD language I would suggest some structured markup
> language such as XML (with XPath, XSLT...) or HTML (4.0, with its
> DOM).

I think he meant programming languages, not database query languages or
markup languages.

Java or C++ would probably be up his alley, depending on exactly what he
wanted to do with that further knowledge.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Human love is often but the encounter of two weaknesses.
\__/ Francois Mauriac
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From aleaxit at yahoo.com  Sun Feb 18 12:03:53 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 12:03:53 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96o0su0mv6@news2.newsguy.com> <3A8FA056.F7E767B4@alcyone.com>
Message-ID: <96oab20vda@news2.newsguy.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3A8FA056.F7E767B4 at alcyone.com...
    [snip]
> > There are several possibilities, but I would suggest SQL as the second
    [snip]
> I think he meant programming languages, not database query languages or
> markup languages.

This is possible, but maybe (as long as we're reading his mind anyway:-)
by the 'programming' adjective (that he didn't use but you think he had
implied) he meant languages *used IN* programming -- and SQL and
XML (particularly the XSLT part thereof) then qualify again:-). That's why
I was suggesting them rather than, say, French, Esperanto, or Latin:-).


> Java or C++ would probably be up his alley, depending on exactly what he
> wanted to do with that further knowledge.

C++, IMHO, is somewhat too complicated even as a second programming
language; unless C++ is a specific requirement, Java is more approchable.


Alex





From anton at home.ogre.nu  Tue Feb 20 07:45:11 2001
From: anton at home.ogre.nu (Anton Sherwood)
Date: 19 Feb 2001 22:45:11 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96o0su0mv6@news2.newsguy.com> <3A8FA056.F7E767B4@alcyone.com> <96oab20vda@news2.newsguy.com>
Message-ID: <96t3pn$3ij$1@home.ogre.nu>

Alex Martelli <aleaxit at yahoo.com> writes
: . . . That's why I was suggesting them [SQL,XML]
: rather than, say, French, Esperanto, or Latin:-).

Eerie!  I haven't even *read* this group before, and already
someone makes a crack about my fourth, second and third languages!

ObP: Today I had my first occasion to use string.join, map, lambda --
all in the same expression.

-- 
Anton Sherwood  --  br0nt0 at p0b0x.com  --  http://ogre.nu/


From andrew at andrewcooke.free-online.co.uk  Sun Feb 18 09:53:47 2001
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Sun, 18 Feb 2001 08:53:47 +0000
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <3A8F8D9B.78B37578@andrewcooke.free-online.co.uk>

Oh boy, what a choice :-)

What do you want to do?  Learn something *completely* different?  Or
explore different ways of looking at Objects?  Or learn a faster
language?  Or...?

For completely different, how about a functional language - there are
dozens, but maybe Scheme/Lisp or OCaml or Haskell?  Or a stack-based
language like Forth?  Or a logical language like Prolog or Oz or
Mercury?

For different OOP, how about Smalltalk or Eiffel?

For something faster, how about C++ (although all the above have
compilers and will be faster than C - what I mean here is a language
that you can use to get a job easilly, and one that, being rather
unpleasant, means you still use Python for the little tasks ;-)

Have a poke around lambda (http://lambda.weblogs.com) for more
ideas/references.

I also wrote a sketchy summary of many different languages (these days
I'm not that proud of it, but it was really written to answer exactly
this question) which you can find at
http://www.andrewcooke.free-online.co.uk/andrew/lang.html

If you post with what you'd like to do next (if you have a particular
project in mind, or what to learn something specific) then we might be
able to narrow the choice.  But thereare hundreds of languages out
there, all great in their own way:
http://directory.google.com/Top/Computers/Programming/Languages/

Have fun,
Andrew

Jim Eaton wrote:
> 
> I am learning Python as a first language, and I have been wondering what
> is a good language to learn after Python?  I'm thinking of either going
> into C or Java but I'm not sure which one because I've heard many
> arguments either way.  Will going into Java first be any detriment to
> learning C later?  Any suggestions would be greatly appreciated.
> 
> Thank you in advance.
> 
> Josh
> cooljage at hotmail.com


From andrew at andrewcooke.free-online.co.uk  Sun Feb 18 09:56:06 2001
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Sun, 18 Feb 2001 08:56:06 +0000
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A8F8D9B.78B37578@andrewcooke.free-online.co.uk>
Message-ID: <3A8F8E26.2D8D948F@andrewcooke.free-online.co.uk>


Andrew Cooke wrote:
[...]
> For something faster, how about C++ (although all the above have
> compilers and will be faster than C - what I mean here is a language
[...]

Sorry - faster than Python!

(and I add "probably")
(and the comment that speed isn't that important - I use and like Python
myself)
(and the comment that Python isn't that slow).

OK?

Andrew


From mwh21 at cam.ac.uk  Sun Feb 18 11:34:04 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Feb 2001 10:34:04 +0000
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <m3ae7ke0v7.fsf@atrus.jesus.cam.ac.uk>

Jim Eaton <eatonalive%qwestinternet.net at pop3.qwestinternet.net> writes:

> I am learning Python as a first language, and I have been wondering
> what is a good language to learn after Python?  I'm thinking of
> either going into C or Java but I'm not sure which one because I've
> heard many arguments either way.  Will going into Java first be any
> detriment to learning C later?  Any suggestions would be greatly
> appreciated.

Depends what you want to do, obviously.  If you're after a job, then
C, C++, Java are good choices.  If you want to broaden your mind, try
scheme or Common Lisp or Haskell or some variant of ML or Smalltalk...

Cheers,
M.

-- 
  SPIDER:  'Scuse me. [scuttles off]
  ZAPHOD:  One huge spider.
    FORD:  Polite though.
                   -- The Hitch-Hikers Guide to the Galaxy, Episode 11


From aleaxit at yahoo.com  Sun Feb 18 11:59:15 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 11:59:15 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <m3ae7ke0v7.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <96oa1e0v4h@news2.newsguy.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3ae7ke0v7.fsf at atrus.jesus.cam.ac.uk...
> Jim Eaton <eatonalive%qwestinternet.net at pop3.qwestinternet.net> writes:
>
> > I am learning Python as a first language, and I have been wondering
> > what is a good language to learn after Python?  I'm thinking of
> > either going into C or Java but I'm not sure which one because I've
> > heard many arguments either way.  Will going into Java first be any
> > detriment to learning C later?  Any suggestions would be greatly
> > appreciated.
>
> Depends what you want to do, obviously.  If you're after a job, then
> C, C++, Java are good choices.  If you want to broaden your mind, try
> scheme or Common Lisp or Haskell or some variant of ML or Smalltalk...

Perfectly correct (though I would not put things in this specific order:-),
but I just want to underline my original suggestion in this light: SQL
does _both_ -- it broadens the mind, being a non-procedural, set
oriented language, *AND* is also a good choice in terms of practical
use and on-the-job usefulness (most programmers use relational DB's,
very few use them _really well_...:-).  My second suggestion, for some
structured-markup language such as XML (with XPath, XLink, XSLT...)
is much in the same vein, although more debatable.


Alex





From bowman at montana.com  Sun Feb 18 17:05:40 2001
From: bowman at montana.com (bowman)
Date: Sun, 18 Feb 2001 09:05:40 -0700
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <m3ae7ke0v7.fsf@atrus.jesus.cam.ac.uk> <96oa1e0v4h@news2.newsguy.com>
Message-ID: <AsSj6.302$8u5.1400@newsfeed.slurp.net>

Alex Martelli <aleaxit at yahoo.com> wrote in message
news:96oa1e0v4h at news2.newsguy.com...
> Perfectly correct (though I would not put things in this specific
order:-),
> but I just want to underline my original suggestion in this light: SQL
> does _both_ -- it broadens the mind

the only thing SQL ever broadened was my boredom. Once the initial concepts
are grasped, which takes about an hour, the rest is endless repetition. I
usually feel like I'm headed for the Augean stables with a very small
shovel.

being a crusty old C programmer, I don't consider RPG a language, either. I
admit many people have had long and lucrative careers dabbling with
databases, but I've also heard of a language, Cobol or something, that is
also widely used.

XML? Perhaps it will amount to something -- someday.





From roy at panix.com  Sun Feb 18 17:22:58 2001
From: roy at panix.com (Roy Smith)
Date: Sun, 18 Feb 2001 11:22:58 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <m3ae7ke0v7.fsf@atrus.jesus.cam.ac.uk> <96oa1e0v4h@news2.newsguy.com> <AsSj6.302$8u5.1400@newsfeed.slurp.net>
Message-ID: <roy-0CD73B.11225818022001@news.panix.com>

"bowman" <bowman at montana.com> wrote:
> the only thing SQL ever broadened was my boredom. Once the initial concepts
> are grasped, which takes about an hour, the rest is endless repetition. I
> usually feel like I'm headed for the Augean stables with a very small
> shovel.

I agree that SQL is about as exciting as a rotting turnip, but the fact is 
that there is a heck of a lot of SQL going on in the world, and it's a good 
skill to know.  What's more, the Python SQL interface makes the two a 
really powerful combination.  I find what works well most of the time is to 
let SQL dig out the data you want, then use Python to manipulate it.

I'm working right now with a product which was historically SQL-based.  The 
stuff we're doing now has evolved almost entirely away from SQL, but we 
still maintain a copy of some data in Oracle for backward-compatability.  
Every so often, while my co-workers are struggling to figure something out, 
I'll impress them by firing up sqlplus, typing "select blah from blah where 
blah blah blah", and getting the answer in 30 seconds.


From aleaxit at yahoo.com  Sun Feb 18 19:57:35 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 19:57:35 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <m3ae7ke0v7.fsf@atrus.jesus.cam.ac.uk> <96oa1e0v4h@news2.newsguy.com> <AsSj6.302$8u5.1400@newsfeed.slurp.net>
Message-ID: <96p62s05r@news1.newsguy.com>

"bowman" <bowman at montana.com> wrote in message
news:AsSj6.302$8u5.1400 at newsfeed.slurp.net...
>
> Alex Martelli <aleaxit at yahoo.com> wrote in message
> news:96oa1e0v4h at news2.newsguy.com...
> > Perfectly correct (though I would not put things in this specific
> order:-),
> > but I just want to underline my original suggestion in this light: SQL
> > does _both_ -- it broadens the mind
>
> the only thing SQL ever broadened was my boredom. Once the initial
concepts
> are grasped, which takes about an hour, the rest is endless repetition. I

I wonder if you're one of the greatest geniuses ever lived (to grasp the
concepts of relational databases in an hour!), or one of the myriad people
who didn't get SQL taught to them _right_.  You _have_ read Celko's
"SQL for Smarties", for example?  And beyond that, do you understand,
say, outer joins and why you might want to use one, the implications
of update-propagation strategies and their interactions with integrity
constraints, or locking and transactioning issues in typical scenarios...?


Alex





From bowman at montana.com  Sun Feb 18 23:05:44 2001
From: bowman at montana.com (bowman)
Date: Sun, 18 Feb 2001 15:05:44 -0700
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <m3ae7ke0v7.fsf@atrus.jesus.cam.ac.uk> <96oa1e0v4h@news2.newsguy.com> <AsSj6.302$8u5.1400@newsfeed.slurp.net> <96p62s05r@news1.newsguy.com>
Message-ID: <9KXj6.2408$8u5.4426@newsfeed.slurp.net>

Alex Martelli <aleaxit at yahoo.com> wrote in message
news:96p62s05r at news1.newsguy.com...
> "SQL for Smarties", for example?  And beyond that, do you understand,
> say, outer joins and why you might want to use one, the implications
> of update-propagation strategies and their interactions with integrity
> constraints, or locking and transactioning issues in typical scenarios...?

was that the left or right outer join? or the less useful, but symmetrical,
full outer join? on good days, I can even tell the difference between 2NF
and BCNF and discuss the subtle joys of update anomalies.

However, database programming still bores me utterly. Fortunately, there are
people in this world who actually seem to enjoy it, as well as business
computing. Personally, I find discussions of the uses of fuzzy logic in
anti-lock brake systems algorithms more entertaining.





From mkx at excite.com  Mon Feb 19 18:10:25 2001
From: mkx at excite.com (mkx at excite.com)
Date: Mon, 19 Feb 2001 12:10:25 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <m3ae7ke0v7.fsf@atrus.jesus.cam.ac.uk> <96oa1e0v4h@news2.newsguy.com> <AsSj6.302$8u5.1400@newsfeed.slurp.net>
Message-ID: <p7k29tsgverrd253tpk0op74f997l4tefh@4ax.com>

On Sun, 18 Feb 2001 09:05:40 -0700, "bowman" <bowman at montana.com>
wrote:

>the only thing SQL ever broadened was my boredom. Once the initial concepts
>are grasped, which takes about an hour, the rest is endless repetition.

If your scope for these "initial concepts" is bound by simple SELECT
queries than perhaps the statement above is accurate. However, it
seems the many people believe that the SQL language is simply what
they see built by the Microsoft Access query builder.

The SQL language has many of the same constructs found in most other
languages (procedures, functions, recursion, etc.) But within the
environment provided by a RDBMS, the SQL language is backed by a lot
of power. Too often, I have seen "experienced" programmers consume
excessive server resources by needlessly executing multiple queries
inside loops of other program languages, because they have not spent
the time to learn that a properly written SQL query can save many
iterations on the "program side."

If one is to develop software that uses a RDBMS, then a full command
of the SQL language is as important as a command of the language used
to develop the application itself. Too often, I have seen the example
where the development staff complains that the "database is too slow",
beacuse their code that builds a hierarchical tree-view display takes
30-60 seconds before the web page can be built. Inspection shows that
the web server is exectuing separate SELECTs for each branch of the
tree, requiring 100's of queries to be executed to build a single
page. Once these programmers see that the same hierarchical structure
can be built with a single database call to a stored procedure, or
recursive query, and completes in a second or two, then they are
interested in really learning the language.

>being a crusty old C programmer, I don't consider RPG a language, either.

What defines a programming language, by your views?


From spamers at must.die  Sun Feb 18 12:21:07 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Sun, 18 Feb 2001 12:21:07 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31@4ax.com>

On 18 Feb 2001 08:00:11 GMT, Jim Eaton
<eatonalive%qwestinternet.net at pop3.qwestinternet.net> wrote:

>I am learning Python as a first language, and I have been wondering what
>is a good language to learn after Python?  I'm thinking of either going
>into C or Java but I'm not sure which one because I've heard many
>arguments either way.  Will going into Java first be any detriment to
>learning C later?  Any suggestions would be greatly appreciated.

A good combination is, IMHO: Python, CommonLisp, C++ and Prolog.  From a
didactic point of view, good second languages are CommonLisp (a more powerful
Python-like language) and Java (an easier to learn C/C++ language).

Remember that there's much more to programming than just learning the syntax
of different languages. If you're new to programming, you should first learn
some basic concepts.  Good beginers books are: "Simply Scheme: Introducing
computer science" and "The Schemer's Guide".  Both use a lisp dialect called
Scheme, that should be easy to undertand (sort of Python with extra
parentheses).





//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From hannah at schlund.de  Mon Feb 19 16:14:35 2001
From: hannah at schlund.de (Hannah Schroeter)
Date: 19 Feb 2001 16:14:35 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31@4ax.com>
Message-ID: <96rd8r$asp$1@c3po.schlund.de>

Hello!

In article <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31 at 4ax.com>,
Fernando Rodr?guez  <spamers at must.die> wrote:
>[...]

>Remember that there's much more to programming than just learning the syntax
>of different languages. If you're new to programming, you should first learn
>some basic concepts.  Good beginers books are: "Simply Scheme: Introducing
>computer science" and "The Schemer's Guide".  Both use a lisp dialect called
>Scheme, that should be easy to undertand (sort of Python with extra
>parentheses).

Then, you should also suggest "Structure and Interpretation of Computer
Programs".

Kind regards,

Hannah.


From cg at gaia.intranet.cdegroot.com  Mon Feb 19 21:08:02 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 19 Feb 2001 21:08:02 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31@4ax.com> <96rd8r$asp$1@c3po.schlund.de>
Message-ID: <96ruf2$ht8$1@gaia.intranet.cdegroot.com>

Hannah Schroeter <hannah at schlund.de> said:
>Then, you should also suggest "Structure and Interpretation of Computer
>Programs".
>
And "Goedel, Escher, Bach", "The Timeless Way of Building", and other
assorted non-programming books that are important for programmers... 

-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From scarblac at pino.selwerd.nl  Tue Feb 20 14:53:33 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 20 Feb 2001 13:53:33 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31@4ax.com> <96rd8r$asp$1@c3po.schlund.de> <96ruf2$ht8$1@gaia.intranet.cdegroot.com>
Message-ID: <slrn994tif.7gk.scarblac@pino.selwerd.nl>

Cees de Groot <cg at gaia.intranet.cdegroot.com> wrote in comp.lang.python:
> Hannah Schroeter <hannah at schlund.de> said:
> >Then, you should also suggest "Structure and Interpretation of Computer
> >Programs".
> 
> And "Goedel, Escher, Bach", "The Timeless Way of Building", and other
> assorted non-programming books that are important for programmers... 
 
Ooh, nice thread. I'll add "Monty Python: Just The Words" (so you can
understand comments), "Alice in Wonderland", "The Hitchhiker's Guide to the
Galaxy", and a dose of recent SF (Stephenson, Vinge...).

-- 
Remco Gerlich


From amk at mira.erols.com  Wed Feb 21 02:48:43 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 21 Feb 2001 01:48:43 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31@4ax.com> <96rd8r$asp$1@c3po.schlund.de> <96ruf2$ht8$1@gaia.intranet.cdegroot.com> <slrn994tif.7gk.scarblac@pino.selwerd.nl>
Message-ID: <slrn9968eu.lf.amk@207-172-36-141.s141.tnt6.ann.va.dialup.rcn.com>

On 20 Feb 2001 13:53:33 GMT, Remco Gerlich <scarblac at pino.selwerd.nl> wrote:
>Ooh, nice thread. I'll add "Monty Python: Just The Words" (so you can
>understand comments), "Alice in Wonderland", "The Hitchhiker's Guide to the
>Galaxy", and a dose of recent SF (Stephenson, Vinge...).

If people want to provide slightly longer descriptions of books, 
I'll happily add them to the Python bookstore (www.amk.ca/bookstore/).

--amk


From spamers at must.die  Tue Feb 20 10:32:26 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Tue, 20 Feb 2001 10:32:26 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31@4ax.com> <96rd8r$asp$1@c3po.schlund.de>
Message-ID: <56n29tcv1cd6tbs0idh1cooavielh2kupd@4ax.com>

On 19 Feb 2001 16:14:35 +0100, hannah at schlund.de (Hannah Schroeter) wrote:


>Then, you should also suggest "Structure and Interpretation of Computer
>Programs".
 
I swear by that book, but I'm not sure if it's a good 1st book on programming.
That's why I suggested something simpler to start with, but if the OP has a
good math background or likes math, then go for it. :-)




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From rpm1deleteme at frontiernet.net  Sun Feb 18 15:40:22 2001
From: rpm1deleteme at frontiernet.net (RPM1)
Date: Sun, 18 Feb 2001 09:40:22 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <96omoi$1jls$1@node17.cwnet.frontiernet.net>

For a job: C, C++, Java, Visual Basic (yuk).  Java would probably be the
easiest transition from Python.  If you are sure you'll want to learn C in
the future, then you should start with C first.  It's not that hard.  It's
actually a smaller language than any of the others.  The drawbacks with C
are no GUI built in and more importantly no safety features.  C's strongest
feature is that you can do anything you want... that's also its biggest
problem.  I think you'll get 'spoiled' if you learn Java first.

Patrick Mullhaupt, (C programmer)



Jim Eaton wrote in message <3A8F81BC.108488CA at pop3.qwestinternet.net>...
>I am learning Python as a first language, and I have been wondering what
>is a good language to learn after Python?  I'm thinking of either going
>into C or Java but I'm not sure which one because I've heard many
>arguments either way.  Will going into Java first be any detriment to
>learning C later?  Any suggestions would be greatly appreciated.
>




From scarblac at pino.selwerd.nl  Sun Feb 18 16:13:44 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 18 Feb 2001 15:13:44 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <slrn98vph8.45u.scarblac@pino.selwerd.nl>

Jim Eaton wrote in comp.lang.python:
> I am learning Python as a first language, and I have been wondering what
> is a good language to learn after Python?  I'm thinking of either going
> into C or Java but I'm not sure which one because I've heard many
> arguments either way.  Will going into Java first be any detriment to
> learning C later?  Any suggestions would be greatly appreciated.

This depends on what you want to do, and on the things you want to do but
can't do in Python.

-- 
Remco Gerlich


From bowman at montana.com  Sun Feb 18 17:21:38 2001
From: bowman at montana.com (bowman)
Date: Sun, 18 Feb 2001 09:21:38 -0700
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <xHSj6.324$8u5.1550@newsfeed.slurp.net>

Jim Eaton <eatonalive%qwestinternet.net at pop3.qwestinternet.net> wrote in
message news:3A8F81BC.108488CA at pop3.qwestinternet.net...
>
> Wlll going into Java first be any detriment to
> learning C later?  Any suggestions would be greatly appreciated.

Java will not be a detriment, but it won't be a great help either. A C
progammer is responsible for memory management and any data structure beyond
a struct. In Python and Java, lists, hashes, tuples and so forth are
available for use out of the box. In C, dynamically constructing a doubly
linked list, as well as iterating it, is a common task that devolves on the
programmer. It is also a task that is commonly screwed up, resulting in
SIGSEGV cores and a familiarity with the debugger of your choice.

This is not an argument that C is any better a language; in most cases the
best language is the one that offers the highest level of abstraction and
still can accomplish the task at hand.

For instance, though I'd quibble about SQL being a language, I certainly use
embedded SQL or a CLI interface to the database engine and wouldn't dream of
re-inventing the interface in C.






From eatonalive%qwestinternet.net at pop3.qwestinternet.net  Sun Feb 18 18:35:50 2001
From: eatonalive%qwestinternet.net at pop3.qwestinternet.net (Jim Eaton)
Date: 18 Feb 2001 17:35:50 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <3A9008A1.6E7DD27F@pop3.qwestinternet.net>

Thank you very much for your advice.  I was also wondering if learning Perl
sometime later would be needed.  I've heard people talk of Python as the
natural successor to Perl and I'd like to know what people feel about this.



From aleaxit at yahoo.com  Sun Feb 18 19:47:22 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 19:47:22 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net>
Message-ID: <96p5ip0314o@news1.newsguy.com>

"Jim Eaton" <eatonalive%qwestinternet.net at pop3.qwestinternet.net> wrote in
message news:3A9008A1.6E7DD27F at pop3.qwestinternet.net...
> Thank you very much for your advice.  I was also wondering if learning
Perl
> sometime later would be needed.  I've heard people talk of Python as the
> natural successor to Perl and I'd like to know what people feel about
this.

IMHO, you'll need to learn Perl if, and only if, you want to go work for
some place that requires it, and/or (roughly equivalent) you need to
maintain or translate an existing body of Perl code, to code for an
environment which only supplies a Perl interpreter, to use a C-coded
module that's only made available for Perl.

Roughly, again IMHO, the same goes for Cobol, Fortran or Javascript --
they don't have much to offer you 'intrinsically' once you know Python,
but some externalities might make it desirable to know one or more
of them -- externalities of the same kind as above described (wanting
to work for a shop that requires the language, to maintain or translate
existing legacy code, to use libraries or environments which will only
support one of these languages).

Other languages do have much to offer you, conceptually and/or
pragmatically.  With Java, C, or C++, you'll be able to extend Python
(Jython with Java, CPython with C or C++) to provide faster versions
of some algorithm or occasional deep system-specific interaction;
any of them will also teach you something about the worth and problems
of compile-time strong (but not _obsessively_ strong:-) typing.  SQL
will teach you to "think in sets" and non-procedurally, and also enable
you to use most effectively that crucial tool, relational databases.
Others may have weaker pragmatics (hard to get actual need for their
use), but good conceptual returns -- e.g., Haskell or ML variants for
functional programming, etc.


Alex





From johann at physics.berkeley.edu  Sun Feb 18 21:03:48 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 18 Feb 2001 12:03:48 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net>
Message-ID: <mtofvz69nf.fsf@astron.berkeley.edu>

Jim Eaton writes:

> Thank you very much for your advice.  I was also wondering if learning Perl
> sometime later would be needed.  I've heard people talk of Python as the
> natural successor to Perl and I'd like to know what people feel about this.

Heh.  I didn't even get a chance to get started.

Clearly, this depends on what you want to do with your new-found
programming power.  If you're interested in web-related work, then
Perl would be a good thing to know, simply because it's very common.
I don't do web-stuff, I do numerics, so I've never been able to make
myself learn it.  (Well, I know enough to use it as a "better awk",
but that's about it.)

If you want to become a better general programmer, Perl will not help
you at all, IMHO.  I don't really think that Java would help, either,
since I'm not sure what it would teach you.

C or C++ would at least help you understand how to manage your own
memory and perhaps give you a better sense of what goes on in the guts
of the machine.  Out of those, I'd recommend learning the basics of C
first, because I find C++ very confusing.  Now, that being said, my
main production code is a Python/SWIG/C++ system, so I like some parts
of C++, but it's an awful mess to learn.

This may be heresy in a c.l.python, but I really wish I'd written the
whole thing in Common Lisp to start with.  Too late now, but my next
project will be entirely CL.  I'm tired of the wall between the
scripting language and the speed language; CL can be both (I get
numeric performance as good as Fortran in Common Lisp using the free
CMUCL compiler), so I'd just use that.

(Python is great for things that I think I will never need that
to-the-metal performance in.  If I can stay in pure Python, I use
that.  I'll just never try to maintain a dual Python/C++ project
again.  Arg.)

One warning about this lisps; you might want to be careful with them.
I never like advocating them.  I absolutely *hated* Scheme when I
first had to deal with it in an AI class, but I've grown fonder of it
with time.  I had to do far too many things in a completely backwards
way; if you're just trying to learn new ways to do things, it's great,
but if you're trying to implement alpha-beta pruning in your checkers
game for the assignment due the next day, you'll come to hate it.
Common Lisp is a lot more practical.

Other languages.  Hm.  I've heard prolog is cool, but I've done all of
my constraint-based programming in lisp.  I looked at Haskel, but it
seemed over-pure.  OCaML is nice, but I never really got into it.
Sather is a great language, but it's dying.  It's like Eiffel, in that
it teaches good OO design, but it has really cool Icon-like iterators
(I would so love it if python got those some how), and it's
contravariant rather than covariant, so it doesn't have to do nearly
as many type-checks and is thereby much faster.

That's about it for my experience.  Sorry for the brain dump.  I just
started out by saying that learning Perl won't teach you anything that
you can't get from Python, so don't bother with it, unless you just
want to make yourself more marketable.  And I stand by that statement.

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From max at alcyone.com  Sun Feb 18 21:42:31 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 12:42:31 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu>
Message-ID: <3A9033B7.D5ADA989@alcyone.com>

Johann Hibschman wrote:

> C or C++ would at least help you understand how to manage your own
> memory and perhaps give you a better sense of what goes on in the guts
> of the machine.  Out of those, I'd recommend learning the basics of C
> first, because I find C++ very confusing.

Learning Standard C++ as Standard C++ as opposed to C with extra
features is a lot more straightforward.  For the new approaches to
learning C++ for its own sake, with emphasis on the modern features
which make it really powerful and surprisingly easy to use, check out
_Accelerated C++_ by Koenig and Moo.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I wonder if this grief will ever let me go
\__/ Sade
    The laws list / http://www.alcyone.com/max/physics/laws/
 Laws, rules, principles, effects, paradoxes, etc. in physics.


From sheila at spamcop.net  Sun Feb 18 22:29:43 2001
From: sheila at spamcop.net (Sheila King)
Date: Sun, 18 Feb 2001 21:29:43 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com>
Message-ID: <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com>

On Sun, 18 Feb 2001 12:42:31 -0800, Erik Max Francis <max at alcyone.com> wrote
in comp.lang.python in article <3A9033B7.D5ADA989 at alcyone.com>:

:Johann Hibschman wrote:
:
:> C or C++ would at least help you understand how to manage your own
:> memory and perhaps give you a better sense of what goes on in the guts
:> of the machine.  Out of those, I'd recommend learning the basics of C
:> first, because I find C++ very confusing.
:
:Learning Standard C++ as Standard C++ as opposed to C with extra
:features is a lot more straightforward.  For the new approaches to
:learning C++ for its own sake, with emphasis on the modern features
:which make it really powerful and surprisingly easy to use, check out
:_Accelerated C++_ by Koenig and Moo.

I must say, that I am shocked at the number of apparent *groans* over C++
language, in this thread.

I teach C++ as a first programming language to high school students, via the
AP Computer Science Curriculum. (The course I teach is supposed to be
equivalent to a first semester University course for CS majors.) I think that
the trick, is to use a carefully selected subset of the C++ language.

Still, I must say that I really enjoy the Python I've been teaching myself the
last few months. I do intend to do a number of projects in Python, eventually.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/




From max at alcyone.com  Sun Feb 18 22:59:53 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 13:59:53 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com>
Message-ID: <3A9045D9.3777F585@alcyone.com>

Sheila King wrote:

> I must say, that I am shocked at the number of apparent *groans* over
> C++
> language, in this thread.
> 
> I teach C++ as a first programming language to high school students,
> via the
> AP Computer Science Curriculum. (The course I teach is supposed to be
> equivalent to a first semester University course for CS majors.) I
> think that
> the trick, is to use a carefully selected subset of the C++ language.

Indeed.  Often C++ is taught strictly as an extension to C, and when
done that way you're essentially learning two languages:  C and Standard
C++.  When learning Standard C++, there's no reason to do it
as-an-extension-to-C.  For only one example, all the details of handling
and processing C strings with the str... functions (which is a big
stumbling block in C or C++-as-C for novice programmers) are totally
unnecessary in an introductory Standard C++ class, because you have
std::string to do all that work for you.  As another example, one
doesn't have to deal with difficult memory management issues when you
have std::vector and other such template classes.

> Still, I must say that I really enjoy the Python I've been teaching
> myself the
> last few months. I do intend to do a number of projects in Python,
> eventually.

Indeed; I use Python for many recreational and professional projects. 
For a major one that's public, my Web game Interstelen is being written
in Python:  http://www.interstelen.com/

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Isn't the best defense always a good attack?
\__/ Ovid
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!


From sheila at spamcop.net  Mon Feb 19 00:06:56 2001
From: sheila at spamcop.net (Sheila King)
Date: Sun, 18 Feb 2001 23:06:56 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <3A9045D9.3777F585@alcyone.com>
Message-ID: <99l09t024i0v1ro976ka7kkuniams6k0qd@4ax.com>

On Sun, 18 Feb 2001 13:59:53 -0800, Erik Max Francis <max at alcyone.com> wrote
in comp.lang.python in article <3A9045D9.3777F585 at alcyone.com>:

:Sheila King wrote:
:
:> I must say, that I am shocked at the number of apparent *groans* over
:> C++
:> language, in this thread.
:> 
:> I teach C++ as a first programming language to high school students,
:> via the
:> AP Computer Science Curriculum. (The course I teach is supposed to be
:> equivalent to a first semester University course for CS majors.) I
:> think that
:> the trick, is to use a carefully selected subset of the C++ language.
:
:Indeed.  Often C++ is taught strictly as an extension to C, and when
:done that way you're essentially learning two languages:  C and Standard
:C++.  When learning Standard C++, there's no reason to do it
:as-an-extension-to-C.  For only one example, all the details of handling
:and processing C strings with the str... functions (which is a big
:stumbling block in C or C++-as-C for novice programmers) are totally
:unnecessary in an introductory Standard C++ class, because you have
:std::string to do all that work for you.  As another example, one
:doesn't have to deal with difficult memory management issues when you
:have std::vector and other such template classes.

Yes, the AP course offers a sting class and a vector class (called apstring
and apvector) which are very similar to the classes in the standard library,
although with fewer features. I have used these classes myself, for programs I
had to write for my coursework at Cal Poly (so I could avoid messing with
C-style strings and arrays).

:> Still, I must say that I really enjoy the Python I've been teaching
:> myself the
:> last few months. I do intend to do a number of projects in Python,
:> eventually.
:
:Indeed; I use Python for many recreational and professional projects. 
:For a major one that's public, my Web game Interstelen is being written
:in Python:  http://www.interstelen.com/

Cool. Neat site. I will have to point it out to some friends.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From ken at lt.com  Mon Feb 19 00:28:00 2001
From: ken at lt.com (Kenneth Loafman)
Date: Sun, 18 Feb 2001 17:28:00 -0600
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com>
Message-ID: <3A905A80.1BC53DB7@lt.com>

Sheila King wrote:
> 
> I teach C++ as a first programming language to high school students, via the
> AP Computer Science Curriculum. (The course I teach is supposed to be
> equivalent to a first semester University course for CS majors.) I think that
> the trick, is to use a carefully selected subset of the C++ language.
> 
> Still, I must say that I really enjoy the Python I've been teaching myself the
> last few months. I do intend to do a number of projects in Python, eventually.

Please also teach the C Standard Library along with C++.  I've seen some
very poor examples of "reinventing the wheel" happen because the student
did not even know the basics of the C Standard Library and its
interaction with the basic elements of C/C++.  Consider one case I found
where the goal was to remove the last character of a line.  Instead of
using something like:

   if (strlen(s)) s[strlen(s)-1] = 0

the (7-year C++ veteran) wrote several lines of code that:

   reversed the string
   trimmed the first char
   re-reversed the string

Considering that he was using M$ C++ CString operations, the code took
several thousand instructions, instead of a just a few.

...Thanks,
...Kenneth



From max at alcyone.com  Mon Feb 19 00:49:51 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 15:49:51 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org>
Message-ID: <3A905F9F.C7D93AA8@alcyone.com>

Kenneth Loafman wrote:

> Please also teach the C Standard Library along with C++.  I've seen
> some
> very poor examples of "reinventing the wheel" happen because the
> student
> did not even know the basics of the C Standard Library and its
> interaction with the basic elements of C/C++.  Consider one case I
> found
> where the goal was to remove the last character of a line.  Instead of
> using something like:
> 
>    if (strlen(s)) s[strlen(s)-1] = 0

If one is actually writing Standard C++, one doesn't need the str... C
functions, since one won't be using NUL-terminated C strings, but rather
std::string.

> the (7-year C++ veteran) wrote several lines of code that:
> 
>    reversed the string
>    trimmed the first char
>    re-reversed the string

If he actually thought that was a viable solution, he wasn't a veteran
of _anything_.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ A man that studieth revenge keeps his own wounds green.
\__/ Francis Bacon
    Product's Quake III Arena Tips / http://www.bosskey.net/
 Tips and tricks from the absolute beginner to the Arena Master.


From qrczak at knm.org.pl  Tue Feb 20 20:33:19 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 20 Feb 2001 19:33:19 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <3A905F9F.C7D93AA8@alcyone.com>
Message-ID: <slrn995hjv.p74.qrczak@qrnik.zagroda>

Sun, 18 Feb 2001 15:49:51 -0800, Erik Max Francis <max at alcyone.com> pisze:

> If one is actually writing Standard C++, one doesn't need the
> str... C functions, since one won't be using NUL-terminated C
> strings, but rather std::string.

Unless he wants to open a file with a given name, or get command
line arguments. Not to mention string literals.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From aleaxit at yahoo.com  Wed Feb 21 00:20:10 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 00:20:10 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <3A905F9F.C7D93AA8@alcyone.com> <slrn995hjv.p74.qrczak@qrnik.zagroda>
Message-ID: <96uu2f0j60@news1.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn995hjv.p74.qrczak at qrnik.zagroda...
> Sun, 18 Feb 2001 15:49:51 -0800, Erik Max Francis <max at alcyone.com> pisze:
>
> > If one is actually writing Standard C++, one doesn't need the
> > str... C functions, since one won't be using NUL-terminated C
> > strings, but rather std::string.
>
> Unless he wants to open a file with a given name, or get command
> line arguments. Not to mention string literals.

These tasks will use C-strings, but Erik's assertions about not
needing "the str... C functions" remains true -- one just builds a
std::string out of every C-string one happens to meet, and, when
a C-string is needed (e.g., in a std::fstream constructor as the
name argument), the .c_str() method of the std::string object
supplies it.  No functionality at all needs strwhatever()...


Alex





From roy at panix.com  Mon Feb 19 01:02:14 2001
From: roy at panix.com (Roy Smith)
Date: Sun, 18 Feb 2001 19:02:14 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org>
Message-ID: <roy-8C4CC0.19021418022001@news.panix.com>

Kenneth Loafman <ken at lt.com> wrote:
> Consider one case I found
> where the goal was to remove the last character of a line.  Instead of
> using something like:
> 
>    if (strlen(s)) s[strlen(s)-1] = 0
> 
> the (7-year C++ veteran) wrote several lines of code that:
> 
>    reversed the string
>    trimmed the first char
>    re-reversed the string

How about:

   if (length = strlen(s))
      s[length-1] = 0;

and cut your instruction count in half :-)


From erno-news at erno.iki.fi  Mon Feb 19 02:18:47 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 19 Feb 2001 03:18:47 +0200
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <roy-8C4CC0.19021418022001@news.panix.com>
Message-ID: <ku4rxrtqq0.fsf@lasipalatsi.fi>

In article <roy-8C4CC0.19021418022001 at news.panix.com>, Roy Smith
<roy at panix.com> writes:

| Kenneth Loafman <ken at lt.com> wrote:
|| if (strlen(s)) s[strlen(s)-1] = 0

| How about:

|    if (length = strlen(s))
|       s[length-1] = 0;

| and cut your instruction count in half :-)

a sensible optimising compiler will generate identical code for both.

  -- erno


From roy at panix.com  Mon Feb 19 02:53:38 2001
From: roy at panix.com (Roy Smith)
Date: Sun, 18 Feb 2001 20:53:38 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <roy-8C4CC0.19021418022001@news.panix.com> <ku4rxrtqq0.fsf@lasipalatsi.fi>
Message-ID: <roy-770A13.20533818022001@news.panix.com>

Erno Kuusela <erno-news at erno.iki.fi> wrote:
> |    if (length = strlen(s))
> |       s[length-1] = 0;
> 
> | and cut your instruction count in half :-)
> 
> a sensible optimising compiler will generate identical code for both.

I will admit that my compiler theory is a bit rusty, but it's hard to see 
how a compiler would know that the strlen function would return the same 
value when called again with the same argument (idempotent?).

In any case, I just tried it on my Linux box (GNU compiler) and on a 
Solaris machine (Sun Workshop compiler).  Neither compiler generated the 
same code for the two versions, regardless of whether optimizing was turned 
on or not.  Both generated two strlen stanzas (see below), with or without 
-O.

The most interesting case was Linux with -O.  I don't know i86 assembler, 
so I'm guessing a bit here, but seems to have in-lined the strlen function, 
eliminating the subroutine call completely.  It still left what looks like 
two code fragments to count the length of the string.


From erno-news at erno.iki.fi  Mon Feb 19 06:02:50 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 19 Feb 2001 07:02:50 +0200
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <roy-8C4CC0.19021418022001@news.panix.com> <ku4rxrtqq0.fsf@lasipalatsi.fi> <roy-770A13.20533818022001@news.panix.com>
Message-ID: <kuzofjs1s5.fsf@lasipalatsi.fi>

In article <roy-770A13.20533818022001 at news.panix.com>, Roy Smith
<roy at panix.com> writes:

| Erno Kuusela <erno-news at erno.iki.fi> wrote:
|| |    if (length = strlen(s))
|| |       s[length-1] = 0;
|| 
|| | and cut your instruction count in half :-)
|| 
|| a sensible optimising compiler will generate identical code for both.

| I will admit that my compiler theory is a bit rusty, but it's hard to see 
| how a compiler would know that the strlen function would return the same 
| value when called again with the same argument (idempotent?).

it's part of the language, so the compiler is allowed to have
special knowledge about it.

| In any case, I just tried it on my Linux box (GNU compiler) and on a 
| Solaris machine (Sun Workshop compiler).  Neither compiler generated the 
| same code for the two versions, regardless of whether optimizing was turned 
| on or not.  Both generated two strlen stanzas (see below), with or without 
| -O.

i guess they are not sensible then ;)

at least the cc/libc combo on linux does have special knowledge about
many functions, including strlen, but i guess this case has not been
considered important enough to bother with.

look in /usr/include/bits/string.h on your linux box and
at <URL:http://puffin.external.hp.com/lxr/
source/gcc/gcc/builtins.c#L1306>...

  -- erno


From ken at lt.com  Mon Feb 19 14:41:40 2001
From: ken at lt.com (Kenneth Loafman)
Date: Mon, 19 Feb 2001 07:41:40 -0600
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <roy-8C4CC0.19021418022001@news.panix.com>
Message-ID: <3A912294.83F3D08D@lt.com>


Roy Smith wrote:
> 
> Kenneth Loafman <ken at lt.com> wrote:
> > Consider one case I found
> > where the goal was to remove the last character of a line.  Instead of
> > using something like:
> >
> >    if (strlen(s)) s[strlen(s)-1] = 0
> >
> > the (7-year C++ veteran) wrote several lines of code that:
> >
> >    reversed the string
> >    trimmed the first char
> >    re-reversed the string
> 
> How about:
> 
>    if (length = strlen(s))
>       s[length-1] = 0;
> 
> and cut your instruction count in half :-)

AArgh!  Been doing Python too long!  You can't assign in an if statement
in Python, so the awkward construct represented the worst of both.  In
retrospect, I'd probably write something like:

    if (s[0])
       s[strlen(s)-1] = 0

...Ken



From max at alcyone.com  Mon Feb 19 17:52:28 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 19 Feb 2001 08:52:28 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <roy-8C4CC0.19021418022001@news.panix.com> <mailman.982590218.10673.python-list@python.org>
Message-ID: <3A914F4C.E0CF4D1C@alcyone.com>

Kenneth Loafman wrote:

> AArgh!  Been doing Python too long!  You can't assign in an if
> statement
> in Python, so the awkward construct represented the worst of both.  In
> retrospect, I'd probably write something like:
> 
>     if (s[0])
>        s[strlen(s)-1] = 0

There's nothing wrong with using assignment in an if conditional,
provided that's what you actually meant.  Good compilers will warn, and
you can typically suppress the warning by adding another layer of
parentheses:

    if ((length = strlen(s)))
        ...

Making assignment illegal in expressions is to prevent _accidental_
cases like this.  If the language allows them and it's intentional,
there's nothing wrong.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Life is painting a picture, not doing a sum.
\__/ Oliver Wendell Holmes, Jr.
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From root at rainerdeyke.com  Mon Feb 19 02:41:36 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Mon, 19 Feb 2001 01:41:36 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org>
Message-ID: <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com>

"Kenneth Loafman" <ken at lt.com> wrote in message
news:mailman.982538961.27262.python-list at python.org...
> Please also teach the C Standard Library along with C++.  I've seen some
> very poor examples of "reinventing the wheel" happen because the student
> did not even know the basics of the C Standard Library and its
> interaction with the basic elements of C/C++.  Consider one case I found
> where the goal was to remove the last character of a line.  Instead of
> using something like:
>
>    if (strlen(s)) s[strlen(s)-1] = 0

Argh!

The following is the correct ways of doing this in C++:

s = std::string(s.begin(), s.end() - 1);

Not using 'std::string' in C++ is analogous to not using the built-in string
type in Python.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From ken at lt.com  Mon Feb 19 14:31:32 2001
From: ken at lt.com (Kenneth Loafman)
Date: Mon, 19 Feb 2001 07:31:32 -0600
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com>
Message-ID: <3A912034.BD25C502@lt.com>

Rainer Deyke wrote:
> 
> "Kenneth Loafman" <ken at lt.com> wrote in message
> news:mailman.982538961.27262.python-list at python.org...
> > Please also teach the C Standard Library along with C++.  I've seen some
> > very poor examples of "reinventing the wheel" happen because the student
> > did not even know the basics of the C Standard Library and its
> > interaction with the basic elements of C/C++.  Consider one case I found
> > where the goal was to remove the last character of a line.  Instead of
> > using something like:
> >
> >    if (strlen(s)) s[strlen(s)-1] = 0
> 
> Argh!
> 
> The following is the correct ways of doing this in C++:
> 
> s = std::string(s.begin(), s.end() - 1);
> 
> Not using 'std::string' in C++ is analogous to not using the built-in string
> type in Python.

They were using CString under M$VC.  I don't think std::string existst
on the M$ compilers.  Would make it too easy to port to the heathen UNIX
boxen <grin>.

...Ken



From aleaxit at yahoo.com  Mon Feb 19 17:10:33 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 19 Feb 2001 17:10:33 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org>
Message-ID: <96rgs602drn@news1.newsguy.com>

"Kenneth Loafman" <ken at lt.com> wrote in message
news:mailman.982589543.7737.python-list at python.org...
    [snip]
> > Not using 'std::string' in C++ is analogous to not using the built-in
string
> > type in Python.
>
> They were using CString under M$VC.  I don't think std::string existst
> on the M$ compilers.  Would make it too easy to port to the heathen UNIX
> boxen <grin>.

Microsoft Visual C++ implements std::string (and the rest of std::) since
release 5.0 (or earlier... I think I remember it being in some late 4.???).

MS bought the implementation provided by P.J.Plauger's company
("dinkumware",
I think I recall).  They don't distribute later bugfixes, that Plauger and
friends made after Microsoft released, but those can be downloaded at the
dinkumware site.

At some point in time, MS also supported (indirectly, through another
smaller company -- "Mainsoft", perhaps?) an MFC version that also made
it possible (if not "easy") to port MFC applications from Win/MSVC++
directly to Unix/proprietary C++ compilers of various Unix box suppliers
(HP, IBM, Dec, and Sun, at least; those were the ones we tried -- and
that's when I learned that Microsoft's C++ was *MUCH* cleaner, and far
closer to the standard, than _many_ other proprietary C++ compilers by
various other operating system vendors -- before then, the only C++
compiler I had used on Unix and Unix-like boxes was gcc, but that was
not supported by 'mainsoft' [if that's the right name]).


Alex






From max at alcyone.com  Mon Feb 19 18:16:33 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 19 Feb 2001 09:16:33 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <96rgs602drn@news1.newsguy.com>
Message-ID: <3A9154F1.4351E80D@alcyone.com>

Alex Martelli wrote:

> (HP, IBM, Dec, and Sun, at least; those were the ones we tried -- and
> that's when I learned that Microsoft's C++ was *MUCH* cleaner, and far
> closer to the standard, than _many_ other proprietary C++ compilers by
> various other operating system vendors -- before then, the only C++
> compiler I had used on Unix and Unix-like boxes was gcc, but that was
> not supported by 'mainsoft' [if that's the right name]).

Which is ironic, since that's the best, most Standard compliant C++
compiler for UNIX out there.  (It certainly has some problems, but
they're all have workarounds and are being addressed in gcc 3.0.)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Granted that I must die, how shall I live?
\__/ Michael Novak
    Alcyone Systems / http://www.alcyone.com/
 Alcyone Systems, San Jose, California.


From aleaxit at yahoo.com  Mon Feb 19 23:39:05 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 19 Feb 2001 23:39:05 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <96rgs602drn@news1.newsguy.com> <3A9154F1.4351E80D@alcyone.com>
Message-ID: <96s78k0a7a@news1.newsguy.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3A9154F1.4351E80D at alcyone.com...
> Alex Martelli wrote:
>
> > (HP, IBM, Dec, and Sun, at least; those were the ones we tried -- and
> > that's when I learned that Microsoft's C++ was *MUCH* cleaner, and far
> > closer to the standard, than _many_ other proprietary C++ compilers by
> > various other operating system vendors -- before then, the only C++
> > compiler I had used on Unix and Unix-like boxes was gcc, but that was
> > not supported by 'mainsoft' [if that's the right name]).
>
> Which is ironic, since that's the best, most Standard compliant C++
> compiler for UNIX out there.  (It certainly has some problems, but
> they're all have workarounds and are being addressed in gcc 3.0.)

I think KCC (KAI's compiler) is/was better (not sure about its status
now that Intel [?] has acquired it).  But mainsoft didn't support that
one, either.


Alex





From fredrik at pythonware.com  Tue Feb 20 01:07:39 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 20 Feb 2001 00:07:39 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <96rgs602drn@news1.newsguy.com> <3A9154F1.4351E80D@alcyone.com> <96s78k0a7a@news1.newsguy.com>
Message-ID: <fzik6.18639$AH6.2540196@newsc.telia.net>

Alex Martelli wrote:
> > Which is ironic, since that's the best, most Standard compliant C++
> > compiler for UNIX out there.  (It certainly has some problems, but
> > they're all have workarounds and are being addressed in gcc 3.0.)
>
> I think KCC (KAI's compiler) is/was better (not sure about its status
> now that Intel [?] has acquired it).

footnote: KCC is just one of many compilers based on the EDG frontend
(http://www.edg.com) -- which is definitely the best C++ frontend ever
written.

(had DEC switched to EDG a bit earlier, I would probably still use C++ ;-)

Cheers /F




From aleaxit at yahoo.com  Tue Feb 20 13:18:59 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 13:18:59 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <96rgs602drn@news1.newsguy.com> <3A9154F1.4351E80D@alcyone.com> <96s78k0a7a@news1.newsguy.com> <fzik6.18639$AH6.2540196@newsc.telia.net>
Message-ID: <96tnbh011gf@news2.newsguy.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:fzik6.18639$AH6.2540196 at newsc.telia.net...
> Alex Martelli wrote:
> > > Which is ironic, since that's the best, most Standard compliant C++
> > > compiler for UNIX out there.  (It certainly has some problems, but
> > > they're all have workarounds and are being addressed in gcc 3.0.)
> >
> > I think KCC (KAI's compiler) is/was better (not sure about its status
> > now that Intel [?] has acquired it).
>
> footnote: KCC is just one of many compilers based on the EDG frontend
> (http://www.edg.com) -- which is definitely the best C++ frontend ever
> written.

Good point.  OTOH, KAI's own traditional strengths are/were entirely
in code generation and optimization -- they are/were the guys whose
Fortran source-to-source optimizing converter completely busted some
parts of the first-generation "SPEC" benchmarks.  Given that (just
about) the only sensible motivation for using C++ (or C or Fortran)
rather than higher-level languages is *PERFORMANCE*, optimization
issues would seem to be pretty germane.

The cheapest way to try out EDG's front-end is no doubt through
www.comeau.com -- they sell their C++-to-C translator, built with
EDG's front-end, for as low as $50 (not sure how the price varies
by platform &c -- $50 is what it would cost for Windows, or Linux
with an Intel CPU); apparently it now also supports C99 (not sure
how complete a subset); plus, a really cool marketing idea, one
can visit http://www.comeaucomputing.com/tryitout/ and enter some
(small but complete) C++ snippet for compilation (no execution, of
course, but you'll be able to see any message the compiler emits).

(It *would* be nice to have a similar arrangement for _Python_ on
some site or other, by the way... although, here, "just compiling"
might not be much use:-).


Alex





From ken at lt.com  Mon Feb 19 19:47:26 2001
From: ken at lt.com (Kenneth Loafman)
Date: Mon, 19 Feb 2001 12:47:26 -0600
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <96rgs602drn@news1.newsguy.com>
Message-ID: <3A916A3E.6CF35FD2@lt.com>

Alex Martelli wrote:
> 
> At some point in time, MS also supported (indirectly, through another
> smaller company -- "Mainsoft", perhaps?) an MFC version that also made
> it possible (if not "easy") to port MFC applications from Win/MSVC++
> directly to Unix/proprietary C++ compilers of various Unix box suppliers
> (HP, IBM, Dec, and Sun, at least; those were the ones we tried -- and
> that's when I learned that Microsoft's C++ was *MUCH* cleaner, and far
> closer to the standard, than _many_ other proprietary C++ compilers by
> various other operating system vendors -- before then, the only C++
> compiler I had used on Unix and Unix-like boxes was gcc, but that was
> not supported by 'mainsoft' [if that's the right name]).

Thanks for the info.  I guess it was just hard to find info on std::, or
at least not obvious.  Never saw M$ examples that mentioned them.  All
examples were based on M$ classes, so I guess that's where I got that
impression.

The rule in my shop now is two part:
  1) throw away the vendors manuals
  2) use only ANSI standard constructs

This solves the problem (for the most part) of programming to some
construct that's unique to the vendor, and with a few additions, such as
named integral types instead of int, char, short, long, etc, you can
solve a lot of the general porting problem.

A lot of that goes away with Python.

...Ken



From root at rainerdeyke.com  Mon Feb 19 17:54:54 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Mon, 19 Feb 2001 16:54:54 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org>
Message-ID: <ydck6.262863$ge4.90996248@news2.rdc2.tx.home.com>

"Kenneth Loafman" <ken at lt.com> wrote in message
news:mailman.982589543.7737.python-list at python.org...
> Rainer Deyke wrote:
> > The following is the correct ways of doing this in C++:
> >
> > s = std::string(s.begin(), s.end() - 1);
> >
> > Not using 'std::string' in C++ is analogous to not using the built-in
string
> > type in Python.
>
> They were using CString under M$VC.  I don't think std::string existst
> on the M$ compilers.  Would make it too easy to port to the heathen UNIX
> boxen <grin>.

I've used 'std::string' under Visual C++.  I'm not familiar with 'CString',
but it would probably not be too hard to convert between real strings and
'CString'.  Something like this:

std::string s(zxczcCString.GetTheStupidCharacterArrayPointer());
s = std::string(s.begin(), s.end() - 1);
zxczcCString = CString(s.c_str());

This would obviously be bad for performance, but at least it minimizes
dependence on proprietary classes.  This is for when the API demands
'CString'.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From sheila at spamcop.net  Mon Feb 19 19:23:24 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 19 Feb 2001 18:23:24 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org>
Message-ID: <82p29t494h8evf1fvpjmr83abqv3enadja@4ax.com>

On Mon, 19 Feb 2001 07:31:32 -0600, Kenneth Loafman <ken at lt.com> wrote in
comp.lang.python in article <mailman.982589543.7737.python-list at python.org>:

:Rainer Deyke wrote:

:> The following is the correct ways of doing this in C++:
:> 
:> s = std::string(s.begin(), s.end() - 1);
:> 
:> Not using 'std::string' in C++ is analogous to not using the built-in string
:> type in Python.
:
:They were using CString under M$VC.  I don't think std::string existst
:on the M$ compilers.  Would make it too easy to port to the heathen UNIX
:boxen <grin>.

I just ran this code on my Win98 box, using MetroWerks CW 5:

#include <iostream>

using namespace std; //introduces namespace std

int main ( void )
{
	std::string s = "cat";
	cout << "Length of s is " << s.length() << endl;
	s = std::string(s.begin(), s.end() - 1);
	cout << "here comes s: \n";
	cout << s << endl;
	return 0;
}

Here is the output:

Length of s is 3
here comes s: 
ca


I think that the standard namespace is available on my Win98 machine.

Scary, huh?

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/





From max at alcyone.com  Tue Feb 20 05:23:28 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 19 Feb 2001 20:23:28 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <82p29t494h8evf1fvpjmr83abqv3enadja@4ax.com>
Message-ID: <3A91F140.27EF253B@alcyone.com>

Sheila King wrote:

> I just ran this code on my Win98 box, using MetroWerks CW 5:
> 
> #include <iostream>
> 
> using namespace std; //introduces namespace std

No, that _uses_ the namespace std.  After that statement, you can use
the identifiers in the std namespace _without_ the std:: prefix.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ We'll have to make our own luck from now on.
\__/ Louis Wu
    7 sisters productions / http://www.7sisters.com/
 Web design for the future.


From sheila at spamcop.net  Wed Feb 21 01:18:22 2001
From: sheila at spamcop.net (Sheila King)
Date: Wed, 21 Feb 2001 00:18:22 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <82p29t494h8evf1fvpjmr83abqv3enadja@4ax.com> <3A91F140.27EF253B@alcyone.com>
Message-ID: <19269t8upp4gpo4jpfga4uqtiogp4j1586@4ax.com>

On Mon, 19 Feb 2001 20:23:28 -0800, Erik Max Francis <max at alcyone.com> wrote
in comp.lang.python in article <3A91F140.27EF253B at alcyone.com>:

:Sheila King wrote:
:
:> I just ran this code on my Win98 box, using MetroWerks CW 5:
:> 
:> #include <iostream>
:> 
:> using namespace std; //introduces namespace std
:
:No, that _uses_ the namespace std.  After that statement, you can use
:the identifiers in the std namespace _without_ the std:: prefix.

Sorry. That comment is automatically inserted by my compiler when I start a
new project. (I could delete or modify it, of course, by I seldom even think
about it.)

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From sheila at spamcop.net  Mon Feb 19 03:10:52 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 19 Feb 2001 02:10:52 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org>
Message-ID: <etv09t4f6o9s4ov6oeo1rq9qldsoh3fdmq@4ax.com>

On Sun, 18 Feb 2001 17:28:00 -0600, Kenneth Loafman <ken at lt.com> wrote in
comp.lang.python in article <mailman.982538961.27262.python-list at python.org>:

:Sheila King wrote:
:> 
:> I teach C++ as a first programming language to high school students, via the
:> AP Computer Science Curriculum. (The course I teach is supposed to be
:> equivalent to a first semester University course for CS majors.) I think that
:> the trick, is to use a carefully selected subset of the C++ language.
:> 
:> Still, I must say that I really enjoy the Python I've been teaching myself the
:> last few months. I do intend to do a number of projects in Python, eventually.
:
:Please also teach the C Standard Library along with C++. 

Um, I don't have time?

: I've seen some
:very poor examples of "reinventing the wheel" happen because the student
:did not even know the basics of the C Standard Library and its
:interaction with the basic elements of C/C++.  Consider one case I found
:where the goal was to remove the last character of a line.  Instead of
:using something like:
:
:   if (strlen(s)) s[strlen(s)-1] = 0
:
:the (7-year C++ veteran) wrote several lines of code that:
:
:   reversed the string
:   trimmed the first char
:   re-reversed the string

My students would never do that. To remove the last character of a string, I
would hope they would do something like this:

s = s.substr(0, s.length()-1);

Or at least, I hope they would. The substr() and length() commands are
something that is included in the apstring class that we use, and I was
recently looking at the standard library, and I believe it also includes these
commands, or something very similar.

I teach them how to work with a variety of classes. And I do tell them about
the standard library. So, when they leave me class, they should be able to
easily learn new classes and utilize them.

:Considering that he was using M$ C++ CString operations, the code took
:several thousand instructions, instead of a just a few.

Hmm. Did you consider the possibility that this was just a function of the
individual involved, and not the fact that he'd been taught C++?

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From johann at physics.berkeley.edu  Mon Feb 19 03:28:53 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 18 Feb 2001 18:28:53 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com>
Message-ID: <mtelwvbe3e.fsf@astron.berkeley.edu>

Sheila King writes:

> On Sun, 18 Feb 2001 12:42:31 -0800, Erik Max Francis <max at alcyone.com> wrote
> in comp.lang.python in article <3A9033B7.D5ADA989 at alcyone.com>:

> I must say, that I am shocked at the number of apparent *groans* over C++
> language, in this thread.

Really?  Any language that encourages

vector<double>::iterator end = vec->end();
vector<double> out(vec->size);
for(vector<double>::const_iterator i = vec->begin(),
    vector<double>::iterator j = out->begin();
    i != end;
    ++i, ++j) {
  *j = (*i)*(*i);
}

as the right way to square a vector is, well, not quite the way I'd
want to do that.  Personally, I prefer something like

(map-into (make-vector (length vec))
          #'(lambda (x) (* x x))
          vec)

The whole STL thing seems incredibly verbose, and there's no way
that's really needed.  I wouldn't want to try to teach that to people.
I guess that's just my preference.


-- 
Johann Hibschman                           johann at physics.berkeley.edu


From max at alcyone.com  Mon Feb 19 04:22:25 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 19:22:25 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu>
Message-ID: <3A909171.C1298A01@alcyone.com>

Johann Hibschman wrote:

> Really?  Any language that encourages
> 
> vector<double>::iterator end = vec->end();
> vector<double> out(vec->size);
                              ^
Oops.

> for(vector<double>::const_iterator i = vec->begin(),
>     vector<double>::iterator j = out->begin();
>     i != end;
>     ++i, ++j) {
>   *j = (*i)*(*i);
> }
> 
> as the right way to square a vector is, well, not quite the way I'd
> want to do that.

You mean generate another vector whose elements contain the squares of
the first vector.  The usual meaning of "squaring a vector" would
involve vector reduction.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Dead men have no victory.
\__/ Euripides
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!


From bowman at montana.com  Mon Feb 19 06:30:22 2001
From: bowman at montana.com (bowman)
Date: Sun, 18 Feb 2001 22:30:22 -0700
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu>
Message-ID: <Xe2k6.3039$8u5.9979@newsfeed.slurp.net>

Johann Hibschman <johann at physics.berkeley.edu> wrote in message
news:mtelwvbe3e.fsf at astron.berkeley.edu...
>
> as the right way to square a vector is, well, not quite the way I'd
> want to do that.  Personally, I prefer something like
>
> (map-into (make-vector (length vec))
>           #'(lambda (x) (* x x))
>           vec)

that looks quite verbose too. my APL is incredibly rusty, but I think about
5 characters  would do the job. :)






From jpet at eskimo.com  Mon Feb 19 06:38:31 2001
From: jpet at eskimo.com (Jeff Petkau)
Date: Mon, 19 Feb 2001 05:38:31 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu>
Message-ID: <rj2k6.2617$%M1.414262@paloalto-snr1.gtei.net>

Johann Hibschman <johann at physics.berkeley.edu> wrote in message
news:mtelwvbe3e.fsf at astron.berkeley.edu...
> Sheila King writes:
> > I must say, that I am shocked at the number of apparent *groans* over
C++
> > language, in this thread.
>
> Really?  Any language that encourages
>
> vector<double>::iterator end = vec->end();
> vector<double> out(vec->size);
> for(vector<double>::const_iterator i = vec->begin(),
>     vector<double>::iterator j = out->begin();
>     i != end;
>     ++i, ++j) {
>   *j = (*i)*(*i);
> }
>
> as the right way to square a vector is, well, not quite the way I'd
> ...
> The whole STL thing seems incredibly verbose, and there's no way
> that's really needed.  I wouldn't want to try to teach that to people.
> I guess that's just my preference.

I believe the current fashion recommends using std:: explicitly. So that
should have been,

std::vector<double>::iterator end = vec->end();
std::vector<double> out(vec->size);
for(std::vector<double>::const_iterator i = vec->begin(),
    std::vector<double>::iterator j = out->begin();
    i != end;
    ++i, ++j) {
  *j = (*i)*(*i);
}

Nice and concise.  It would take twice as many lines in Python:

"spam spam spam spam spam spam spam spam spam spam"
out = [i*i for i in vec]
"""spam spam spam spam spam spam spam
spam spam spam spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam
spam spam Spam spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam SPAM SPAM spam spam
spam spam spam spam spam spam spam spam spam spam spam
spam spam out = map(lambda x: x*x,vec) spam spam
spam spam spam spam spam spam spam spam spam spam spam
spam spam spam spam spam spam spam spam spam spam"""

--Jeff Petkau (jpet at eskimo.com)





From max at alcyone.com  Mon Feb 19 07:19:20 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 22:19:20 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu> <rj2k6.2617$%M1.414262@paloalto-snr1.gtei.net>
Message-ID: <3A90BAE8.E09BB5FC@alcyone.com>

Jeff Petkau wrote:

> I believe the current fashion recommends using std:: explicitly.

It's required if the compiler supports namespaces as the Standard says
it should.
-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ You cannot step into the same river once.
\__/ Cratylus
    Alcyone Systems' Daily Planet / http://www.alcyone.com/planet.html
 A new, virtual planet, every day.


From sheila at spamcop.net  Mon Feb 19 07:41:16 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 19 Feb 2001 06:41:16 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu> <rj2k6.2617$%M1.414262@paloalto-snr1.gtei.net> <3A90BAE8.E09BB5FC@alcyone.com>
Message-ID: <nlf19tge339tjh0eoiigvsimndjm45b990@4ax.com>

On Sun, 18 Feb 2001 22:19:20 -0800, Erik Max Francis <max at alcyone.com> wrote
in comp.lang.python in article <3A90BAE8.E09BB5FC at alcyone.com>:

:Jeff Petkau wrote:
:
:> I believe the current fashion recommends using std:: explicitly.
:
:It's required if the compiler supports namespaces as the Standard says
:it should.

You know, I use a compiler that supports namespaces. I use CodeWarrior Pro 5,
from MetroWerks. And we just put, near the top of the program, and usually
after the #include directives:

using namespace std;

and I do not have to put things like

std::cout << "some statement\n";

I just put 

cout << "some statement\n";

It was explained to me, that by putting the statement

using namespace std;

in the program, I avoid the need to have to put the scope on each of these
elements? (I know this is very far off topic now, for this group...)

Here are further references on namespaces:

http://www.mvps.org/windev/cpp/nspaces.html
http://www.cplusplus.com/doc/tutorial/tut5-2.html
http://www.glenmccl.com/ns_comp.htm
http://www.cog.brown.edu/~rodrigo/c++mini-course/Namespaces.html


--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From spamers at must.die  Mon Feb 19 12:38:49 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Mon, 19 Feb 2001 12:38:49 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu> <rj2k6.2617$%M1.414262@paloalto-snr1.gtei.net> <3A90BAE8.E09BB5FC@alcyone.com> <nlf19tge339tjh0eoiigvsimndjm45b990@4ax.com>
Message-ID: <v5p19t0ph7o333decr2eqlrigbksph632r@4ax.com>

On Mon, 19 Feb 2001 06:41:16 GMT, Sheila King <sheila at spamcop.net> wrote:


>cout << "some statement\n";
>
>It was explained to me, that by putting the statement
>
>using namespace std;
>
>in the program, I avoid the need to have to put the scope on each of these
>elements? (I know this is very far off topic now, for this group...)

Yes, it's like the difference between
import string
string.join((blah, blah))

and

from string import *
join((blah, blah))

There, now we're back on topic! :-)

PS Just don't put the "using" thing in header files and you'll be fine.




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From bowman at montana.com  Mon Feb 19 15:47:42 2001
From: bowman at montana.com (bowman)
Date: Mon, 19 Feb 2001 07:47:42 -0700
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu> <rj2k6.2617$%M1.414262@paloalto-snr1.gtei.net> <3A90BAE8.E09BB5FC@alcyone.com> <nlf19tge339tjh0eoiigvsimndjm45b990@4ax.com>
Message-ID: <5xlk6.1189$va7.2463@newsfeed.slurp.net>

Sheila King <sheila at spamcop.net> wrote in message
news:nlf19tge339tjh0eoiigvsimndjm45b990 at 4ax.com...
> It was explained to me, that by putting the statement
>
> using namespace std;
>
> in the program, I avoid the need to have to put the scope on each of these
> elements? (I know this is very far off topic now, for this group...)

it is the same as

import foo
from foo import bar
from foo import *

in the last case, and in the case of 'using namespace std' , there is the
potential for some variable or function in the imported module to conflict
with a name you are using in your module.

howver, if you say foo.bar() or std::ends, you have precisely and explicitly
chosen the module. Not a big deal for casual programming but when the line
count gets into the tens of thousands or more, and a number of programmers
are developing modules independently, sooner or later there will be two
bar's, and which one too use will require further resolution. The worse case
will be a weak compiler that will assume the wrong one and emit a warning.
Sad but true, but in large projects involving a great deal of older code,
one tends to ignore a lot of warnings, as many are innocuous.







From sheila at spamcop.net  Mon Feb 19 06:54:19 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 19 Feb 2001 05:54:19 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu>
Message-ID: <llc19tkvjjf31m8ejqpe8ppm4640ku98kh@4ax.com>

On 18 Feb 2001 18:28:53 -0800, Johann Hibschman <johann at physics.berkeley.edu>
wrote in comp.lang.python in article <mtelwvbe3e.fsf at astron.berkeley.edu>:

:Sheila King writes:
:
:> On Sun, 18 Feb 2001 12:42:31 -0800, Erik Max Francis <max at alcyone.com> wrote
:> in comp.lang.python in article <3A9033B7.D5ADA989 at alcyone.com>:
:
:> I must say, that I am shocked at the number of apparent *groans* over C++
:> language, in this thread.
:
:Really?  Any language that encourages
:
:vector<double>::iterator end = vec->end();
:vector<double> out(vec->size);
:for(vector<double>::const_iterator i = vec->begin(),
:    vector<double>::iterator j = out->begin();
:    i != end;
:    ++i, ++j) {
:  *j = (*i)*(*i);
:}

I really don't understand that code, probably because I don't teach the STL. I
can see that you're creating some new vector of doubles, called out, the same
size as the vector vec. And for some reason, you're using two iterators to do
your for-loop, though I can't see why. We use only two templated classes in
the course I teach (remember, I'm only teaching a one semester course spread
out over a year), and we don't do them that way. The templates I'm familiar
with are the apvector, which I believe is quite similar to the STL vector
class, and the apmatrix, which is similar to the STL matrix class.

To take the contents of one vector, and square each entry and put it in
another vector, I would probably do this:

apvector<double> out( vec.length());
for ( int i= 0; i < vec.length(); i++)
    out[i]= vec[i]*vec[i]

I think that the code I list above is equivalent to yours?

:as the right way to square a vector is, well, not quite the way I'd
:want to do that.  Personally, I prefer something like
:
:(map-into (make-vector (length vec))
:          #'(lambda (x) (* x x))
:          vec)

My Python isn't that good, either. I haven't learned the lambda stuff, yet.
Still, I never was claiming any sort of comparison between Python and C++.
Matter of fact, I thought I said (in another message) that I really LIKE and
sort of prefer Python. The question was, what is a good second language to
learn. I don't see why C++ would be a bad choice, although I never argued in
favor of it, either.

:The whole STL thing seems incredibly verbose, and there's no way
:that's really needed.  I wouldn't want to try to teach that to people.
:I guess that's just my preference.

Well, I don't know about the STL in great detail. I've not been using nor
teaching it. I suppose, I should eventually learn it. Well, when the need
comes along. However, my spouse codes in C++ for a living (works for/worked
for Rockwell/JPL/Boeing) and he thinks your example code is too verbose, and
suggested this:

vector<double> in;
vector<double> out;
:
:
:
for (int i=0; i<in.size(); i++) 
 out[i]=in[i]*in[i];

instead, which is very similar to my code, which I suggest.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/




From johann at physics.berkeley.edu  Mon Feb 19 08:04:51 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 18 Feb 2001 23:04:51 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu> <llc19tkvjjf31m8ejqpe8ppm4640ku98kh@4ax.com>
Message-ID: <mtbsrzgnl8.fsf@astron.berkeley.edu>

First off, everyone, I'm sorry I opened my big fat mouth.
Now I'm just trying to run away gracefully.  :-)

Sheila King writes:

> :vector<double>::iterator end = vec->end();
> :vector<double> out(vec->size);
> :for(vector<double>::const_iterator i = vec->begin(),
> :    vector<double>::iterator j = out->begin();
> :    i != end;
> :    ++i, ++j) {
> :  *j = (*i)*(*i);
> :}

> I really don't understand that code, probably because I don't teach
> the STL. I can see that you're creating some new vector of doubles,
> called out, the same size as the vector vec.

Right.

> And for some reason, you're using two iterators to do your for-loop,
> though I can't see why.

Well, one's formally a constant iterator, and the other's not, so it
could be important if the vector vec were declared constant.
Likewise, it avoids having to call operator[] and doing the full
base+offset math, if we assume the compiler's not smart enough to
optimize the loop.


> To take the contents of one vector, and square each entry and put it in
> another vector, I would probably do this:

> apvector<double> out( vec.length());
> for ( int i= 0; i < vec.length(); i++)
>     out[i]= vec[i]*vec[i]

> I think that the code I list above is equivalent to yours?

Well, mine's a bit more efficient, since it doesn't have to maintain
the loop variable i, it doesn't have to call the function vec.length()
at each iteration, and it doesn't have to do the full operator[]
calculation.  It just steps along both of them.

But phooey on that!  It's bit-twiddling like that which drives me mad
while writing C++.  :-(


> My Python isn't that good, either. I haven't learned the lambda stuff, yet.
> Still, I never was claiming any sort of comparison between Python and C++.
> Matter of fact, I thought I said (in another message) that I really LIKE and
> sort of prefer Python. The question was, what is a good second language to
> learn. I don't see why C++ would be a bad choice, although I never argued in
> favor of it, either.

Again, I'm sorry, you just stumbled into a pet micro-managing
optimization peeve of mine, and I wasn't strong enough to resist
posting.


> vector<double> in;
> vector<double> out;
> :
> :
> :
> for (int i=0; i<in.size(); i++) 
>  out[i]=in[i]*in[i];

Yeah, that works.  I forget how the initialization checks for vector
work.  In addition, vector may even be checking each index i to see if
it's in range.  Etc.  In practice, most of my C++ code looks like
that, because the optimized way is just so ugly.

I'll be quiet now.

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From max at alcyone.com  Mon Feb 19 08:16:42 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 23:16:42 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu> <llc19tkvjjf31m8ejqpe8ppm4640ku98kh@4ax.com> <mtbsrzgnl8.fsf@astron.berkeley.edu>
Message-ID: <3A90C85A.ECF15D1C@alcyone.com>

Johann Hibschman wrote:

> Well, mine's a bit more efficient, since it doesn't have to maintain
> the loop variable i, it doesn't have to call the function vec.length()
> at each iteration, and it doesn't have to do the full operator[]
> calculation.  It just steps along both of them.
> 
> But phooey on that!  It's bit-twiddling like that which drives me mad
> while writing C++.  :-(

The real reason for going through all the trouble to write that all out
is now that you're using Standard C++ iterators, they will work with
_any_ Standard container.  Not just std::vector, but anyone at all.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ My life was better before I knew you.
\__/ Edith Wharton (to Morton Fullerton)
    REALpolitik / http://www.realpolitik.com/
 Get your own customized newsfeed online in realtime ... for free!


From sheila at spamcop.net  Mon Feb 19 09:32:39 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 19 Feb 2001 08:32:39 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mtelwvbe3e.fsf@astron.berkeley.edu> <llc19tkvjjf31m8ejqpe8ppm4640ku98kh@4ax.com> <mtbsrzgnl8.fsf@astron.berkeley.edu>
Message-ID: <egm19tgkc4gb344a4k3s0brr6gco8vt3ui@4ax.com>

On 18 Feb 2001 23:04:51 -0800, Johann Hibschman <johann at physics.berkeley.edu>
wrote in comp.lang.python in article <mtbsrzgnl8.fsf at astron.berkeley.edu>:

:Yeah, that works.  I forget how the initialization checks for vector
:work.  In addition, vector may even be checking each index i to see if
:it's in range.  Etc.  In practice, most of my C++ code looks like
:that, because the optimized way is just so ugly.

Yes, vector does check to see if the index in is range.

:I'll be quiet now.

Well, OK, then.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From tanzer at swing.co.at  Mon Feb 19 09:07:10 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Mon, 19 Feb 2001 09:07:10 +0100
Subject: What to do after Python? 
In-Reply-To: Your message of "Sun, 18 Feb 2001 21:29:43 GMT."
             <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> 
Message-ID: <m14UlLi-000wcEC@swing.co.at>

Sheila King <sheila at spamcop.net> wrote:

> I must say, that I am shocked at the number of apparent *groans* over C++
> language, in this thread.
> 
> I teach C++ as a first programming language to high school students, via the
> AP Computer Science Curriculum. (The course I teach is supposed to be
> equivalent to a first semester University course for CS majors.) I think that
> the trick, is to use a carefully selected subset of the C++ language.

I think you are missing at least two issues:

- The carefully selected subset isn't going to help your students much
  when they have to start reading/extending other people's code.

- In C++, there are so many low-level issues the programmer has to
  take care of that it distracts from the problem to be solved -- not
  very helpful for learning how to program.

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From sheila at spamcop.net  Mon Feb 19 09:37:00 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 19 Feb 2001 08:37:00 GMT
Subject: What to do after Python?
References: <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982570403.27694.python-list@python.org>
Message-ID: <ijm19tg60megg8pu25julaapa8cvdqvbjr@4ax.com>

On Mon, 19 Feb 2001 09:07:10 +0100, tanzer at swing.co.at (Christian Tanzer)
wrote in comp.lang.python in article
<mailman.982570403.27694.python-list at python.org>:

:
:Sheila King <sheila at spamcop.net> wrote:
:
:> I must say, that I am shocked at the number of apparent *groans* over C++
:> language, in this thread.
:> 
:> I teach C++ as a first programming language to high school students, via the
:> AP Computer Science Curriculum. (The course I teach is supposed to be
:> equivalent to a first semester University course for CS majors.) I think that
:> the trick, is to use a carefully selected subset of the C++ language.
:
:I think you are missing at least two issues:
:
:- The carefully selected subset isn't going to help your students much
:  when they have to start reading/extending other people's code.

Sheesh, I thought I made it clear, this was a BEGINNING, first course in
programming! They aren't going out and getting jobs the semester after my
course ends. (Actually, some of them have, but that's quite unusual.) They can
learn the rest, later, at a more appropriate point in their comp
sci/programming education.

:- In C++, there are so many low-level issues the programmer has to
:  take care of that it distracts from the problem to be solved -- not
:  very helpful for learning how to program.

I never suggested that C++ was the best choice for a first programming
language. In fact, I personally don't care for it, myself (as a first
language). However, the course that I am teaching requires that it be taught
in that language. For more information about the AP Computer Science program,
go here:
http://www.collegeboard.org/ap/computer-science/

I didn't write the course. I just teach it.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/




From dsh8290 at rit.edu  Wed Feb 21 05:16:03 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 20 Feb 2001 23:16:03 -0500
Subject: What to do after Python?
In-Reply-To: <ijm19tg60megg8pu25julaapa8cvdqvbjr@4ax.com>; from sheila@spamcop.net on Mon, Feb 19, 2001 at 08:37:00AM +0000
References: <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982570403.27694.python-list@python.org> <ijm19tg60megg8pu25julaapa8cvdqvbjr@4ax.com>
Message-ID: <20010220231603.C23613@harmony.cs.rit.edu>

On Mon, Feb 19, 2001 at 08:37:00AM +0000, Sheila King wrote:
| On Mon, 19 Feb 2001 09:07:10 +0100, tanzer at swing.co.at (Christian Tanzer)
| wrote in comp.lang.python in article
| <mailman.982570403.27694.python-list at python.org>:
| 
| :
| :Sheila King <sheila at spamcop.net> wrote:
| :
| :> I must say, that I am shocked at the number of apparent *groans* over C++
| :> language, in this thread.
| :> 
| :> I teach C++ as a first programming language to high school students, via the

What part of the country do you teach in?

| :> AP Computer Science Curriculum. (The course I teach is supposed to be
| :> equivalent to a first semester University course for CS majors.) I think that
| :> the trick, is to use a carefully selected subset of the C++ language.
| :
| :- In C++, there are so many low-level issues the programmer has to
| :  take care of that it distracts from the problem to be solved -- not
| :  very helpful for learning how to program.
| 
| I never suggested that C++ was the best choice for a first programming
| language. In fact, I personally don't care for it, myself (as a first

I agree with these, C++ is too low-level to beginning.  Something like
Eiffel, Java or Python (the last shall be first ...) would be better
suited, IMO.

| language). However, the course that I am teaching requires that it be taught
| in that language. For more information about the AP Computer Science program,
| go here:
| http://www.collegeboard.org/ap/computer-science/
| 
| I didn't write the course. I just teach it.

Someone ought to point out to the AP committee(s) the pros and cons of
the various languages, then they would decide that Python is better
;-).


My high school taught BASIC and Pascal on Apple IIe computers until
the year I graduated (I never found the time to take them anyways).
The year after I graduated (1999) they upgraded to Pentium computers
with Windows (95 or 98 I think) and are now offering (MSV) C++ and VB.
(Both with Intro- and Advanced- courses).  My brother and his friend
took the Intro to C++ course (20 weeks, half-year) last year.  I don't
think they should call it C++ because they barely got throuh if,
while, for, and functions and didn't mention classes at all.  I highly
doubt they used any of the C library stuff (like str functions).  They
didn't use new/delete or malloc/free either.  I think the only C++
specific feature they used was cout and cin.  Actually, I think it is
a bit unfair and misleading to tell the students they are learning C++
when they don't even mention 90% of the language to them.  It would be
really cool if they taught a nice, clean, small and very useful
language like Python instead, but ...

-D


PS.  this is/was a very interesting thread to read




From sheila at spamcop.net  Thu Feb 22 03:43:18 2001
From: sheila at spamcop.net (Sheila King)
Date: Thu, 22 Feb 2001 02:43:18 GMT
Subject: What to do after Python?
References: <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982570403.27694.python-list@python.org> <ijm19tg60megg8pu25julaapa8cvdqvbjr@4ax.com> <mailman.982765660.17876.python-list@python.org>
Message-ID: <2ts89tss0mrcoo6a6v54rl8liq6m7uhbpn@4ax.com>

On Tue, 20 Feb 2001 23:16:03 -0500, D-Man <dsh8290 at rit.edu> wrote in
comp.lang.python in article <mailman.982765660.17876.python-list at python.org>:

:| :Sheila King <sheila at spamcop.net> wrote:
 
:| :> I teach C++ as a first programming language to high school students, via the
:
:What part of the country do you teach in?

Orange County, California.

:| :> AP Computer Science Curriculum. (The course I teach is supposed to be
:| :> equivalent to a first semester University course for CS majors.) I think that
:| :> the trick, is to use a carefully selected subset of the C++ language.
:| :
:| :- In C++, there are so many low-level issues the programmer has to
:| :  take care of that it distracts from the problem to be solved -- not
:| :  very helpful for learning how to program.
:| 
:| I never suggested that C++ was the best choice for a first programming
:| language. In fact, I personally don't care for it, myself (as a first
:
:I agree with these, C++ is too low-level to beginning.  Something like
:Eiffel, Java or Python (the last shall be first ...) would be better
:suited, IMO.

Whatever. The original question in this thread wasn't even about a
recommendation for a first programming language. It was about a recommendation
for a second language, after having learned Python. I would assert, that after
learning Python, one (among many) appropriate choices for a second language,
would be C++ (as others have also recommended).

:| language). However, the course that I am teaching requires that it be taught
:| in that language. For more information about the AP Computer Science program,
:| go here:
:| http://www.collegeboard.org/ap/computer-science/
:| 
:| I didn't write the course. I just teach it.
:
:Someone ought to point out to the AP committee(s) the pros and cons of
:the various languages, then they would decide that Python is better
:;-).

I assume you put the winking smiley, because you know that's been done to
death? (I run one mailing list for APCS teachers, and subscribe to another,
and have read many of the language wars on the lists over the last few years.)
To see me suggest/ask about others opinions of Python, read this search result
from the archives of my mailing list:
http://www.escribe.com/education/apcsteachers/search.html?query=Python

See professors from Rice University suggest Scheme as a first programming
language, with this search result of my mailing list:
http://www.escribe.com/education/apcsteachers/search.html?query=Scheme&mt=&th=&c=30

The official College Board list, doesn't have public archives, but alternate
languages have been discussed even more on that list. Let me also point out,
that at many schools, the AP Computer Science course is not a first
programming course, but often a second or third programming course.

C++ remains the official language for the exams to be given in May 2001, 2002,
and 2003. Beginning in May 2004, until the foreseeable future, the programming
language will be Java.

:My brother and his friend
:took the Intro to C++ course (20 weeks, half-year) last year.  I don't
:think they should call it C++ because they barely got throuh if,
:while, for, and functions and didn't mention classes at all. 

I can't speak for what others schools do. A semester course in C++ for a first
programming course certainly doesn't sound like enough time to me.

Here is what I covered with my students before Xmas break:
Karel++ Chaps. 1-5
(see http://csis.pace.edu/~bergin/karel.html )
Introduces Object/class/inheritance. Loops, conditionals, defining new
classes. That was the first 1.5 months of my class. In the next two months I
covered (teaching the C++ language) all the basic types, conditionals, boolean
expressions, arithmetic expressions, all the different looping structures,
functions that return a value and void functions, nested loops, and a brief
intro to the apstring class and a brief intro to value vs. reference
parameters.

Before the end of the first semester, I also introduced some additional
classes from the Marine Biology Case Study that I have to teach.
http://www.collegeboard.org/ap/computer-science/html/case_study.html

Since the end of the first semester (Jan), we've covered file I/O, the
apvector class, structs, enumeration types and just started writing our own
classes yesterday. My students will learn all of the A topics listed on this
page:
http://www.thinkspot.net/materdei/apcompsci/APCS%20Topic%20Outline.html

before the end of April. (I have a few studying independently, who will also
learn the AB topics, as well.)

I think one point of the AP committee that should be mentioned, is that the
course is not advertised as a C++ course (it is not intended to teach the
language), but rather as a Computer SCIENCE course, thus the emphasis on
things like program design, sorting algorithms and the like.

: I highly
:doubt they used any of the C library stuff (like str functions).  They
:didn't use new/delete or malloc/free either.

I hardly think that memory management is an appropriate topic for a one
semester, first programming course. Also, malloc and free are not C++. They
are C. new/delete are only used with pointers??? And other dynamically
allocated types?

: Actually, I think it is
:a bit unfair and misleading to tell the students they are learning C++
:when they don't even mention 90% of the language to them.

Do you know for a fact, that that is what the students were told? Or did you
only hear the students' version of what they thought they were told? (These
are not always one and the same, says the 18-year veteran high school
teacher.) In any case, I cannot really have any meaningful comment on what
isolated schools do. I suppose that if they really told the kids they were
"teaching them C++" in such a way as to make the kids believe, that in one
semester, they would know the entire C++ language (something I find hard to
imagine), then that wasn't very nice.

:  It would be
:really cool if they taught a nice, clean, small and very useful
:language like Python instead, but ...

You will see me suggest some of the exact same things on the archives of my
mailing list. I don't recall anywhere suggesting that I recommended that C++
be the first programming language of choice. I merely said, that it could be
done. I stated that only in the context of justifying it as a reasonable
choice for a second language, for someone who had first learned Python.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From cg at gaia.intranet.cdegroot.com  Mon Feb 19 21:12:46 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 19 Feb 2001 21:12:46 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com>
Message-ID: <96runu$hvg$1@gaia.intranet.cdegroot.com>

Sheila King  <sheila at spamcop.net> said:
>I must say, that I am shocked at the number of apparent *groans* over C++
>language, in this thread.
>
*groan*. I learnt C++ by installing a, err, not-so-official AT&T cfront source
tape on a Xenix box and hacking away until the damn thing ran, and really did
my best to learn it. It's /not/ a real OO language. But it is sort of nice for
applying OO to low-level/system programming, and because it was meant to be
just that, it's OK. Just don't go writing apps in it (but the same holds
for C).

>I think that
>the trick, is to use a carefully selected subset of the C++ language.
>
Yup. And that might be workable in a classroom project, but forget it for
commercial development. Better start with a good language right away. For
teaching OO, I'd use Smalltalk. For teaching general programming, Python is a
nice choice because it is the most "neutral" language I know (neutral w.r.t.
to OO, procedural, functional). Although chances are I'd use Smalltalk
anyway :-)



-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From rcalco at cortechs.com  Sun Feb 18 21:51:00 2001
From: rcalco at cortechs.com (Bob Calco)
Date: Sun, 18 Feb 2001 15:51:00 -0500
Subject: What to do after Python?
In-Reply-To: <mtofvz69nf.fsf@astron.berkeley.edu>
Message-ID: <HBEJKOJGHBOIBHLHPMKGCELNCKAA.rcalco@cortechs.com>

# Other languages.  Hm.  I've heard prolog is cool, but I've done all of
# my constraint-based programming in lisp.  I looked at Haskel, but it
# seemed over-pure.  OCaML is nice, but I never really got into it.
# Sather is a great language, but it's dying.  It's like Eiffel, in that
# it teaches good OO design, but it has really cool Icon-like iterators
# (I would so love it if python got those some how), and it's
# contravariant rather than covariant, so it doesn't have to do nearly
# as many type-checks and is thereby much faster.

Check out Mozart, an implementation of the Oz language, at
www.mozart-oz.org. It's very fringe in the sense that not very many folks
know about it but its one of the few (IMHO) successful hybrid languages
(from an engineering perspective, not a commercial perspective -- it hasn't
had enough exposure yet to make it to the "big time"), with emphasis on
concurrent programming, constraints-based programming, mobile agents, and
the like; plus, it's cross platform like Java. Its power lies in being able
to code each piece of an application according to the paradigm that best
fits the solution (functional, object-oriented, declarative, etc.) all in
the same language. I find it very "natural" for all sorts of applications,
particulary expert systems, if your into that.

It's funny; I stumbled into it looking for classical music MP3's and
information about famous composers on the web, believe it or not!

Sincerely,

Bob Calco

# -----Original Message-----
# From: python-list-admin at python.org
# [mailto:python-list-admin at python.org]On Behalf Of Johann Hibschman
# Sent: Sunday, February 18, 2001 3:04 PM
# To: python-list at python.org
# Subject: Re: What to do after Python?
#
#
# Jim Eaton writes:
#
# > Thank you very much for your advice.  I was also wondering if
# learning Perl
# > sometime later would be needed.  I've heard people talk of Python as the
# > natural successor to Perl and I'd like to know what people feel
# about this.
#
# Heh.  I didn't even get a chance to get started.
#
# Clearly, this depends on what you want to do with your new-found
# programming power.  If you're interested in web-related work, then
# Perl would be a good thing to know, simply because it's very common.
# I don't do web-stuff, I do numerics, so I've never been able to make
# myself learn it.  (Well, I know enough to use it as a "better awk",
# but that's about it.)
#
# If you want to become a better general programmer, Perl will not help
# you at all, IMHO.  I don't really think that Java would help, either,
# since I'm not sure what it would teach you.
#
# C or C++ would at least help you understand how to manage your own
# memory and perhaps give you a better sense of what goes on in the guts
# of the machine.  Out of those, I'd recommend learning the basics of C
# first, because I find C++ very confusing.  Now, that being said, my
# main production code is a Python/SWIG/C++ system, so I like some parts
# of C++, but it's an awful mess to learn.
#
# This may be heresy in a c.l.python, but I really wish I'd written the
# whole thing in Common Lisp to start with.  Too late now, but my next
# project will be entirely CL.  I'm tired of the wall between the
# scripting language and the speed language; CL can be both (I get
# numeric performance as good as Fortran in Common Lisp using the free
# CMUCL compiler), so I'd just use that.
#
# (Python is great for things that I think I will never need that
# to-the-metal performance in.  If I can stay in pure Python, I use
# that.  I'll just never try to maintain a dual Python/C++ project
# again.  Arg.)
#
# One warning about this lisps; you might want to be careful with them.
# I never like advocating them.  I absolutely *hated* Scheme when I
# first had to deal with it in an AI class, but I've grown fonder of it
# with time.  I had to do far too many things in a completely backwards
# way; if you're just trying to learn new ways to do things, it's great,
# but if you're trying to implement alpha-beta pruning in your checkers
# game for the assignment due the next day, you'll come to hate it.
# Common Lisp is a lot more practical.
#
# Other languages.  Hm.  I've heard prolog is cool, but I've done all of
# my constraint-based programming in lisp.  I looked at Haskel, but it
# seemed over-pure.  OCaML is nice, but I never really got into it.
# Sather is a great language, but it's dying.  It's like Eiffel, in that
# it teaches good OO design, but it has really cool Icon-like iterators
# (I would so love it if python got those some how), and it's
# contravariant rather than covariant, so it doesn't have to do nearly
# as many type-checks and is thereby much faster.
#
# That's about it for my experience.  Sorry for the brain dump.  I just
# started out by saying that learning Perl won't teach you anything that
# you can't get from Python, so don't bother with it, unless you just
# want to make yourself more marketable.  And I stand by that statement.
#
# --Johann
#
# --
# Johann Hibschman                           johann at physics.berkeley.edu
# --
# http://mail.python.org/mailman/listinfo/python-list
#




From scarblac at pino.selwerd.nl  Mon Feb 19 01:29:19 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 19 Feb 2001 00:29:19 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net>
Message-ID: <slrn990q2j.4si.scarblac@pino.selwerd.nl>

Jim Eaton wrote in comp.lang.python:
> Thank you very much for your advice.  I was also wondering if learning Perl
> sometime later would be needed.  I've heard people talk of Python as the
> natural successor to Perl and I'd like to know what people feel about this.
 
They are really similar languages if you look at what they can do, but
Python tries to be small and clean and elegant, while Perl tries to be
powerful and flexible and huge. (the difference is perl doesn't try to be
small or clean, but python does also try to be powerful and flexible ;)).

When Python is your first language, learning Perl as a second one isn't very
logical since you're not really able to attack new problems - you just get
the option of using another language for things you could already solve
easily. (Generalizing here, but...)

Learning C, for instance, means you can now attack all those problems that
Python is too slow to handle, for instance by making modules in C that you
can call from Python. Java and C++ are important if you're looking for
world, *a lot* of real work application programming is done in those two.

Other languages like Haskell, Prolog, Lisp etc are a good for thing to
learn because they are programming in a completely different way. It
broadens the mind.

-- 
Remco Gerlich


From cg at gaia.intranet.cdegroot.com  Sun Feb 18 19:53:18 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 18 Feb 2001 19:53:18 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <96p5mu$3eq$1@gaia.intranet.cdegroot.com>

Jim Eaton  <eatonalive at qwestinternet.net> said:
>I am learning Python as a first language, and I have been wondering what
>is a good language to learn after Python?  I'm thinking of either going
>into C or Java but I'm not sure which one because I've heard many
>arguments either way.  Will going into Java first be any detriment to
>learning C later?  Any suggestions would be greatly appreciated.
>
You should learn at least a new programming language per year to keep
your mind fresh, so it doesn't really matter. 

Python is nice for a general view of programming: it does OO, procedural
and functional programming reasonably well. I'd advise C as a second
language because it forces you to deal with the computer behind all 
that code, which tends to give you a good perspective for other
language. Smalltalk is probably the best general language around if
you think OO is cool - watch out, lot of people never want anything
else after learning it. I'd sniff at Haskell and/or Scheme as well,
but you probably want to do quite a bit of programming and reading
before these languages start to make sense, and even then... 

If you want a job, decide whether you are pro or contra M$ and sign
up for a VB or Java course depending on the outcome. Needless to say
that these languages won't teach you a lot of new stuff compared to
the ones I mentioned above. You probably shouldn't take them as second
language because they are likely to teach you a lot of bad habits.



-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From roy at panix.com  Sun Feb 18 20:13:58 2001
From: roy at panix.com (Roy Smith)
Date: Sun, 18 Feb 2001 14:13:58 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com>
Message-ID: <roy-6DE44A.14135818022001@news.panix.com>

cg at cdegroot.com wrote:
> I'd advise C as a second language because it forces you to deal with the 
> computer behind all that code, which tends to give you a good 
> perspective for other language.

Actually, the problem with C is that it hides too much of the computer 
behind all that code.  I spent a few years writing C on alphas, and never 
even found out how many registers the danged thing had.  If you really want 
to find out what the hardware is all about, learn assembler, which is 
perilously close to handing a soldering iron to a software guy.

Another alternative would be to learn lisp.  It may not get you a better 
job, but will give you the ability to write a better python major mode for 
emacs :-)

I don't suppose I could interest you in fortran?


From cg at gaia.intranet.cdegroot.com  Sun Feb 18 21:17:25 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 18 Feb 2001 21:17:25 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com>
Message-ID: <96pakl$4fq$1@gaia.intranet.cdegroot.com>

Roy Smith  <roy at panix.com> said:
>Actually, the problem with C is that it hides too much of the computer 
>behind all that code.  I spent a few years writing C on alphas, and never 
>even found out how many registers the danged thing had. 

Err, lots? What I meant was that C forces you to think about memory
issues, pointers, and such close-to-hardware stuff without forcing you to
to actually dive into a particular CPU (ah, those days when "register"
had measurable effects...). Much closer, and you'll indeed do assembly
(or machine code - back in the 70's I didn't have an assembler for the
C64 so I did everything in machine code inside Basic DATA statements
;-)). Probably every programmer should learn assembly as well, but then,
C programming often drives you under the hood anyway (dirty work like
picking coredumps apart).

A fun way to learn assembler is to play core wars...

>I don't suppose I could interest you in fortran?

Well, I've been following this advise of learning a language-per-year myself
for the last decade or so, and I think some dusty corner of my brains has
Fortran knowledge (dig, dig - ah: there it is. I ported PDP-11 Fortran/77 to
MS-DOS in 1986 :-)).

But since my languager-per-year adagium let me execute "learn Smalltalk"
last year, I'm an ST convert, so sorry: no interest.

-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From cjc26 at nospam.cornell.edu  Wed Feb 21 03:56:54 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Wed, 21 Feb 2001 02:56:54 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com> <96pakl$4fq$1@gaia.intranet.cdegroot.com>
Message-ID: <W7Gk6.99345$v.8244887@typhoon.nyroc.rr.com>

* Cees de Groot <cg at gaia.intranet.cdegroot.com> menulis:
| 
| ...Much closer, and you'll indeed do assembly
| (or machine code - back in the 70's I didn't have an assembler for the
| C64

Yeah, that's because the C64 didn't even exist yet :)  It came out in
1983 IIRC.  The Commodore PET models were around in the 70s, though.


| so I did everything in machine code inside Basic DATA statements
| ;-)).

I can't remember how many hundreds of hours I must've spent typing in
program listings from Ahoy! magazine, only to have the damn thing not
run because I mistyped a number in one of the DATA statements.
Yeah, those were the days...:)

169-is-decimal-for-LDA-ly y'rs, Cliff


-- 
Cliff Crawford                 http://www.people.cornell.edu/pages/cjc26/
"huh?  Are you all there?  You missing some key parts up there?  Becoming  
a English major or an minor?   Do you need a annual report?  Are you slow
today, eh?  Did someone sever you parietal lobe?" - Mark


From amk at mira.erols.com  Wed Feb 21 04:11:10 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 21 Feb 2001 03:11:10 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com> <96pakl$4fq$1@gaia.intranet.cdegroot.com> <W7Gk6.99345$v.8244887@typhoon.nyroc.rr.com>
Message-ID: <slrn996d9h.123.amk@207-172-112-118.s118.tnt4.ann.va.dialup.rcn.com>

On Wed, 21 Feb 2001 02:56:54 GMT, 
	Cliff Crawford <cjc26 at nospam.cornell.edu> wrote:
>169-is-decimal-for-LDA-ly y'rs, Cliff

20 D2 FF was JSR $FFD2, which was the machine's printing routine.
Yup, been there, done that.  I really regret throwing out all my
magazines from that era; many of the games would be fun to implement
as Python scripts these days.

(Resolute CBM techies may be interested in the online Transactor
archive at http://www.csbruce.com/pub/transactor/ .  Transactor was a
Canadian CBM magazine that was ferociously technical.  Regret throwing
*those* out, too. <sigh>)

--amk
 


From sheila at spamcop.net  Sun Feb 18 21:31:06 2001
From: sheila at spamcop.net (Sheila King)
Date: Sun, 18 Feb 2001 20:31:06 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com>
Message-ID: <76c09t8omgga576dlsmnaapmqd8is3mjlt@4ax.com>

I realize that you were writing this, all tongue in cheek, but a newbie who
asked the original question, might not pick up on that. The next thing you
know, he will be asking, "What are registers?" and wondering why he should
know how many a machine has.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/


On Sun, 18 Feb 2001 14:13:58 -0500, Roy Smith <roy at panix.com> wrote in
comp.lang.python in article <roy-6DE44A.14135818022001 at news.panix.com>:

:Actually, the problem with C is that it hides too much of the computer 
:behind all that code.  I spent a few years writing C on alphas, and never 
:even found out how many registers the danged thing had.  If you really want 
:to find out what the hardware is all about, learn assembler, which is 
:perilously close to handing a soldering iron to a software guy.
:
:Another alternative would be to learn lisp.  It may not get you a better 
:job, but will give you the ability to write a better python major mode for 
:emacs :-)
:
:I don't suppose I could interest you in fortran?



From bowman at montana.com  Sun Feb 18 22:39:06 2001
From: bowman at montana.com (bowman)
Date: Sun, 18 Feb 2001 14:39:06 -0700
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com> <76c09t8omgga576dlsmnaapmqd8is3mjlt@4ax.com>
Message-ID: <dlXj6.2313$8u5.3054@newsfeed.slurp.net>

Sheila King <sheila at spamcop.net> wrote in message
news:76c09t8omgga576dlsmnaapmqd8is3mjlt at 4ax.com...
> I realize that you were writing this, all tongue in cheek, but a newbie
who
> asked the original question, might not pick up on that. The next thing you
> know, he will be asking, "What are registers?" and wondering why he should
> know how many a machine has.

I recently exposed a coworker to the joys of 8051 programming. It gets even
better when the newbie realizes not only does he have to be familiar with
the registers, he also needs to know each bit in every SFR does, as well as
knowing which SFRs are actually implemented in that particular chip.

Or even better, i threw together an AVR disassembler in Python last month --
one has to get intimate with what each bit in the word means.






From scarblac at pino.selwerd.nl  Mon Feb 19 10:46:45 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 19 Feb 2001 09:46:45 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com> <76c09t8omgga576dlsmnaapmqd8is3mjlt@4ax.com>
Message-ID: <slrn991qno.4si.scarblac@pino.selwerd.nl>

Sheila King <sheila at spamcop.net> wrote in comp.lang.python:
> I realize that you were writing this, all tongue in cheek, but a newbie who
> asked the original question, might not pick up on that. The next thing you
> know, he will be asking, "What are registers?" and wondering why he should
> know how many a machine has.

Heck, next thing he might even learn something! I tell you, these assembler
guys are dangerous.

-- 
Remco Gerlich


From cobrien at Radix.Net  Mon Feb 19 00:30:45 2001
From: cobrien at Radix.Net (Cary O'Brien)
Date: 18 Feb 2001 18:30:45 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com>
Message-ID: <96plv5$gqg$1@saltmine.radix.net>

In article <roy-6DE44A.14135818022001 at news.panix.com>,
Roy Smith  <roy at panix.com> wrote:
>cg at cdegroot.com wrote:
>> I'd advise C as a second language because it forces you to deal with the 
>> computer behind all that code, which tends to give you a good 
>> perspective for other language.
>
>Actually, the problem with C is that it hides too much of the computer 
>behind all that code.  I spent a few years writing C on alphas, and never 
>even found out how many registers the danged thing had.  If you really want 
>to find out what the hardware is all about, learn assembler, which is 
>perilously close to handing a soldering iron to a software guy.
>

Your problem is that you C compiler was too good.  I learned C and
68000 assembly at the same time with a stupid C compiler.  Stupid
meaning after a while we knew exactly what the C compiler would do.
Often (in the days of vt220 terminals) two of us would work together
debugging, one single-stepping through assembly, and one reading the C
code and keeping in sync.  If you said register you meant register.
Plus if you did you loops just right you could get the 68010 cpu to go
into wicked-fast-loop-mode.

>Another alternative would be to learn lisp.  It may not get you a better 
>job, but will give you the ability to write a better python major mode for 
>emacs :-)
>

Prolog is kinda different.  Plus I like the SQL suggestion.  Go buy
Bruce's PostgreSQL book and go to town.

-- cary

>I don't suppose I could interest you in fortran?










From roy at panix.com  Mon Feb 19 01:04:24 2001
From: roy at panix.com (Roy Smith)
Date: Sun, 18 Feb 2001 19:04:24 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com> <96plv5$gqg$1@saltmine.radix.net>
Message-ID: <roy-4D734B.19042418022001@news.panix.com>

cobrien at Radix.Net (Cary O'Brien) wrote:
> Your problem is that you C compiler was too good.  I learned C and
> 68000 assembly at the same time with a stupid C compiler.

Actually, I learned C and pdp-11 assembler at the same time :-)


From garry at sage.att.com  Tue Feb 20 17:15:16 2001
From: garry at sage.att.com (Garry Hodgson)
Date: Tue, 20 Feb 2001 16:15:16 GMT
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com> <96plv5$gqg$1@saltmine.radix.net> <roy-4D734B.19042418022001@news.panix.com>
Message-ID: <3A929814.825AA36D@sage.att.com>

Roy Smith wrote:
> 
> cobrien at Radix.Net (Cary O'Brien) wrote:
> > Your problem is that you C compiler was too good.  I learned C and
> > 68000 assembly at the same time with a stupid C compiler.
> 
> Actually, I learned C and pdp-11 assembler at the same time :-)

i used to bristle when a colleague of mine insisted that
c *was* assembly language.  years later, i realized he was 
correct.  c++ is a rather powerful macro-assembler.


-- 
Garry Hodgson                   Once in a while
Senior Hacker                   you can get shown the light
Software Innovation Services    in the strangest of places
AT&T Labs                       if you look at it right


From aahz at panix.com  Tue Feb 20 18:20:32 2001
From: aahz at panix.com (Aahz Maruch)
Date: 20 Feb 2001 09:20:32 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96plv5$gqg$1@saltmine.radix.net> <roy-4D734B.19042418022001@news.panix.com> <3A929814.825AA36D@sage.att.com>
Message-ID: <96u910$ld5$1@panix6.panix.com>

In article <3A929814.825AA36D at sage.att.com>,
Garry Hodgson  <garry at sage.att.com> wrote:
>Roy Smith wrote:
>> cobrien at Radix.Net (Cary O'Brien) wrote:
>>>
>>> Your problem is that you C compiler was too good.  I learned C and
>>> 68000 assembly at the same time with a stupid C compiler.
>> 
>> Actually, I learned C and pdp-11 assembler at the same time :-)
>
>i used to bristle when a colleague of mine insisted that
>c *was* assembly language.  years later, i realized he was 
>correct.  c++ is a rather powerful macro-assembler.

More precisely, C is a cross-platform assembley language.  It's not
perfectly cross-platform, by any means, but it's pretty damn close.
Consider Python, for example....
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From embed at geocities.com  Mon Feb 19 22:58:31 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 19 Feb 2001 16:58:31 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <2Hgk6.109$TW.1264@tor-nn1.netcom.ca>

"Jim Eaton" <eatonalive%qwestinternet.net at pop3.qwestinternet.net> wrote in
message news:3A8F81BC.108488CA at pop3.qwestinternet.net...
> I am learning Python as a first language, and I have been wondering what
> is a good language to learn after Python?

Delphi (aka ObjectPascal).

Compared to Delphi and Python, everything else just plain sucks.   :-)

Opinionately-yrs,

Warren




From dsh8290 at rit.edu  Wed Feb 21 05:05:06 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 20 Feb 2001 23:05:06 -0500
Subject: What to do after Python?
In-Reply-To: <3A8F81BC.108488CA@pop3.qwestinternet.net>; from eatonalive%qwestinternet.net@pop3.qwestinternet.net on Sun, Feb 18, 2001 at 08:00:11AM +0000
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <20010220230506.B23613@harmony.cs.rit.edu>

On Sun, Feb 18, 2001 at 08:00:11AM +0000, Jim Eaton wrote:
| I am learning Python as a first language, and I have been wondering what
| is a good language to learn after Python?  I'm thinking of either going
| into C or Java but I'm not sure which one because I've heard many
| arguments either way.  Will going into Java first be any detriment to
| learning C later?  Any suggestions would be greatly appreciated.

How do you use python?  As an OO language (class) or a procedural
language (def without class)?  C is not naturally OO, Java forces
everything to be inside a class.  Java will ease you into curly
braces, semicolons, and static type-checking while still maintaining
automatic memory management.  It is possible to write OO code in C
(see GTK+/GNOME for an example), but it takes a lot more work.  C is
also less forgiving allowing you to stomp on memory (and leak it)
however it pleases you.  Another thing to consider, Java will print a
traceback if an exception goes uncaught.  C will simply dump a core
file if you do something _utterly_ bad (SEGV, Bus Error,
Divide-by-zero).  If you simply overrun your array bounds, C will
SEGV/Bus Error only if you happen to be lucky.  More likely than not,
it will happily overwrite whatever happened to be there.

C will be more useful to you if you are interested in systems level
programming.  Both C and Java can be used to extend python (CPython
and Jython respectively).  Java is rather similar to C++ except it has
garbage collection, array bounds checking, automatic pointer
dereferencing, and no standalone functions (like C).  If you want to
learn C++, I would recommend learning Java first since it is more
forgiving, and informative when you do something bad (at runtime).

Other people have mentioned other languages that are interesting and
useful to better understand language design, etc, but I haven't seen
used in industry much.

If you want a pure OO language with a different approach to access
modifiers and generics, try Eiffel.  I think those are 2 things Eiffel
did very well (and C++/Java do very poorly, it is irrelevant in
Python).

HTH,
-D




From cborn at swbell.net  Sun Feb 18 09:10:33 2001
From: cborn at swbell.net (cborn at swbell.net)
Date: Sun, 18 Feb 2001 08:10:33 -0000
Subject: Using win32Com with Internet Explorer
Message-ID: <96o01p+ahvv@eGroups.com>

I am fairly new to Python and Com programming so hopefully
this will be simple.

Here is a script I have written which attempts to extract
links from a webpage. I have used makepy to generate Python
IE support object (Msie ActiveX control Module 1.0).



import win32com.client

# Get the browser object
ie = win32com.client.Dispatch("InternetExplorer.Application")
# open a page
ie.Navigate("http://www.python.org")

i = 0
# Get links from IE's Document Object
for x in ie.Document.links:
        #Print links
 	print ie.Document.links(i)
 	i = i +1



When I run the script in PythonWin this is what I get:



Traceback (most recent call last):
  File "C:\Python20\Pythonwin\pywin\framework\scriptutils.py", line 
301, in RunScript
    exec codeObject in __main__.__dict__
  File "C:\Python20\CB\test.py", line 12, in ?
    for x in ie.Document.links:
  File "C:\Python20\win32com\client\__init__.py", line 340, in 
__getattr__
    return apply(self._ApplyTypes_, args)
  File "C:\Python20\win32com\client\__init__.py", line 334, in 
_ApplyTypes_
    return self._get_good_object_(apply(self._oleobj_.InvokeTypes, 
(dispid, 0, wFlags, retType, argTypes) + args), user, resultCLSID)
com_error: (-2147352567, 'Exception occurred.', (0, None, None, None, 
0, -2147467259), None)


However, when I enter the script interactively 1 line at a time it 
works. I am sure there is some small detail that I am 
missing. Can anyone help me?

Thanks

Charles Bornstein
cborn at swbell.net




From MarkH at ActiveState.com  Sun Feb 18 23:52:06 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 18 Feb 2001 22:52:06 GMT
Subject: Using win32Com with Internet Explorer
References: <mailman.982483879.3236.python-list@python.org>
Message-ID: <3A90513C.3090809@ActiveState.com>

cborn at swbell.net wrote:

> I am fairly new to Python and Com programming so hopefully
> this will be simple.
> 
> Here is a script I have written which attempts to extract
> links from a webpage. I have used makepy to generate Python
> IE support object (Msie ActiveX control Module 1.0).

...

> However, when I enter the script interactively 1 line at a time it 
> works. I am sure there is some small detail that I am 
> missing. Can anyone help me?

Probably you need to wait for IE to load the page.  There was an example of that posted a few days ago here that used events - basically you wait until IE tells you the page is loaded.

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From fuess at att.net  Mon Feb 19 16:51:27 2001
From: fuess at att.net (David Fuess)
Date: Mon, 19 Feb 2001 15:51:27 GMT
Subject: Using win32Com with Internet Explorer
References: <mailman.982483879.3236.python-list@python.org> <3A90513C.3090809@ActiveState.com>
Message-ID: <42g29ts8rmaepg5gfqovvi07mo468mfb50@4ax.com>

Pardon my directness, but isn't it rather foolish to go messing around
with IE and COM when Python provides libraries for this purpose. I've
implemented many crawlers using urllib, httplib, and htmllib to fetch,
parse, and follow URL threads.

Dave

On Sun, 18 Feb 2001 22:52:06 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>cborn at swbell.net wrote:
>
>> I am fairly new to Python and Com programming so hopefully
>> this will be simple.
>> 
>> Here is a script I have written which attempts to extract
>> links from a webpage. I have used makepy to generate Python
>> IE support object (Msie ActiveX control Module 1.0).
>
>...
>
>> However, when I enter the script interactively 1 line at a time it 
>> works. I am sure there is some small detail that I am 
>> missing. Can anyone help me?
>
>Probably you need to wait for IE to load the page.  There was an example of that posted a few days ago here that used events - basically you wait until IE tells you the page is loaded.
>
>Mark.



From mbel44 at dial.pipex.net  Mon Feb 19 19:31:05 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Mon, 19 Feb 2001 18:31:05 +0000
Subject: Using win32Com with Internet Explorer
References: <mailman.982483879.3236.python-list@python.org> <3A90513C.3090809@ActiveState.com> <42g29ts8rmaepg5gfqovvi07mo468mfb50@4ax.com>
Message-ID: <5hp29toe6qi2sijbofa25g5e883bmr8983@4ax.com>

David Fuess <fuess at att.net> wrote:

>Pardon my directness, but isn't it rather foolish to go messing around
>with IE and COM when Python provides libraries for this purpose. I've
>implemented many crawlers using urllib, httplib, and htmllib to fetch,
>parse, and follow URL threads.

For many cases the IE-based code is better. For example, it uses IE's
proxy configuration and local cache.

If you need to redistribute a script to someone else, then a working
IE configuration is probably something you can rely on.

Toby Dickenson
tdickenson at geminidataloggers.com


From fuess at att.net  Wed Feb 21 05:25:39 2001
From: fuess at att.net (David Fuess)
Date: Wed, 21 Feb 2001 04:25:39 GMT
Subject: Using win32Com with Internet Explorer
References: <mailman.982483879.3236.python-list@python.org> <3A90513C.3090809@ActiveState.com> <42g29ts8rmaepg5gfqovvi07mo468mfb50@4ax.com> <5hp29toe6qi2sijbofa25g5e883bmr8983@4ax.com>
Message-ID: <fag69t4bd3osevq7iidah1mrehn8jlgjid@4ax.com>

Now trhat I've had a day to think about it, I must respectfully
disagree. I think the one, and only, thing you can count on is the
Pyrhon libraries. The urllib has a FancyOpener that understands
proxies and is simple and stable. IE is neither simple nor stable
particularly as you never know what version will be installed on the
target machine. Further, httplib and htmllib give you complete control
over the decoding of the web pages.

I use the same code on Windows,  UNIX, and Macintosh systems. Neither
UNIX nor the Mac have IE, yet all were capable of retrieving and
decoding web pages with no changes in the code. If you write in COM
you are restricted to Windows platforms only.

Dave

On Mon, 19 Feb 2001 18:31:05 +0000, Toby Dickenson
<mbel44 at dial.pipex.net> wrote:

>David Fuess <fuess at att.net> wrote:
>
>>Pardon my directness, but isn't it rather foolish to go messing around
>>with IE and COM when Python provides libraries for this purpose. I've
>>implemented many crawlers using urllib, httplib, and htmllib to fetch,
>>parse, and follow URL threads.
>
>For many cases the IE-based code is better. For example, it uses IE's
>proxy configuration and local cache.
>
>If you need to redistribute a script to someone else, then a working
>IE configuration is probably something you can rely on.
>
>Toby Dickenson
>tdickenson at geminidataloggers.com



From fredrik at pythonware.com  Wed Feb 21 07:54:26 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 06:54:26 GMT
Subject: Using win32Com with Internet Explorer
References: <mailman.982483879.3236.python-list@python.org> <3A90513C.3090809@ActiveState.com> <42g29ts8rmaepg5gfqovvi07mo468mfb50@4ax.com> <5hp29toe6qi2sijbofa25g5e883bmr8983@4ax.com>
Message-ID: <CCJk6.14722$Qb7.2357893@newsb.telia.net>

Toby Dickenson wrote:
> For many cases the IE-based code is better. For example, it uses IE's
> proxy configuration

so does Python

Cheers /F




From mbel44 at dial.pipex.net  Wed Feb 21 14:09:43 2001
From: mbel44 at dial.pipex.net (Toby Dickenson)
Date: Wed, 21 Feb 2001 13:09:43 +0000
Subject: Using win32Com with Internet Explorer
References: <mailman.982483879.3236.python-list@python.org> <3A90513C.3090809@ActiveState.com> <42g29ts8rmaepg5gfqovvi07mo468mfb50@4ax.com> <5hp29toe6qi2sijbofa25g5e883bmr8983@4ax.com> <CCJk6.14722$Qb7.2357893@newsb.telia.net>
Message-ID: <c0f79tk6u54aoquo9t7oog6vtc5c4oacqt@4ax.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote:

>Toby Dickenson wrote:
>> For many cases the IE-based code is better. For example, it uses IE's
>> proxy configuration
>
>so does Python

Yes, it does make a good effort (thanks Mark!)

However, many proxy configurations are implemented using a javascript
function that computes the proxy configuration. 


On Wed, 21 Feb 2001 04:25:39 GMT, David Fuess <fuess at att.net> wrote:

>I use the same code on Windows,  UNIX, and Macintosh systems. Neither
>UNIX nor the Mac have IE, yet all were capable of retrieving and
>decoding web pages with no changes in the code.

Im pretty sure it wont work on my current machine, without manual
proxy configuration.

>If you write in COM
>you are restricted to Windows platforms only.

Yes, it depends on your requirements. I didnt intent to suggest that
IE/COM is always superior to httplib.

My current http client project uses COM/IE if it can, but falls back
to httplib. Both modes of operation have been necessary at some time.



From fredrik at pythonware.com  Wed Feb 21 19:06:32 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 18:06:32 GMT
Subject: Using win32Com with Internet Explorer
References: <mailman.982483879.3236.python-list@python.org> <3A90513C.3090809@ActiveState.com> <42g29ts8rmaepg5gfqovvi07mo468mfb50@4ax.com> <5hp29toe6qi2sijbofa25g5e883bmr8983@4ax.com> <CCJk6.14722$Qb7.2357893@newsb.telia.net> <pbd79t0tm5psdg714cv4mifko9k95f6uck@4ax.com>
Message-ID: <IsTk6.14833$Qb7.2369744@newsb.telia.net>

Wolfgang Strobl wrote:
> >so does Python
>
> Huh? Python uses IE's proxy configuration? How?

by looking in the registry:

    HKEY_CURRENT_USER\Software\Microsoft\Windows\
        CurrentVersion\Internet Settings\ProxyEnable, ProxyServer

Cheers /F




From 1126-810 at online.de  Sun Feb 18 10:18:08 2001
From: 1126-810 at online.de (Stefan Witzgall)
Date: Sun, 18 Feb 2001 10:18:08 +0100
Subject: play_animals and MacPython
Message-ID: <1126-810-F3D9B6.10180818022001@news.online.de>

Hello,

I tried to run the sample program "play_animals" found in "Python for 
the unexperienced". Now I get an error "unslicecable object str"...
Ok, str is None, therefore I can't get some part of it.

Has anyone run this on his machine with python, is this a MacPython 
specific prob, how to get around?

Thank you for any hints.

Bye, Stefan


From c941520 at alinga.newcastle.edu.au  Sun Feb 18 11:30:22 2001
From: c941520 at alinga.newcastle.edu.au (Ben de Luca)
Date: Sun, 18 Feb 2001 21:30:22 +1100
Subject: file access
Message-ID: <3a8e53e2$0$25475$7f31c96c@news01.syd.optusnet.com.au>

i am building a database in python its quite small only a few hundred
entries? i gues dictionaries are the best way to store it in memory

when i write the data to a file whats the best way? I guess there is no way
to sumb the dictonaries strait to disk? i will have to parse it?




From fredrik at pythonware.com  Sat Feb 17 12:00:55 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 17 Feb 2001 11:00:55 GMT
Subject: file access
References: <3a8e53e2$0$25475$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <HRsj6.17589$AH6.2422794@newsc.telia.net>

Ben de Luca wrote:
> when i write the data to a file whats the best way? I guess there is no way
> to sumb the dictonaries strait to disk? i will have to parse it?

http://www.python.org/doc/current/lib/module-pickle.html
http://www.python.org/doc/current/lib/module-marshal.html

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From ruediger.maehl at dzsh.de  Tue Feb 20 14:32:02 2001
From: ruediger.maehl at dzsh.de (Ruediger Maehl)
Date: Tue, 20 Feb 2001 14:32:02 +0100
Subject: file access
References: <3a8e53e2$0$25475$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <newscache$6j529g$71d$1@www.dzsh.de>

Ben de Luca schrieb in Nachricht
<3a8e53e2$0$25475$7f31c96c at news01.syd.optusnet.com.au>...
>when i write the data to a file whats the best way? I guess there is no way
>to sumb the dictonaries strait to disk? i will have to parse it?
>
You can write it to disk like:

# build a dictionary
a = {}
a["AAA"] = 1
a["BBB"] = 2
a["CCC"] = 3

# writing out
f = open("dict.tmp", "w")
f.write(repr(a))
f.close()

# reading in
f = open("dict.tmp", "r")
b = eval(f.read())
f.close()
print type(b), b


HTH
R?diger




From parker at gol.com  Wed Feb 21 23:40:14 2001
From: parker at gol.com (Ian Parker)
Date: Wed, 21 Feb 2001 22:40:14 GMT
Subject: file access
References: <3a8e53e2$0$25475$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <Wke8QAA1OEl6Ew40@gol.com>

In article <3a8e53e2$0$25475$7f31c96c at news01.syd.optusnet.com.au>, Ben
de Luca <c941520 at alinga.newcastle.edu.au> writes
>i am building a database in python its quite small only a few hundred
>entries? i gues dictionaries are the best way to store it in memory
>
>when i write the data to a file whats the best way? I guess there is no way
>to sumb the dictonaries strait to disk? i will have to parse it?
>
>

You may be able to use the "anydbm' module which gives you dictionary-
like access to files.  There's a restriction -  the keys must be strings

import anydbm
        # create the database
db = anydbm.open("data.db", "c")
        # set some keys and values
db["where"] = "here"
db["when"] = "now"
        # close it - it's now safely on disk
db.close()

And when you want to access that data again perhaps in another program,
in read-only or in read/write, just open the database accordingly and
use it like a dictionary

db = anydbm.open("data.db", "r")
for key in db.keys(): print key,":", db[key]

I guess the overhead might be high, but it's so convenient.

Regards

-- 
Ian Parker


From spamers at must.die  Sun Feb 18 12:00:35 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Sun, 18 Feb 2001 12:00:35 +0100
Subject: Remove directories from os.listdir output
Message-ID: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>

Hi!

	I need to filter out all the diretories of the list returned by
os.listdir(). How can I tell if an entry of that list is a dir or a regular
file? O:-)

TIA




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From stephen_purcell at yahoo.com  Sun Feb 18 12:39:10 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sun, 18 Feb 2001 12:39:10 +0100
Subject: Remove directories from os.listdir output
In-Reply-To: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>; from spamers@must.die on Sun, Feb 18, 2001 at 12:00:35PM +0100
References: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>
Message-ID: <20010218123910.A21374@freedom.puma-ag.com>

Fernando Rodr?guez wrote:
> Hi!
> 
> 	I need to filter out all the diretories of the list returned by
> os.listdir(). How can I tell if an entry of that list is a dir or a regular
> file? O:-)

Use os.path.isdir():

>>> import os
>>> def listdir_files_only(directory):
...     contents = os.listdir(directory)
...     return filter(lambda f,d=directory: not os.path.isdir(os.path.join(d,f)), contents)
... 
>>> listdir_files_only('/tmp')
['python-sHAOTT', 'python-sHASlX', 'python-sHATtS']


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From fredrik at pythonware.com  Sun Feb 18 12:47:27 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 11:47:27 GMT
Subject: Remove directories from os.listdir output
References: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>
Message-ID: <jDOj6.18081$AH6.2469355@newsc.telia.net>

Fernando Rodr?guez wrote:
> I need to filter out all the diretories of the list returned by
> os.listdir(). How can I tell if an entry of that list is a dir or a regular
> file? O:-)

os.path.isdir(filename)
os.path.isfile(filename)

e.g.

for file in os.listdir(mydir):
    fullname = os.path.join(mydir, file)
    if os.path.isfile(fullname):
        ...

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From scarblac at pino.selwerd.nl  Sun Feb 18 17:00:56 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 18 Feb 2001 16:00:56 GMT
Subject: Remove directories from os.listdir output
References: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>
Message-ID: <slrn98vs9o.4ab.scarblac@pino.selwerd.nl>

Fernando Rodr?guez <spamers at must.die> wrote in comp.lang.python:
> 	I need to filter out all the diretories of the list returned by
> os.listdir(). How can I tell if an entry of that list is a dir or a regular
> file? O:-)

With os.path.isdir(file).

-- 
Remco Gerlich


From fabrice.n at home.com  Sun Feb 18 14:01:58 2001
From: fabrice.n at home.com (shaka)
Date: Sun, 18 Feb 2001 13:01:58 GMT
Subject: 3D animation
Message-ID: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com>


Ya, I am planning to write a program to present a project at school.  I need
a language that will allow me to do 3D animations and other animation.  And
I am struggling between writting the program in java or using python.
Can someone help me?  If python is the ultimate choice then can you give me
some instructions on what libraries I should use, because I am quite a
beginner here>

Thanks.

Shaka. Z




From aleaxit at yahoo.com  Sun Feb 18 14:55:39 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sun, 18 Feb 2001 14:55:39 +0100
Subject: 3D animation
References: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com>
Message-ID: <96ok7l017uq@news2.newsguy.com>

"shaka" <fabrice.n at home.com> wrote in message
news:aJPj6.171612$KP3.43764577 at news3.rdc1.on.home.com...
>
> Ya, I am planning to write a program to present a project at school.  I
need
> a language that will allow me to do 3D animations and other animation.
And
> I am struggling between writting the program in java or using python.
> Can someone help me?  If python is the ultimate choice then can you give
me
> some instructions on what libraries I should use, because I am quite a
> beginner here>

Start at http://www.alice.org/ -- a totally free 3D animation framework,
with authoring tool, scriptable in either Python or Java, originally by
University of Virginia and now by Carnegie-Mellon University.  This will
put Python and Java on a totally even playing field, and I think you'll
find that Python is preferable for this use -- but if you already know both
languages you'll be able to try both and make up your own mind!-)


Alex





From rtexier at elikya.com  Mon Feb 19 11:24:33 2001
From: rtexier at elikya.com (Romuald Texier)
Date: Mon, 19 Feb 2001 10:24:33 +0000
Subject: 3D animation
References: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com> <96ok7l017uq@news2.newsguy.com>
Message-ID: <96qoop$6qb$1@news.irisa.fr>

Alex Martelli wrote:



> Start at http://www.alice.org/ -- a totally free 3D animation framework,

It is Microsoft only, so if Shaka choose Java of Python to be portable, it 
is definitively not the good choice.

Regards,

-- 
Romuald Texier


From bryan at eevolved.com  Sun Feb 18 15:32:40 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Sun, 18 Feb 2001 14:32:40 GMT
Subject: 3D animation
References: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com>
Message-ID: <c2Rj6.148997$Pm2.2652810@news20.bellglobal.com>

shaka wrote:

> 
> 
> Ya, I am planning to write a program to present a project at school.  I
> need
> a language that will allow me to do 3D animations and other animation. 
> And I am struggling between writting the program in java or using python.
> Can someone help me?  If python is the ultimate choice then can you give
> me some instructions on what libraries I should use, because I am quite a
> beginner here>
> 
> Thanks.
> 
> Shaka. Z
> 
> 

Honestly couldn't tell you if Java has anything better for this purpose but 
I can tell you python has what you need.

First you need a real GUI toolkit. Try:

http://sourceforge.net/projects/foxgui/
Make sure to compile with OpenGL support.

with the python bindings here:

http://sourceforge.net/projects/fxpy/

Grab the openGL bindings:

http://sourceforge.net/projects/pyopengl/

Now you should be set to look at the funky GL tests that come in the FXPy 
bindings. This should give you a great example of what you can do easily in 
python.

Good luck.
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"This is one of the hardest lessons for humans to learn. We cannot admit 
that things might be neither good nor evil, neither cruel nor kind, but 
simply callous - indifferent to all suffering, lacking all purpose." -- 
Richard Dawkins



From clee at gnwy100.wuh.wustl.edu  Sun Feb 18 16:52:10 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 18 Feb 2001 09:52:10 -0600
Subject: 3D animation
References: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com>
Message-ID: <7uvgq8at05.fsf@gnwy100.wuh.wustl.edu>

>>>>> "shaka" == shaka  <fabrice.n at home.com> writes:

 
    shaka> Ya, I am planning to write a program to present a project at
    shaka> school.  I need a language that will allow me to do 3D
    shaka> animations and other animation.

I would recommend:

PyOpenGL:  for OpenGL scripting
           requires that you know or are willing to learn OpenGL.
           http://pyopengl.sourceforge.net/
           

VPython:   for easy visualization of physical systems.  Very cool.  Looks good
           for beginners.
           http://virtualphoton.pc.cc.cmu.edu/projects/visual/
           (link seems to not be working at the moment. check google)


You might also want to look at pygame (http://pygame.seul.org) for
producing multimedia animations.  For really heavy duty visualization,
there's also the visualization toolkit (VTK at
http://www.kitware.com/vtk.html) and IBM's Open Data Explorer (OpenDx at
http://www.opendx.org), both of which have python bindings.

-christopher


From phlip_cpp at my-deja.com  Sun Feb 18 16:55:56 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 15:55:56 GMT
Subject: 3D animation
References: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com>
Message-ID: <96orac$brv@dispatch.concentric.net>

Proclaimed shaka from the mountaintops:

> Ya, I am planning to write a program to present a project at school.  I
> need
> a language that will allow me to do 3D animations and other animation. 
> And I am struggling between writting the program in java or using python.
> Can someone help me?  If python is the ultimate choice then can you give
> me some instructions on what libraries I should use, because I am quite a
> beginner here>

Lose your self in the universe of Blender, and you'l never need non-virtual 
reality again:

        http://www.blender.nl

But be warned: The system's top level is all scripted in this dumb trendy 
language whose name escapes me at the moment.

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  "Probably one of the toughest times in anyone's life is when
       you have to kill someone you love because they'r the Devil"
                                     --Emo Phillips  --


From dave at dave.org.uk  Sun Feb 18 14:16:49 2001
From: dave at dave.org.uk (Dave Cross)
Date: Sun, 18 Feb 2001 13:16:49 +0000
Subject: Weird Language Features
Message-ID: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>

[Please watch the replies on this message as it's heavily
cross-posted]

I'm doing some comparisons on programming language features and I'd be
very interested to know how you would handle the following scenarios
in your programming language of choice.

1/ The programmer calls a function that doesn't actually exist within
the application (or libraries). Is the a feature whereby the
programmer can create a "catch-all" function which is called in cases
like these? Can this function examine the list of existing functions
and call the most appropriate one? Or create a new function on the fly
and install it into the application?

2/ Can ou filter the input source code before compilation (or
interpretation) in some way so that language keywords could be changed
for other strings? Imagine you wanted to allow someone to program your
language of choice in, say, French. How would you go about translating
French keywords into ones that the compiler (or interpreter) could
understand. What if the translation wasn't one-to-one or fixed? Could
you put enough intelligence into the translator so that it could
handle certain strings differently depending on where they appeared in
the source code?


If you're wondering why I'm inventing these bizarre scenarios, it's
for a paper I'm writing for this year's Perl Conference.  Perl does
have these features (see the AUTOLOAD function and source filters) and
I'm interested in seeing how widespread they are in other languages.

Of course, if you'd like to tell me just why you consider it's a good
thing that your language of choice doesn't have these features, then
I'd be only too happy to hear that too.

I'd just like to make it clear that I'm not interested in getting into
"my language is better than your language" types of flamewars. I'm
certainly not trying to argue that Perl is better than other languages
for having these features.

Thanks for your time.

Dave...

-- 
<http://www.dave.org.uk>  SMS: sms at dave.org.uk


From nem00u at cs.nott.ac.uk  Sun Feb 18 17:58:04 2001
From: nem00u at cs.nott.ac.uk (Neil Madden)
Date: Sun, 18 Feb 2001 16:58:04 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <3A8FFF1C.1FFEADE7@cs.nott.ac.uk>


Dave Cross wrote:
> 
> [Please watch the replies on this message as it's heavily
> cross-posted]
> 
> I'm doing some comparisons on programming language features and I'd be
> very interested to know how you would handle the following scenarios
> in your programming language of choice.
> 
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?

Tcl has the "unknown" procedure that is called when the interpreter
doesn't recognize a command. From the Tcl help pages:

"If the Tcl interpreter encounters a command name for which there is not
a defined command, then Tcl checks for the existence of a command named
unknown. If there is no such command, then the interpreter returns an
error. If the unknown command exists, then it is invoked with arguments
consisting of the fully-substituted name and arguments for the original
non-existent command. The unknown command typically does things like
searching through library directories for a command procedure with the
name cmdName, or expanding abbreviated command names to full-length, or
automatically executing unknown commands as sub-processes. In some cases
(such as expanding abbreviations) unknown will change the original
command slightly and then (re-)execute it. The result of the unknown
command is used as the result for the original non-existent command."
(<URL: http://dev.scriptics.com/man/tcl8.4/TclCmd/unknown.htm>)

It is left up to the application programmer to implement this, so you
could have an unknown procedure that does the things you mention above.

See also: <URL: http://mini.net/cgi-bin/wikit/495.html>) for some
examples.


> 
> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could
> you put enough intelligence into the translator so that it could
> handle certain strings differently depending on where they appeared in
> the source code?

There are no reserved words in Tcl, so all keywords can be overwritten
or renamed. Using the unknown command above, you can put as much
intelligence as you want into interpreting a command. Although, this
seems like a fairly dangerous thing to do, as any packages or extensions
you used would have to know that the keywords had all been changed! But
once again, you could allow the [unknown] handler to map commands onto
their new names.

> 
> If you're wondering why I'm inventing these bizarre scenarios, it's
> for a paper I'm writing for this year's Perl Conference.  Perl does
> have these features (see the AUTOLOAD function and source filters) and
> I'm interested in seeing how widespread they are in other languages.
> 
> Of course, if you'd like to tell me just why you consider it's a good
> thing that your language of choice doesn't have these features, then
> I'd be only too happy to hear that too.
> 
> I'd just like to make it clear that I'm not interested in getting into
> "my language is better than your language" types of flamewars. I'm
> certainly not trying to argue that Perl is better than other languages
> for having these features.
> 
> Thanks for your time.
> 
> Dave...
> 
> --
> <http://www.dave.org.uk>  SMS: sms at dave.org.uk

-- 
--------------------------------------------------------------
Neil Madden.
neil at tallniel.co.uk                       nem00u at cs.nott.ac.uk
http://www.tallniel.co.uk     http://www.cs.nott.ac.uk/~nem00u
--------------------------------------------------------------


From nem00u at cs.nott.ac.uk  Mon Feb 19 00:29:27 2001
From: nem00u at cs.nott.ac.uk (Neil Madden)
Date: Sun, 18 Feb 2001 23:29:27 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <3A8FFF1C.1FFEADE7@cs.nott.ac.uk>
Message-ID: <3A905AD7.152F6626@cs.nott.ac.uk>

BTW. I think that the problems you have posed are very much "scripting"
solutions rather than system programming solutions. In other words,
scripting languages (perl, python, tcl) are likely to be able to solve
these fairly easily, whereas compiled languages are going to find it
harder. If you posted scenarios involving intensive maths computation
the results might be the other way around :-). If you are writing a
paper, compare like to like...

Neil Madden wrote:
> 
> Dave Cross wrote:
> >
> > [Please watch the replies on this message as it's heavily
> > cross-posted]
> >
> > I'm doing some comparisons on programming language features and I'd be
> > very interested to know how you would handle the following scenarios
> > in your programming language of choice.
> >
> > 1/ The programmer calls a function that doesn't actually exist within
> > the application (or libraries). Is the a feature whereby the
> > programmer can create a "catch-all" function which is called in cases
> > like these? Can this function examine the list of existing functions
> > and call the most appropriate one? Or create a new function on the fly
> > and install it into the application?
> 
> Tcl has the "unknown" procedure that is called when the interpreter
> doesn't recognize a command. From the Tcl help pages:
> 
> "If the Tcl interpreter encounters a command name for which there is not
> a defined command, then Tcl checks for the existence of a command named
> unknown. If there is no such command, then the interpreter returns an
> error. If the unknown command exists, then it is invoked with arguments
> consisting of the fully-substituted name and arguments for the original
> non-existent command. The unknown command typically does things like
> searching through library directories for a command procedure with the
> name cmdName, or expanding abbreviated command names to full-length, or
> automatically executing unknown commands as sub-processes. In some cases
> (such as expanding abbreviations) unknown will change the original
> command slightly and then (re-)execute it. The result of the unknown
> command is used as the result for the original non-existent command."
> (<URL: http://dev.scriptics.com/man/tcl8.4/TclCmd/unknown.htm>)
> 
> It is left up to the application programmer to implement this, so you
> could have an unknown procedure that does the things you mention above.
> 
> See also: <URL: http://mini.net/cgi-bin/wikit/495.html>) for some
> examples.
> 
> >
> > 2/ Can ou filter the input source code before compilation (or
> > interpretation) in some way so that language keywords could be changed
> > for other strings? Imagine you wanted to allow someone to program your
> > language of choice in, say, French. How would you go about translating
> > French keywords into ones that the compiler (or interpreter) could
> > understand. What if the translation wasn't one-to-one or fixed? Could
> > you put enough intelligence into the translator so that it could
> > handle certain strings differently depending on where they appeared in
> > the source code?
> 
> There are no reserved words in Tcl, so all keywords can be overwritten
> or renamed. Using the unknown command above, you can put as much
> intelligence as you want into interpreting a command. Although, this
> seems like a fairly dangerous thing to do, as any packages or extensions
> you used would have to know that the keywords had all been changed! But
> once again, you could allow the [unknown] handler to map commands onto
> their new names.
> 
> >
> > If you're wondering why I'm inventing these bizarre scenarios, it's
> > for a paper I'm writing for this year's Perl Conference.  Perl does
> > have these features (see the AUTOLOAD function and source filters) and
> > I'm interested in seeing how widespread they are in other languages.
> >
> > Of course, if you'd like to tell me just why you consider it's a good
> > thing that your language of choice doesn't have these features, then
> > I'd be only too happy to hear that too.
> >
> > I'd just like to make it clear that I'm not interested in getting into
> > "my language is better than your language" types of flamewars. I'm
> > certainly not trying to argue that Perl is better than other languages
> > for having these features.
> >
> > Thanks for your time.
> >
> > Dave...
> >
> > --
> > <http://www.dave.org.uk>  SMS: sms at dave.org.uk
> 
> --
> --------------------------------------------------------------
> Neil Madden.
> neil at tallniel.co.uk                       nem00u at cs.nott.ac.uk
> http://www.tallniel.co.uk     http://www.cs.nott.ac.uk/~nem00u
> --------------------------------------------------------------

-- 
--------------------------------------------------------------
Neil Madden.
neil at tallniel.co.uk                       nem00u at cs.nott.ac.uk
http://www.tallniel.co.uk     http://www.cs.nott.ac.uk/~nem00u
--------------------------------------------------------------


From fellowsd at cs.man.ac.uk  Mon Feb 19 11:49:03 2001
From: fellowsd at cs.man.ac.uk (Donal K. Fellows)
Date: Mon, 19 Feb 2001 10:49:03 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <3A8FFF1C.1FFEADE7@cs.nott.ac.uk>
Message-ID: <3A90FA1F.BE0B7277@cs.man.ac.uk>

Neil Madden wrote:
> Dave Cross wrote:
>> 2/ Can ou filter the input source code before compilation (or
>> interpretation) in some way so that language keywords could be changed
>> for other strings? Imagine you wanted to allow someone to program your
>> language of choice in, say, French. How would you go about translating
>> French keywords into ones that the compiler (or interpreter) could
>> understand. What if the translation wasn't one-to-one or fixed? Could
>> you put enough intelligence into the translator so that it could
>> handle certain strings differently depending on where they appeared in
>> the source code?
> 
> There are no reserved words in Tcl, so all keywords can be overwritten
> or renamed. Using the unknown command above, you can put as much
> intelligence as you want into interpreting a command. Although, this
> seems like a fairly dangerous thing to do, as any packages or extensions
> you used would have to know that the keywords had all been changed! But
> once again, you could allow the [unknown] handler to map commands onto
> their new names.

It'd be better to rewrite the [source] command as well, since then you
could apply the transformations at code-loading time as well, and if you
were doing something as static as supporting a second language for the
words of the program, then (as long as the literal strings used for the
two languages were distinct) you could easily use interpreter aliasing
instead.

Donal.
-- 
Donal K. Fellows    http://www.cs.man.ac.uk/~fellowsd/    fellowsd at cs.man.ac.uk
-- If somebody in a suit or a uniform can sit you down and have you believe
   every word he says, you've just surrendered your license to be a thinking
   human being and traded it in for a zombie suit. -- mhoye at prince.carleton.ca


From Robert_Hand at notspammehotmail.com  Sun Feb 18 18:09:20 2001
From: Robert_Hand at notspammehotmail.com (Robert Hand)
Date: Sun, 18 Feb 2001 09:09:20 -0800
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <3iTj6.1813$E41.169931@news.uswest.net>

> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?

If you call a function that doesn't exist, I don't think it will compile.
To examine
an object and it's functions I think you use the reflection library (haven't
messed with it myself)

>
> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could
> you put enough intelligence into the translator so that it could
> handle certain strings differently depending on where they appeared in
> the source code?
>
This would be an application to replace strings.  Probably impossible due to
abbreviations and wierd stuff.
Unless you want to type in each translation, instead of automating it.
Java does have some neat internationalization features for replacing
messages, menu options etc.
Pretty advanced/
>
> If you're wondering why I'm inventing these bizarre scenarios, it's
> for a paper I'm writing for this year's Perl Conference.  Perl does
> have these features (see the AUTOLOAD function and source filters) and
> I'm interested in seeing how widespread they are in other languages.
>
Really?  Thats great!  If you can do it in perl, you can probably implement
it in java.
I'm not a perl expert.  Is the perl source code translation a function of
perl or a separate app?
if it's a separate app, handling ascii input, give me a link.

> Of course, if you'd like to tell me just why you consider it's a good
> thing that your language of choice doesn't have these features, then
> I'd be only too happy to hear that too.
>
I think as far as internationalization goes, java is prety mature.
Unfortunately, it's
2000 miles from me to the next non-english speaking nation.  No opportunity
to pursue it.


> I'd just like to make it clear that I'm not interested in getting into
> "my language is better than your language" types of flamewars. I'm
> certainly not trying to argue that Perl is better than other languages
> for having these features.
nope.  I must start a flame.  I am compelled to by satan.  Bill Gates is our
savior.  All hail the great redmond giant! HAIL or be consumed! Oh come on,
please hail.  Please.  I'll give you a free copy of interdev.  I'll but your
company and give you free coca-cola.  Fine then.  If you won't praise Bill
I'll just go pout until you do....


> <http://www.dave.org.uk>  SMS: sms at dave.org.uk




From dave at dave.org.uk  Sun Feb 18 18:24:19 2001
From: dave at dave.org.uk (Dave Cross)
Date: Sun, 18 Feb 2001 17:24:19 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <3iTj6.1813$E41.169931@news.uswest.net>
Message-ID: <c7109to8ccq3r4gmpqq1eqf9io8co0vhum@4ax.com>

On Sun, 18 Feb 2001 09:09:20 -0800, "Robert Hand"
<Robert_Hand at notspammehotmail.com> wrote:

>>
>> If you're wondering why I'm inventing these bizarre scenarios, it's
>> for a paper I'm writing for this year's Perl Conference.  Perl does
>> have these features (see the AUTOLOAD function and source filters) and
>> I'm interested in seeing how widespread they are in other languages.
>>
>Really?  Thats great!  If you can do it in perl, you can probably implement
>it in java.
>I'm not a perl expert.  Is the perl source code translation a function of
>perl or a separate app?
>if it's a separate app, handling ascii input, give me a link.

It's a separate library that gets to see the source code before it's
compiled. For an (admittedly extreme) example of what you can do see:

<http://www.csse.monash.edu.au/~damian/papers/HTML/Perligata.html>

Which discusses a filter allowing you to write Perl programs in Latin!

Cheers,

Dave...

-- 
<http://www.dave.org.uk>  SMS: sms at dave.org.uk


From zala_ at hotmail.com  Sun Feb 18 18:35:00 2001
From: zala_ at hotmail.com (Les F.)
Date: Sun, 18 Feb 2001 10:35:00 -0700
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <3iTj6.1813$E41.169931@news.uswest.net>
Message-ID: <ujC7xDdmAHA.1232@tkmsftngp05>

Non-english speaking nations?
If you live on the east coast of USA:
St Pierre Miglon (France)
Quebec (wants to be France)
Cuba (Cuba)
Florida (wants to be Cuba)
Louisianna (displaced, disgruntled, former Acadians)
Middle America:
Mexico (Mexico)
New Mexico (Mexico)
Canada (eh)
West Coast:
Mexico (Mexico)
Vancouver (Hong Kong)

Les F ;-)


Mexico (Mexico)
Robert Hand <Robert_Hand at notspammehotmail.com> wrote in message
news:3iTj6.1813$E41.169931 at news.uswest.net...
> > 1/ The programmer calls a function that doesn't actually exist within
> > the application (or libraries). Is the a feature whereby the
> > programmer can create a "catch-all" function which is called in cases
> > like these? Can this function examine the list of existing functions
> > and call the most appropriate one? Or create a new function on the fly
> > and install it into the application?
>
> If you call a function that doesn't exist, I don't think it will compile.
> To examine
> an object and it's functions I think you use the reflection library
(haven't
> messed with it myself)
>
> >
> > 2/ Can ou filter the input source code before compilation (or
> > interpretation) in some way so that language keywords could be changed
> > for other strings? Imagine you wanted to allow someone to program your
> > language of choice in, say, French. How would you go about translating
> > French keywords into ones that the compiler (or interpreter) could
> > understand. What if the translation wasn't one-to-one or fixed? Could
> > you put enough intelligence into the translator so that it could
> > handle certain strings differently depending on where they appeared in
> > the source code?
> >
> This would be an application to replace strings.  Probably impossible due
to
> abbreviations and wierd stuff.
> Unless you want to type in each translation, instead of automating it.
> Java does have some neat internationalization features for replacing
> messages, menu options etc.
> Pretty advanced/
> >
> > If you're wondering why I'm inventing these bizarre scenarios, it's
> > for a paper I'm writing for this year's Perl Conference.  Perl does
> > have these features (see the AUTOLOAD function and source filters) and
> > I'm interested in seeing how widespread they are in other languages.
> >
> Really?  Thats great!  If you can do it in perl, you can probably
implement
> it in java.
> I'm not a perl expert.  Is the perl source code translation a function of
> perl or a separate app?
> if it's a separate app, handling ascii input, give me a link.
>
> > Of course, if you'd like to tell me just why you consider it's a good
> > thing that your language of choice doesn't have these features, then
> > I'd be only too happy to hear that too.
> >
> I think as far as internationalization goes, java is prety mature.
> Unfortunately, it's
> 2000 miles from me to the next non-english speaking nation.  No
opportunity
> to pursue it.
>
>
> > I'd just like to make it clear that I'm not interested in getting into
> > "my language is better than your language" types of flamewars. I'm
> > certainly not trying to argue that Perl is better than other languages
> > for having these features.
> nope.  I must start a flame.  I am compelled to by satan.  Bill Gates is
our
> savior.  All hail the great redmond giant! HAIL or be consumed! Oh come
on,
> please hail.  Please.  I'll give you a free copy of interdev.  I'll but
your
> company and give you free coca-cola.  Fine then.  If you won't praise Bill
> I'll just go pout until you do....
>
>
> > <http://www.dave.org.uk>  SMS: sms at dave.org.uk
>
>




From agold at bga.com  Sun Feb 18 19:03:21 2001
From: agold at bga.com (Arthur H. Gold)
Date: Sun, 18 Feb 2001 12:03:21 -0600
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <3iTj6.1813$E41.169931@news.uswest.net> <ujC7xDdmAHA.1232@tkmsftngp05>
Message-ID: <3A900E69.6FCCAD03@bga.com>

"Les F." wrote:
> 
> Non-english speaking nations?
> If you live on the east coast of USA:
> St Pierre Miglon (France)
> Quebec (wants to be France)
> Cuba (Cuba)
> Florida (wants to be Cuba)
> Louisianna (displaced, disgruntled, former Acadians)
Ey! Y'left out Brooklyn! Fuhgeddaboudit!
--ag
> Middle America:
> Mexico (Mexico)
> New Mexico (Mexico)
> Canada (eh)
> West Coast:
> Mexico (Mexico)
> Vancouver (Hong Kong)
> 
> Les F ;-)
> 
> Mexico (Mexico)
> Robert Hand <Robert_Hand at notspammehotmail.com> wrote in message
> news:3iTj6.1813$E41.169931 at news.uswest.net...
> > > 1/ The programmer calls a function that doesn't actually exist within
> > > the application (or libraries). Is the a feature whereby the
> > > programmer can create a "catch-all" function which is called in cases
> > > like these? Can this function examine the list of existing functions
> > > and call the most appropriate one? Or create a new function on the fly
> > > and install it into the application?
> >
> > If you call a function that doesn't exist, I don't think it will compile.
> > To examine
> > an object and it's functions I think you use the reflection library
> (haven't
> > messed with it myself)
> >
> > >
> > > 2/ Can ou filter the input source code before compilation (or
> > > interpretation) in some way so that language keywords could be changed
> > > for other strings? Imagine you wanted to allow someone to program your
> > > language of choice in, say, French. How would you go about translating
> > > French keywords into ones that the compiler (or interpreter) could
> > > understand. What if the translation wasn't one-to-one or fixed? Could
> > > you put enough intelligence into the translator so that it could
> > > handle certain strings differently depending on where they appeared in
> > > the source code?
> > >
> > This would be an application to replace strings.  Probably impossible due
> to
> > abbreviations and wierd stuff.
> > Unless you want to type in each translation, instead of automating it.
> > Java does have some neat internationalization features for replacing
> > messages, menu options etc.
> > Pretty advanced/
> > >
> > > If you're wondering why I'm inventing these bizarre scenarios, it's
> > > for a paper I'm writing for this year's Perl Conference.  Perl does
> > > have these features (see the AUTOLOAD function and source filters) and
> > > I'm interested in seeing how widespread they are in other languages.
> > >
> > Really?  Thats great!  If you can do it in perl, you can probably
> implement
> > it in java.
> > I'm not a perl expert.  Is the perl source code translation a function of
> > perl or a separate app?
> > if it's a separate app, handling ascii input, give me a link.
> >
> > > Of course, if you'd like to tell me just why you consider it's a good
> > > thing that your language of choice doesn't have these features, then
> > > I'd be only too happy to hear that too.
> > >
> > I think as far as internationalization goes, java is prety mature.
> > Unfortunately, it's
> > 2000 miles from me to the next non-english speaking nation.  No
> opportunity
> > to pursue it.
> >
> >
> > > I'd just like to make it clear that I'm not interested in getting into
> > > "my language is better than your language" types of flamewars. I'm
> > > certainly not trying to argue that Perl is better than other languages
> > > for having these features.
> > nope.  I must start a flame.  I am compelled to by satan.  Bill Gates is
> our
> > savior.  All hail the great redmond giant! HAIL or be consumed! Oh come
> on,
> > please hail.  Please.  I'll give you a free copy of interdev.  I'll but
> your
> > company and give you free coca-cola.  Fine then.  If you won't praise Bill
> > I'll just go pout until you do....
> >
> >
> > > <http://www.dave.org.uk>  SMS: sms at dave.org.uk
> >
> >

-- 
Artie Gold, Austin, TX  (finger the cs.utexas.edu account
for more info)
mailto:agold at bga.com or mailto:agold at cs.utexas.edu
--
Verbing weirds language.


From binary at eton.powernet.co.uk  Sun Feb 18 18:42:35 2001
From: binary at eton.powernet.co.uk (Richard Heathfield)
Date: Sun, 18 Feb 2001 17:42:35 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <3A90098B.5EBB3993@eton.powernet.co.uk>

Dave Cross wrote:
> 
> [Please watch the replies on this message as it's heavily
> cross-posted]

Noted.

> 
> I'm doing some comparisons on programming language features and I'd be
> very interested to know how you would handle the following scenarios
> in your programming language of choice.
> 
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries).

In ISO conforming C, he can't. End of story.

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings?

Again in ISO conforming C...

#define ALIAS original

e.g.

#define voiture auto /* :-) */


-- 
Richard Heathfield
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
K&R answers, C books, etc: http://users.powernet.co.uk/eton


From pfaffben at msu.edu  Sun Feb 18 18:44:28 2001
From: pfaffben at msu.edu (Ben Pfaff)
Date: 18 Feb 2001 12:44:28 -0500
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <874rxrj37n.fsf@pfaffben.user.msu.edu>

Dave Cross <dave at dave.org.uk> writes:

> I'm doing some comparisons on programming language features and I'd be
> very interested to know how you would handle the following scenarios
> in your programming language of choice.
> 
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries).

Undefined behavior in C.

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings?

#define keyword replacement
Also causes undefined behavior in at least some circumstances.
-- 
"The expression isn't unclear *at all* and only an expert could actually
 have doubts about it"
--Dan Pop


From erno-news at erno.iki.fi  Sun Feb 18 19:15:47 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 18 Feb 2001 20:15:47 +0200
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <kubsrzuab0.fsf@lasipalatsi.fi>

In article <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n at 4ax.com>, Dave Cross
<dave at dave.org.uk> writes:

| 1/ The programmer calls a function that doesn't actually exist within
| the application (or libraries). Is the a feature whereby the
| programmer can create a "catch-all" function which is called in cases
| like these? Can this function examine the list of existing functions
| and call the most appropriate one? Or create a new function on the fly
| and install it into the application?

python: i can't think of a way to do this for functions,
but you can easily do it for methods - __getattr__ is the magic
method.

see <URL:http://www.python.org/doc/current/ref/attribute-access.html>.

| 2/ Can you filter the input source code before compilation

python does not allow you to do this.

| Of course, if you'd like to tell me just why you consider it's a good
| thing that your language of choice doesn't have these features, then
| I'd be only too happy to hear that too.

as far as i'm concerned language features are guilty (of being
superfluous) before proven innocent :).

  -- erno


From johann at physics.berkeley.edu  Sun Feb 18 21:16:31 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 18 Feb 2001 12:16:31 -0800
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <mtk86n6928.fsf@astron.berkeley.edu>

Dave Cross writes:

> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries).

It's easy to trap this error in Common Lisp, but it's hard to set up a
"general replacement" routine.  You can, generally, trap the error for
a given block of code, then attempt to guess a replacement function,
but I don't think you can then resume the calculation.

However, if you're reading in user/scripting code, it's an easy thing
to pre-process it, looking for unknown functions, then replace those
symbols with your guess.  It just has to be a pre-processing step, not
a run-time step.  i.e. you have to do (eval (process (read-all
"file"))), rather than just (load "file").

If the program's not making an error, the package system should be
able to handle it.

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings?

Pathetically easy in any of the lisp-derivative languages.  This is
where having a code form that looks like data really shines; you just
read the code in, manipulate it at will, then execute it afterwards.
Use the same (eval (process (read-all "file"))).

> If you're wondering why I'm inventing these bizarre scenarios, it's
> for a paper I'm writing for this year's Perl Conference.

Have fun.  I do suspect that Common Lisp whups Perl in this respect,
though.  :-)

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From neelk at alum.mit.edu  Sun Feb 18 22:15:37 2001
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 18 Feb 2001 21:15:37 GMT
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <mtk86n6928.fsf@astron.berkeley.edu>
Message-ID: <slrn990ime.a2b.neelk@alum.mit.edu>

[Crossposting trimmed]

On 18 Feb 2001 12:16:31 -0800, Johann Hibschman <johann at physics.berkeley.edu>
wrote:
>Dave Cross writes:
>
>> 1/ The programmer calls a function that doesn't actually exist within
>> the application (or libraries).
>
> It's easy to trap this error in Common Lisp, but it's hard to set up a
> "general replacement" routine.  You can, generally, trap the error for
> a given block of code, then attempt to guess a replacement function,
> but I don't think you can then resume the calculation.

I'm not a serious CL user, but aren't its exceptions restartable? That
would permit you to do this. It would be really nice if Python's
exception system supported this. Then I wouldn't have to choose
between raising an exception and having my constructor functions
behave algebraically.

> Have fun.  I do suspect that Common Lisp whups Perl in this respect,
> though.  :-)

Yeah. :)


Neel


From hannah at schlund.de  Mon Feb 19 18:01:23 2001
From: hannah at schlund.de (Hannah Schroeter)
Date: 19 Feb 2001 18:01:23 +0100
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <mtk86n6928.fsf@astron.berkeley.edu> <slrn990ime.a2b.neelk@alum.mit.edu>
Message-ID: <96rjh3$9c3$1@c3po.schlund.de>

Hello!

In article <slrn990ime.a2b.neelk at alum.mit.edu>,
Neelakantan Krishnaswami <neelk at alum.mit.edu> wrote:
>[...]

>I'm not a serious CL user, but aren't its exceptions restartable? That
>would permit you to do this. It would be really nice if Python's
>exception system supported this. Then I wouldn't have to choose
>between raising an exception and having my constructor functions
>behave algebraically.

No, they aren't.

CL differentiates between exceptions (called "conditions").

You can define handlers for them. A handler can either (as the name
says) handle the condition, i.e. do something and then non-locally
transfer control somewhere using go/return/throw/invoke-restart.
(throw has nothing to do with conditions besides that!), decline
to handle (just return), then the next most handler is invoked in turn,
or defer the decision (signal another condition/resignal the same
condition/enter the debugger/...).

Separate from that are definitions of restart points (which can
be used to handle a condition with invoke-restart). However
the normal function call doesn't establish any restarts which can
help in continuing after e.g. installing an autoloaded function
definition.

>[...]

Kind regards,

Hannah.


From elflord at panix.com  Sun Feb 18 23:28:52 2001
From: elflord at panix.com (Donovan Rebbechi)
Date: 18 Feb 2001 22:28:52 GMT
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <slrn990j54.pcq.elflord@panix6.panix.com>

On Sun, 18 Feb 2001 13:16:49 +0000, Dave Cross wrote:

>1/ The programmer calls a function that doesn't actually exist within
>the application (or libraries). Is the a feature whereby the
>programmer can create a "catch-all" function which is called in cases
>like these? Can this function examine the list of existing functions
>and call the most appropriate one? Or create a new function on the fly
>and install it into the application?

Posting this from comp.lang.c++. In C++, the object model itself does 
not support this.

In both C and C++, one uses dynamic loading to get this kind of behaviour.
Indeed, interpreters written in C and C++ tend to rely heavily on dynamic
loading.

To get the kind of functionality you're talking about, one would probably
store a function table in a map/tree and one could put a "default"
method in there that acts as a catch all.

However, typically one uses exceptions to deal with failed requests, the
idea being that if a request fails, the caller is in a better position 
than the library author to work out what to do.

>2/ Can ou filter the input source code before compilation (or
>interpretation) in some way so that language keywords could be changed
>for other strings? Imagine you wanted to allow someone to program your
>language of choice in, say, French. How would you go about translating
>French keywords into ones that the compiler (or interpreter) could
>understand. What if the translation wasn't one-to-one or fixed? Could
>you put enough intelligence into the translator so that it could
>handle certain strings differently depending on where they appeared in
>the source code?

One could do this with macros but it's a fairly primitive solution ...

>  Perl does
>have these features (see the AUTOLOAD function and source filters) and
>I'm interested in seeing how widespread they are in other languages.

... of course one could always write a preprocessor in perl (-;

>Of course, if you'd like to tell me just why you consider it's a good
>thing that your language of choice doesn't have these features, then
>I'd be only too happy to hear that too.

C and C++ are standardised (as opposed to implementation defined) languages,
the former is meant to be small, the latter is big enough as it is. 
Introducing support for this kind of thing in the language (either
C or C++) would probably be a mistake IMO.

-- 
Donovan Rebbechi * http://pegasus.rutgers.edu/~elflord/ * 
elflord at panix dot com


From clayberg at instantiations.com  Mon Feb 19 03:40:35 2001
From: clayberg at instantiations.com (Eric Clayberg)
Date: Sun, 18 Feb 2001 21:40:35 -0500
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <96q137$q3d$1@nntp9.atl.mindspring.net>

"Dave Cross" <dave at dave.org.uk> wrote in message
news:nnhv8t4obq30ljfgseplp7pi4khhsf9g9n at 4ax.com...
>
> [Please watch the replies on this message as it's heavily
> cross-posted]
>
> I'm doing some comparisons on programming language features and I'd be
> very interested to know how you would handle the following scenarios
> in your programming language of choice.
>
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?

Smalltalk can do all of the above very easily. This general technique is
used to create generic proxies, for example. I have also played around with
"fault tolerant" apps that could automatically detect and correct spelling
errors in function calls (as a lark; never in production code).

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could
> you put enough intelligence into the translator so that it could
> handle certain strings differently depending on where they appeared in
> the source code?

Smalltalk can also do the above very easily. Everything is Smalltalk is an
object - including the parser, the compiler, the compiled methods and the
method source. Setting up a pre-processor or modifying the actual parser is
very easy to do. Since all of Smalltalk's control structures are built in
Smalltalk itself, you can easily alias any function name or operator to
another. Enhancing the "syntax" and adding your own control structures is
also trivial.

> If you're wondering why I'm inventing these bizarre scenarios, it's
> for a paper I'm writing for this year's Perl Conference.  Perl does
> have these features (see the AUTOLOAD function and source filters) and
> I'm interested in seeing how widespread they are in other languages.
>
> Of course, if you'd like to tell me just why you consider it's a good
> thing that your language of choice doesn't have these features, then
> I'd be only too happy to hear that too.
>
> I'd just like to make it clear that I'm not interested in getting into
> "my language is better than your language" types of flamewars. I'm
> certainly not trying to argue that Perl is better than other languages
> for having these features.
>
> Thanks for your time

Thanks for the interesting thread.

-Eric




From hannah at schlund.de  Mon Feb 19 18:08:06 2001
From: hannah at schlund.de (Hannah Schroeter)
Date: 19 Feb 2001 18:08:06 +0100
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <96q137$q3d$1@nntp9.atl.mindspring.net>
Message-ID: <96rjtm$2q$1@c3po.schlund.de>

Hello!

[F'up]

In article <96q137$q3d$1 at nntp9.atl.mindspring.net>,
Eric Clayberg <clayberg at instantiations.com> wrote:
>[...]

[syntax changes through e.g. preprocessing]

>Smalltalk can also do the above very easily. Everything is Smalltalk is an
>object - including the parser, the compiler, the compiled methods and the
>method source. Setting up a pre-processor or modifying the actual parser is
>very easy to do. Since all of Smalltalk's control structures are built in
>Smalltalk itself, you can easily alias any function name or operator to
>another. Enhancing the "syntax" and adding your own control structures is
>also trivial.

Partly yes, but only partly. Access to parser/compiler innards isn't
standardized in any way, much less than Common Lisp's readtable and
macro facilities.

Of course, when it just comes to method names (selectors), it's
easy to modify/extend the standard classes.

>[...]

Kind regards,

Hannah.


From sdm7g at Virginia.EDU  Mon Feb 19 19:30:52 2001
From: sdm7g at Virginia.EDU (Steven D. Majewski)
Date: Mon, 19 Feb 2001 13:30:52 -0500
Subject: Weird Language Features
In-Reply-To: <96q137$q3d$1@nntp9.atl.mindspring.net>
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <96q137$q3d$1@nntp9.atl.mindspring.net>
Message-ID: <Pine.NXT.4.21.0102191313180.231-100000@localhost.virginia.edu>

In objective-c, the functional/procedural part of the language is
just ANSI C, so, with respect to undefined functions, the answer
would be the same as C: no. 

The object-oriented feature of objective-C are based on smalltalk,
so with respect to methods, it's like smalltalk. Any class can
implement forwardInvocation: (as well as some other introspective
methods) to pass methods it doesn't understand to a proxy or delegate.

(I believe you can even redefine this method for NSObject itself, so
 that all subclasses will inherit this new implementation. )

The python<->objective-c bridge in PyObjC uses this in it's
implementation. There are proxy objects sitting between python
classes and their objective-c equivalents. Some common transformations
are built in, so NSStrings and NSArrays can be treated as python
sequences, but in general, python methods are name-mangled into
objective-c method names by translating colons to/from underscores,
and the proxies forward them (depending on which direction you're
going -- objective-C calling Python or Python calling objective-c.) 


-- Steve Majewski <sdm7g at Virginia.EDU>


On Sun, 18 Feb 2001, Eric Clayberg wrote:

> "Dave Cross" <dave at dave.org.uk> wrote in message
> news:nnhv8t4obq30ljfgseplp7pi4khhsf9g9n at 4ax.com...
> >
> > [Please watch the replies on this message as it's heavily
> > cross-posted]
> >
> > I'm doing some comparisons on programming language features and I'd be
> > very interested to know how you would handle the following scenarios
> > in your programming language of choice.
> >
> > 1/ The programmer calls a function that doesn't actually exist within
> > the application (or libraries). Is the a feature whereby the
> > programmer can create a "catch-all" function which is called in cases
> > like these? Can this function examine the list of existing functions
> > and call the most appropriate one? Or create a new function on the fly
> > and install it into the application?
> 
> Smalltalk can do all of the above very easily. This general technique is
> used to create generic proxies, for example. I have also played around with
> "fault tolerant" apps that could automatically detect and correct spelling
> errors in function calls (as a lark; never in production code).
> 



From JeffH at ActiveState.com  Mon Feb 19 18:42:47 2001
From: JeffH at ActiveState.com (Jeffrey Hobbs)
Date: Mon, 19 Feb 2001 09:42:47 -0800
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <3A915B17.1E54882B@ActiveState.com>

Dave Cross wrote:
	...
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?

For Tcl, yes this is a standard feature, set in the 'unknown' procedure,
which can be redefined by the user.  The default behavior is to use it
for lazy-loading of new procedures.  It has also been used to handle
completely new syntax structures in Tcl.

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could

You could build this with any language using message catalogs or the
like.  It's very easy with interpreted languages, and Tcl for one has
the facilities (helper modules) built in.

-- 
  Jeff Hobbs                     The Tcl Guy
  Senior Developer               http://www.ActiveState.com/


From shaleh at valinux.com  Mon Feb 19 19:52:44 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Mon, 19 Feb 2001 10:52:44 -0800
Subject: Weird Language Features
In-Reply-To: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>; from dave@dave.org.uk on Sun, Feb 18, 2001 at 01:16:49PM +0000
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <20010219105244.B29164@valinux.com>

On Sun, Feb 18, 2001 at 01:16:49PM +0000, Dave Cross wrote:
> 
> [Please watch the replies on this message as it's heavily
> cross-posted]
> 
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?
>

python will throw an exception, the code could then decide that the exception
was due to a missing function and decide how to deal with it.  I suspect this
could be useful in really restricted areas, but in general code it strikes me
as horribly confusing.
 
> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could
> you put enough intelligence into the translator so that it could
> handle certain strings differently depending on where they appeared in
> the source code?

perligata makes a LOT of assumptions for you.  it is not really latin.  I
presume it is one of the poster children for this in perl.  its fun, but a toy.

even if you could code essential language words in natural language Foo, things
like if and for, there is a wealth of code out there NOT in Foo.  So even if
I use perligata for the base of my code, when I use any other module I have to
be able to speak the language the module was written in.  So if I try to
import a module written by your friend in French, I now have to know French to
figure out the function and variable names.  As much as we hate it, a common
language is good.

As for python, the interpreter, keyword and parser are available, I suspect
this could be done with a little work.




From sss at ed.com  Wed Feb 21 12:59:12 2001
From: sss at ed.com (sss)
Date: Wed, 21 Feb 2001 12:59:12 +0100
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <3A93AD90.F2C03540@ed.com>

Ruby

Dave Cross wrote:

>
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?
>

"method_missing(...)" does it, either per class or when  overwriting
Object>>method_missing(...) globally


>
> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could
> you put enough intelligence into the translator so that it could
> handle certain strings differently depending on where they appeared in
> the source code?

'alias' allows you to create an alias for any method of any class.
Language keywords such as class, module, while , etc. cannot be aliased
though.

Clemens



From jschmitt at vmlabs.com  Thu Feb 22 05:19:05 2001
From: jschmitt at vmlabs.com (John Schmitt)
Date: Wed, 21 Feb 2001 20:19:05 -0800
Subject: Weird Language Features
Message-ID: <008F0A63472BD311AF9800104BCD102561CCA6@minirex.vmlabs.com>

Doesn't Applescript do something like this?  I recall browsing the docs
where they mentioned that you can program Applescript in more than one
dialect.  I guessed from the docs that you can write Applescript program in
English, French, and Japanese.  If you write an Applescript program in one
language (ie English), change your dialect (ie to French) and load the
program in your editor again, the source code will show up in the new
language  (ie French).  Do I understand that correctly?

John

> -----Original Message-----
> From: Donal K. Fellows [mailto:fellowsd at cs.man.ac.uk]
> Sent: Monday, February 19, 2001 2:49 AM
> To: python-list at python.org
> Subject: Re: Weird Language Features
> [...]
> It'd be better to rewrite the [source] command as well, since then you
> could apply the transformations at code-loading time as well, 
> and if you
> were doing something as static as supporting a second language for the
> words of the program, then (as long as the literal strings 
> used for the
> two languages were distinct) you could easily use interpreter aliasing
> instead.



From fellowsd at cs.man.ac.uk  Thu Feb 22 10:54:16 2001
From: fellowsd at cs.man.ac.uk (Donal K. Fellows)
Date: Thu, 22 Feb 2001 09:54:16 +0000
Subject: Weird Language Features
In-Reply-To: <008F0A63472BD311AF9800104BCD102561CCA6@minirex.vmlabs.com>
	(message from John Schmitt on Wed, 21 Feb 2001 20:19:05 -0800)
References: <008F0A63472BD311AF9800104BCD102561CCA6@minirex.vmlabs.com>
Message-ID: <E14VsS0-0003n9-00@athene>

John Schmitt <jschmitt at vmlabs.com>
> Doesn't Applescript do something like this?  I recall browsing the docs
> where they mentioned that you can program Applescript in more than one
> dialect.  I guessed from the docs that you can write Applescript program in
> English, French, and Japanese.  If you write an Applescript program in one
> language (ie English), change your dialect (ie to French) and load the
> program in your editor again, the source code will show up in the new
> language  (ie French).  Do I understand that correctly?

It's a Mac, so I've no idea!  I have heard rumours to this effect
though.

However, to me this says that the raw program text is a binary form
that is never actually displayed to anyone in the normal course of
things.  Furthermore, this auto-translation is deeply unlikely to
affect the bits of the program that usually need attention; strings
shown to users (and possibly the names of identifiers in the program
if you need it to be maintained by someone with a different native
language.)  Somehow I'm fairly sure that AppleScript can't change the
language of those automatically (since no-one else in the world can
either, as many manuals are ample demonstration of...  :^)

A downside of this is that it makes it difficult for people in
different countries to discuss applescript properly, since a correct
example fragment for one person might be complete gibberish for
another.  You'd have to pass structured documents about instead of
just plain text messages...  :^(

Donal.
-- 
Donal K. Fellows, Department of Computer Science, University of Manchester, UK.
(work) fellowsd at cs.man.ac.uk     Tel: +44-161-275-6137  (preferred email addr.)
(home) donal at ugglan.demon.co.uk  Tel: +44-1274-401017   Mobile: +44-7957-298955
http://www.cs.man.ac.uk/~fellowsd/  (Don't quote my .sig; I've seen it before!)



From me at nospam.net  Sat Feb 24 21:13:23 2001
From: me at nospam.net (Scottie)
Date: Sat, 24 Feb 2001 12:13:23 -0800
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <mailman.982608802.22541.python-list@python.org>
Message-ID: <t9g5fd3qc51df@corp.supernews.com>

The "Mesa" language (used as a system language at Xerox Parc in the
70s) allowed this same fully general exception/interrupt structure.
their experience, after a number of years, was that it did not work out
so well.  Finally, they looked at all of their code, found almost all of
the uses of "fix and continue" could be easily converted to the same
kind of mechanism that Python now uses.  Of the few (five, as I recall)
remaining instances, one had a bug.  So, the last few were rewritten
and the feature was removed from the language.

"Sean 'Shaleh' Perry" <shaleh at valinux.com> wrote in message
news:mailman.982608802.22541.python-list at python.org...
> On Sun, Feb 18, 2001 at 01:16:49PM +0000, Dave Cross wrote:
> > ...1/ The programmer calls a function that doesn't actually exist within
> > the application (or libraries). Is the a feature whereby the
> > programmer can create a "catch-all" function which is called in cases
> > like these? Can this function examine the list of existing functions
> > and call the most appropriate one? Or create a new function on the fly
> > and install it into the application?
> ...python will throw an exception, the code could then decide that the
> exception was due to a missing function and decide how to deal with it.
> I suspect this could be useful in really restricted areas, but in general
> code it strikes me as horribly confusing.

-Scott





From c941520 at alinga.newcastle.edu.au  Sun Feb 18 14:40:56 2001
From: c941520 at alinga.newcastle.edu.au (Ben de Luca)
Date: Mon, 19 Feb 2001 00:40:56 +1100
Subject: pointing at who?
Message-ID: <3a8e808a$0$25513$7f31c96c@news01.syd.optusnet.com.au>

hmm i am trying to find out how the assignmnet in this situation would
occure, is talkingdata[1] the same as (talkingTo[CorrectNumber])[1] in this
situation in essense i want to wait for the switch to flip? can i do this
other wise i will have to use another message or starve


talkingData=[sender,threading.Event(),''] #create working data

data=[] #create final data

self.talkingToLock.acquire() #one hand in the cookie jar at a time
self.talkingTo.append(talkingData) #put in the cookies
self.talkingToLock.release()#shut the lid

talkingData[1].wait() #wait for response






From tim.one at home.com  Sat Feb 17 19:47:13 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 17 Feb 2001 13:47:13 -0500
Subject: pointing at who?
In-Reply-To: <3a8e808a$0$25513$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEAPJAAA.tim.one@home.com>

[Ben de Luca]
> hmm i am trying to find out how the assignmnet in this situation would
> occure, is talkingdata[1] the same as
> (talkingTo[CorrectNumber])[1] in this
> situation

I'm doing my best to break this at sentence boundaries, but you didn't make
it easy <wink>.  The answer is probably "yes".

> in essense i want to wait for the switch to flip? can i do this
> other wise i will have to use another message or starve
>
>
> talkingData=[sender,threading.Event(),''] #create working data

Cool.

> data=[] #create final data

You never use this again, so unclear why it's here.

> self.talkingToLock.acquire() #one hand in the cookie jar at a time
> self.talkingTo.append(talkingData) #put in the cookies
> self.talkingToLock.release()#shut the lid

You didn't tell us anything about self.talkingTo.  Assuming it's a list.  If
so, list.append is atomic and you don't really need the acquire/release pair
around it.

> talkingData[1].wait() #wait for response

After list.append(x), and regardless of x,

    list[-1] is x

is true.  So, ya, for some value of i,

    self.talkingTo[i] is talkingData

is true, i.e. they're exactly the same object.

BTW, drop the fiddly list and create a little SharedData (whatever) class so
you can at least *name* these fields.  Meaningless little integers will come
back to bite you some day.

they-breed-in-the-dark-and-some-have-very-sharp-edges-ly y'rs  - tim




From slhath at home.com  Sun Feb 18 15:31:40 2001
From: slhath at home.com (Scott Hathaway)
Date: Sun, 18 Feb 2001 14:31:40 GMT
Subject: tkinter
Message-ID: <g1Rj6.90426$bb.1166840@news1.rdc1.tx.home.com>

With tkinter, is it possible to put an icon in the Windows system tray for
your app?  If so, where can I find an example?

Thanks,
Scott




From spamers at must.die  Sun Feb 18 15:43:01 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Sun, 18 Feb 2001 15:43:01 +0100
Subject: formatting numbers for output
Message-ID: <9sfv8t45ve3dpql39rmp7082gibbicf56b@4ax.com>

Hi!

	How can I format a number so it pretty prints:

15,456.6 instead of 15456.6 ?

TIA




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From phlip_cpp at my-deja.com  Sun Feb 18 16:46:13 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 15:46:13 GMT
Subject: formatting numbers for output
References: <9sfv8t45ve3dpql39rmp7082gibbicf56b@4ax.com>
Message-ID: <96oqo5$bld@dispatch.concentric.net>

Proclaimed Fernando Rodr?guez from the mountaintops:

> Hi!
> 
> How can I format a number so it pretty prints:
> 
> 15,456.6 instead of 15456.6 ?

While you are at it, ensure the program prints 15.456,6 in parts of Europe 
& certain other places.

(Seriously, "15 456.6" with just a space is a happy median.)

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  My opinions are those of your employer  --


From emile at fenx.com  Sun Feb 18 18:44:18 2001
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 18 Feb 2001 09:44:18 -0800
Subject: formatting numbers for output
References: <9sfv8t45ve3dpql39rmp7082gibbicf56b@4ax.com>
Message-ID: <96p1nt$milta$1@ID-11957.news.dfncis.de>

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)]
on win32
Type "copyright", "credits" or "license" for more
information.
IDLE 0.6 -- press F1 for help
>>> import locale
>>> locale.setlocale(locale.LC_ALL,"")
'English_United States.1252'
>>> print locale.format("%8.2f" , 4123.45, 1)
 4,123.45
>>>

--

Emile van Sebille
emile at fenx.com
-------------------


"Fernando Rodr?guez" <spamers at must.die> wrote in message
news:9sfv8t45ve3dpql39rmp7082gibbicf56b at 4ax.com...
> Hi!
>
> How can I format a number so it pretty prints:
>
> 15,456.6 instead of 15456.6 ?
>
> TIA
>
>
>
>
> //-----------------------------------------------
> // Fernando Rodriguez Romero
> //
> // frr at mindless dot com
> //------------------------------------------------




From tjg at exceptionalminds.com  Sun Feb 18 23:14:33 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Sun, 18 Feb 2001 14:14:33 -0800
Subject: formatting numbers for output
In-Reply-To: <9sfv8t45ve3dpql39rmp7082gibbicf56b@4ax.com>; from spamers@must.die on Sun, Feb 18, 2001 at 03:43:01PM +0100
References: <9sfv8t45ve3dpql39rmp7082gibbicf56b@4ax.com>
Message-ID: <20010218141433.F862@trufflehunter.avalongroup.net>

This is something you have to do yourself. I've attached a little module
I wrote to do it. It's not too flexible, it only deals with the US format,
but it works pretty well. I created it for a Tkinter app that needed to 
display pretty numbers, so it actually goes both ways unformatted to
formatted and formatted to unformatted.

On Sun, Feb 18, 2001 at 03:43:01PM +0100, Fernando Rodr?guez wrote:
> Hi!
> 
> 	How can I format a number so it pretty prints:
> 
> 15,456.6 instead of 15456.6 ?
> 

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  33 days  2:23 hours ago<<
-------------- next part --------------
#! /usr/bin/env python
#
import string

############################################################
#
# Name: commanumber()
#
# Purpose:  To format a number with commas and possibly a
#           dollar sign.
#
# Arguments:    n  = the number to be converted
#               dp = number of decimal places (defaults to 2)
#               ds = dollar sign (1|0) (defaults to $)
#  
def commanumber(n, dp=2, ds=1):
    
    if not n:
        return ''       # If None then bail out here
        
    if type(n) == type('x'):
        if n == '':
            return ''   # If an empty string then bail out here.
        n = string.atof(n)
    
    m = '%0.*f' % (dp, n)
    
    d = string.split(m, '.')    # Split at the decimal point
    r = list(d[0])
    
    # It looks ugly but it really isn't. A couple of really nice Pythonisms
    # make it work right. The first is list insertion, and the second is integer
    # division.
    #
    # the insertion point is calculated based on the counter item, plus the a left
    # shift for each ',' already inserted the ((x/3)-1)
    #
    for x in range(3, len(r), 3):
        r[(x+((x/3)-1))*(-1):(x+((x/3)-1))*(-1)] = [',']
    
    if ds:
        s = '$'
    else:
        s = ''          # Rebuild the string from the list
        
    for i in r:
        s = s + i
        
    if len(d) == 2: # Check to see if we have a decimal portion to add
        return s + '.' + d[1]
    else:
        return s
    
    
############################################################
#
# Name: stripfmt()
#
# Purpose:  Strip all formatting from a prettified number
#
# Arguments:    n = the number to strip
#  
def stripfmt(n):
    n = string.replace(n, ',', '')  #remove commas
    n = string.replace(n, '$', '')  #remove dollar signs.
    return n
    
    
if __name__ == '__main__':
    test = 12345.346
    
    print 'number           -->       ' + `test`
    print 'commanumber(test)-->       ' + commanumber(test)
    print 'commanumber(test,1)-->     ' + commanumber(test,1)
    print 'commanumber(test,2)-->     ' + commanumber(test,2)
    print 'commanumber(test,3)-->     ' + commanumber(test,3)
    print 'commanumber(test,4)-->     ' + commanumber(test,4)

From sholden at holdenweb.com  Wed Feb 28 20:43:22 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 28 Feb 2001 14:43:22 -0500
Subject: formatting numbers for output
References: <9sfv8t45ve3dpql39rmp7082gibbicf56b@4ax.com>
Message-ID: <eBcn6.5973$zL.267891@e420r-atl1.usenetserver.com>

"Fernando Rodr?guez" <spamers at must.die> wrote in message
news:9sfv8t45ve3dpql39rmp7082gibbicf56b at 4ax.com...
> Hi!
>
> How can I format a number so it pretty prints:
>
> 15,456.6 instead of 15456.6 ?
>
> TIA
>
Take a look at the locale module.

regards
 Steve





From glen.mettler at home.com  Sun Feb 18 15:58:17 2001
From: glen.mettler at home.com (glen mettler)
Date: Sun, 18 Feb 2001 14:58:17 GMT
Subject: New Guy question
Message-ID: <1103_982505663@cc462845-a>

I just downloaded Python.  I am trying to start the windows version and nothing happens.  The EXE files available are PYTHON.EXE, 
PYTHONW.EXE and W9XPOPEN.EXE.  

PYTHON.EXE  gives me a DOS window.  I want the windows version.  What should I do?

Thanks,

Glen




From fredrik at pythonware.com  Sun Feb 18 16:08:44 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 15:08:44 GMT
Subject: New Guy question
References: <1103_982505663@cc462845-a>
Message-ID: <0ARj6.18119$AH6.2476545@newsc.telia.net>

"glen mettler" wrote:
> PYTHON.EXE  gives me a DOS window.  I want the
> windows version.  What should I do?

python.exe is the *console* version.

if you, for some reason, think that a program using the Windows
console cannot possible be a Windows program, try IDLE, Python-
Win, or any other Python IDE (check google for pointers).

Cheers /F




From glen.mettler at home.com  Sun Feb 18 16:16:31 2001
From: glen.mettler at home.com (glen mettler)
Date: Sun, 18 Feb 2001 15:16:31 GMT
Subject: New guy question
Message-ID: <1104_982506707@cc462845-a>

I just downloaded Python.  I am trying to start the windows version and nothing happens.  The EXE files available are PYTHON.EXE, 
PYTHONW.EXE and W9XPOPEN.EXE.  
PYTHON.EXE  gives me a DOS window.  I want the windows version.  What should I do?
Thanks,

Glen






From glenfant at equod.com.nospam  Sun Feb 18 17:13:50 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Sun, 18 Feb 2001 17:13:50 +0100
Subject: New guy question
References: <1104_982506707@cc462845-a>
Message-ID: <96os0e$84i$1@reader1.imaginet.fr>

What distro (BeOpen, Activestate, Pythonware) ?
What version (1.5.2, 2.0) ?

AFAIK, after installing python 2.0 from the official site www.python.org you
should have IDLE in the Python group when clicking "start>program".
You can have a free 100% Windows IDE from Activestate (www.activestate.com).
python.exe is the interactive Python console.
pythonw.exe is the run-time for Windows.
With this, you can run python scripts you write with any text editor.

"glen mettler" <glen.mettler at home.com> a ?crit dans le message news:
1104_982506707 at cc462845-a...
> I just downloaded Python.  I am trying to start the windows version and
nothing happens.  The EXE files available are PYTHON.EXE,
> PYTHONW.EXE and W9XPOPEN.EXE.
> PYTHON.EXE  gives me a DOS window.  I want the windows version.  What
should I do?
> Thanks,
>
> Glen
>
>
>
>



From abouf066 at aix2.uottawa.ca  Sun Feb 18 16:18:17 2001
From: abouf066 at aix2.uottawa.ca (abouf066 at aix2.uottawa.ca)
Date: Sun, 18 Feb 2001 10:18:17 -0500
Subject: What IPC accessible from Python
Message-ID: <Pine.A41.3.96.1010218101440.21000A-100000@aix2.uottawa.ca>

Greetings, 

	I need to spawn a process within a python application and I need
to have some communication between the child and the parent. Sockets are
'sort of heavy' for my application. I was thinking to use a shared memory
or a pipe but I did not find any documentation (example) to understand how
this would be used.

Any pointers to this matter would be greatly appreciated. 


Thanks a lot

_____________________________________________________________
							
					LaBoufarikoise




From erno-news at erno.iki.fi  Sun Feb 18 17:56:18 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 18 Feb 2001 18:56:18 +0200
Subject: What IPC accessible from Python
References: <Pine.A41.3.96.1010218101440.21000A-100000@aix2.uottawa.ca>
Message-ID: <kuhf1sszf1.fsf@lasipalatsi.fi>

In article
<Pine.A41.3.96.1010218101440.21000A-100000 at aix2.uottawa.ca>,
<abouf066 at aix2.uottawa.ca> writes:

| Greetings, 
| 	I need to spawn a process within a python application and I need
| to have some communication between the child and the parent. Sockets are
| 'sort of heavy' for my application. I was thinking to use a shared memory
| or a pipe but I did not find any documentation (example) to understand how
| this would be used.

os.pipe() works just like pipe(2) in the unix c syscall api.

no sysv shared memory (at least in the standard library, there
may be 3rd party modules), but the mmap shared memory may work
(with the mmap module in python >= 2.0).

shared memory feels a lot heavier than pipes or unix domain sockets to
me...

  -- erno


From claird at starbase.neosoft.com  Mon Feb 19 15:17:48 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Feb 2001 08:17:48 -0600
Subject: What IPC accessible from Python
References: <Pine.A41.3.96.1010218101440.21000A-100000@aix2.uottawa.ca> <kuhf1sszf1.fsf@lasipalatsi.fi>
Message-ID: <010D92FB2EBD6C65.87C5B1D9E9635DEA.37DEA6894CD83C3A@lp.airnews.net>

In article <kuhf1sszf1.fsf at lasipalatsi.fi>,
Erno Kuusela  <erno-news at erno.iki.fi> wrote:
>In article
><Pine.A41.3.96.1010218101440.21000A-100000 at aix2.uottawa.ca>,
><abouf066 at aix2.uottawa.ca> writes:
>
>| Greetings, 
>| 	I need to spawn a process within a python application and I need
>| to have some communication between the child and the parent. Sockets are
>| 'sort of heavy' for my application. I was thinking to use a shared memory
>| or a pipe but I did not find any documentation (example) to understand how
>| this would be used.
>
>os.pipe() works just like pipe(2) in the unix c syscall api.
>
>no sysv shared memory (at least in the standard library, there
>may be 3rd party modules), but the mmap shared memory may work
>(with the mmap module in python >= 2.0).
>
>shared memory feels a lot heavier than pipes or unix domain sockets to
>me...
>
>  -- erno

Seconded.  That is, shared memory had it all over
sockets ten years ago, but a lot of work has been
done since then to improve implementations, and
socket programming is what I generally recommend
nowadays.  There are even stacks that use shmem
for loopback.

There's nothing particularly wrong with shmem; it's
just not exposed or exercised as it once was.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From gst at sysfrog.org  Sun Feb 18 19:38:35 2001
From: gst at sysfrog.org (Guenther Starnberger)
Date: 18 Feb 2001 19:38:35 +0100
Subject: What IPC accessible from Python
References: <Pine.A41.3.96.1010218101440.21000A-100000@aix2.uottawa.ca>
Message-ID: <87k86nsuok.fsf@nirvana.sysfrog.org>

<abouf066 at aix2.uottawa.ca> writes:

> I need to spawn a process within a python application and I need >
> to have some communication between the child and the parent. Sockets
> are > 'sort of heavy' for my application. I was thinking to use a
> shared memory > or a pipe but I did not find any documentation
> (example) to understand how > this would be used.

try the popen2 module.

import popen2
foo=popen2.Popen3(command,1,1024)
foo.tochild.write("hello")

/gst

-- 
sysfrog.org


From mikholbor at mail.ru  Sun Feb 18 16:32:57 2001
From: mikholbor at mail.ru (Oleg B. Mikhaylov)
Date: Sun, 18 Feb 2001 17:32:57 +0200
Subject: Hey all, make easy money with me
Message-ID: <96or0o$1121$1@News.TM.Odessa.UA>

Hey,

I just joined a FREE site, TargetShop.com, that lets you earn $10,000 just
for referring
your friends!

Go to TargetShop.com
http://www.shopping4points.com/Default.asp?refId=2133869 and
register for Membership using my Referral Number 2133869. If you sign up,
before May 1st,
I will be eligible to receive $50 payable in cash or applied to the purchase
of TargetShop.com
stock (if and when they complete an IPO, subject to the Terms and Conditions
of the Referral
Program).

After you register, you will be able to earn easy money just by sending
email to refer
your friends. Each time someone you referred, in turn, refers someone, you
will make money!
We can both make thousands of dollars in no time at all! As TargetShop.com
Members, we are
entitled to discounts at some of the Web's best e-commerce sites using the
TargetPoints we
get when we register and take advantage of site features, like the
Playground.

So, go sign up, earn big bucks, and enjoy your Membership!
http://www.shopping4points.com/Default.asp?refId=2133869.

Talk to you later,

Oleg




From gleonid at actcom.co.il  Sun Feb 18 17:10:56 2001
From: gleonid at actcom.co.il (Leonid Gluhovsky)
Date: Sun, 18 Feb 2001 16:10:56 GMT
Subject: [Q] how to protect python program from decompilation
Message-ID: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>

Hello.  

The company I work for has a program written in Python, and wants to
ship it to customers in such a way that even if the customers are Python
gurus they will not be able to decompile Python bytecode and arrive at
readable source.

Is it possible?  How to do it?

Many thanks in advance,

leonid




From moshez at zadka.site.co.il  Sun Feb 18 17:35:48 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 18 Feb 2001 18:35:48 +0200 (IST)
Subject: [Q] how to protect python program from decompilation
In-Reply-To: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
Message-ID: <20010218163548.6C57EA840@darjeeling.zadka.site.co.il>

On Sun, 18 Feb 2001 16:10:56 GMT, Leonid Gluhovsky <gleonid at actcom.co.il> wrote:
> The company I work for has a program written in Python, and wants to
> ship it to customers in such a way that even if the customers are Python
> gurus they will not be able to decompile Python bytecode and arrive at
> readable source.
> 
> Is it possible?  How to do it?

No it's not. No way will withstand a true guru. Note that the same goes
for C/C++/Java. Use a license.

sick-of-this-question-ly y'rs, Z.
-- 
"I'll be ex-DPL soon anyway so I'm        |LUKE: Is Perl better than Python?
looking for someplace else to grab power."|YODA: No...no... no. Quicker,
   -- Wichert Akkerman (on debian-private)|      easier, more seductive.
For public key, finger moshez at debian.org  |http://www.{python,debian,gnu}.org



From erno-news at erno.iki.fi  Sun Feb 18 17:40:09 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 18 Feb 2001 18:40:09 +0200
Subject: [Q] how to protect python program from decompilation
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
Message-ID: <kulmr4t05y.fsf@lasipalatsi.fi>

In article <newscache$8sny8g$bdk$1 at lnews.actcom.co.il>, Leonid
Gluhovsky <gleonid at actcom.co.il> writes:

| Hello.  
| The company I work for has a program written in Python, and wants to
| ship it to customers in such a way that even if the customers are Python
| gurus they will not be able to decompile Python bytecode and arrive at
| readable source.

| Is it possible?  How to do it?

it's not possible (for python or for any other language).

  -- erno


From aahz at panix.com  Sun Feb 18 18:09:05 2001
From: aahz at panix.com (Aahz Maruch)
Date: 18 Feb 2001 09:09:05 -0800
Subject: [Q] how to protect python program from decompilation
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
Message-ID: <96ovjh$4nt$1@panix3.panix.com>

In article <newscache$8sny8g$bdk$1 at lnews.actcom.co.il>,
Leonid Gluhovsky  <gleonid at actcom.co.il> wrote:
>
>The company I work for has a program written in Python, and wants to
>ship it to customers in such a way that even if the customers are Python
>gurus they will not be able to decompile Python bytecode and arrive at
>readable source.
>
>Is it possible?  How to do it?

The only way to do it is to physically secure the code in a way that's
inaccessible to the user's computer.  For example, one could ship a card
with a CPU; realistically, though, the only serious way to do it with
current technology is to force your users to access your web site, where
the program resides.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From gleonid at actcom.co.il  Sun Feb 18 18:18:54 2001
From: gleonid at actcom.co.il (Leonid Gluhovsky)
Date: Sun, 18 Feb 2001 17:18:54 GMT
Subject: [Q] how to protect python program from decompilation
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il> <mailman.982514238.21934.python-list@python.org>
Message-ID: <newscache$ixqy8g$mwo$1@lnews.actcom.co.il>

Moshe Zadka <moshez at zadka.site.co.il> wrote:

> sick-of-this-question-ly y'rs, Z.

I honestly tried to locate tips on this problem before posting :)
Could you please let me know where I can find prior discussion of this?

> No it's not. No way will withstand a true guru. Note that the same goes
> for C/C++/Java.  

OK.

> Use a license.

I am not in a position to decide on this matter.  Let's say I am looking
for a way to make it prohibitively hard for the customer to decompile
the byte code.  I agree that complete protection is impossible. 

Right now we are considering the following scheme: build a Python
with opcodes in Include/opcode.h reshuffled, and without the dis module
in it; use this interpreter to byte compile our code; use its Tools/freeze
to package our program into executable and ship it to customers. 

What are the holes in this approach?  What is a better approach?

Thank you,

leonid




From moshez at zadka.site.co.il  Sun Feb 18 18:45:11 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 18 Feb 2001 19:45:11 +0200 (IST)
Subject: [Q] how to protect python program from decompilation
In-Reply-To: <newscache$ixqy8g$mwo$1@lnews.actcom.co.il>
References: <newscache$ixqy8g$mwo$1@lnews.actcom.co.il>, <newscache$8sny8g$bdk$1@lnews.actcom.co.il> <mailman.982514238.21934.python-list@python.org>
Message-ID: <20010218174511.3BD21A840@darjeeling.zadka.site.co.il>

On Sun, 18 Feb 2001 17:18:54 GMT, Leonid Gluhovsky <gleonid at actcom.co.il> wrote:
 
> I am not in a position to decide on this matter.  Let's say I am looking
> for a way to make it prohibitively hard for the customer to decompile
> the byte code.  I agree that complete protection is impossible. 

If management has asked you to find a problem, they should be able to
hear that the problem has no solution. Any protection scheme can be broken,
and most (hard!) ones can be broken by someone in less then a day. Consider
whether a day of someone else's time is worth more then (at least) an
hour of your time.

> Right now we are considering the following scheme: build a Python
> with opcodes in Include/opcode.h reshuffled, and without the dis module
> in it; use this interpreter to byte compile our code; use its Tools/freeze
> to package our program into executable and ship it to customers. 
> 
> What are the holes in this approach?  What is a better approach?

That an idiot with a debugger can see what each opcode does enough to remap
the opcodes, and then read from your program the freezed strings using
a tool I can probably write in Python in 15 minutes.
So, if I were to break your program (and I'm not -- I'm using only software
I have the source for at home), I estimate it would take me about one hour.
It's probably as good approach as any, but it's also probably as bad
as any too.

-- 
"I'll be ex-DPL soon anyway so I'm        |LUKE: Is Perl better than Python?
looking for someplace else to grab power."|YODA: No...no... no. Quicker,
   -- Wichert Akkerman (on debian-private)|      easier, more seductive.
For public key, finger moshez at debian.org  |http://www.{python,debian,gnu}.org



From mwh21 at cam.ac.uk  Sun Feb 18 18:52:07 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Feb 2001 17:52:07 +0000
Subject: [Q] how to protect python program from decompilation
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il> <mailman.982514238.21934.python-list@python.org> <newscache$ixqy8g$mwo$1@lnews.actcom.co.il>
Message-ID: <m37l2nev5k.fsf@atrus.jesus.cam.ac.uk>

Leonid Gluhovsky <gleonid at actcom.co.il> writes:

> Right now we are considering the following scheme: build a Python
> with opcodes in Include/opcode.h reshuffled, and without the dis module
> in it; use this interpreter to byte compile our code; use its Tools/freeze
> to package our program into executable and ship it to customers. 
> 
> What are the holes in this approach?  What is a better approach?

Sounds like a reasonable (effort on your part)/(effort on their part
tradeoff).  It would still probably be crackable (by statistical
analysis of byteocde of decompiling the assembly of your ceval.c), but
it would be a lot of work.

Cheers,
M.

-- 
  I also feel it essential to note, [...], that Description Logics,
  non-Monotonic Logics, Default Logics and Circumscription Logics can
  all collectively go suck a cow. Thank you.
              -- http://advogato.org/person/Johnath/diary.html?start=4


From skip at mojam.com  Sun Feb 18 19:18:50 2001
From: skip at mojam.com (Skip Montanaro)
Date: Sun, 18 Feb 2001 12:18:50 -0600 (CST)
Subject: [Q] how to protect python program from decompilation
In-Reply-To: <newscache$ixqy8g$mwo$1@lnews.actcom.co.il>
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
	<mailman.982514238.21934.python-list@python.org>
	<newscache$ixqy8g$mwo$1@lnews.actcom.co.il>
Message-ID: <14992.4618.346376.319196@beluga.mojam.com>


    Leonid> Right now we are considering the following scheme: build a
    Leonid> Python with opcodes in Include/opcode.h reshuffled, and without
    Leonid> the dis module in it; use this interpreter to byte compile our
    Leonid> code; use its Tools/freeze to package our program into
    Leonid> executable and ship it to customers.

Well, a determined person could simply locate the real dis, and start (using
trial and error, perhaps), to decipher your byte code, adjusting dis.py as
they go along.  For example, you can't simply shuffle all your byte codes,
there are two cut points, HAVE_ARGUMENT and EXTENDED_ARG.  Below
HAVE_ARGUMENT, none of the opcodes take arguments, so only opcodes <
HAVE_ARGUMENT could be shuffled amongst each other.  Another clue, all
functions are compiled with RETURN None at the end.  Without modifying the
compiler or running an optimizer of some sort, those will always be there.
Those are just starter clues that came off the top of my head.  I'm sure
there are other clues in the bytecode.  In addition, a determined person
could step through eval_code2 to see what it does.  Even at the assembler
level if necessary.

-- 
Skip Montanaro (skip at mojam.com)
Support Mojam & Musi-Cal: http://www.musi-cal.com/sponsor.shtml
(847)971-7098



From tim.one at home.com  Sun Feb 18 19:42:04 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 18 Feb 2001 13:42:04 -0500
Subject: [Q] how to protect python program from decompilation
In-Reply-To: <newscache$ixqy8g$mwo$1@lnews.actcom.co.il>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEDGJAAA.tim.one@home.com>

[Leonid Gluhovsky]
> ...
> Right now we are considering the following scheme: build a Python
> with opcodes in Include/opcode.h reshuffled, and without the dis module
> in it; use this interpreter to byte compile our code; use its Tools/freeze
> to package our program into executable and ship it to customers.
>
> What are the holes in this approach?

The different sections of the Python eval loop are easy to recognize in
machine language.  It wouldn't take a Python mini-guru more than an hour to
reverse-engineer your permutation by just watching the eval loop in a
machine-language debugger, noting which opcodes jump to which sections of
the eval loop.  In fact, they only have to do that for *one* opcode.  Then
they know where in memory the eval loop case-stmt jump table (generated by
the C compiler) lives, and the entire permutation can be read off directly
from that.  Then they can alter the std dis.py accordingly to do an exact
symbolic disassembly of your whole code base.

> What is a better approach?

If there were one, I expect Microsoft would do that instead of just
prohibiting reverse-engineering in their licenses.  It's not Python gurus
you have to worry about so much, it's machine-language gurus.  That's why
this has little to do with Python.  You can't hide anything from them,
unless (as Aahz suggested) you ship your own CPU card that doesn't allow
tracing instructions.

You can make it more *unpleasant* by running your code through a name
obfuscator first, e.g. systematically replace all vrbl instances of "i" by a
fixed 20-character randomly generated identifier ("AAAA75WRWLKJLKJ3lS9Q"),
and so on.  Certain branches of the US Govt do that before sending in
compiler bug reports, as if nobody can recognize an FFT from its structure
alone <0.9 wink>.

It's much better to assume your code *will* be reverse-engineered (simply
because it will be).  Then the focus shifts to dreaming up ways to prove
infringement of your license.  An effective way is (no kidding) to build in
subtle bugs.  If a competitor develops the same set of subtle bugs later,
they're holding a smoking gun.

for-the-same-reason-when-i-was-growing-up-the-regional-map-showed-
    a-river-in-my-neighborhood-that-never-existed-ly y'rs  - tim




From tim.one at home.com  Sun Feb 18 18:59:17 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 18 Feb 2001 12:59:17 -0500
Subject: [Q] how to protect python program from decompilation
In-Reply-To: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEDEJAAA.tim.one@home.com>

[Leonid Gluhovsky]
> The company I work for has a program written in Python, and wants to
> ship it to customers in such a way that even if the customers are Python
> gurus they will not be able to decompile Python bytecode and arrive at
> readable source.
>
> Is it possible?

Not a chance, but that's got little to do with Python.  If you let me run
any code whatsoever on my box, I can get a trace of the CPU instructions
executed, and that suffices for reasonable reconstruction of the executed
paths in the source code given enough effort.  If I know the code was
originally written in Python, or C, or C++, or Java, ..., it makes it that
much easier (because any given compiler is pretty consistent in the way it
generates code for various language constructs), but it isn't essential.
For example, it's easy to reconstruct your hierarchy of subroutines just by
noting when hardware call and return sequences occur, and from and to which
addresses.

The ironic thing is that the more obscure you try to make it, the more
crackers will consider it a worthy challenge to reverse-engineer it.  Hiding
a thing in plain view can be more effective in the end!

> How to do it?

Look at it this way:  a computer CPU only speaks French.  You have some
technical text written in English, and your mgmt wants you to translate it
into French in such a way that nobody can get back an approximation to the
English you started with.  Then they don't understand the technical issues
(perhaps they're under the illusion that *only* a CPU can understand
French), or they're on powerful drugs, or both.  This is an exact analogy,
because French is such a low-level language <wink>.

any-language-without-"potato"-is-low-level-by-american-definition-ly
    y'rs  - tim




From gst at sysfrog.org  Sun Feb 18 19:33:47 2001
From: gst at sysfrog.org (Guenther Starnberger)
Date: 18 Feb 2001 19:33:47 +0100
Subject: [Q] how to protect python program from decompilation
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
Message-ID: <87ofvzsuwk.fsf@nirvana.sysfrog.org>

Leonid Gluhovsky <gleonid at actcom.co.il> writes:

> The company I work for has a program written in Python, and wants to
> ship it to customers in such a way that even if the customers are
> Python gurus they will not be able to decompile Python bytecode and
> arrive at readable source.

you can try to convert your python modules with Python2C into c code
and distribute the compiled libraries.

http://www.mudlib.org/~rassilon/p2c/

cu
/gst

-- 
sysfrog.org


From aahz at panix.com  Sun Feb 18 20:43:49 2001
From: aahz at panix.com (Aahz Maruch)
Date: 18 Feb 2001 11:43:49 -0800
Subject: [Q] how to protect python program from decompilation
References: <mailman.982521799.21630.python-list@python.org>
Message-ID: <96p8ll$lq5$1@panix2.panix.com>

In article <mailman.982521799.21630.python-list at python.org>,
Tim Peters <tim.one at home.com> wrote:
>
>If there were one, I expect Microsoft would do that instead of just
>prohibiting reverse-engineering in their licenses.  It's not Python gurus
>you have to worry about so much, it's machine-language gurus.  That's why
>this has little to do with Python.  You can't hide anything from them,
>unless (as Aahz suggested) you ship your own CPU card that doesn't allow
>tracing instructions.

Just to be clear, this solution only protects you from software hackers.
Hardware hackers can still break just about anything.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From moshez at zadka.site.co.il  Sun Feb 18 21:20:56 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun, 18 Feb 2001 22:20:56 +0200 (IST)
Subject: [Q] how to protect python program from decompilation
In-Reply-To: <96p8ll$lq5$1@panix2.panix.com>
References: <96p8ll$lq5$1@panix2.panix.com>, <mailman.982521799.21630.python-list@python.org>
Message-ID: <20010218202056.66C66A840@darjeeling.zadka.site.co.il>

On 18 Feb 2001 11:43:49 -0800, aahz at panix.com (Aahz Maruch) wrote:
> Just to be clear, this solution only protects you from software hackers.
> Hardware hackers can still break just about anything.

Well, hardware can be much nastier. Have something which destroys itself
when opened forcefully. Not really beyond the state of art today to do
such things, but it is costly. Of course, *really* good hardware hackers
can break that too, but I don't think there are too many of those.
OTOH, script-kiddies with debuggers and too much free time are a dime
a dozen.
-- 
"I'll be ex-DPL soon anyway so I'm        |LUKE: Is Perl better than Python?
looking for someplace else to grab power."|YODA: No...no... no. Quicker,
   -- Wichert Akkerman (on debian-private)|      easier, more seductive.
For public key, finger moshez at debian.org  |http://www.{python,debian,gnu}.org



From cfelling at iae.nl  Sun Feb 18 21:02:40 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 18 Feb 2001 21:02:40 +0100
Subject: [Q] how to protect python program from decompilation
References: <mailman.982521799.21630.python-list@python.org>
Message-ID: <96p9p0$4g1$1@animus.fel.iae.nl>

Tim Peters <tim.one at home.com> wrote:
...
> It's much better to assume your code *will* be reverse-engineered (simply
> because it will be).  Then the focus shifts to dreaming up ways to prove
> infringement of your license.  An effective way is (no kidding) to build in
> subtle bugs.  If a competitor develops the same set of subtle bugs later,
> they're holding a smoking gun.

Oh, that's why closed source software is so buggy compared to open
software, they're merely protecting their code-base:)

> for-the-same-reason-when-i-was-growing-up-the-regional-map-showed-
>     a-river-in-my-neighborhood-that-never-existed-ly y'rs  - tim

The same holds for dictionaries.

-- 
groetjes, carel


From gleonid at actcom.co.il  Mon Feb 19 12:51:30 2001
From: gleonid at actcom.co.il (Leonid Gluhovsky)
Date: Mon, 19 Feb 2001 11:51:30 GMT
Subject: [Q] how to protect python program from decompilation
References: <96p8ll$lq5$1@panix2.panix.com> <mailman.982521799.21630.python-list@python.org> <mailman.982527740.11857.python-list@python.org>
Message-ID: <newscache$yf609g$myo$1@lnews.actcom.co.il>

Many thanks for all the responses!  They are most enlightening and useful.

leonid



From rturpin at do.not.use  Mon Feb 19 20:04:25 2001
From: rturpin at do.not.use (Russell Turpin)
Date: Mon, 19 Feb 2001 13:04:25 -0600
Subject: [Q] how to protect python program from decompilation -- a better 
 idea
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il>
Message-ID: <3A916E39.DE870D93@do.not.use>

An optimizing JIT byte-code compiler would do a far better 
job of hiding Python source than any of the ideas presented
so far. The resulting executable would be no easier to 
decompile than a program made with C. The more optimization 
the compiler does, the harder it would be to decompile the
results. Ideally, the compiler would maximally execute the 
Python program before freezing it into an executable. 
("Maximal," in this sense, means performing all execution 
that does NOT depend on outside data or stimulus.)

And think of the performance win! Leonid, I think your company 
should get started on this right away.

Regards,
Russell


From not.this at seebelow.org  Tue Feb 20 04:52:53 2001
From: not.this at seebelow.org (Grant Griffin)
Date: Mon, 19 Feb 2001 21:52:53 -0600
Subject: [Q] how to protect python program from decompilation
References: <mailman.982519221.9789.python-list@python.org>
Message-ID: <3A91EA15.571F0A32@seebelow.org>

Tim Peters wrote:
> 
> [Leonid Gluhovsky]
> > The company I work for has a program written in Python, and wants to
> > ship it to customers in such a way that even if the customers are Python
> > gurus they will not be able to decompile Python bytecode and arrive at
> > readable source.
> >
> > Is it possible?
> 
> Not a chance, but that's got little to do with Python.  If you let me run
> any code whatsoever on my box, I can get a trace of the CPU instructions
> executed, and that suffices for reasonable reconstruction of the executed
> paths in the source code given enough effort.

That last part is the key: "given enough effort".  From a strictly
economic point of view, you want to make them spend more money
decompiling it than they would spend writing it themselves from
scratch.  But bored teenagers aside (who defy all economic theory), this
is virtually always the case: the economic value of most software
doesn't derive from any special "trade secrets" it embodies; it's
usually pretty easy to figure out basically how it works (Perl's parser
notwithstanding <wink>).  Instead, the economic value of most software
derives just from the "hard work" it took to create the (finished)
software package--including not just "design and coding" but also (and
perhaps more significantly) things like testing, documentation,
packaging, and marketing.

Although there isn't much actual "security" in any sort of compilation
process, I think it _is_ highly successful in "keeping honest people
honest".  In this case, the Python-to-exe converters probably have that
effect.  In any event, unless one has created a software gizmo whose
algorithms are so profoundly new and amazing that nobody can possibly
figure out how it basically works just by seeing it in action, there's
not much point in worrying about the limitations of any sort of
compilation-based "security".

open-source-software-carries-this-to-its-logical-conclusion-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From zessin at decus.de  Sun Feb 18 17:23:26 2001
From: zessin at decus.de (zessin at decus.de)
Date: Sun, 18 Feb 2001 17:23:26 +0100
Subject: OpenVMS import (was Re: Windows/Cygwin/MacOSX import (was RE:
         python-dev summary, 2001-02-01 - 2001-02-15)
Message-ID: <009F7D57.F76B21F7.2@decus.de>

Cameron Laird wrote:
>In article <mailman.982310784.27748.python-list at python.org>,
>Tim Peters <tim.one at home.com> wrote:
>>[Michael Hudson]
>>> ...
>>>    * Imports on case-insensitive file systems *
>>>
>>>  There was quite some discussion about how to handle imports on a
>>>  case-insensitive file system (eg. on Windows).  I didn't follow the
>>>  details, but Tim Peters is on the case (sorry), so I'm confident it
>>>  will get sorted out.
>>
>>You can be sure the whitespace will be consistent, anyway <wink>.
>			.
>			.
>			.
>>them is ugly.  We're already supporting (and will continue to support)
>>PYTHONCASEOK for their benefit, but they don't deserve multiple hacks in
>>2001.
>>
>>Flame at will.
>>
>>or-flame-at-tim-your-choice-ly y'rs  - tim
>
>1.  Thanks.  Along with all the other benefits, I find
>    this explanation FAR more entertaining than anything
>    network television broadcasts (although nearly as
>    tendentious as "The West Wing").
>2.  I hope a few OS/400 and OpenVMS refugees convert and
>    walk through the door soon.  *That* would make for a
>    nice dose of fun.

Let's see if I can explain the OpenVMS part.
I'll do so by walking over Tim's text. (I'll step carefully over it.
 I don't intend to destroy it, Tim ;-)

] Here's the scoop:  file systems vary across platforms in whether or not they
] preserve the case of filenames, and in whether or not the platform C library
] file-opening functions do or don't insist on case-sensitive matches:
]
]
]                        case-preserving   case-destroying
]                     +-------------------+------------------+
]    case-sensitive   | most Unix flavors | brrrrrrrrrr      |
]                     +-------------------+------------------+
]    case-insensitive | Windows           | some unfortunate |
]                     | MacOSX HFS+       | network schemes  |
]                     | Cygwin            |                  |
                      |                   | OpenVMS          |
]                     +-------------------+------------------+

Phew.
I'm glad we're only 'unfortunate' and not in the 'brrrrrrrrrr' section ;-)

] In the upper left box, if you create "fiLe" it's stored as "fiLe", and only
] open("fiLe") will open it (open("file") will not, nor will the 14 other
] variations on that theme).

] In the lower right box, if you create "fiLe", there's no telling what it's
] stored as-- but most likely as "FILE" --and any of the 16 obvious variations
] on open("FilE") will open it.

>>> f = open ('fiLe', 'w')

$ directory f*

Directory DSA3:[PYTHON.PYTHON-2_1A2CVS.VMS]

FILE.;1

>>> f = open ('filE', 'r')
>>> f
<open file 'filE', mode 'r' at 3E03B8>
>>>

This is on the default file system (ODS-2). Only very recent versions of
OpenVMS Alpha (V7.2 and up) support the ODS-5 FS that has Windows-like
behaviour (case-preserving,case-insensitive), but many sites don't use it
(yet). Also, there are many older versions running in the field that don't
get upgraded any time soon.

] The lower left box is a mix:  creating "fiLe" stores "fiLe" in the platform
] directory, but you don't have to match case when opening it; any of the 16
] obvious variations on open("FILe") work.

Same here.

] What's proposed is to change the semantics of Python "import" statements,
] and there *only* in the lower left box.
]
] Support for MaxOSX HFS+, and for Cygwin, is new in 2.1, so nothing is
] changing there.  What's changing is Windows behavior.  Here are the current
] rules for import on Windows:
] 
] 1. Despite that the filesystem is case-insensitive, Python insists on
]    a case-sensitive match.  But not in the way the upper left box works:
]    if you have two files, FiLe.py and file.py on sys.path, and do
] 
]        import file
] 
]    then if Python finds FiLe.py first, it raises a NameError.  It does
]    *not* go on to find file.py; indeed, it's impossible to import any
]    but the first case-insensitive match on sys.path, and then only if
]    case matches exactly in the first case-insensitive match.

For OpenVMS I have just changed 'import.c':
  MatchFilename() and some code around it is not executed.

] 2. An ugly exception:  if the first case-insensitive match on sys.path
]    is for a file whose name is entirely in upper case (FILE.PY or
]    FILE.PYC or FILE.PYO), then the import silently grabs that, no matter
]    what mixture of case was used in the import statement.  This is
]    apparently to cater to miserable old filesystems that really fit in
]    the lower right box.  But this exception is unique to Windows, for
]    reasons that may or may not exist <frown>.

I guess that is Windows-specific code?
Something to do with 'allcaps8x3()'?

] 3. And another exception:  if the envar PYTHONCASEOK exists, Python
]    silently grabs the first case-insensitive match of any kind.

The check is in 'check_case()', but there is no OpenVMS implementation (yet).

] So these Windows rules are pretty complicated, and neither match the Unix
] rules nor provide semantics natural for the native filesystem.  That makes
] them hard to explain to Unix *or* Windows users.  Nevertheless, they've
] worked fine for years, and in isolation there's no compelling reason to
] change them.

] However, that was before the MacOSX HFS+ and Cygwin ports arrived.  They
] also have case-preserving case-insensitive filesystems, but the people doing
] the ports despised the Windows rules.  Indeed, a patch to make HFS+ act like
] Unix for imports got past a reviewer and into the code base, which
] incidentally made Cygwin also act like Unix (but this met the unbounded
] approval of the Cygwin folks, so they sure didn't complain -- they had
] patches of their own pending to do this, but the reviewer for those balked).
] 
] At a higher level, we want to keep Python consistent, and I in particular
] want Python to do the same thing on *all* platforms with case-preserving
] case-insensitive filesystems.  Guido too, but he's so sick of this argument
] don't ask him to confirm that <0.9 wink>.

What are you thinking about the 'unfortunate / OpenVMS' group ?
Hey, it could be worse, could be 'brrrrrrrrrr'...

] The proposed new semantics for the lower left box:
] 
] A. If the PYTHONCASEOK envar exists, same as before:  silently accept
]    the first case-insensitive match of any kind; raise ImportError if
]    none found.
] 
] B. Else search sys.path for the first case-sensitive match; raise
]    ImportError if none found.
] 
] #B is the same rule as is used on Unix, so this will improve cross-platform
] portability.  That's good.  #B is also the rule the Mac and Cygwin folks
] want (and wanted enough to implement themselves, multiple times, which is a
] powerful argument in PythonLand).  It can't cause any existing
] non-exceptional Windows import to fail, because any existing non-exceptional
] Windows import finds a case-sensitive match first in the path -- and it
] still will.  An exceptional Windows import currently blows up with a
] NameError or ImportError, in which latter case it still will, or in which
] former case will continue searching, and either succeed or blow up with an
] ImportError.
] 
] #A is needed to cater to case-destroying filesystems mounted on Windows, and
] *may* also be used by people so enamored of "natural" Windows behavior that
] they're willing to set an envar to get it.  That's their problem <wink>.  I
] don't intend to implement #A for Unix too, but that's just because I'm not
] clear on how I *could* do so efficiently (I'm not going to slow imports
] under Unix just for theoretical purity).
] 
] The potential damage is here:  #2 (matching on ALLCAPS.PY) is proposed to be
] dropped.  Case-destroying filesystems are a vanishing breed, and support for
] them is ugly.  We're already supporting (and will continue to support)
] PYTHONCASEOK for their benefit, but they don't deserve multiple hacks in
] 2001.

Would using unique names be an acceptable workaround?

] Flame at will.
] 
] or-flame-at-tim-your-choice-ly y'rs  - tim

No flame intended. Not at will and not at tim.

>-- 
>
>Cameron Laird <claird at NeoSoft.com>
>Business:  http://www.Phaseit.net
>Personal:  http://starbase.neosoft.com/~claird/home.html

--
Uwe Zessin



From slenhartsprint at earthlink.net  Sun Feb 18 17:39:21 2001
From: slenhartsprint at earthlink.net (Steve Lenhart)
Date: Sun, 18 Feb 2001 16:39:21 GMT
Subject: PIL Install - I will worship you
Message-ID: <ZUSj6.1711$DS3.159530@newsread1.prod.itd.earthlink.net>

Python 2.0 on Windows 2000. How do I setup the Python Image Library? It
shouldn't be this hard. I've downloaded and unzipped into a folder called
Imaging-1.1.1 in the Python folder. Now what? This isn't easy for a
beginner. I will worship you if you tell me what to do next. Thx




From syring at email.com  Sun Feb 18 18:06:38 2001
From: syring at email.com (Karl M. Syring)
Date: Sun, 18 Feb 2001 18:06:38 +0100
Subject: PIL Install - I will worship you
References: <ZUSj6.1711$DS3.159530@newsread1.prod.itd.earthlink.net>
Message-ID: <96p1fm$lrkpu$1@ID-7529.news.dfncis.de>

"Steve Lenhart" <slenhartsprint at earthlink.net> schrieb im Newsbeitrag
news:ZUSj6.1711$DS3.159530 at newsread1.prod.itd.earthlink.net...
> Python 2.0 on Windows 2000. How do I setup the Python Image Library? It
> shouldn't be this hard. I've downloaded and unzipped into a folder called
> Imaging-1.1.1 in the Python folder. Now what? This isn't easy for a
> beginner. I will worship you if you tell me what to do next. Thx

To not miss this opportunity: you can get the _imaging.dll and pil111py.zip
from Fredrik Lundh's site at
http://w1.132.telia.com/~u13208596/pil/index.htm.

Karl M. Syring




From fredrik at pythonware.com  Tue Feb 20 10:45:35 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 20 Feb 2001 09:45:35 GMT
Subject: PIL Install - I will worship you
References: <ZUSj6.1711$DS3.159530@newsread1.prod.itd.earthlink.net>
Message-ID: <31rk6.14592$Qb7.2329539@newsb.telia.net>

Steve Lenhart wrote:
> Python 2.0 on Windows 2000. How do I setup the Python Image Library? It
> shouldn't be this hard. I've downloaded and unzipped into a folder called
> Imaging-1.1.1 in the Python folder

what exactly have you downloaded and unzipped?  the
source distribution?

if you don't have a compiler, you're probably better off
downloading a binary kit:

    http://www.pythonware.com/products/pil

get the right version, unzip it into your Python installation
directory (e.g. c:\python20), and off you go...

Cheers /F




From ransen_spam_me_not at nemo.it  Sun Feb 18 17:41:39 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Sun, 18 Feb 2001 16:41:39 GMT
Subject: Done the tutorial, now which book? (Python 2)
Message-ID: <3a90ed92.1349768@news.newsguy.com>

Hello All,

Done the HTML tutorial, now is there a book? (Python 2)

I mean a book which goes through examples and excercises
and things like that? (Rather than a manual). For background
I'll tell you that I am a C++ graphics programmer.

Amazon gave me this list:

  Learning Python ~Mark Lutz, David Ascher 

  Python Programming on Win32 ~Mark Hammond, Andy Robinson 

  Programming Python ~Mark Lutz 

  Python and Tkinter Programming ~John E. Grayson 
 
  The Quick Python Book ~Daryl D. Harms, et al 
  
  Sams Teach Yourself Python in 24 Hours ~Ivan Van Laningham 
  
  Programming with PYTHON ~Tim Alton 
  
  Core Python Programming ~Wesley J. Chun 
  
  Python Developer's Handbook ~Andre Lessa 
  
  Learning to Program Using Python ~Alan Gauld 



--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From clee at gnwy100.wuh.wustl.edu  Mon Feb 19 04:23:52 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 18 Feb 2001 21:23:52 -0600
Subject: Done the tutorial, now which book? (Python 2)
References: <3a90ed92.1349768@news.newsguy.com>
Message-ID: <7uk86n73uf.fsf@gnwy100.wuh.wustl.edu>

>>>>> "Owen" == Owen F Ransen <ransen_spam_me_not at nemo.it> writes:

    Owen> Hello All, Done the HTML tutorial, now is there a book? (Python
    Owen> 2)

I haven't read all of the books on your list, but I would rank "Learning
Python" by Lutz and Ascher among the best language books I've ever read.  I
felt like it clearly conveyed python's "philosophy" and quickly gave me a
feel for how to solve problems in python.  If I were only going to go for
one book, I'd choose that.

Python Programming on Win32 is also excellent if you're using Windows,
especially if you're interested in doing COM servers, scripting Excel and
the like.  "Programming Python" by Lutz is coming out in a second
edition--make sure to get the new version if you decide to buy it.  It
seems like a more advanced text---I found the first edition helpful when
starting my first C python extension.

-C


From ransen_spam_me_not at nemo.it  Mon Feb 19 16:31:53 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Mon, 19 Feb 2001 15:31:53 GMT
Subject: Done the tutorial, now which book? (Python 2)
References: <3a90ed92.1349768@news.newsguy.com> <7uk86n73uf.fsf@gnwy100.wuh.wustl.edu>
Message-ID: <3a923bbb.4385473@news.newsguy.com>

On 18 Feb 2001 21:23:52 -0600, Christopher Lee
<clee at gnwy100.wuh.wustl.edu> wrote:

>>>>>> "Owen" == Owen F Ransen <ransen_spam_me_not at nemo.it> writes:
>
>    Owen> Hello All, Done the HTML tutorial, now is there a book? (Python
>    Owen> 2)
>I haven't read all of the books on your list, but I would rank "Learning
>Python" by Lutz and Ascher among the best language books I've ever read.
Ok.

>Python Programming on Win32 is also excellent if you're using Windows,
I'm aiming for Win and Mac so this is probably not much
use to me.


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From ws-news at gmx.at  Mon Feb 19 17:51:35 2001
From: ws-news at gmx.at (Werner Schiendl)
Date: Mon, 19 Feb 2001 17:51:35 +0100
Subject: Done the tutorial, now which book? (Python 2)
References: <3a90ed92.1349768@news.newsguy.com> <7uk86n73uf.fsf@gnwy100.wuh.wustl.edu> <3a923bbb.4385473@news.newsguy.com>
Message-ID: <982601746.924551@newsmaster-04.atnet.at>

> >I haven't read all of the books on your list, but I would rank "Learning
> >Python" by Lutz and Ascher among the best language books I've ever read.
> Ok.

I agree with that, used the book from a comparable situation than you are
probably (C and C++ programmer)
Found it very well written and it does not describe each single option that
can be taken from the docs


>
> >Python Programming on Win32 is also excellent if you're using Windows,
> I'm aiming for Win and Mac so this is probably not much
> use to me.
>
>

I think this on is rather COM loaded, the exact reason while I am going to
read it soon...

hth
werner




From mertz at gnosis.cx  Sun Feb 18 19:25:23 2001
From: mertz at gnosis.cx (Dr. David Mertz)
Date: Sun, 18 Feb 2001 13:25:23 -0500
Subject: Book reviews
Message-ID: <TOBk6kKkXYXF092yn@bellatlantic.net>                         

I have written, for IBM developerWorks, a comparative review of most of
the books below.  Actually, I wrote it quite a while ago by now... and
publishers have... delays.  Look for it on IBM soon, and in the
meanwhile, feel free to read:

  http://gnosis.cx/cgi/txt2html.cgi?source=http://gnosis.cx/publish/programming/charming_python_12.txt

Yours, David...

-------- Forwarded message --------
Newsgroups: comp.lang.python
Date: Sun, 18 Feb 2001 16:41:39 GMT
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Subject: Done the tutorial, now which book? (Python 2)

I mean a book which goes through examples and excercises
and things like that? (Rather than a manual). For background
I'll tell you that I am a C++ graphics programmer.

  Learning Python ~Mark Lutz, David Ascher
  Python Programming on Win32 ~Mark Hammond, Andy Robinson
  Programming Python ~Mark Lutz
  Python and Tkinter Programming ~John E. Grayson
  The Quick Python Book ~Daryl D. Harms, et al
  Sams Teach Yourself Python in 24 Hours ~Ivan Van Laningham
  Programming with PYTHON ~Tim Alton
  Core Python Programming ~Wesley J. Chun
  Python Developer's Handbook ~Andre Lessa
  Learning to Program Using Python ~Alan Gauld




From jdries at mail.com  Sun Feb 18 19:43:03 2001
From: jdries at mail.com (Jan Dries)
Date: Sun, 18 Feb 2001 19:43:03 +0100
Subject: Book reviews
References: <TOBk6kKkXYXF092yn@bellatlantic.net>
Message-ID: <3A9017B7.EA372E34@mail.com>


"Dr. David Mertz" wrote:
> 
> I have written, for IBM developerWorks, a comparative review of most of
> the books below.  Actually, I wrote it quite a while ago by now... 

You mean, you will write it in some distant time in the future. That is,
is the part of the world where I live, January 20001 is still some time
ahead. By that time, we will have seen the Y10K problem. Imagine that.

Regards,
Jan



From ransen_spam_me_not at nemo.it  Mon Feb 19 19:06:34 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Mon, 19 Feb 2001 18:06:34 GMT
Subject: Book reviews
References: <mailman.982521200.18821.python-list@python.org>
Message-ID: <3a933d77.4830008@news.newsguy.com>

On Sun, 18 Feb 2001 13:25:23 -0500, "Dr. David Mertz"
<mertz at gnosis.cx> wrote:

>I have written, for IBM developerWorks, a comparative review of most of
>the books below.  Actually, I wrote it quite a while ago by now... and
>publishers have... delays.  Look for it on IBM soon, and in the
>meanwhile, feel free to read:
>
>  http://gnosis.cx/cgi/txt2html.cgi?source=http://gnosis.cx/publish/programming/charming_python_12.txt
>
>Yours, David...

Thanks for your reviews. It sounds to me that 
"The Quick Python Book" is probably my best 
initial bet, given that I program already....



--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From kwoeltje at mail.mcg.edu  Tue Feb 20 00:01:11 2001
From: kwoeltje at mail.mcg.edu (Keith F. Woeltje)
Date: Mon, 19 Feb 2001 23:01:11 GMT
Subject: Book reviews
References: <mailman.982521200.18821.python-list@python.org> <3a933d77.4830008@news.newsguy.com>
Message-ID: <3A91A5F6.B1285199@mail.mcg.edu>

FYI, the second edition of "Programming Python" is supposed to be out in
March.
>KFW

"Owen F. Ransen" wrote:
> 
> On Sun, 18 Feb 2001 13:25:23 -0500, "Dr. David Mertz"
> <mertz at gnosis.cx> wrote:
> 
> >I have written, for IBM developerWorks, a comparative review of most of
> >the books below.  Actually, I wrote it quite a while ago by now... and
> >publishers have... delays.  Look for it on IBM soon, and in the
> >meanwhile, feel free to read:
> >
> >  http://gnosis.cx/cgi/txt2html.cgi?source=http://gnosis.cx/publish/programming/charming_python_12.txt
> >
> >Yours, David...
> 
> Thanks for your reviews. It sounds to me that
> "The Quick Python Book" is probably my best
> initial bet, given that I program already....
> 
> --
> Owen F. Ransen
> http://www.ransen.com/
> Home of Gliftic & Repligator Image Generators


From danielk at aracnet.com  Sun Feb 18 19:34:42 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 18 Feb 2001 10:34:42 -0800
Subject: Determining EOF character
Message-ID: <66409tk72epolackapla0egk4907k9a9hi@4ax.com>

When managing child processes (like with popen2.popen2) I have occasion to do

	instream.readlines()

to get all lines in a list rather than

	instream.readline()

which gets the next line terminated with 'os.linesep'.

On Windows the child process needs to send a ^Z in response to readlines() and
on UNIX it needs to send ^D. I don't know what this is on MAC.

What I would like to do is to tell the child process what to use as the EOF
character so that my code is platform independent. I have searched thru the
'os' and 'sys' modules for something like a 'os.eof' or 'sys.eof' with no joy.

I am currently using '#ifdef WIN32' to determine this in the child process but
I would like to move this up to the controlling Python program.

If this exists please let me know where. If not, how do I submit an enhancement
request?

Thank you,
Daniel Klein


From max at alcyone.com  Sun Feb 18 20:35:41 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 11:35:41 -0800
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com>
Message-ID: <3A90240D.384ABA3E@alcyone.com>

Daniel Klein wrote:

> When managing child processes (like with popen2.popen2) I have
> occasion to do
> 
>         instream.readlines()
> 
> to get all lines in a list rather than
> 
>         instream.readline()
> 
> which gets the next line terminated with 'os.linesep'.
> 
> On Windows the child process needs to send a ^Z in response to
> readlines() and
> on UNIX it needs to send ^D. I don't know what this is on MAC.

This isn't the way you should signal end of file.  You should signal it
the client closing his end of the pipe.  In fact, what you suggest above
may work on Windows, but it won't work on UNIX -- under UNIX, an end of
file isn't indicated with a sentinel character in the stream, but rather
with an out-of-bounds indicator (the underlying file structure has an
end-of-file flag).  As I recall, Windows/DOS does much the same thing,
but also treats a ^Z as an explicit end-of-file indicator.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I just don't know why they're shooting at us.
\__/ Capt. Benjamin "Hawkeye" Pierce
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From danielk at aracnet.com  Sun Feb 18 23:08:10 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 18 Feb 2001 14:08:10 -0800
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com>
Message-ID: <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com>

On Sun, 18 Feb 2001 11:35:41 -0800, Erik Max Francis <max at alcyone.com> wrote:

>Daniel Klein wrote:
>> On Windows the child process needs to send a ^Z in response to
>> readlines() and
>> on UNIX it needs to send ^D. I don't know what this is on MAC.
>
>This isn't the way you should signal end of file.  You should signal it
>the client closing his end of the pipe.  In fact, what you suggest above
>may work on Windows, but it won't work on UNIX -- under UNIX, an end of
>file isn't indicated with a sentinel character in the stream, but rather
>with an out-of-bounds indicator (the underlying file structure has an
>end-of-file flag).  As I recall, Windows/DOS does much the same thing,
>but also treats a ^Z as an explicit end-of-file indicator.

I'm not sure what you mean by "You should signal it the client closing his end
of the pipe". This is an interactive session with the client and server in
constant communication. If I close the pipe, I can do no further processing.

How do I specify this 'out-of-bounds' indicator?

Dan


From timr at probo.com  Mon Feb 19 03:44:21 2001
From: timr at probo.com (Tim Roberts)
Date: Sun, 18 Feb 2001 18:44:21 -0800
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com>
Message-ID: <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com>

Daniel Klein <danielk at aracnet.com> wrote:

>On Sun, 18 Feb 2001 11:35:41 -0800, Erik Max Francis <max at alcyone.com> wrote:
>
>>Daniel Klein wrote:
>>> On Windows the child process needs to send a ^Z in response to
>>> readlines() and
>>> on UNIX it needs to send ^D. I don't know what this is on MAC.
>>
>>This isn't the way you should signal end of file.  You should signal it
>>the client closing his end of the pipe.  In fact, what you suggest above
>>may work on Windows, but it won't work on UNIX -- under UNIX, an end of
>>file isn't indicated with a sentinel character in the stream, but rather
>>with an out-of-bounds indicator (the underlying file structure has an
>>end-of-file flag).
>
>I'm not sure what you mean by "You should signal it the client closing his end
>of the pipe". This is an interactive session with the client and server in
>constant communication. If I close the pipe, I can do no further processing.

I don't understand.  If the character you're waiting for doesn't terminate
the connection, then why use an O/S-dependent character at all?  Why not
just choose some arbitrary control character specific to your application.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From danielk at aracnet.com  Mon Feb 19 05:30:35 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 18 Feb 2001 20:30:35 -0800
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com>
Message-ID: <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com>

On Sun, 18 Feb 2001 18:44:21 -0800, Tim Roberts <timr at probo.com> wrote:

>Daniel Klein <danielk at aracnet.com> wrote:
>
>>On Sun, 18 Feb 2001 11:35:41 -0800, Erik Max Francis <max at alcyone.com> wrote:
>>
>>>Daniel Klein wrote:
>>>> On Windows the child process needs to send a ^Z in response to
>>>> readlines() and
>>>> on UNIX it needs to send ^D. I don't know what this is on MAC.
>>>
>>>This isn't the way you should signal end of file.  You should signal it
>>>the client closing his end of the pipe.  In fact, what you suggest above
>>>may work on Windows, but it won't work on UNIX -- under UNIX, an end of
>>>file isn't indicated with a sentinel character in the stream, but rather
>>>with an out-of-bounds indicator (the underlying file structure has an
>>>end-of-file flag).
>>
>>I'm not sure what you mean by "You should signal it the client closing his end
>>of the pipe". This is an interactive session with the client and server in
>>constant communication. If I close the pipe, I can do no further processing.
>
>I don't understand.  If the character you're waiting for doesn't terminate
>the connection, then why use an O/S-dependent character at all?  Why not
>just choose some arbitrary control character specific to your application.

I don't want to terminate the connection, I only want to know when the server
(child) process has stopped sending. If I terminate (close) the connection, the
server process becomes orphaned.

Here is a simple scenario of what I am trying to do:  Python (the parent
process) starts the child process with 'popen2.popen'. Let's assume that the
file handles are called 'instream' and 'outstream'.

1) parent sends a request to the child via outstream

2) child stdin receives the request and performs some action which results in a
variable number of lines (say less than 10) of output which are sent to stdout.

3) parent receives the lines via instream, analyzes the lines for it's next
action. Go to step 1.

This interaction should be continuous until the parent process sends a 'quit'
command to the child, at which point the child shuts itself down and the parent
proceeds to close both handles (instream and outstream).

If the client and server were only sending one line at a time, it would be
simple to do with readline() and write():

1) parent send line
2) child recv line
3) child send line
4) parent recv line
5 go to 1)

But, as illustrated above, the server (child) process can sometimes send more
than one line and it is not possible to determin in advance how many lines. To
be honest, the server process _does_ know how many lines it is sending except
when an error occurs. It is these errors that produce an indeterminate number
of lines which get sent back to the parent (child stdout --> parent instream).

On Windows, all I need to do is send a ^Z to indicate (end of transmission).

Dan



From qrczak at knm.org.pl  Mon Feb 19 22:14:02 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 19 Feb 2001 21:14:02 GMT
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com> <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com>
Message-ID: <slrn99334p.2fk.qrczak@qrnik.zagroda>

Sun, 18 Feb 2001 20:30:35 -0800, Daniel Klein <danielk at aracnet.com> pisze:

> I don't want to terminate the connection, I only want to know when
> the server (child) process has stopped sending.

You must invent your own way of signalling this. A way which does not
interfere with normal data. There is no other system-wide concept of
stopping sending than closing the connection.

When you press ^D on a Unix terminal, it is not sent in the stream.
It only flushes the line without '\n' at the end. If it was pressed
at the beginning of a line (or after a previous ^D), the read()
syscall returns 0, which is interpreted as the end of file.

The file is not physically closed - the process can read further.
But I think it works only for terminals and such signal cannot be
sent through a pipe. The read end returns 0 only when the write end
closes its pipe handle, and no character is treated specially.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From grante at visi.com  Tue Feb 20 00:27:21 2001
From: grante at visi.com (Grant Edwards)
Date: Mon, 19 Feb 2001 23:27:21 GMT
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com> <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com> <slrn99334p.2fk.qrczak@qrnik.zagroda>
Message-ID: <slrn993auo.2bo.grante@isis.visi.com>

In article <slrn99334p.2fk.qrczak at qrnik.zagroda>, Marcin 'Qrczak' Kowalczyk wrote:
>Sun, 18 Feb 2001 20:30:35 -0800, Daniel Klein <danielk at aracnet.com> pisze:
>
>> I don't want to terminate the connection, I only want to know when
>> the server (child) process has stopped sending.
>
>You must invent your own way of signalling this. A way which does not
>interfere with normal data. There is no other system-wide concept of
>stopping sending than closing the connection.
>
>When you press ^D on a Unix terminal, it is not sent in the stream.
>It only flushes the line without '\n' at the end. If it was pressed
>at the beginning of a line (or after a previous ^D), the read()
>syscall returns 0, which is interpreted as the end of file.

IIRC, it's the tty driver that interprets the ^D and generates an EOF
condition.  The rest of Unix doesn't know ^D from your uncle Bob.

>The file is not physically closed - the process can read further.
>But I think it works only for terminals and such signal cannot be
>sent through a pipe. The read end returns 0 only when the write end
>closes its pipe handle, and no character is treated specially.

-- 
Grant Edwards                   grante             Yow!  Ha ha  Ha ha Ha ha
                                  at               Ha Ha Ha Ha -- When will I
                               visi.com            EVER stop HAVING FUN?!!


From chris.gonnerman at usa.net  Tue Feb 20 03:47:22 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Mon, 19 Feb 2001 20:47:22 -0600
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com> <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com> <slrn99334p.2fk.qrczak@qrnik.zagroda> <slrn993auo.2bo.grante@isis.visi.com>
Message-ID: <006e01c09ae7$798731c0$a100000a@local>

----- Original Message -----
From: "Grant Edwards" <grante at visi.com>
Subject: Re: Determining EOF character


> In article <slrn99334p.2fk.qrczak at qrnik.zagroda>, Marcin 'Qrczak'
Kowalczyk wrote:
> >When you press ^D on a Unix terminal, it is not sent in the stream.
> >It only flushes the line without '\n' at the end. If it was pressed
> >at the beginning of a line (or after a previous ^D), the read()
> >syscall returns 0, which is interpreted as the end of file.
>
> IIRC, it's the tty driver that interprets the ^D and generates an EOF
> condition.  The rest of Unix doesn't know ^D from your uncle Bob.

That's what he said.

So why does closing the pipe to signal EOF not work?  I've written a great
many programs
based on that.  The writer closes the pipe and terminates, and the reader
gets EOF and
continues operation.  What am I missing here?






From jwbnews at scandaroon.com  Tue Feb 20 06:11:41 2001
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Mon, 19 Feb 2001 21:11:41 -0800
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com> <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com> <slrn99334p.2fk.qrczak@qrnik.zagroda> <slrn993auo.2bo.grante@isis.visi.com> <mailman.982636942.2852.python-list@python.org>
Message-ID: <jwbnews-081BEF.21114119022001@news.olympus.net>

In article <mailman.982636942.2852.python-list at python.org>, "Chris 
Gonnerman" <chris.gonnerman at usa.net> wrote:

> 
> So why does closing the pipe to signal EOF not work?  I've written a great
> many programs
> based on that.  The writer closes the pipe and terminates, and the reader
> gets EOF and
> continues operation.  What am I missing here?

He wants to signal the end of each chunk of data, leaving the pipe open 
for the next chunk.  I think he needs to define and use an end-of-chunk 
sentinal.

  --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From grante at visi.com  Tue Feb 20 19:07:12 2001
From: grante at visi.com (Grant Edwards)
Date: Tue, 20 Feb 2001 18:07:12 GMT
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com> <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com> <slrn99334p.2fk.qrczak@qrnik.zagroda> <slrn993auo.2bo.grante@isis.visi.com> <mailman.982636942.2852.python-list@python.org> <jwbnews-081BEF.21114119022001@news.olympus.net>
Message-ID: <knyk6.500$S7.49088@ruti.visi.com>

In article <jwbnews-081BEF.21114119022001 at news.olympus.net>, John W. Baxter wrote:
>
>> So why does closing the pipe to signal EOF not work?  I've
>> written a great many programs based on that.  The writer closes
>> the pipe and terminates, and the reader gets EOF and continues
>> operation.  What am I missing here?
>
>He wants to signal the end of each chunk of data, leaving the pipe open 
>for the next chunk.  I think he needs to define and use an end-of-chunk 
>sentinal.

Often, the easiest thing to do is to pass information in ASCII
and use '\n' as the chunk delimiter.  That way all of the
normal library routines just "do the right thing".

If you don't want to use ASCII, you can use some sort of binary
protocol (e.g. STX/ETX framing).

-- 
Grant Edwards                   grante             Yow!  Look into my eyes and
                                  at               try to forget that you have
                               visi.com            a Macy's charge card!


From danielk at aracnet.com  Tue Feb 20 15:12:30 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Tue, 20 Feb 2001 06:12:30 -0800
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com> <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com> <slrn99334p.2fk.qrczak@qrnik.zagroda> <slrn993auo.2bo.grante@isis.visi.com> <mailman.982636942.2852.python-list@python.org>
Message-ID: <3jt49toov10igcedlj14g336k0c4srsr8a@4ax.com>

On Mon, 19 Feb 2001 20:47:22 -0600, "Chris Gonnerman" <chris.gonnerman at usa.net>
wrote:

>So why does closing the pipe to signal EOF not work?  I've written a great many programs
>based on that.  The writer closes the pipe and terminates, and the reader gets EOF and
>continues operation.  What am I missing here?

John Baxter is correct, but to explain a bit further, the server process could
have an entire environment set up; closing the pipe would lose that
environment. If I closed the pipe it would be like initializing the same
objects for each iteration of a loop, ie

while 1:
	foo = Bar()
	dead = Parrot()
	spam = Eggs()
	# rest of processing using 'foo', 'dead', 'spam'

rather than

foo = Bar()
dead = Parrot()
spam = Eggs()
while 1:
	# etc.

To be honest, the whole purpose for this was to catch and report unexpected
server errors that could not be controlled by the server process. In addition,
I had thought to be able to use the 'readlines()' method to recv more than one
line at a time. For now, since I can't do this, I have coded around the problem
by using only 'readline()' and ensuring that if the server process sends more
than one line of data to the client, it will gracefully (rather than violently)
shut itself down rather than just 'hang' waiting for an EOF. I do this by
always sending an 'expected' reply to the client; if the client does not see
this, it takes the aforementioned action.

I think you can put this thread to rest at this point. Thanks for all your
input, this has been a great learning experience for me. :^)

Daniel Klein
Portland OR USA


From anandg at cs.utexas.edu  Sun Feb 18 20:06:46 2001
From: anandg at cs.utexas.edu (Anand Ganesh)
Date: Sun, 18 Feb 2001 13:06:46 -0600
Subject: newbie : python path setting on linux
Message-ID: <Pine.LNX.4.21.0102181246300.7695-100000@sneem.cs.utexas.edu>

Hi, 
	I downloaded the python gtk bindings and compiled the module.
I've a set of files in the directory ~/tmp/pygtk-0.6.6. Now I try to 
run the examples in ~/tmp/pygtk-0.6.6/examples/simple say simple1.py.
It says :
ImportError: No module named _gtk

Now, I tried fiddling with PYTHONPATH and PYTHONHOME. I set PYTHONHOME to
~/tmp/pygtk-0.6.6 and PYTHONPATH to ./ (b'cos PYTHONHOME is prefixed ?).
Anyways, didnt work. Same error. Infact printing sys.path I find that it
doesnt have ~/tmp/pygtk-0.6.6 on it at all.

I'm totally confused. I dont have root perms to install the gtk module
in the standard places...any help appreciated.

Thanks,
Anand.




From max at alcyone.com  Sun Feb 18 20:37:38 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 11:37:38 -0800
Subject: newbie : python path setting on linux
References: <Pine.LNX.4.21.0102181246300.7695-100000@sneem.cs.utexas.edu>
Message-ID: <3A902482.D1FF0683@alcyone.com>

Anand Ganesh wrote:

>         I downloaded the python gtk bindings and compiled the module.
> I've a set of files in the directory ~/tmp/pygtk-0.6.6. Now I try to
> run the examples in ~/tmp/pygtk-0.6.6/examples/simple say simple1.py.
> It says :
> ImportError: No module named _gtk
> 
> Now, I tried fiddling with PYTHONPATH and PYTHONHOME. I set PYTHONHOME
> to
> ~/tmp/pygtk-0.6.6 and PYTHONPATH to ./ (b'cos PYTHONHOME is prefixed
> ?).
> Anyways, didnt work. Same error. Infact printing sys.path I find that
> it
> doesnt have ~/tmp/pygtk-0.6.6 on it at all.

Because you didn't add it to it.  PYTHONHOME indicates where the main
Python directory resides (if different from the compiled-in default, if
it's in your path); PYTHONPATH adds additional search paths.  You should
add the directory which the actual gtk module resides in to PYTHONPATH,
not PYTHONHOME.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I just don't know why they're shooting at us.
\__/ Capt. Benjamin "Hawkeye" Pierce
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From phlip_cpp at my-deja.com  Sun Feb 18 20:23:14 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 19:23:14 GMT
Subject: What's up with enscript for syntax highlighting?
Message-ID: <96p7f2$bqi@dispatch.concentric.net>

Noh Typ:

Here's a command for GNU enscript 1.6.1:

enscript --pretty-print=python Kozmik.py --language=html 
--output=Kozmik.py.html 

The output ain't pretty. It uses no colors.

How do I convert Python to syntax highlighted HTML? Is there a class 
library available somewhere on www.python.org (perhaps under "hilite" or 
something? or is there a better version of enscript?

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  Personally qualified to snub Mensa  --


From phlip_cpp at my-deja.com  Sun Feb 18 20:29:56 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 19:29:56 GMT
Subject: What's up with enscript for syntax highlighting?
References: <96p7f2$bqi@dispatch.concentric.net>
Message-ID: <96p7rk$sub@dispatch.concentric.net>

Proclaimed Phlip from the mountaintops:

> Is there a class
> library available somewhere on www.python.org (perhaps under "hilite" or
> something?

It's in ClassModules.py you dumb f**k - can't you tell by the name?

        http://www.lemburg.com/files/python/doc.py.html

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  Have a :-) day  --


From thomas at xs4all.net  Sun Feb 18 22:13:36 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 18 Feb 2001 22:13:36 +0100
Subject: What's up with enscript for syntax highlighting?
In-Reply-To: <96p7rk$sub@dispatch.concentric.net>; from phlip_cpp@my-deja.com on Sun, Feb 18, 2001 at 07:29:56PM +0000
References: <96p7f2$bqi@dispatch.concentric.net> <96p7rk$sub@dispatch.concentric.net>
Message-ID: <20010218221336.C14292@xs4all.nl>

On Sun, Feb 18, 2001 at 07:29:56PM +0000, Phlip wrote:
> Proclaimed Phlip from the mountaintops:

> > Is there a class
> > library available somewhere on www.python.org (perhaps under "hilite" or
> > something?

> It's in ClassModules.py you dumb f**k - can't you tell by the name?

That's totally uncalled for. Being a newbie might be a grave error, but one
everyone commits at some point in life. Furthermore, RTFM is much more
effective if you do it gently and make them feel nicely embarassed, rather
than having them just say 'well, fuck you too' when reading the first
insult, and not learn a thing.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!





From tim.one at home.com  Sun Feb 18 22:41:13 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 18 Feb 2001 16:41:13 -0500
Subject: What's up with enscript for syntax highlighting?
In-Reply-To: <20010218221336.C14292@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEECJAAA.tim.one@home.com>

> On Sun, Feb 18, 2001 at 07:29:56PM +0000, Phlip wrote:
> > Proclaimed Phlip from the mountaintops:
>
>>> Is there a class
>>> library available somewhere on www.python.org (perhaps under
>>> "hilite" or something?
>
>> It's in ClassModules.py you dumb f**k - can't you tell by the name?

[Thomas Wouters]
> That's totally uncalled for. ...

Yes.  Still, Philip *was* flaming himself.  I was hoping he'd follow it up
with a post admonishing himself for being so gauche, and then defend himself
to himself, and after a long bitter battle eventually make up and hug
himself for all to see and learn from.  I guess he did that offline <wink>.

civility-is-mandatory-but-sanity-is-optional-ly y'rs  - tim




From phlip_cpp at my-deja.com  Sun Feb 18 22:32:06 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 21:32:06 GMT
Subject: What's up with enscript for syntax highlighting?
References: <96p7f2$bqi@dispatch.concentric.net> <96p7rk$sub@dispatch.concentric.net> <mailman.982530861.24974.python-list@python.org>
Message-ID: <96pf0m$sur@dispatch.concentric.net>

Proclaimed Thomas Wouters from the mountaintops:

> On Sun, Feb 18, 2001 at 07:29:56PM +0000, Phlip wrote:
>> Proclaimed Phlip from the mountaintops:
> 
>> > Is there a class
>> > library available somewhere on www.python.org (perhaps under "hilite"
>> > or something?
> 
>> It's in ClassModules.py you dumb f**k - can't you tell by the name?
> 
> That's totally uncalled for. Being a newbie might be a grave error, but
> one everyone commits at some point in life. Furthermore, RTFM is much more
> effective if you do it gently and make them feel nicely embarassed, rather
> than having them just say 'well, fuck you too' when reading the first
> insult, and not learn a thing.

Thanks. I'l try to keep that in mind the next time I flame myself.

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  I'l have my Web site call your Web site...  --


From amk at mira.erols.com  Tue Feb 20 01:54:43 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 20 Feb 2001 00:54:43 GMT
Subject: What's up with enscript for syntax highlighting?
References: <96p7f2$bqi@dispatch.concentric.net> <96p7rk$sub@dispatch.concentric.net> <mailman.982530861.24974.python-list@python.org> <96pf0m$sur@dispatch.concentric.net>
Message-ID: <slrn993gtk.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>

On 18 Feb 2001 21:32:06 GMT, Phlip <phlip_cpp at my-deja.com> wrote:
>Proclaimed Thomas Wouters from the mountaintops:
>> one everyone commits at some point in life. Furthermore, RTFM is much more
>> effective if you do it gently and make them feel nicely embarassed, rather
>> than having them just say 'well, fuck you too' when reading the first
>> insult, and not learn a thing.
>
>Thanks. I'l try to keep that in mind the next time I flame myself.

Isn't comp.lang.python a great newsgroup? :)

--amk



From phlip_cpp at my-deja.com  Sun Feb 18 20:51:55 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 19:51:55 GMT
Subject: What's up with enscript for syntax highlighting?
References: <96p7f2$bqi@dispatch.concentric.net>
Message-ID: <96p94r$sub@dispatch.concentric.net>

Proclaimed Phlip from the mountaintops:

> Here's a command for GNU enscript 1.6.1:
> 
> enscript --pretty-print=python Kozmik.py --language=html
> --output=Kozmik.py.html
> 
> The output ain't pretty. It uses no colors.

Per this exellent brand-new site I've just discovered called 
http://groups.google.com, those brave enough to survive the terrors of 
'gvim' can click on its menu items Syntax -> Convert to HTML to get some 
color in their life.

Man, putting all the USENET archives in one Web site's a killer idea. Why 
hasn't anyone ever thought of that before? ;-)

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  The first few lines of code must "hook" the
      computer, and make it "care" about the program  --


From phlip_cpp at my-deja.com  Sun Feb 18 21:16:11 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 20:16:11 GMT
Subject: SNIPPET: Pipe Graphviz's graph geometry into Pmw.ScrolledCanvas
Message-ID: <96paib$7j@dispatch.concentric.net>

Not Hyp:

Thanks for all that help on the colorized HTML problem - this newsgroup's a 
real lifesaver, huh?

I have just finished wrapping www.graphviz.org's 'dot' language in Python. 
You define a graph (a set of nodes connected by edges) as Node objects in 
Python, and 'dot' determines how to display them asthetically with the 
edges as curves and the vertices as ovals with labels in them. /Tres/ 
cutsey.

The code looks a little bit like this:

http://freehosting1.at.webjump.com/5e5d5afd9/ph/phlip-webjump/Grapher.py.html

http://freehosting1.at.webjump.com/5e5d5afd9/ph/phlip-webjump/Grapher.py

That's off my new Web site:

        http://phlip.webjump.com

The jiggling baloney's a little heavy, but it was the first free site I 
could find that didn't already have a "phlip". Poseurs.

After Grapher.py translates the Nodes to the most primitive example of 
'dot' language, 'dot' reads this and writes either a plain text 
representation, or a PostScript file. It can probably do other formats too.

More Grapher.py code then reads the plain text representation, extracts all 
the beautiful screen coordinates from it, associates them with the Nodes 
they came from, and paints these on a Python MegaWidgets Pmw.ScrolledCanvas 
suitable for framing.

The source code illustrates Design Patterns (but I forget which ones), 
de-coupling (sort of), dictionaries, lists, and other goodies.

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  DARE to resist drug-war profiteering  --


From phlip_cpp at my-deja.com  Sun Feb 18 22:33:01 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 21:33:01 GMT
Subject: SNIPPET: Pipe Graphviz's graph geometry into Pmw.ScrolledCanvas
References: <96paib$7j@dispatch.concentric.net>
Message-ID: <96pf2d$38@freepress.concentric.net>

Proclaimed Phlip from the mountaintops:
> 
http://freehosting1.at.webjump.com/5e5d5afd9/ph/phlip-webjump/Grapher.py.html
> 
> http://freehosting1.at.webjump.com/5e5d5afd9/ph/phlip-webjump/Grapher.py

Sorry - those click-thru's don't work. Jeeze good free help is hard to find!

> That's off my new Web site:
> 
>         http://phlip.webjump.com

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  Whoever dies with the most features wins  --


From stevemak at softhome.net  Sun Feb 18 22:38:54 2001
From: stevemak at softhome.net (Steve Mak)
Date: Sun, 18 Feb 2001 21:38:54 GMT
Subject: Pls Help! Need binary search tree module
Message-ID: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com>

I was wondering if anyone had a simple binary search tree module I could
use? I need to write a program that counts the number of occurrences of each
word in a file, and outputs the word and corresponding counts
alphabetically. And it is required that one of the data structure is a
binary search tree.

 Eg: if the inputs is:

"How are you. How are you. How are you."

the output is:

are 3
how 3
you 3

Thanks in advance!!!!!!!!

JJ





From fredrik at pythonware.com  Sun Feb 18 22:51:12 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 21:51:12 GMT
Subject: Pls Help! Need binary search tree module
References: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com>
Message-ID: <ktXj6.18191$AH6.2496311@newsc.telia.net>

Steve Mak wrote:
> I was wondering if anyone had a simple binary search tree module I could
> use? I need to write a program that counts the number of occurrences of each
> word in a file, and outputs the word and corresponding counts
> alphabetically. And it is required that one of the data structure is a
> binary search tree.

are you sure your teacher doesn't read comp.lang.python?

I'm pretty sure you'll learn more if you do it yourself, but you
could take a look at the bisect.py module (it's in Python's Lib
directory).

Cheers /F




From phlip_cpp at my-deja.com  Sun Feb 18 23:00:43 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 22:00:43 GMT
Subject: Pls Help! Need binary search tree module
References: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com>
Message-ID: <96pgmb$b4o@dispatch.concentric.net>

Proclaimed Steve Mak from the mountaintops:

> I was wondering if anyone had a simple binary search tree module I could
> use? I need to write a program that counts the number of occurrences of
> each word in a file, and outputs the word and corresponding counts
> alphabetically. And it is required that one of the data structure is a
> binary search tree.
> 
>  Eg: if the inputs is:
> 
> "How are you. How are you. How are you."
> 
> the output is:
> 
> are 3
> how 3
> you 3

What's wrong with using a dictionary object? That supplies a hash or 
something for you behind a clean interface. Is this for a class? Look up 
'has_key' to get started.

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  Personally qualified to snub Mensa  --


From stevemak at softhome.net  Sun Feb 18 23:28:22 2001
From: stevemak at softhome.net (Steve Mak)
Date: Sun, 18 Feb 2001 22:28:22 GMT
Subject: ATTN: Phlip
References: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com> <96pgmb$b4o@dispatch.concentric.net>
Message-ID: <3A904EAB.2AA272A4@softhome.net>

Hey Phlip

    Yes, it's for class....the dictionary object would make things ALOT
easier, but they want us to use a bst, so I can't help it! Can you help me?






From phlip_cpp at my-deja.com  Mon Feb 19 01:15:27 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 19 Feb 2001 00:15:27 GMT
Subject: ATTN: Phlip
References: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com> <96pgmb$b4o@dispatch.concentric.net> <3A904EAB.2AA272A4@softhome.net>
Message-ID: <96poiv$crf@dispatch.concentric.net>

Proclaimed Steve Mak from the mountaintops:

> Hey Phlip
> 
>     Yes, it's for class....the dictionary object would make things ALOT
> easier, but they want us to use a bst, so I can't help it! Can you help
> me?

Okay. Try this:

class Node:
        def __init__ (self, left, right, label):
                (self.left, self.right, self.label) = (left, right, label)
                return

a = Node (None, None, "a")
g = Node (None, None, "g")
p = Node (None, None, "p")
z = Node (None, None, "z")
f = Node (a, g, "f")
u = Node (p, z, "u")
o = Node (f, u, "o")

There. That's a tree balanced by alpha order. But it can't insert or 
rebalance by itself. But adding an in-order traversal method would be 
trivial.

The best thing now would be to write a Unit Test that expresses what your 
interface should do.

        http://c2.com/cgi/wiki?CodeUnitTestFirst

Then you run it over and over again until there are no error messages or 
assertion exceptions:

def testTree ():
        t = Tree ()
        assert (t.count () == 0)
        leaf = t.addLeaf ("leaf")
        assert (t.count () == 1)
        got = t.getLeaf ("leaf")
        assert (got == leaf)
        ...

Doing it test-first like this keeps the hugest projects locked on track.

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  Have a :-) day  -


From gregj at pobox.com  Mon Feb 19 08:07:02 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 19 Feb 2001 07:07:02 GMT
Subject: Pls Help! Need binary search tree module
References: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com>
Message-ID: <3a90c1c4.48617037@news.potlnd1.or.home.com>

On Sun, 18 Feb 2001 21:38:54 GMT, "Steve Mak" <stevemak at softhome.net>
wrote:

>I was wondering if anyone had a simple binary search tree module I could
>use? I need to write a program that counts the number of occurrences of each
>word in a file, and outputs the word and corresponding counts
>alphabetically. And it is required that one of the data structure is a
>binary search tree.
>
> Eg: if the inputs is:
>
>"How are you. How are you. How are you."
>
>the output is:
>
>are 3
>how 3
>you 3

If you are working on an assignment and must use a binary search tree,
I refer you to Sedgewick's book "Algorithms." Translating
pointer-based trees to Python will challenge your Python skills. Hint:
a Python list can contain ANY kind of object reference, including
references to other lists. I implemented Sedgewick's algorithms in
Python (from his C versions) in about 30 minutes.

The typical Python way to do what you describe uses a dictionary with
the words as keys as a counter as the value. For example:

# insert/count words using dictionary
words = ['your', 'list', 'of', 'words']
wc = {}
for w in words:
    if wc.has_key(w):
        wc[w] += 1
    else:
        wc[w] = 1

# show counts by word in alpha order
words = wc.keys()
words.sort()
for w in words:
    print "%s: %d" % (w, wc[w])


Greg Jorgensen
Deschooling Society
Portland, Oregon USA


From othello at javanet.com  Fri Feb 23 03:22:06 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Thu, 22 Feb 2001 21:22:06 -0500
Subject: Pls Help! Need binary search tree module
References: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com>
Message-ID: <3A95C94E.FBDD0C7@javanet.com>

Steve Mak wrote:

> I was wondering if anyone had a simple binary search tree module I could
> use? I need to write a program that counts the number of occurrences of each
> word in a file, and outputs the word and corresponding counts
> alphabetically. And it is required that one of the data structure is a
> binary search tree.
>

Here's a list oriented, binary counting tree for you:

def add( n, elem ):
    if n == None:
        n = [ elem, 1, None, None ]
    elif n[0] == elem:
        n[1] += 1
    else:
        tgt = 2     # choose left side
        if elem > n[0]:
            tgt = 3 # choose right side
        n[tgt] = add( n[tgt], elem )
    return n

n = None
for w in "the man and the moon and the tree".split():
    n = add(n,w)
print n


Raymond



From seandc at att.net  Sun Feb 18 22:39:35 2001
From: seandc at att.net (Sean Doull-Connolly)
Date: Sun, 18 Feb 2001 21:39:35 GMT
Subject: Tkinter Widgets Needed
Message-ID: <3A9040CE.33B00019@att.net>

Does anyone know of widgets I can use to
build data-entry tools for a database:

1. Number Entry

   Passed parameters:
       Given value (as text string)
       Length
       Number of decimals
   Displays
       The given value with the decimal
          point in the correct spot
   Accepts
       Only legal input (eg 0-9 and .)

   Example:
       ne = NumberEntry(MasterFrame, ('9999.99', 10, 2))
       ....
       returnval = ne.get()

2. Date Entry

   Passed parameters:
       Given value (as text string)
       Century ON/OFF
   Displays
       The given value with the date
          correctly formatted
   Accepts
       Only legal input (eg 0-9 and /)
          and parses the date (Zeller?) to
          ensure date is real

   Example:
       de = DateEntry(MasterFrame, ('20010218', ON))
       ....
       returnval = de.get()

Thanks.

Sean
seandc at att.net


From phlip_cpp at my-deja.com  Mon Feb 19 18:10:05 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: Mon, 19 Feb 2001 09:10:05 -0800
Subject: Tkinter Widgets Needed
References: <3A9040CE.33B00019@att.net>
Message-ID: <Otck6.7$n27.13605@news.pacbell.net>

Sean Doull-Connolly wrote:

> 1. Number Entry

> 2. Date Entry

Python MegaWidgets has a zillion full-featured Tkinter controls, including 
a formatted entry field that can do numbers and dates.

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  Proud victim of the dreaded boomerang effect  --


From Mike_B at T-Online.de  Sun Feb 18 23:21:44 2001
From: Mike_B at T-Online.de (Michael Bauer)
Date: Sun, 18 Feb 2001 23:21:44 +0100
Subject: How to determine own IP
Message-ID: <96phtr$rar$07$1@news.t-online.com>

Hi *

I'm stuck with a kind of dumb problem. I'm implementing a project based on 
a distributed, decentralised network. Communication between nodes is 
established through xml-rpc.
For a new host to gather some other members of the network, it sends out a 
PING that is passed on through the net. Each host receiving the PING 
returns a PONG. Since the PONG is not sent directly to the new host but 
passed back through all hosts that passed on the PING, i need to store the 
IP of the answering hosts inside the PONG package.
Now, to the real problem :-)
If a host has more than one network adapter, how do i know which IP should 
i put inside  the PONG package? 
Specifically, when dealing with dial-up users who are also connected to a 
LAN, gethostbyname(gethostname()) always returns their internal adress.

Any ideas?
        


From thomas at xs4all.net  Sun Feb 18 23:51:12 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 18 Feb 2001 23:51:12 +0100
Subject: How to determine own IP
In-Reply-To: <96phtr$rar$07$1@news.t-online.com>; from Mike_B@T-Online.de on Sun, Feb 18, 2001 at 11:21:44PM +0100
References: <96phtr$rar$07$1@news.t-online.com>
Message-ID: <20010218235112.E14292@xs4all.nl>

On Sun, Feb 18, 2001 at 11:21:44PM +0100, Michael Bauer wrote:

> I'm stuck with a kind of dumb problem. I'm implementing a project based on 
> a distributed, decentralised network. Communication between nodes is 
> established through xml-rpc.
> For a new host to gather some other members of the network, it sends out a 
> PING that is passed on through the net. Each host receiving the PING 
> returns a PONG. Since the PONG is not sent directly to the new host but 
> passed back through all hosts that passed on the PING, i need to store the 
> IP of the answering hosts inside the PONG package.
> Now, to the real problem :-)
> If a host has more than one network adapter, how do i know which IP should 
> i put inside  the PONG package? 
> Specifically, when dealing with dial-up users who are also connected to a 
> LAN, gethostbyname(gethostname()) always returns their internal adress.

And don't forget multi-homed machines (multiple network cards/ip-addresses)
and NAT/IP-Masquerading. You can't always trust what a machine thinks is its
own ipaddress.

> Any ideas?

Have a 'WHOAMI' command/function/packet that's answered by the directly
connected node(s). If you want highly-connected nodes, consider treating
each node's response as unique for that node (so you can work both 'inside'
and 'outside' a masqueraded network) and always use it for communication
that came through the node.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From Mike_B at T-Online.de  Mon Feb 19 00:21:09 2001
From: Mike_B at T-Online.de (Michael Bauer)
Date: Mon, 19 Feb 2001 00:21:09 +0100
Subject: How to determine own IP
References: <96phtr$rar$07$1@news.t-online.com> <mailman.982536739.17799.python-list@python.org>
Message-ID: <96pld8$s5t$05$1@news.t-online.com>

Thomas Wouters wrote:

> On Sun, Feb 18, 2001 at 11:21:44PM +0100, Michael Bauer wrote:
>> Now, to the real problem :-)
>> If a host has more than one network adapter, how do i know which IP
>> should
>> i put inside  the PONG package?
>> Specifically, when dealing with dial-up users who are also connected to a
>> LAN, gethostbyname(gethostname()) always returns their internal adress.
> 
> And don't forget multi-homed machines (multiple network
> cards/ip-addresses) and NAT/IP-Masquerading. You can't always trust what a
> machine thinks is its own ipaddress.
> 
>> Any ideas?
> 
> Have a 'WHOAMI' command/function/packet that's answered by the directly
> connected node(s). If you want highly-connected nodes, consider treating
> each node's response as unique for that node (so you can work both
> 'inside' and 'outside' a masqueraded network) and always use it for
> communication that came through the node.

Great! 

Sometimes when you think something is really hard to solve, it turns out to 
be quite obvious ;-)

Thank you, Mike

-- 
        


From bryan at eevolved.com  Mon Feb 19 07:14:47 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Mon, 19 Feb 2001 06:14:47 GMT
Subject: How to determine own IP
References: <96phtr$rar$07$1@news.t-online.com>
Message-ID: <rR2k6.153122$Pm2.2783213@news20.bellglobal.com>

Michael Bauer wrote:

> Hi *
> 
> I'm stuck with a kind of dumb problem. I'm implementing a project based on
> a distributed, decentralised network. Communication between nodes is
> established through xml-rpc.
> For a new host to gather some other members of the network, it sends out a
> PING that is passed on through the net. Each host receiving the PING
> returns a PONG. Since the PONG is not sent directly to the new host but
> passed back through all hosts that passed on the PING, i need to store the
> IP of the answering hosts inside the PONG package.
> Now, to the real problem :-)
> If a host has more than one network adapter, how do i know which IP should
> i put inside  the PONG package?
> Specifically, when dealing with dial-up users who are also connected to a
> LAN, gethostbyname(gethostname()) always returns their internal adress.
> 
> Any ideas?
>        

Umm, I won't pretend to understand your entire protocol from this brief 
description but I do applaud your attempt and your choice of programming 
language. However, from what I understand, it sounds like your scheme won't 
scale well with many users due to the high number of ping/pongs.

Learn from the mistakes of others:
http://www.monkey.org/~dugsong/mirror/gnutella.html

Good luck and I apologize if this problem does not apply to your protocol.
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"This is one of the hardest lessons for humans to learn. We cannot admit 
that things might be neither good nor evil, neither cruel nor kind, but 
simply callous - indifferent to all suffering, lacking all purpose." -- 
Richard Dawkins



From Mike_B at T-Online.de  Mon Feb 19 10:57:05 2001
From: Mike_B at T-Online.de (Michael Bauer)
Date: Mon, 19 Feb 2001 10:57:05 +0100
Subject: How to determine own IP
References: <96phtr$rar$07$1@news.t-online.com> <rR2k6.153122$Pm2.2783213@news20.bellglobal.com>
Message-ID: <96qqlk$er0$05$1@news.t-online.com>

Bryan Mongeau wrote:

> Michael Bauer wrote:
> 
> > [distributed network, PING and PONG]
> 
> Umm, I won't pretend to understand your entire protocol from this brief
> description but I do applaud your attempt and your choice of programming
> language. However, from what I understand, it sounds like your scheme
> won't scale well with many users due to the high number of ping/pongs.
> 
> Learn from the mistakes of others:
> http://www.monkey.org/~dugsong/mirror/gnutella.html

Yeah, i already know this document. It actually does apply to our project 
but with a much lesser impact since we took some measures to reduce the 
amount of PING and PONG going through the network.
Another problem with gnutella seems to be the shere number of search 
requests a user kicks off, while we produce network traffic rather seldom.
 
> Good luck and I apologize if this problem does not apply to your protocol.

Thank you, i appreciated your reply :-)

Mike

-- 
excuse my poor English...       


From bryan at eevolved.com  Tue Feb 20 03:44:59 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Tue, 20 Feb 2001 02:44:59 GMT
Subject: How to determine own IP
References: <96phtr$rar$07$1@news.t-online.com> <rR2k6.153122$Pm2.2783213@news20.bellglobal.com> <96qqlk$er0$05$1@news.t-online.com>
Message-ID: <LSkk6.259580$f36.10175244@news20.bellglobal.com>

Michael Bauer wrote:

> Bryan Mongeau wrote:
> 
> > Michael Bauer wrote:
> > 
> > > [distributed network, PING and PONG]
> > 
> > Umm, I won't pretend to understand your entire protocol from this brief
> > description but I do applaud your attempt and your choice of programming
> > language. However, from what I understand, it sounds like your scheme
> > won't scale well with many users due to the high number of ping/pongs.
> > 
> > Learn from the mistakes of others:
> > http://www.monkey.org/~dugsong/mirror/gnutella.html
> 
> Yeah, i already know this document. It actually does apply to our project
> but with a much lesser impact since we took some measures to reduce the
> amount of PING and PONG going through the network.
> Another problem with gnutella seems to be the shere number of search
> requests a user kicks off, while we produce network traffic rather seldom.
>  
> > Good luck and I apologize if this problem does not apply to your
> > protocol.
> 
> Thank you, i appreciated your reply :-)
> 
> Mike
> 

I'd be interested in learning more about your protocol and network. Do you 
have anything up yet? Docs would be nice... I say this because I think it's 
about time someone uses python for this application. Could be the killer 
app that propells python into the mainstream.

-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"Scientific beliefs are supported by evidence, and they get results. Myths 
and faiths are not and do not" -- Richard Dawkins



From Mike_B at T-Online.de  Tue Feb 20 23:54:54 2001
From: Mike_B at T-Online.de (Michael Bauer)
Date: Tue, 20 Feb 2001 23:54:54 +0100
Subject: How to determine own IP
References: <96phtr$rar$07$1@news.t-online.com> <rR2k6.153122$Pm2.2783213@news20.bellglobal.com> <96qqlk$er0$05$1@news.t-online.com> <LSkk6.259580$f36.10175244@news20.bellglobal.com>
Message-ID: <96usmk$107$05$1@news.t-online.com>

Bryan Mongeau wrote:

> I'd be interested in learning more about your protocol and network. Do you
> have anything up yet? Docs would be nice... I say this because I think
> it's about time someone uses python for this application. Could be the
> killer app that propells python into the mainstream.

Well, there is nothing available yet for the masses, but as soon as we 
finished a first release-able version i will announce it here, since we 
indeed think we might be building a killer-app :-)

cu Mike

-- 
Please excuse my poor English...


From donn at oz.net  Mon Feb 19 07:25:53 2001
From: donn at oz.net (Donn Cave)
Date: 19 Feb 2001 06:25:53 GMT
Subject: How to determine own IP
References: <96phtr$rar$07$1@news.t-online.com>
Message-ID: <96qe9h$962$0@216.39.151.169>

Quoth Michael Bauer <Mike_B at T-Online.de>:
...
| If a host has more than one network adapter, how do i know which IP should 
| i put inside  the PONG package? 
| Specifically, when dealing with dial-up users who are also connected to a 
| LAN, gethostbyname(gethostname()) always returns their internal adress.

If you have a connected socket s, s.getsockname() will return the
IP address and port number of the local endpoint.

	Donn Cave, donn at oz.net


From Mike_B at T-Online.de  Mon Feb 19 10:58:48 2001
From: Mike_B at T-Online.de (Michael Bauer)
Date: Mon, 19 Feb 2001 10:58:48 +0100
Subject: How to determine own IP
References: <96phtr$rar$07$1@news.t-online.com> <96qe9h$962$0@216.39.151.169>
Message-ID: <96qqoq$er0$05$2@news.t-online.com>

Donn Cave wrote:

> Quoth Michael Bauer <Mike_B at T-Online.de>:
> ...
> | If a host has more than one network adapter, how do i know which IP
> | should
> | i put inside  the PONG package?
> | Specifically, when dealing with dial-up users who are also connected to
> | a LAN, gethostbyname(gethostname()) always returns their internal
> | adress.
> 
> If you have a connected socket s, s.getsockname() will return the
> IP address and port number of the local endpoint.

Hmm, seems even better than the first solution!

Thanks, Mike


-- 
Excuse my poor english...
        


From akuchlin at mems-exchange.org  Sun Feb 18 23:47:33 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 18 Feb 2001 17:47:33 -0500
Subject: How widespread is NIS support?
Message-ID: <3d1ysv8v7e.fsf@ute.cnri.reston.va.us>

Does anyone have an idea how widespread NIS support is these days?
Patch #103544 changes the 2.1 setup.py script to compile the NIS
module on any Unix platform, and I'm wondering if that's too
ambitious.  Is anyone out there using a Unix *without* NIS?  (Try 'man
yp_get_default_domain' and see if you get anything, for a start.)  Are
the NIS functions part of libc, or in a different library (such as
libnsl on Solaris). 

--amk


From jwbnews at scandaroon.com  Mon Feb 19 01:19:40 2001
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Sun, 18 Feb 2001 16:19:40 -0800
Subject: How widespread is NIS support?
References: <3d1ysv8v7e.fsf@ute.cnri.reston.va.us>
Message-ID: <jwbnews-0B4EEE.16194018022001@news.olympus.net>

In article <3d1ysv8v7e.fsf at ute.cnri.reston.va.us>, Andrew Kuchling 
<akuchlin at mems-exchange.org> wrote:

> Does anyone have an idea how widespread NIS support is these days?
> Patch #103544 changes the 2.1 setup.py script to compile the NIS
> module on any Unix platform, and I'm wondering if that's too
> ambitious.  Is anyone out there using a Unix *without* NIS?  (Try 'man
> yp_get_default_domain' and see if you get anything, for a start.)  Are
> the NIS functions part of libc, or in a different library (such as
> libnsl on Solaris). 
> 
> --amk

We don't run NIS on our farm of a few BSDi machines and several Linux 
(Red Hat, which is beginning to look like "unfortunately" to me, but 
we'll see how 7.1 is).

I don't run it on my laptop's Linux (safely in 6.2 at the moment), I 
don't thing the boss runs it on his.

Whether or not we could compile it, I don't know.  But if the code in 
Python expects NIS to actually *do* something, we'll have a problem.

  --John

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From akuchlin at mems-exchange.org  Mon Feb 19 16:28:05 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 19 Feb 2001 10:28:05 -0500
Subject: How widespread is NIS support?
References: <3d1ysv8v7e.fsf@ute.cnri.reston.va.us> <jwbnews-0B4EEE.16194018022001@news.olympus.net>
Message-ID: <3dy9v27kvu.fsf@ute.cnri.reston.va.us>

"John W. Baxter" <jwbnews at scandaroon.com> writes:
> Whether or not we could compile it, I don't know.  But if the code in 
> Python expects NIS to actually *do* something, we'll have a problem.

glibc 2.x includes NIS support (in libnsl), so it's available on my
Red Hat 6.2 machine, and probably on your 7.0 boxes, too.  Luckily,
the NIS test is smart enough to figure out if NIS is enabled, and
skips the test if it isn't.  I'm simply concerned whether the module
will compile everywhere.

--amk



From mark at chem.uwa.edu.au  Mon Feb 19 22:38:23 2001
From: mark at chem.uwa.edu.au (Mark C Favas)
Date: 19 Feb 01 21:38:23 GMT
Subject: How widespread is NIS support?
References: <3d1ysv8v7e.fsf@ute.cnri.reston.va.us> <jwbnews-0B4EEE.16194018022001@news.olympus.net> <3dy9v27kvu.fsf@ute.cnri.reston.va.us>
Message-ID: <mark.982618703@declan>

[Andrew Kuchling wonders whether the nis module will compile everywhere on Unix]

>the NIS test is smart enough to figure out if NIS is enabled, and
>skips the test if it isn't.  I'm simply concerned whether the module
>will compile everywhere.

Well, at least ithe version of nis in 2.1a2 compiles without warnings on Tru64
Unix, V4.0F with Compaq's C compiler, and does the right thing with
"make test" (since nis is installed but not running):
test_nis
test test_nis skipped --  internal NIS server or client error

and when run manually:
python Lib/test/test_nis.py
nis.maps()
Traceback (most recent call last):
  File "Lib/test/test_nis.py", line 10, in ?
    raise TestFailed, msg
test_support.TestFailed: internal NIS server or client error

As mentioned in previously, nis support on this platform is in libc.

--
Email  - mark at chem.uwa.edu.au      ,-_|\                           Mark C Favas
Phone  - +61 9 380 3482           /     \               Department of Chemistry
Fax    - +61 9 380 1005      ---> *_,-._/   The University of Western Australia
                                       v                               Nedlands
Loc    - 31.97 S, 115.81 E                               Western Australia 6009


From mwh21 at cam.ac.uk  Mon Feb 19 01:25:32 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Feb 2001 00:25:32 +0000
Subject: How widespread is NIS support?
References: <3d1ysv8v7e.fsf@ute.cnri.reston.va.us>
Message-ID: <m33ddbv7r7.fsf@atrus.jesus.cam.ac.uk>

Andrew Kuchling <akuchlin at mems-exchange.org> writes:

> Does anyone have an idea how widespread NIS support is these days?
> Patch #103544 changes the 2.1 setup.py script to compile the NIS
> module on any Unix platform, and I'm wondering if that's too
> ambitious.  Is anyone out there using a Unix *without* NIS?  (Try 'man
> yp_get_default_domain' and see if you get anything, for a start.)  Are
> the NIS functions part of libc, or in a different library (such as
> libnsl on Solaris). 

Well;

$ man yp_get_default_domain
No manual entry for yp_get_default_domain
$ uname -a
Linux atrus.jesus.cam.ac.uk 2.2.14-1.1.0 #1 Thu Jan 6 05:12:58 EST 2000 i686 unknown

And nismodule.c didn't build the last time I tried it... I suppose I
could install the yp-tools rpm, but I'm not in any hurry to...

Cheers,
M.

-- 
  First time I've gotten a programming job that required a drug
  test. I was worried they were going to say "you don't have enough
  LSD in your system to do Unix programming".          -- Paul Tomblin
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From mark at chem.uwa.edu.au  Mon Feb 19 08:49:37 2001
From: mark at chem.uwa.edu.au (Mark C Favas)
Date: 19 Feb 01 07:49:37 GMT
Subject: How widespread is NIS support?
References: <3d1ysv8v7e.fsf@ute.cnri.reston.va.us>
Message-ID: <mark.982568977@declan>

[Andrew Kuchling asks about NIS support on Unix]

>Patch #103544 changes the 2.1 setup.py script to compile the NIS
>module on any Unix platform, and I'm wondering if that's too
>ambitious.  Is anyone out there using a Unix *without* NIS?  (Try 'man
>yp_get_default_domain' and see if you get anything, for a start.)  Are
>the NIS functions part of libc, or in a different library (such as
>libnsl on Solaris). 

I think most variants have NIS support. In Tru64 (nee Digital Unix or OSF1),
the NIS functions are part of libc, with include files in
/usr/include/rpcsvc/: yp_prot.h, ypclnt.h, ypdefs.h, yppasswd.h, ypsym.h,
ypupdate_prot.h, ypv1_prot.h and ypxfrd.h.
--
Email  - mark at chem.uwa.edu.au      ,-_|\                           Mark C Favas
Phone  - +61 9 380 3482           /     \               Department of Chemistry
Fax    - +61 9 380 1005      ---> *_,-._/   The University of Western Australia
                                       v                               Nedlands
Loc    - 31.97 S, 115.81 E                               Western Australia 6009


From andymac at bullseye.apana.org.au  Mon Feb 19 22:23:15 2001
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Tue, 20 Feb 2001 08:23:15 +1100 (EDT)
Subject: How widespread is NIS support?
In-Reply-To: <3d1ysv8v7e.fsf@ute.cnri.reston.va.us>
Message-ID: <Pine.OS2.4.32.0102200800580.382-100000@CENTRAL>

On 18 Feb 2001, Andrew Kuchling wrote:

> Does anyone have an idea how widespread NIS support is these days?
> Patch #103544 changes the 2.1 setup.py script to compile the NIS
> module on any Unix platform, and I'm wondering if that's too
> ambitious.  Is anyone out there using a Unix *without* NIS?  (Try 'man
> yp_get_default_domain' and see if you get anything, for a start.)  Are
> the NIS functions part of libc, or in a different library (such as
> libnsl on Solaris).

Be aware (you probably are...) that there's NIS and NIS+.

FreeBSD supports NIS (though not NIS+) but doesn't have the
'yp_get_default_domain' man page you refer to.  I think the other BSDs and
most Linux distros are in the same boat.

>From memory (some time ago...) an older HP-UX system (8.0?) had NIS
support, so I would expect most commercial Unixen to have this.

Given the above, 'yp_get_default_domain' may be a NIS+ call, although
FreeBSD's nis (actually yp(4)) man page notes that many NIS functions
don't have man pages - the *BSDs & Linux using a freely distributable
implementation of NIS.

AFAIK, NIS+ is Solaris only, but its possible that SGI and HP may have
licensed it as well (I have access to no system of either of these latter
two to check).  I'm not aware of any freely available/distributable
implemention of NIS+.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From aahz at panix.com  Tue Feb 20 06:16:12 2001
From: aahz at panix.com (Aahz Maruch)
Date: 19 Feb 2001 21:16:12 -0800
Subject: How widespread is NIS support?
References: <mailman.982639343.12125.python-list@python.org>
Message-ID: <96suis$2c1$1@panix3.panix.com>

In article <mailman.982639343.12125.python-list at python.org>,
Andrew MacIntyre  <andymac at bullseye.apana.org.au> wrote:
>
>FreeBSD supports NIS (though not NIS+) but doesn't have the
>'yp_get_default_domain' man page you refer to.  I think the other BSDs and
>most Linux distros are in the same boat.

Interesting.  My Panix account (NetBSD 1.4.3) has that man page, but
you're right about FreeBSD and the Linux on starship.python.net.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Laws that restrict speech are an abomination for a simple reason.
They're like a poison gas; they seem to be a good weapon to vanquish
your enemy, but the wind has a way of shifting."  -- Ira Glasser


From amk at mira.erols.com  Wed Feb 21 03:28:22 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 21 Feb 2001 02:28:22 GMT
Subject: How widespread is NIS support?
References: <mailman.982639343.12125.python-list@python.org>
Message-ID: <slrn996ap9.k4.amk@207-172-112-118.s118.tnt4.ann.va.dialup.rcn.com>

On Tue, 20 Feb 2001 08:23:15 +1100 (EDT), 
	Andrew MacIntyre <andymac at bullseye.apana.org.au> wrote:
>FreeBSD supports NIS (though not NIS+) but doesn't have the
>'yp_get_default_domain' man page you refer to.  I think the other BSDs and
>most Linux distros are in the same boat.

CNRI uses NIS+ internally.  A while ago, when we first began replacing
Solaris machines with Linux boxes, Greg Ward slammed his head against
configuring NIS+ support; there's a HOWTO that's far out of date, and
various hacked pieces of software to install.  Getting NIS working is
a matter of changing a few config files on current Linux versions; we
never managed to get NIS+ working at all.  

I'd be surprised if the module supports NIS+ at all.  Well, let's be
brave; for beta1 I'll enable NIS by default on all platforms, and see
where it breaks.

--amk


From bulbul at ucla.edu  Mon Feb 19 01:45:48 2001
From: bulbul at ucla.edu (Leston Buell)
Date: Sun, 18 Feb 2001 16:45:48 -0800
Subject: can't open older anydbm file
Message-ID: <3A906CBC.1060009@ucla.edu>

Under Python 1.4 which was running on my web host's server, i created a 
database file using anydbm. Using whichdb.whichdb on it, i learn that it 
is a dbhash file, which i understand is what bsddb uses. Now my account 
has been moved to a new server, which has python 1.5.2. Now i can no 
longer open the file. When i try in an interactive session, i get this:

 >>> import anydbm
 >>> db = anydbm.open( 'address.db', 'r' )
Traceback (innermost last):
   File "<stdin>", line 1, in ?
   File "/usr/lib/python1.5/anydbm.py", line 86, in open
     return mod.open(file, flag, mode)
   File "/usr/lib/python1.5/dbhash.py", line 8, in open
     return bsddb.hashopen(file, flag, mode)
bsddb.error: (22, 'Invalid argument')

?Is there anything i can do to open this file or to recover my data?

Thanks,

Leston



From phd at phd.pp.ru  Mon Feb 19 10:24:09 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 19 Feb 2001 12:24:09 +0300 (MSK)
Subject: can't open older anydbm file
In-Reply-To: <3A906CBC.1060009@ucla.edu>
Message-ID: <Pine.LNX.4.30.0102191223270.24446-100000@fep132.fep.ru>

On Sun, 18 Feb 2001, Leston Buell wrote:
> bsddb.error: (22, 'Invalid argument')
>
> ?Is there anything i can do to open this file or to recover my data?

   Either db_dump, db_dump185 or relink bsddb with older libs.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From costas at springmail.com  Mon Feb 19 04:34:55 2001
From: costas at springmail.com (Costas Menico)
Date: Mon, 19 Feb 2001 03:34:55 GMT
Subject: Is there an ISAPI interface for Python?
Message-ID: <3a90940f.2436689@News.CIS.DFN.DE>

Does anyone know if there is an ISAPI interface DLL that works with
Python? I am using IIS4.0 and Windows NT.

Thanks



From syver at NOSPAMcyberwatcher.com  Mon Feb 19 08:39:42 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Mon, 19 Feb 2001 08:39:42 +0100
Subject: Is there an ISAPI interface for Python?
References: <3a90940f.2436689@News.CIS.DFN.DE>
Message-ID: <W44k6.2533$X_2.64479@news1.oke.nextra.no>

I haven't found one, but I was working on one as a little side project. It
was starting to work out OK, but then I managed to screw it up one late
evening.

On the other hand the ASP ISAPI dll supports python, but I guess you were
talking about a custom ISAPI dll, weren't you?

The ISAPI extension I was making was supposed to mimic Python CGI as
accurately as possible so that you could use Python CGI scripts with it
without changing the .py file. Most of the dll revolves around handling one
python interpreter per thread and setting up the environment for these to
match the CGI enironment.

If you really need it I could see if I could get the project back on line.

"Costas Menico" <costas at springmail.com> wrote in message
news:3a90940f.2436689 at News.CIS.DFN.DE...
> Does anyone know if there is an ISAPI interface DLL that works with
> Python? I am using IIS4.0 and Windows NT.
>
> Thanks
>




From costas at springmail.com  Tue Feb 20 13:41:33 2001
From: costas at springmail.com (Costas Menico)
Date: Tue, 20 Feb 2001 12:41:33 GMT
Subject: Is there an ISAPI interface for Python?
References: <3a90940f.2436689@News.CIS.DFN.DE> <W44k6.2533$X_2.64479@news1.oke.nextra.no>
Message-ID: <3a92643f.610397@News.CIS.DFN.DE>

With everything Python hasy I was suprised to see that there isnt a
published ISAPI for Python. I have an ISAPI that I wrote for something
else. I was hoping to find a good and tested one for Python :) without
having to modify mine. May I ask you a couple of questions?

What method did you use to interface the ISAPI to Python? Which CGI
module are you using for testing? Also can you handle multiple
requests at the same time?

Maybe I could modify the one I have to do this.

thanks

costas

"Syver Enstad" <syver at NOSPAMcyberwatcher.com> wrote:

>I haven't found one, but I was working on one as a little side project. It
>was starting to work out OK, but then I managed to screw it up one late
>evening.
>
>On the other hand the ASP ISAPI dll supports python, but I guess you were
>talking about a custom ISAPI dll, weren't you?
>
>The ISAPI extension I was making was supposed to mimic Python CGI as
>accurately as possible so that you could use Python CGI scripts with it
>without changing the .py file. Most of the dll revolves around handling one
>python interpreter per thread and setting up the environment for these to
>match the CGI enironment.
>
>If you really need it I could see if I could get the project back on line.
>
>"Costas Menico" <costas at springmail.com> wrote in message
>news:3a90940f.2436689 at News.CIS.DFN.DE...
>> Does anyone know if there is an ISAPI interface DLL that works with
>> Python? I am using IIS4.0 and Windows NT.
>>
>> Thanks
>>
>
>



From costas at springmail.com  Mon Feb 19 05:22:34 2001
From: costas at springmail.com (Costas Menico)
Date: Mon, 19 Feb 2001 04:22:34 GMT
Subject: Why ASP <%@ %> does not work
Message-ID: <3a909e7f.5108679@News.CIS.DFN.DE>

Hi,

 I am trying to execute the following page in PWS which contains
Python ASP. However it does not execute to display the "text".
 I have installed the latest  ActivePython .  Is there some registry
setting that I should be dong? I know <SCRIPT language=Python> works.

<html>
<head></head>
<body>
<h1>Querying database</h1>
<%@ LANGUAGE = Python%>
<%
text="abc"
%>
<%= text%>
</body>
</html>

Any ideas?
Thanks


From yartz at imaginet.fr  Mon Feb 19 08:06:02 2001
From: yartz at imaginet.fr (Yann Schwartz)
Date: Mon, 19 Feb 2001 08:06:02 +0100
Subject: Why ASP <%@ %> does not work
References: <3a909e7f.5108679@News.CIS.DFN.DE>
Message-ID: <47F7E45B059247C9.6C3B173F0635E5FD.35F50673E3AFCE68@lp.airnews.net>

On Mon, 19 Feb 2001 04:22:34 GMT, costas at springmail.com (Costas
Menico) wrote:

>
>Hi,
>
> I am trying to execute the following page in PWS which contains
>Python ASP. However it does not execute to display the "text".
> I have installed the latest  ActivePython .  Is there some registry
>setting that I should be dong? I know <SCRIPT language=Python> works.
>
><html>
><head></head>
><body>
><h1>Querying database</h1>
><%@ LANGUAGE = Python%>
><%
>text="abc"
>%>
><%= text%>
></body>
></html>
>
>Any ideas?
>Thanks

The compilation directive
<%@LANGUAGE=FOO%>
must be the first line of the ASP script (even before any HTML). Using
Python or not has nothing to do with it.

Your page should look like :

><%@ LANGUAGE = Python%>
><html>
><head></head>
><body>
><h1>Querying database</h1>
><%
>text="abc"
>%>
><%= text%>
></body>
></html>


From gregj at pobox.com  Mon Feb 19 08:18:13 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Mon, 19 Feb 2001 07:18:13 GMT
Subject: Why ASP <%@ %> does not work
References: <3a909e7f.5108679@News.CIS.DFN.DE>
Message-ID: <3a90c719.49981769@news.potlnd1.or.home.com>

On Mon, 19 Feb 2001 04:22:34 GMT, costas at springmail.com (Costas
Menico) wrote:

> I am trying to execute the following page in PWS which contains
>Python ASP. However it does not execute to display the "text".
> I have installed the latest  ActivePython .  Is there some registry
>setting that I should be dong? I know <SCRIPT language=Python> works.
>
><html>
><head></head>
><body>
><h1>Querying database</h1>
><%@ LANGUAGE = Python%>
><%
>text="abc"
>%>
><%= text%>
></body>
></html>

Does your file have a .asp extension? It should.

Are you sure the Python is executing in the server, and not in the
client (which may be the same machine in your case)? Does this work:

<html>
<head></head>
<body>
<script language=Python runat=server>
Response.Write("<H2>hello, world</H2>")
</script>
</body>
</html>

If it does, then this should work, too:

<%@ language=Python %>
<html>
<head></head>
<body>
<%
Response.Write("<H2>hello, world</H2>")
%>
</body>
</html>

It works for me without any trouble.

Greg Jorgensen
Deschooling Society
Portland, Oregon USA


From costas at springmail.com  Tue Feb 20 13:53:40 2001
From: costas at springmail.com (Costas Menico)
Date: Tue, 20 Feb 2001 12:53:40 GMT
Subject: Why ASP <%@ %> does not work
References: <3a909e7f.5108679@News.CIS.DFN.DE> <3a90c719.49981769@news.potlnd1.or.home.com>
Message-ID: <3a9267c3.1510596@News.CIS.DFN.DE>

That worked. Thanks to both of you for the help. Does this also mean
that it uses the ASP/ISAPI interface and CGI is not really needed? 

gregj at pobox.com (Greg Jorgensen) wrote:

>On Mon, 19 Feb 2001 04:22:34 GMT, costas at springmail.com (Costas
>Menico) wrote:
>
>> I am trying to execute the following page in PWS which contains
>>Python ASP. However it does not execute to display the "text".
>> I have installed the latest  ActivePython .  Is there some registry
>>setting that I should be dong? I know <SCRIPT language=Python> works.
>>
>><html>
>><head></head>
>><body>
>><h1>Querying database</h1>
>><%@ LANGUAGE = Python%>
>><%
>>text="abc"
>>%>
>><%= text%>
>></body>
>></html>
>
>Does your file have a .asp extension? It should.
>
>Are you sure the Python is executing in the server, and not in the
>client (which may be the same machine in your case)? Does this work:
>
><html>
><head></head>
><body>
><script language=Python runat=server>
>Response.Write("<H2>hello, world</H2>")
></script>
></body>
></html>
>
>If it does, then this should work, too:
>
><%@ language=Python %>
><html>
><head></head>
><body>
><%
>Response.Write("<H2>hello, world</H2>")
>%>
></body>
></html>
>
>It works for me without any trouble.
>
>Greg Jorgensen
>Deschooling Society
>Portland, Oregon USA



From costas at springmail.com  Mon Feb 19 05:48:12 2001
From: costas at springmail.com (Costas Menico)
Date: Mon, 19 Feb 2001 04:48:12 GMT
Subject: Bug: Bad link in help file
Message-ID: <3a90a547.6844905@News.CIS.DFN.DE>

This link is messed up and I cant get to the scripting setup info.

D:\Python20\win32comext\axscript\demos\client\ie\demo.htm

if you click on Introduction it keeps you on the same page...\

costas


From jurgen.defurne at philips.com  Mon Feb 19 09:12:40 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Mon, 19 Feb 2001 09:12:40 +0100
Subject: Huuuuge memory leak (third attempt)
Message-ID: <0056900015996553000002L032*@MHS>

Thanks, everybody,

for the comments on the multiple assignment. Unfortunately,
that was the least of my problems.

I would urge everybody who runs Python 2.0 under Windows NT,
to try the following script :

<Begin code>
#

# External modules
import os
import sys

from time import *

if len(sys.argv) == 1:
    while 1:
	# Doesn't matter if P_NOWAIT or P_DETACH, tried 'em both
	os.spawnl(os.P_DETACH, 'd:\\python20\\python', 'python', 'cron.py', 'check')
elif sys.argv[1] == 'check':
    pass
<End code>

This is basically the core of a Python cron daemon (for Win), which is unfortunately unusable
at the moment because of the memory leak spawn seems to generate.

Who maintains the Win version of Python ? Maybe I should mail this to that person ?

Jurgen


From fredrik at pythonware.com  Mon Feb 19 10:03:43 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Feb 2001 09:03:43 GMT
Subject: Huuuuge memory leak (third attempt)
References: <mailman.982569802.25543.python-list@python.org>
Message-ID: <Pj5k6.18246$AH6.2511331@newsc.telia.net>

jurgen.defurne at philips.com wrote:
> I would urge everybody who runs Python 2.0 under Windows NT,
> to try the following script:

It's not leaking memory, it's leaking processes.  By design.

On my old slow Windows box, your program creates 50 new pro-
cesses per second.  Each process needs about 0.12 CPU seconds
to finish.  Each Python process uses around 800k of non-shared
memory.  You can figure the rest out yourself...

Cheers /F




From jepler at inetnebr.com  Tue Feb 20 00:32:56 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Mon, 19 Feb 2001 23:32:56 GMT
Subject: Huuuuge memory leak (third attempt)
References: <mailman.982569802.25543.python-list@python.org>
Message-ID: <slrn993b96.d33.jepler@potty.housenet>

On Mon, 19 Feb 2001 09:12:40 +0100, jurgen.defurne at philips.com
 <jurgen.defurne at philips.com> wrote:
>Thanks, everybody,
>
>for the comments on the multiple assignment. Unfortunately,
>that was the least of my problems.

I would suggest entering this bug in the bug tracking system at
	http://sourceforge.net/projects/python/
This may be a genuine bug.

Another thing you could do is try the equivalent program written in
C on Windows NT, to see if it exhibits the problem.  It is possible
that this is a bug in Windows, not in Python.

I don't know what the equivalent of "/bin/false" is in Windows NT, but
does the result change if a more trivial program is executed?

Running the following code on Linux/Python 2.0, I soon get an error
due to the number of "zombie" processes.  "zombie" processes are those
whose exit codes have not been read by the waitpid() or wait() functions
in the os module.  Is something like "waitpid()" necessary/possible under
NT?  Adding waitpid() lets it run seemingly forever without error.

import os, time

while 1:
        pid = os.spawnv(os.P_NOWAIT, "false", "/bin/false")
        #print pid, os.waitpid(pid, 0)

Jeff


From jurgen.defurne at philips.com  Tue Feb 20 14:54:57 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Tue, 20 Feb 2001 14:54:57 +0100
Subject: Huuuuge memory leak (third attempt)
Message-ID: <0056900016050184000002L042*@MHS>

If I were using Linux/Python 2.0, I would preferably use fork()/exec().
Unfortunately, that is not the case.

The MS documentation about spawn does not mention anything about
waiting for termination of the proces. Besides, the _cwait call, which
acts on the process handle, is not implemented in Python, os.waitpid()
is only available on U*ix systems.

So, yes, I will try to write the same functionality in C, and look what that
gives.

Thank you very much.

Jurgen




jepler at inetnebr.com@SMTP at python.org on 20/02/2001 14:31:50
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	Re: Huuuuge memory leak (third attempt)
Classification:	

On Mon, 19 Feb 2001 09:12:40 +0100, jurgen.defurne at philips.com
 <jurgen.defurne at philips.com> wrote:
>Thanks, everybody,
>
>for the comments on the multiple assignment. Unfortunately,
>that was the least of my problems.

I would suggest entering this bug in the bug tracking system at
	http://sourceforge.net/projects/python/
This may be a genuine bug.

Another thing you could do is try the equivalent program written in
C on Windows NT, to see if it exhibits the problem.  It is possible
that this is a bug in Windows, not in Python.

I don't know what the equivalent of "/bin/false" is in Windows NT, but
does the result change if a more trivial program is executed?

Running the following code on Linux/Python 2.0, I soon get an error
due to the number of "zombie" processes.  "zombie" processes are those
whose exit codes have not been read by the waitpid() or wait() functions
in the os module.  Is something like "waitpid()" necessary/possible under
NT?  Adding waitpid() lets it run seemingly forever without error.

import os, time

while 1:
        pid = os.spawnv(os.P_NOWAIT, "false", "/bin/false")
        #print pid, os.waitpid(pid, 0)

Jeff
--
http://mail.python.org/mailman/listinfo/python-list





From fredrik at pythonware.com  Mon Feb 19 09:24:03 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Feb 2001 08:24:03 GMT
Subject: Way to return a dictionary from re.findall?
References: <nqs09t040ckdut25j311ibi19i6dsouukg@4ax.com>
Message-ID: <DK4k6.18241$AH6.2510468@newsc.telia.net>

"Veronica Esquire" wrote:
> Is there a way to get back a dictionary of findall results that is
> based on the regex group names in python's re module?

no.  use a loop.

Cheers /F




From akhar at videotron.ca  Mon Feb 19 09:43:37 2001
From: akhar at videotron.ca (akhar)
Date: Mon, 19 Feb 2001 00:43:37 -0800
Subject: custom gui?
Message-ID: <xw2k6.1848$zw6.48239@wagner.videotron.net>

How would I go about doing my own GUI? By that I mean what library do I use
pyQT, pyGTK, pyWxWindows or tkinter? and By custom I mean changing where the
open and close buttons are (appearance), how windows open and shut
(location), have clipboards ... I am prepared to program but don<t know
where to start .
regards
stephane




From claird at starbase.neosoft.com  Mon Feb 19 15:10:40 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Feb 2001 08:10:40 -0600
Subject: custom gui?
References: <xw2k6.1848$zw6.48239@wagner.videotron.net>
Message-ID: <D884484C20316192.91F96B7FD730D3E0.2E009C3A2A9237DA@lp.airnews.net>

In article <xw2k6.1848$zw6.48239 at wagner.videotron.net>,
akhar <akhar at videotron.ca> wrote:
>How would I go about doing my own GUI? By that I mean what library do I use
>pyQT, pyGTK, pyWxWindows or tkinter? and By custom I mean changing where the
>open and close buttons are (appearance), how windows open and shut
>(location), have clipboards ... I am prepared to program but don<t know
>where to start .
			.
			.
			.
It depends.  I find your question ambiguous.  One possi-
bility is to buy or borrow John Grayson's celebrated book
<URL: http://www.manning.com/Grayson/ > and learn from it.
While it uses Tkinter, much of the instruction applies 
also with other bindings.  Is that the sort of answer
you're after?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From akhar at videotron.ca  Tue Feb 20 04:11:41 2001
From: akhar at videotron.ca (akhar)
Date: Mon, 19 Feb 2001 19:11:41 -0800
Subject: custom gui?
References: <xw2k6.1848$zw6.48239@wagner.videotron.net>
Message-ID: <bLik6.3517$zw6.74776@wagner.videotron.net>

What I meant was to be able to get away from the traditional square windows
and layout of frames.
Stephane
akhar <akhar at videotron.ca> wrote in message
news:xw2k6.1848$zw6.48239 at wagner.videotron.net...
> How would I go about doing my own GUI? By that I mean what library do I
use
> pyQT, pyGTK, pyWxWindows or tkinter? and By custom I mean changing where
the
> open and close buttons are (appearance), how windows open and shut
> (location), have clipboards ... I am prepared to program but don<t know
> where to start .
> regards
> stephane
>
>




From claird at starbase.neosoft.com  Wed Feb 21 01:04:40 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 20 Feb 2001 18:04:40 -0600
Subject: custom gui?
References: <xw2k6.1848$zw6.48239@wagner.videotron.net> <bLik6.3517$zw6.74776@wagner.videotron.net>
Message-ID: <1F1B8A7CFF73AFD4.03A50AAFC4AC473B.6EA500FCB7E4FD46@lp.airnews.net>

In article <bLik6.3517$zw6.74776 at wagner.videotron.net>,
akhar <akhar at videotron.ca> wrote:
>What I meant was to be able to get away from the traditional square windows
>and layout of frames.
			.
			.
			.
How far away?  I'm not getting it yet.  Is there
a positive example of what you're after?  Do you
want a virtual-reality sort of toolkit, or just
one that makes it easy to draw blobby sorts of
things?  Or are you saying that anything that's
not Visual Basic interests you?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From boud at rempt.xs4all.nl  Wed Feb 21 07:15:22 2001
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 21 Feb 2001 06:15:22 GMT
Subject: custom gui?
References: <xw2k6.1848$zw6.48239@wagner.videotron.net> <bLik6.3517$zw6.74776@wagner.videotron.net> <1F1B8A7CFF73AFD4.03A50AAFC4AC473B.6EA500FCB7E4FD46@lp.airnews.net>
Message-ID: <96vmdq$f5$1@news1.xs4all.nl>

Cameron Laird <claird at starbase.neosoft.com> wrote:
> 			.
> How far away?  I'm not getting it yet.  Is there
> a positive example of what you're after?  Do you
> want a virtual-reality sort of toolkit, or just
> one that makes it easy to draw blobby sorts of
> things?  Or are you saying that anything that's
> not Visual Basic interests you?

I get the feeling he's after the kind of things modern mp3-players
do - windows in the shape of a snail and things like that.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From claird at starbase.neosoft.com  Wed Feb 21 17:41:43 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Feb 2001 10:41:43 -0600
Subject: custom gui?
References: <xw2k6.1848$zw6.48239@wagner.videotron.net> <bLik6.3517$zw6.74776@wagner.videotron.net> <1F1B8A7CFF73AFD4.03A50AAFC4AC473B.6EA500FCB7E4FD46@lp.airnews.net> <96vmdq$f5$1@news1.xs4all.nl>
Message-ID: <33AD0B12D8AC8640.93E4CFDD593CF2E8.A6B598462FA61951@lp.airnews.net>

In article <96vmdq$f5$1 at news1.xs4all.nl>,
Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>> 			.
>I get the feeling he's after the kind of things modern mp3-players
>do - windows in the shape of a snail and things like that.
			.
			.
			.
Yes.  What facilities are and will be available for
those who want windows with funny geometry?  The im-
mediate answer, of course, is that motivated people
can synthesize windows as canvas polygons.  I bet
that satisfies few, though.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From boud at rempt.xs4all.nl  Wed Feb 21 17:57:13 2001
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 21 Feb 2001 16:57:13 GMT
Subject: custom gui?
References: <xw2k6.1848$zw6.48239@wagner.videotron.net> <bLik6.3517$zw6.74776@wagner.videotron.net> <1F1B8A7CFF73AFD4.03A50AAFC4AC473B.6EA500FCB7E4FD46@lp.airnews.net> <96vmdq$f5$1@news1.xs4all.nl> <33AD0B12D8AC8640.93E4CFDD593CF2E8.A6B598462FA61951@lp.airnews.net>
Message-ID: <970s19$3bo$1@news1.xs4all.nl>

Cameron Laird <claird at starbase.neosoft.com> wrote:

> Yes.  What facilities are and will be available for
> those who want windows with funny geometry?  The im-
> mediate answer, of course, is that motivated people
> can synthesize windows as canvas polygons.  I bet
> that satisfies few, though.


I doubt that there are cross-platform facilities available; but
perhaps the shape extensions for X could be wrapped? Myself, I really
loathe windows without decent borders.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From akhar at videotron.ca  Mon Feb 19 09:44:03 2001
From: akhar at videotron.ca (akhar)
Date: Mon, 19 Feb 2001 00:44:03 -0800
Subject: custom gui?
Message-ID: <xw2k6.1849$zw6.48239@wagner.videotron.net>

How would I go about doing my own GUI? By that I mean what library do I use
pyQT, pyGTK, pyWxWindows or tkinter? and By custom I mean changing where the
open and close buttons are (appearance), how windows open and shut
(location), have clipboards ... I am prepared to program but don<t know
where to start .
regards
stephane






From wware at world.std.com  Mon Feb 19 16:16:18 2001
From: wware at world.std.com (Will Ware)
Date: Mon, 19 Feb 2001 15:16:18 GMT
Subject: custom gui?
References: <xw2k6.1849$zw6.48239@wagner.videotron.net>
Message-ID: <G90Fr6.1Et@world.std.com>

akhar (akhar at videotron.ca) wrote:
> How would I go about doing my own GUI? By that I mean what library do I use
> pyQT, pyGTK, pyWxWindows or tkinter? and By custom I mean changing where the
> open and close buttons are (appearance), how windows open and shut
> (location), have clipboards ... I am prepared to program but don<t know
> where to start .
> regards
> stephane

If you're willing to use GTK then you might want to look into a
GUI-designer-thing called Glade. It allows you to plunk down GUI
objects (windows, buttons, canvases, sliders, etc) and it saves
the layout as an XML file (if you aren't familiar with XML, think
of it as a generalization of HTML). There have been several efforts
to write translators that convert this XML file into Python code
suitable for use with PyGTK:
http://glc.sourceforge.net/
http://www.baypiggies.org/stubby.py
http://www.fcoutant.freesurf.fr/gladepyc.html



-- 
-----------------------------------+---------------------
 22nd century: Esperanto, geodesic | Will Ware
 domes, hovercrafts, metric system | wware at world.std.com


From bryan at eevolved.com  Tue Feb 20 03:59:18 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Tue, 20 Feb 2001 02:59:18 GMT
Subject: custom gui?
References: <xw2k6.1849$zw6.48239@wagner.videotron.net>
Message-ID: <a4lk6.259644$f36.10175244@news20.bellglobal.com>

akhar wrote:

> How would I go about doing my own GUI? By that I mean what library do I
> use pyQT, pyGTK, pyWxWindows or tkinter? and By custom I mean changing
> where the open and close buttons are (appearance), how windows open and
> shut (location), have clipboards ... I am prepared to program but don<t
> know where to start .
> regards
> stephane
> 
> 
> 
> 

One more that may be hidden:

FOX: (Free objects for X)
http://sourceforge.net/projects/fxpy

Having used all of the above GUI kits, I would recommend pyQT if you never 
intend to make Windows sofware. Otherwise, its licensing is such as to 
preclude you from ever releasing free software on that platform. Fox is 
LGPL and will do everything you need. It's just a little trickier than in 
QT. 

Check out the "themes" example in the pyQT bundle. That should give you a 
kickstart.

-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"Scientific beliefs are supported by evidence, and they get results. Myths 
and faiths are not and do not" -- Richard Dawkins



From phys137 at mailcity.com  Tue Feb 20 05:07:20 2001
From: phys137 at mailcity.com (phys137 at mailcity.com)
Date: Tue, 20 Feb 2001 15:07:20 +1100
Subject: custom gui?
References: <xw2k6.1849$zw6.48239@wagner.videotron.net>
Message-ID: <96sqh9$os0$1@mail.pl.unisys.com>

as in thread 'python 3000'

1. It seems that there are several Python IDEs in which many smart people
invested plenty of effort. We have WinPython, IDLE (do we still have it ?
;-)), ActiveState is coming up with Komodo (build on Mozilla) and I probably
missed a few here - apologies. Plus we have commercial ones.

2. Competition is good, splitting efforts probably less good. And there's a
fine line between them.

3. My abbreviated personal view about technology directions: Win will rule
the desktop, not only because of it's current popularity - but because it
has component model.  Linux doesn't have a native object model - which
creates scalability (in size and functionality) problems for applications.
Java on Linux will provide the component model but only for the Java world,
not Linux.

4. Good and lasting IDE is about modularity and ease of extension. While I
applaud the efforts of ActiveState I'm not sure that using Mozilla gives
enough fuel for the long run.

5. www.netbeans.org - please, please, please read the whitepaper on that!

6. Essentially this is an open source IDE for Java - but can be used for
most anything else. It's highly modular - you can customize/replace editor
or file system etc (see 5). There is some fantastic stuff there as well as
plenty of good design. Jython folks obviously are already having a peek at
it (naturally) - but this could be used for general Python, too. And used on
both platforms - NT and UNIX.

7. I'm not sure how well Netbeans functionality can be blended with
winpython. but then, i suspect that winpython functionality will blend with
.NET anyway.



"akhar" <akhar at videotron.ca> wrote in message
news:xw2k6.1849$zw6.48239 at wagner.videotron.net...
> How would I go about doing my own GUI? By that I mean what library do I
use
> pyQT, pyGTK, pyWxWindows or tkinter? and By custom I mean changing where
the
> open and close buttons are (appearance), how windows open and shut
> (location), have clipboards ... I am prepared to program but don<t know
> where to start .
> regards
> stephane
>
>
>
>




From thorben.schulte at dcc-asia.de  Mon Feb 19 11:09:04 2001
From: thorben.schulte at dcc-asia.de (Thorben Schulte)
Date: Mon, 19 Feb 2001 11:09:04 +0100
Subject: zope & python : handle ZObject in the OSF
Message-ID: <96qrc1$1bp$00$1@news.t-online.com>

Hi all!

I have a small problem:

I would like to write an external method to move a zope-object from one
folder  to another.
My problem ist that i have no idear how to read and write object stored in
the OSF.

The backgroud is that "old" documents should move to an archive folder.

THX a lot, hope you will unterstand my english.

Greeting from Hannover,
Thorben




From robin at jessikat.fsnet.co.uk  Mon Feb 19 11:27:41 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Mon, 19 Feb 2001 10:27:41 +0000
Subject: Cyclops 2.0
Message-ID: <BruR9XAdUPk6EwD3@jessikat.demon.co.uk>

I suppose a lot of the need for T Peters' Cyclops module has gone away
with gc in Python 2.0, but Cyclops provide a convenient framework for
the analysis. Is there any equivalent for 2.0? If I just use Cyclops is
it telling me a lot more than the gc.garbage list?
-- 
Robin Becker


From loewis at informatik.hu-berlin.de  Tue Feb 20 13:48:54 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 20 Feb 2001 13:48:54 +0100
Subject: Cyclops 2.0
References: <BruR9XAdUPk6EwD3@jessikat.demon.co.uk>
Message-ID: <p6q4rxpmseh.fsf@informatik.hu-berlin.de>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> I suppose a lot of the need for T Peters' Cyclops module has gone away
> with gc in Python 2.0, but Cyclops provide a convenient framework for
> the analysis. Is there any equivalent for 2.0? If I just use Cyclops is
> it telling me a lot more than the gc.garbage list?

I have a patch on SF which adds a function gc.getreferents, which
finds all objects that refer to a given object. That is quite helpful
in finding the cause of a cycle.

Regards,
Martin


From tim_one at email.msn.com  Wed Feb 21 07:06:57 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 21 Feb 2001 01:06:57 -0500
Subject: Cyclops 2.0
In-Reply-To: <BruR9XAdUPk6EwD3@jessikat.demon.co.uk>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEODJAAA.tim_one@email.msn.com>

[Robin Becker]
> I suppose a lot of the need for T Peters' Cyclops module has gone away
> with gc in Python 2.0,

Depends on whether you care whether you're creating cycles.  There's less
need to care now, but objects with __del__ methods in cycles still aren't
reclaimed by magic.  I personally care about the latter only to the extent
that I would consider it to be a bug in my code if I ever created such a
thing ("a bug" because I intend never *to* create such a thing).

> but Cyclops provide a convenient framework for the analysis. Is
> there any equivalent for 2.0?

Sure:  Cyclops <0.9 wink>.  Nothing in 2.0 or 2.1 broke it, AFAIK.

> If I just use Cyclops is it telling me a lot more than the
> gc.garbage list?

Depends on *how* you use Cyclops.  It has methods for delivering many kinds
of info, from grouping by connection type to partitioning into strongly
connected components.  It can also be taught how to chase new kinds of
objects without delving into C.  gc.garbage is just a list, and of the kinds
of objects gc knows how to chase on its own.  If "just a list" is all you
need, cool.




From rich_somerfield at tertio.com  Mon Feb 19 11:33:43 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Mon, 19 Feb 2001 10:33:43 -0000
Subject: Calling C from Python
Message-ID: <dE6k6.30366$mO6.567706@news6-win.server.ntlworld.com>

Hi,
I have written a Game Engine in Python (main allow simulation of Board Games
using a config file as the definition of the game). The problem is, is that
the AI section is too slow.  As the code needs to be generic (e.g. cant
specialise as the game type etc... is unknown until runtime - from config
file), i have been unable to optimise the code further.

I want to use Python as my main langauge and have certain code (where
performance is essential) written in C.  I have looked at the doc's and cant
get a feel for it.  Can anyone point me in the direction of some sample code
(complete - not segments!), or give me a little helping hand?

Cheers
Rich.




From dkuhlman at rexx.com  Mon Feb 19 18:57:18 2001
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Mon, 19 Feb 2001 17:57:18 -0000
Subject: Calling C from Python
References: <dE6k6.30366$mO6.567706@news6-win.server.ntlworld.com>
Message-ID: <982605440.846771@rexx.com>

Section 1.1 of "Extending and Embedding the Python Interpreter"
contains a simple but complete sample.

You can also produce more complicated examples with SWIG, which
will generate the wrapper functions for your C functions.  You can
find SWIG at http://www.swig.org.

Also, in the source distribution, look at Demo/embed/demo.c.  It
demonstrates both (1) embedding Python in a C application and (2)
extending that embedded Python interpreter with a module
implemented in C.  It's the second task that you seem to be asking
for.

Hope this helps.

  - Dave



Rich Somerfield <rich_somerfield at tertio.com> wrote:
> 
> Hi,
> I have written a Game Engine in Python (main allow simulation of Board Games
> using a config file as the definition of the game). The problem is, is that
> the AI section is too slow.  As the code needs to be generic (e.g. cant
> specialise as the game type etc... is unknown until runtime - from config
> file), i have been unable to optimise the code further.
> 
> I want to use Python as my main langauge and have certain code (where
> performance is essential) written in C.  I have looked at the doc's and cant
> get a feel for it.  Can anyone point me in the direction of some sample code
> (complete - not segments!), or give me a little helping hand?
> 
> Cheers
> Rich.
> 
> 

-- 
Dave Kuhlman
dkuhlman at rexx.com


From rich_somerfield at tertio.com  Tue Feb 20 12:01:11 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Tue, 20 Feb 2001 11:01:11 -0000
Subject: Calling C from Python
References: <dE6k6.30366$mO6.567706@news6-win.server.ntlworld.com> <982605440.846771@rexx.com>
Message-ID: <D7sk6.1419$MN.14213@news2-win.server.ntlworld.com>

Thanks for the info.

I attempted to complete the Calling Python from C [Demo/embed/demo.c] but
couldnt get it to compile.  Could u please help me with the command?
I was trying [on Unix] :
cc -o new

new.c -I/opt/Python-2.0/Include -I/opt/Python-2.0 -L/opt/Python-2.0 -L/opt/P
ython-2.0/Lib -L/opt/Python-2.0/Modules -lpython2.0

Where new.c is a copy of the demo.c file.

Regards
Rich




Dave Kuhlman <dkuhlman at rexx.com> wrote in message
news:982605440.846771 at rexx.com...
>
> Section 1.1 of "Extending and Embedding the Python Interpreter"
> contains a simple but complete sample.
>
> You can also produce more complicated examples with SWIG, which
> will generate the wrapper functions for your C functions.  You can
> find SWIG at http://www.swig.org.
>
> Also, in the source distribution, look at Demo/embed/demo.c.  It
> demonstrates both (1) embedding Python in a C application and (2)
> extending that embedded Python interpreter with a module
> implemented in C.  It's the second task that you seem to be asking
> for.
>
> Hope this helps.
>
>   - Dave
>
>
>
> Rich Somerfield <rich_somerfield at tertio.com> wrote:
> >
> > Hi,
> > I have written a Game Engine in Python (main allow simulation of Board
Games
> > using a config file as the definition of the game). The problem is, is
that
> > the AI section is too slow.  As the code needs to be generic (e.g. cant
> > specialise as the game type etc... is unknown until runtime - from
config
> > file), i have been unable to optimise the code further.
> >
> > I want to use Python as my main langauge and have certain code (where
> > performance is essential) written in C.  I have looked at the doc's and
cant
> > get a feel for it.  Can anyone point me in the direction of some sample
code
> > (complete - not segments!), or give me a little helping hand?
> >
> > Cheers
> > Rich.
> >
> >
>
> --
> Dave Kuhlman
> dkuhlman at rexx.com




From chris.gonnerman at usa.net  Wed Feb 21 04:35:51 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Tue, 20 Feb 2001 21:35:51 -0600
Subject: Calling C from Python
References: <dE6k6.30366$mO6.567706@news6-win.server.ntlworld.com> <982605440.846771@rexx.com> <D7sk6.1419$MN.14213@news2-win.server.ntlworld.com>
Message-ID: <007c01c09bb7$6e44d8a0$a100000a@local>

----- Original Message -----
From: "Rich Somerfield" <rich_somerfield at tertio.com>
Subject: Re: Calling C from Python


> I attempted to complete the Calling Python from C [Demo/embed/demo.c] but
> couldnt get it to compile.  Could u please help me with the command?
> I was trying [on Unix] :
> cc -o new
>
>
new.c -I/opt/Python-2.0/Include -I/opt/Python-2.0 -L/opt/Python-2.0 -L/opt/P
> ython-2.0/Lib -L/opt/Python-2.0/Modules -lpython2.0

I assume this was originally one line.  What does cc tell you?

For that matter, what "Unix" are you using?  It makes a big difference.





From chris.gonnerman at usa.net  Wed Feb 21 14:33:01 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Wed, 21 Feb 2001 07:33:01 -0600
Subject: Calling C from Python
References: <OFEA3EB6C8.339AA653-ON802569FA.003069BC@LocalDomain>
Message-ID: <000f01c09c0a$d44a0280$a100000a@local>

----- Original Message -----
From: <Rich_Somerfield at tertio.com>
Subject: Re: Calling C from Python


> Thanks for the reply.
>
> My OS is SunOS 5.6.  The message I get from CC (I am using Sun Workshop
> compiler (CC: WorkShop Compilers 5.0 98/12/15 C++ 5.0) is :

Ur.  Not familiar with SunOS anymore...

> CC -o new

new.c -I/opt/Python-2.0/Include -I/opt/Python-2.0 -L/opt/Python-2.0/Modules 
-L/opt/Python-2.0 -L/opt/Python-2.0/Lib -lpython2.0
>
> Undefined                       first referenced
>  symbol                             in file
> socket
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> getpeername
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> recv
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> getprotobyname
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> accept
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> gethostbyname_r
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> send
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> bind
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> setsockopt
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> getservbyname
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> getsockopt
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> inet_addr
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> sendto
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> inet_ntoa
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> shutdown
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> getsockname
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> gethostbyaddr_r
/opt/Python-2.0/libpython2.0.a(socketmodule.o)

This batch of errors is complaining that socket functionality is missing.  I
find this most puzzling.
On Linux, I don't need to load any additional libraries for socket support,
but I suppose on
SunOS you might need to.

> dlsym
/opt/Python-2.0/libpython2.0.a(dynload_shlib.o)  (symbol belongs to implicit
dependency /usr/lib/libdl.so.1)
> dlopen
/opt/Python-2.0/libpython2.0.a(dynload_shlib.o)  (symbol belongs to implicit
dependency /usr/lib/libdl.so.1)
> __eprintf
/opt/Python-2.0/libpython2.0.a(longobject.o)
> dlerror
/opt/Python-2.0/libpython2.0.a(dynload_shlib.o)  (symbol belongs to implicit
dependency /usr/lib/libdl.so.1)
> recvfrom
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> listen
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> connect
/opt/Python-2.0/libpython2.0.a(socketmodule.o)
> ld: fatal: Symbol referencing errors. No output written to new
>
> Do i need to include all of these library files or should the compiler
pick them up from the specified lib directory in the compile command?  Is
these
>  are specific list of libs that I need to include to compile for Python?

I'm not sure you need ALL of those options but they shouldn't hurt.  It
looks like you need an
additional -l... option or two, but I can't say for sure.

I hope there is a SunOS user out there reading this...

> Regards,
> RS.





From mwh21 at cam.ac.uk  Wed Feb 21 15:08:16 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 21 Feb 2001 14:08:16 +0000
Subject: Calling C from Python
References: <dE6k6.30366$mO6.567706@news6-win.server.ntlworld.com>
Message-ID: <m3bsrwt9gv.fsf@atrus.jesus.cam.ac.uk>

"Rich Somerfield" <rich_somerfield at tertio.com> writes:

> I have written a Game Engine in Python (main allow simulation of
> Board Games using a config file as the definition of the game). The
> problem is, is that the AI section is too slow.  As the code needs
> to be generic (e.g. cant specialise as the game type etc... is
> unknown until runtime - from config file), i have been unable to
> optimise the code further.
> 
> I want to use Python as my main langauge and have certain code
> (where performance is essential) written in C.  I have looked at the
> doc's and cant get a feel for it.  Can anyone point me in the
> direction of some sample code (complete - not segments!), or give me
> a little helping hand?

Have a look at:

http://www-jcsu.jesus.cam.ac.uk/~mwh21/hacks/ext/defining-new-types.html

It's not nearly finished yet, but you might find it helpful.

Cheers,
M.

-- 
  Darned confusing, unless you have that magic ingredient coffee, of
  which I can pay you Tuesday for a couple pounds of extra-special
  grind today.                           -- John Mitchell, 11 Jan 1999


From ungrzr2 at ubatxbat.pbz  Sun Feb 25 03:27:48 2001
From: ungrzr2 at ubatxbat.pbz ({-- Rot13 - Hateme)
Date: 25 Feb 2001 02:27:48 GMT
Subject: Calling C from Python
References: <dE6k6.30366$mO6.567706@news6-win.server.ntlworld.com>
Message-ID: <9053610F6YouAreNotMeYouKnow@205.252.144.34>

rich_somerfield at tertio.com (Rich Somerfield) wrote in
<dE6k6.30366$mO6.567706 at news6-win.server.ntlworld.com>: 

>Hi,
>I have written a Game Engine in Python (main allow simulation of Board
>Games using a config file as the definition of the game). The problem
>is, is that the AI section is too slow.  As the code needs to be generic
>(e.g. cant specialise as the game type etc... is unknown until runtime -
>from config file), i have been unable to optimise the code further.
>

I just found this dl module in python manual. I think it
is need for python2. I am very excited about this function!
It basically let you dynamically load a library and use it!
It is a unix module.

So no wrapper is needed for simple C function access. I 
waited it for so long!

Thank you very much for this module!


From thomas at xs4all.net  Sun Feb 25 21:31:36 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 25 Feb 2001 21:31:36 +0100
Subject: Calling C from Python
In-Reply-To: <9053610F6YouAreNotMeYouKnow@205.252.144.34>; from ungrzr2@ubatxbat.pbz on Sun, Feb 25, 2001 at 02:27:48AM +0000
References: <dE6k6.30366$mO6.567706@news6-win.server.ntlworld.com> <9053610F6YouAreNotMeYouKnow@205.252.144.34>
Message-ID: <20010225213136.I16781@xs4all.nl>

On Sun, Feb 25, 2001 at 02:27:48AM +0000, {-- Rot13 - Hateme wrote:

> I just found this dl module in python manual. I think it
> is need for python2. I am very excited about this function!
> It basically let you dynamically load a library and use it!
> It is a unix module.
> So no wrapper is needed for simple C function access. I 
> waited it for so long!

The 'dl' module has been around for a long time, at least since Python 1.1,
though it might have been named different in that distant a past. Note that
the 'dl' module isn't enabled by default, at least not until the upcoming
Python 2.1. The 'dl' module is also not terribly secure or pythonic: misuse
can easily lead to crashes or weird behaviour. And it won't work on systems
where a pointer, a normal integer and a long integer aren't of the same
size.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From guettli at interface-business.de  Mon Feb 19 12:22:29 2001
From: guettli at interface-business.de (Thomas Guettler)
Date: Mon, 19 Feb 2001 12:22:29 +0100
Subject: Newbie's Questions
Message-ID: <20010219122229.A28726@gurkensalat.interface-business.de>

Yesterday I had time to start learning python. I like it, after a half
hour I wasn't missing the "{..}" anymore. Here are some questions I
have:

Is there a way of checking the program when I start it.  So typos are
detected emidiately, and not when I enter the function.

Is there something like "use strict" in perl? I like to declare my
variables before I use them. (Again, to recognize typos)

Is there a way of reading a single character from stdin without
waiting for the return-key? (At best without curses)

Is there something like "exit" in C? Yes, I could just raise an
exception, but I prefere a way without a stacktrace.

-- 
Thomas Guettler
Office: <guettli at interface-business.de> www.interface-business.de
Private:<guettli at gmx.de>  http://yi.org/guettli




From aleaxit at yahoo.com  Mon Feb 19 13:26:31 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 19 Feb 2001 13:26:31 +0100
Subject: Newbie's Questions
References: <mailman.982581805.9060.python-list@python.org>
Message-ID: <96r3dn01u60@news1.newsguy.com>

"Thomas Guettler" <guettli at interface-business.de> wrote in message
news:mailman.982581805.9060.python-list at python.org...
> Yesterday I had time to start learning python. I like it, after a half

Bravo!-)

> hour I wasn't missing the "{..}" anymore. Here are some questions I
> have:
>
> Is there a way of checking the program when I start it.  So typos are
> detected emidiately, and not when I enter the function.

The syntax of the whole script IS checked when you start it,
since it's internally compiled then; this will catch many
typos -- e.g., if you write:
    fi x>0:
rather than
    if x>0:
a SyntaxError will be generated immediately -- but not, of
course, all, e.g., if you write:
    x=y=z
rather than
    x=y+z
(also a frequent occurrence, since + and = are shifted
version of each other on many keyboards), this will not be
detected until you run suitable tests, as both forms happen
to be syntactically correct.

There is no way to even check for typos _that will
generate exceptions when the function is run_, without
actually running the function; e.g.,
    a=b/c
where one meant to write
    a=b/d
and c happens to be 0, will be caught when the function
is run (the division-by-zero attempt will raise an
exception), but not before then.  And similarly if c
happens to refer to a string (you can't divide by a
string), or if it happens to be unbound at that point;
only actual execution, not compilation, will reveal it.


> Is there something like "use strict" in perl? I like to declare my
> variables before I use them. (Again, to recognize typos)

There are no declarations in Python (the 'global' statement
_might_ be thought of as an exception to this assertion...).
If you're using a variable that's not bound at all, at run
time, or that is bound to a string where a number is needed
at that point, etc, etc, you'll get runtime exceptions; no
way to have the compiler diagnose it beforehand.


> Is there a way of reading a single character from stdin without
> waiting for the return-key? (At best without curses)

On Windows, only, the msvcrt module offers functions getch
(to read a single keypress, without waiting for a return, and
without echoing) and getche (ditto, with echoing); but they
don't work on standard-input (i.e., they are immune to stdin
redirections), but, rather, on the actual keyboard.

On Unix, you can use the low-level module termios, or the
higher-level one tty that in turn uses termios, to set the
mode to 'raw' for file-descriptor 0 (standard input).


> Is there something like "exit" in C? Yes, I could just raise an
> exception, but I prefere a way without a stacktrace.

Module sys contains function exit, which in turn raises a
SystemExit exception.  Either calling sys.exit(), or raising
SystemExit directly, will NOT generate a stacktrace -- but
the exception can be caught, so that try/finally and
try/except semantics are preserved.  If you need a way
to exit in an emergency, *without* proper/normal actions
on termination (no flushing of buffers, no finally blocks,
etc, etc), you can do that with function _exit of module
os, but that's a pretty-rare need ("should normally only
be used in the child process after a fork()" is what the
docs say, but os.fork is only available on Unix platforms,
while os._exit is also available on Windows).


Alex





From caumont at semaphora.com  Mon Feb 19 12:40:46 2001
From: caumont at semaphora.com (Laurent Caumont)
Date: Mon, 19 Feb 2001 12:40:46 +0100
Subject: minimum size python interpreter
Message-ID: <96r0ph$n0$1@wanadoo.fr>

Hello,

Where can I fund a minimum size Python Interpreter for Windows?

Thanks.




From claird at starbase.neosoft.com  Mon Feb 19 15:06:44 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 19 Feb 2001 08:06:44 -0600
Subject: minimum size python interpreter
References: <96r0ph$n0$1@wanadoo.fr>
Message-ID: <5B5C5CB509993043.252F85F5C2FE4A13.0E096A770F212271@lp.airnews.net>

In article <96r0ph$n0$1 at wanadoo.fr>,
Laurent Caumont <caumont at semaphora.com> wrote:
>
>Hello,
>
>Where can I fund a minimum size Python Interpreter for Windows?
			.
			.
			.
For what sense of "minimum"?  Perhaps "deeply embedded
python" <URL: http://www.abo.fi/~iporres/python/ >
will interest you.  Another possibility is to retrieve
a Py1.1 for DOS--is that the sort of thing you're after?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From info at pythonware.com  Mon Feb 19 12:44:14 2001
From: info at pythonware.com (PythonWare)
Date: Mon, 19 Feb 2001 12:44:14 +0100
Subject: ANN: SRE 2.1a2 (separate distribution)
Message-ID: <012f01c09a69$49bfe410$0900a8c0@SPIFF>

(trying again)
 
    Secret Labs' Regular Expression Engine (SRE) is the regular expression
    engine shipped with Python 1.6 and 2.0.  The latest SRE version (from
    Python 2.1a2) is now available separately:

        http://www.pythonware.com/products/sre

    Changes from the 2.0 version include a couple of bug fixes, and
    improved portability; this version works with Python 1.5.2 and
    later versions.

    The distribution includes library source code, and prebuilt Windows
    binaries for Python 1.5.2 and 2.0.

enjoy,
the pythonware team

    "Secret Labs -- makers of fine pythonware since 1997."





From aahz at panix.com  Mon Feb 19 18:01:53 2001
From: aahz at panix.com (Aahz Maruch)
Date: 19 Feb 2001 09:01:53 -0800
Subject: ANN: SRE 2.1a2 (separate distribution)
References: <mailman.982599001.13732.clpa-moderators@python.org>
Message-ID: <96rji1$18o$1@panix2.panix.com>

[originally posted in c.l.py.announce]

In article <mailman.982599001.13732.clpa-moderators at python.org>,
PythonWare <info at pythonware.com> wrote:
> 
>    Secret Labs' Regular Expression Engine (SRE) is the regular expression
>    engine shipped with Python 1.6 and 2.0.  The latest SRE version (from
>    Python 2.1a2) is now available separately:
>
>        http://www.pythonware.com/products/sre
>
>    Changes from the 2.0 version include a couple of bug fixes, and
>    improved portability; this version works with Python 1.5.2 and
>    later versions.

Very nice!  You are to be commended for your preservation of backwards
portability.  One thing that isn't clear from your announcement or web
site: how does the Unicode stuff work in Python 1.5.2?
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From fredrik at pythonware.com  Mon Feb 19 23:20:54 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Feb 2001 22:20:54 GMT
Subject: ANN: SRE 2.1a2 (separate distribution)
References: <mailman.982599001.13732.clpa-moderators@python.org> <96rji1$18o$1@panix2.panix.com>
Message-ID: <a%gk6.14543$Qb7.2318844@newsb.telia.net>

Aahz Maruch wrote:
> Very nice!  You are to be commended for your preservation of backwards
> portability.

thanks!

> One thing that isn't clear from your announcement or web
> site: how does the Unicode stuff work in Python 1.5.2?

not at all -- when compiled for 1.5.2, the core engine only
supports 8-bit strings (or compatible objects, like memory
mappings and arrays)

Cheers /F




From info at pythonware.com  Mon Feb 19 12:52:02 2001
From: info at pythonware.com (PythonWare)
Date: Mon, 19 Feb 2001 12:52:02 +0100
Subject: ANN: PIL 1.1.1 binaries for Windows
Message-ID: <014001c09a6a$72fbf660$0900a8c0@SPIFF>

at last:

   Secret Labs now offer "official" PIL builds for Windows.
   The binary kit is based on 1.1.1, and includes JPEG and
   PNG support, optional Tkinter support, and a patch for
   PNG transparency support:

        http://www.pythonware.com/products/pil

   Binary kits are available for Python 1.5.2, Python 2.0,
   and Python 2.1a2 (experimental).  The Tkinter binding
   requires Tcl/Tk 8.2.3 for the 1.5.2 versions, and 8.3.2
   for the 2.X versions.

   The patched PNG driver can be downloaded separately
   from http://effbot.org.

enjoy,
the pythonware team

    "Secret Labs -- makers of fine pythonware since 1997."




From dale at out-think.NOSPAMco.uk  Mon Feb 19 12:59:39 2001
From: dale at out-think.NOSPAMco.uk (Dale Strickland-Clark)
Date: Mon, 19 Feb 2001 11:59:39 +0000
Subject: Scanners and OCR engines
Message-ID: <fa229top5sn0364089508fhf1jss4kbqvg@4ax.com>

We need to automate some document management and I'm looking for a scanner and an OCR engine that I
can automate. 

A Python interface would be preferable but COM would be fine. Failing that a C API would do and we
could put our own object model around it.

Does anyone know of any automation friendly scanners and OCR engines?

TIA.
--
Dale Strickland-Clark
Out-Think Ltd
Business Technology Consultants


From astuart at NO.mira.SPAMnet  Mon Feb 19 13:08:05 2001
From: astuart at NO.mira.SPAMnet (Andrew)
Date: Mon, 19 Feb 2001 22:08:05 +1000
Subject: newbie - POPLIB, MIMETOOLS and SMTPLIB - aaarrrghh!
Message-ID: <3a90fe94.a405@newton.pacific.net.au>

Hi there

sorry for being such a newbie.

In summary, I am trying to write a script which reads a POP mailbox,
extracts the "subject" line from each message, and emails a message back to
the sender, echoing the "subject" line.  Weird problems everywhere!

I can get POPLIB to read emails, but I can't get MIMETOOLS to read the
emails - so there's no way for me to get the subject line out.
I can get SMTPLIB to send emails, but I can;t work out how to include a
subject line in the email.

I've looked for days throughout the Internet for solutions to these
seemingly simple problems but I can't find anything.

Why would it be so hard for MIMETOOLS to read messages that come from
POPLIB?  Surely they are designed to work together?

Why so hard to include a subject line using SMTPLIB?

Any help that can be offered by anyone is deeply appreciated.  I feel like a
total goof.  What am I missing about this?

Thanks in advance

Andrew Stuart
astuart at mira.net









From jhauser at ifm.uni-kiel.de  Mon Feb 19 12:37:31 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 19 Feb 2001 12:37:31 +0100
Subject: newbie - POPLIB, MIMETOOLS and SMTPLIB - aaarrrghh!
References: <3a90fe94.a405@newton.pacific.net.au>
Message-ID: <87d7ceykck.fsf@lisboa.ifm.uni-kiel.de>

"Andrew" <astuart at NO.mira.SPAMnet> writes:

> Hi there
> 
> sorry for being such a newbie.
> 
> In summary, I am trying to write a script which reads a POP mailbox,
> extracts the "subject" line from each message, and emails a message back to
> the sender, echoing the "subject" line.  Weird problems everywhere!
> 
> I can get POPLIB to read emails, but I can't get MIMETOOLS to read the
> emails - so there's no way for me to get the subject line out.
> I can get SMTPLIB to send emails, but I can;t work out how to include a
> subject line in the email.
> 
> I've looked for days throughout the Internet for solutions to these
> seemingly simple problems but I can't find anything.
> 
> Why would it be so hard for MIMETOOLS to read messages that come from
> POPLIB?  Surely they are designed to work together?
I'm not clear, why you want to use mimetools to read a plain email.

An example for this is included in the poplib.py file
        a = POP3(TESTSERVER)
        print a.getwelcome()
        a.user(TESTACCOUNT)
        a.pass_(TESTPASSWORD)
        a.list()
        (numMsgs, totalSize) = a.stat()
        for i in range(1, numMsgs + 1):
                (header, msg, octets) = a.retr(i)
                print "Message ", `i`, ':'
                for line in msg:
                        print '   ' + line
                print '-----------------------'
        a.quit()

If this does not work some more information about your errors would be
interesting.

> 
> Why so hard to include a subject line using SMTPLIB?
For an example with subject line try this

>>> s=smtplib.SMTP("localhost")
>>> print s.sendmail.__doc__

HTH,
__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From barry at digicool.com  Tue Feb 20 17:09:46 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Tue, 20 Feb 2001 11:09:46 -0500
Subject: newbie - POPLIB, MIMETOOLS and SMTPLIB - aaarrrghh!
References: <3a90fe94.a405@newton.pacific.net.au>
Message-ID: <14994.38602.283549.763001@anthem.wooz.org>

>>>>> "A" == Andrew  <astuart at NO.mira.SPAMnet> writes:

    A> Why would it be so hard for MIMETOOLS to read messages that
    A> come from POPLIB?  Surely they are designed to work together?

I'm not familar with poplib, but I'd be interested to know if mimelib
works well with it (and if not what you'd suggest adding to mimelib).

mimelib 0.2 can be found at:

    http://barry.wooz.org/software/pyware.html

-Barry



From mix77 at usa.net  Mon Feb 19 13:10:34 2001
From: mix77 at usa.net (Mix)
Date: Mon, 19 Feb 2001 14:10:34 +0200
Subject: PIL (Python Image Library)
Message-ID: <3A910D3A.74602C0@usa.net>

I am trying to manipulate (rotate) an image using PIL(Python
Image Library). My problem is that I cannot get PIL to work
form a browser, i.e. in my cgi script I have

     import Image

and this gives the error :

        Traceback (innermost last):
              File "/home/mixo/faxserv/cgi-bin/t1.py", line 3, in ?
                  import Image
         ImportError: No module named Image

What is causing the error? If I run the script "t1.py" as any
valid user on the machine from the command line , I get disried
result.I  know PIL has been installed as is if I run the Python i
nterpreter I can manipulate images using PIL.
I




From rich_somerfield at tertio.com  Tue Feb 20 12:18:17 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Tue, 20 Feb 2001 11:18:17 -0000
Subject: PIL (Python Image Library)
References: <3A910D3A.74602C0@usa.net>
Message-ID: <Ensk6.1145$5n4.11539@news6-win.server.ntlworld.com>

Sounds like a problem with the PYTHONPATH, maybe this is not used in CGI
scripts?  I dont know but u might have to specify the full path of the
modules that u r attempting to import.  Or maybe a cheat would be to copy
all of the python scripts that u want to import into the CGI-BIN directory
(or the Python Interpreter directory).

Regards
Rich

Mix <mix77 at usa.net> wrote in message news:3A910D3A.74602C0 at usa.net...
> I am trying to manipulate (rotate) an image using PIL(Python
> Image Library). My problem is that I cannot get PIL to work
> form a browser, i.e. in my cgi script I have
>
>      import Image
>
> and this gives the error :
>
>         Traceback (innermost last):
>               File "/home/mixo/faxserv/cgi-bin/t1.py", line 3, in ?
>                   import Image
>          ImportError: No module named Image
>
> What is causing the error? If I run the script "t1.py" as any
> valid user on the machine from the command line , I get disried
> result.I  know PIL has been installed as is if I run the Python i
> nterpreter I can manipulate images using PIL.
> I
>
>




From jurgen.defurne at philips.com  Mon Feb 19 13:56:08 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Mon, 19 Feb 2001 13:56:08 +0100
Subject: Followup : Huuuuge memory leak (third attempt)
Message-ID: <0056890021971266000002L962*@MHS>

I checked using Python 2.1b2 (or whatever), the leak remains.

If it should be the case that this is a Windows NT bug, then I have
another reason to loathe MS.

It would have been real nice if I could have distributed Python for
Windows NT (easier to install than Cygwin) and created a system
so that the users themselves could install it on their machine and
update scripts from a central point. Since running scheduled jobs is
extremely important, a cron.py and atd.py daemon would be
mandatory. Unfortunately, due to this memory leak, this isn't a real
stable solution.

Jurgen


From mwh21 at cam.ac.uk  Mon Feb 19 14:08:26 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Feb 2001 13:08:26 +0000
Subject: Followup : Huuuuge memory leak (third attempt)
References: <mailman.982587321.31432.python-list@python.org>
Message-ID: <m3y9v2u8fp.fsf@atrus.jesus.cam.ac.uk>

jurgen.defurne at philips.com writes:

> I checked using Python 2.1b2 (or whatever), the leak remains.

What you posted before was in effect a mild fork bomb.  So you have
yet to convince anyone here that there is a real leak.  Try again?

Cheers,
M.

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme


From fredrik at pythonware.com  Mon Feb 19 14:27:21 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Feb 2001 13:27:21 GMT
Subject: Followup : Huuuuge memory leak (third attempt)
References: <mailman.982587321.31432.python-list@python.org>
Message-ID: <Za9k6.18261$AH6.2517513@newsc.telia.net>

jurgen.defurne at philips.com wrote:
> If it should be the case that this is a Windows NT bug, then I have
> another reason to loathe MS.

since this is a bug in your program, does that mean that
we can all loathe you?

Cheers /F




From scarblac at pino.selwerd.nl  Mon Feb 19 18:41:30 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 19 Feb 2001 17:41:30 GMT
Subject: Followup : Huuuuge memory leak (third attempt)
References: <mailman.982587321.31432.python-list@python.org>
Message-ID: <slrn992mhu.5th.scarblac@pino.selwerd.nl>

jurgen.defurne at philips.com wrote in comp.lang.python:
> I checked using Python 2.1b2 (or whatever), the leak remains.

Fredrik also said something like this, when you posted the code fragment:
what would you expect, if you started up programs and detached from them in
an infinite loop? That *would* fill up your memory real fast, wouldn't it?

-- 
Remco Gerlich


From jurgen.defurne at philips.com  Mon Feb 19 14:51:42 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Mon, 19 Feb 2001 14:51:42 +0100
Subject: Followup : Huuuuge memory leak (third attempt)
Message-ID: <0056900016013514000002L042*@MHS>

Since in the second version, the main loop sleeps long enough to make
sure that the spawned script is finished before another one is released,
this bomb has been defused, but the memory still mounts every time
a process is spawned.

Jurgen




mwh21 at cam.ac.uk@SMTP at python.org on 19/02/2001 14:36:56
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	Re: Followup : Huuuuge memory leak (third attempt)
Classification:	

jurgen.defurne at philips.com writes:

> I checked using Python 2.1b2 (or whatever), the leak remains.

What you posted before was in effect a mild fork bomb.  So you have
yet to convince anyone here that there is a real leak.  Try again?

Cheers,
M.

--
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme
--
http://mail.python.org/mailman/listinfo/python-list





From jurgen.defurne at philips.com  Mon Feb 19 14:52:52 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Mon, 19 Feb 2001 14:52:52 +0100
Subject: Followup : Huuuuge memory leak (third attempt)
Message-ID: <0056900016013594000002L042*@MHS>

---------------------- Forwarded by Jurgen Defurne/BRG/CE/PHILIPS on 19/02/2001 14:41 ---------------------------


Jurgen Defurne
19/02/2001 14:36
To:	sda at webde-ag.de@SMTP 
cc:	 
Subject:	Re: Followup : Huuuuge memory leak (third attempt)  
Classification:	Unclassified


One needs Administrator rights : this is absolutely no option here,
since the IT-management likes to play 'Big Brother is watching you'.
They Do Not Want ANYBODY to play Administrator on PC's.

(I thought that this would mentioned one day)

Jurgen




From fredrik at pythonware.com  Mon Feb 19 15:08:04 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Feb 2001 14:08:04 GMT
Subject: Followup : Huuuuge memory leak (third attempt)
References: <mailman.982590241.10789.python-list@python.org>
Message-ID: <8N9k6.18267$AH6.2517900@newsc.telia.net>

jurgen.defurne at philips.com wrote:
> Since in the second version, the main loop sleeps long enough to make
> sure that the spawned script is finished before another one is released,

doesn't look like it's sleeping:

if len(sys.argv) == 1:
    while 1:
        # Doesn't matter if P_NOWAIT or P_DETACH, tried 'em both
        os.spawnl(os.P_DETACH, 'd:\\python20\\python', 'python', 'cron.py', 'check')
        # no sleep here
elif sys.argv[1] == 'check':
    pass

adding a time.sleep(0.2) makes it work better -- and after 10 minutes,
the main process hasn't grown a single byte on my box...

Cheers /F




From jurgen.defurne at philips.com  Mon Feb 19 15:49:38 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Mon, 19 Feb 2001 15:49:38 +0100
Subject: Followup : Huuuuge memory leak (third attempt)
Message-ID: <0056900016016857000002L072*@MHS>

Is your box Win NT or Win95 ?
Are you using Task Manager ?

I monitor my processes and the used memory. It is true,
the memory of the process does not increase, but the used
memory does. That is why I more and more suspect that it
is an OS problem with the spawn calls.

Jurgen




fredrik at pythonware.com@SMTP at python.org on 19/02/2001 15:35:55
Please respond to fredrik at effbot.org@SMTP 
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	Re: Followup : Huuuuge memory leak (third attempt)
Classification:	

jurgen.defurne at philips.com wrote:
> Since in the second version, the main loop sleeps long enough to make
> sure that the spawned script is finished before another one is released,

doesn't look like it's sleeping:

if len(sys.argv) == 1:
    while 1:
        # Doesn't matter if P_NOWAIT or P_DETACH, tried 'em both
        os.spawnl(os.P_DETACH, 'd:\\python20\\python', 'python', 'cron.py', 'check')
        # no sleep here
elif sys.argv[1] == 'check':
    pass

adding a time.sleep(0.2) makes it work better -- and after 10 minutes,
the main process hasn't grown a single byte on my box...

Cheers /F


--
http://mail.python.org/mailman/listinfo/python-list





From maxm at normik.dk  Mon Feb 19 14:22:41 2001
From: maxm at normik.dk (=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?=)
Date: Mon, 19 Feb 2001 14:22:41 +0100
Subject: Is there a Python module for writing Midi Files
Message-ID: <7BD10B680501D411B9DF009027E06F32197C03@exchange>

I have looked all the usual places, Python.org, Google, the starship, etc.
But it is not possible to finde one. But I guess that it must exist
somewhere in someones private library.

I want to use Python for algorithmic composition, and then write to a Midi
file for importing into the reason software synthesizer for easy tweaking of
the sound.

Regards Max M



From wware at world.std.com  Mon Feb 19 16:07:05 2001
From: wware at world.std.com (Will Ware)
Date: Mon, 19 Feb 2001 15:07:05 GMT
Subject: Is there a Python module for writing Midi Files
References: <mailman.982589062.5445.python-list@python.org>
Message-ID: <G90FBt.Jz1@world.std.com>

=?iso-8859-1?Q?Max_M=F8ller_Rasmussen?= (maxm at normik.dk) wrote:
> I want to use Python for algorithmic composition, and then write
> to a Midi file...

I wrote some classes for reading MIDI files. I thought about adding
write methods to them but didn't need it at the time. Still, they
might be better than starting from nothing.
http://groups.google.com/groups?th=a11ca7899a622abb&rnum=7&seld=944315225&ic=1
Good luck.

-- 
-----------------------------------+---------------------
 22nd century: Esperanto, geodesic | Will Ware
 domes, hovercrafts, metric system | wware at world.std.com


From guettli at interface-business.de  Mon Feb 19 16:22:19 2001
From: guettli at interface-business.de (Thomas Guettler)
Date: Mon, 19 Feb 2001 16:22:19 +0100
Subject: Is there a Python module for writing Midi Files
In-Reply-To: <7BD10B680501D411B9DF009027E06F32197C03@exchange>; from maxm@normik.dk on Mon, Feb 19, 2001 at 02:22:41PM +0100
References: <7BD10B680501D411B9DF009027E06F32197C03@exchange>
Message-ID: <20010219162219.E28726@gurkensalat.interface-business.de>

On Mon, Feb 19, 2001 at 02:22:41PM +0100, Max M?ller Rasmussen wrote:
> I have looked all the usual places, Python.org, Google, the starship, etc.
> But it is not possible to finde one. But I guess that it must exist
> somewhere in someones private library.
> 
> I want to use Python for algorithmic composition, and then write to a Midi
> file for importing into the reason software synthesizer for easy tweaking of
> the sound.

try midithing at: http://quitte.de/midithing.html

-- 
Thomas Guettler
Office: <guettli at interface-business.de> www.interface-business.de
Private:<guettli at gmx.de>  http://yi.org/guettli




From arildh at stud.cs.uit.no  Mon Feb 19 14:27:16 2001
From: arildh at stud.cs.uit.no (Arild Hansen)
Date: Mon, 19 Feb 2001 14:27:16 +0100
Subject: Starting a third party application
Message-ID: <Pine.LNX.4.21.0102191423250.15777-100000@arildh.pasta.cs.uit.no>

I'm trying to make a small module that executes a third party application
(for example gcalc or whatever). I do this using
os.system("applicationname &") and then the program enters a while loop
waiting for some external signal. When this signal arrives, I want the
program to make a sys.exit(1) and terminate itself AS WELL AS the third
party application it started. The code looks something like:

os.system("applicationame &")
while continue:
	if signal:
		continue = FALSE


My problem is that when the program gets the signal and terminates, the
third party application is still running. How do I solve this problem?

All help greatly appreciated,

A. Hansen



From johnm at cs.uit.no  Mon Feb 19 16:10:06 2001
From: johnm at cs.uit.no (johnm at cs.uit.no)
Date: 19 Feb 2001 16:10:06 +0100
Subject: Starting a third party application
References: <Pine.LNX.4.21.0102191423250.15777-100000@arildh.pasta.cs.uit.no>
Message-ID: <hv4rxqzp2p.fsf@johnmnb.cs.uit.no>

Arild Hansen <arildh at stud.cs.uit.no> writes:

> I'm trying to make a small module that executes a third party application
> (for example gcalc or whatever). I do this using
> os.system("applicationname &") and then the program enters a while loop
> waiting for some external signal. When this signal arrives, I want the
> program to make a sys.exit(1) and terminate itself AS WELL AS the third
> party application it started. The code looks something like:
> 
> os.system("applicationame &")
> while continue:
> 	if signal:
> 		continue = FALSE
> 
> 
> My problem is that when the program gets the signal and terminates, the
> third party application is still running. How do I solve this problem?

Maybe you should take a look at fork, exec, kill and wait (or spawn if
you use one of the MS Windows variants). This example should do
something similar:

import os
import time

child = os.fork()
if child:
    print "blabla"
    time.sleep(2)
    os.kill(child, 3) # or some other signal
    os.wait(child)
else:
    print "Child, starting /bin/sleep"
    os.execv("/bin/sleep", ["/bin/sleep", "32"])
    os.exit(-1) # Shouldn't happen, but in case exec fails



From arildh at stud.cs.uit.no  Tue Feb 20 09:31:36 2001
From: arildh at stud.cs.uit.no (Arild Hansen)
Date: Tue, 20 Feb 2001 09:31:36 +0100
Subject: Starting a third party application
In-Reply-To: <hv4rxqzp2p.fsf@johnmnb.cs.uit.no>
References: <Pine.LNX.4.21.0102191423250.15777-100000@arildh.pasta.cs.uit.no> <hv4rxqzp2p.fsf@johnmnb.cs.uit.no>
Message-ID: <Pine.LNX.4.21.0102200928070.17555-100000@arildh.pasta.cs.uit.no>

Thx, worked just fine! I guess the "cs.uit.no" means that I didn't have to
look too far to get an answer :) 

A. Hansen


On 19 Feb 2001 johnm at cs.uit.no wrote:

> Arild Hansen <arildh at stud.cs.uit.no> writes:
> 
> > I'm trying to make a small module that executes a third party application
> > (for example gcalc or whatever). I do this using
> > os.system("applicationname &") and then the program enters a while loop
> > waiting for some external signal. When this signal arrives, I want the
> > program to make a sys.exit(1) and terminate itself AS WELL AS the third
> > party application it started. The code looks something like:
> > 
> > os.system("applicationame &")
> > while continue:
> > 	if signal:
> > 		continue = FALSE
> > 
> > 
> > My problem is that when the program gets the signal and terminates, the
> > third party application is still running. How do I solve this problem?
> 
> Maybe you should take a look at fork, exec, kill and wait (or spawn if
> you use one of the MS Windows variants). This example should do
> something similar:
> 
> import os
> import time
> 
> child = os.fork()
> if child:
>     print "blabla"
>     time.sleep(2)
>     os.kill(child, 3) # or some other signal
>     os.wait(child)
> else:
>     print "Child, starting /bin/sleep"
>     os.execv("/bin/sleep", ["/bin/sleep", "32"])
>     os.exit(-1) # Shouldn't happen, but in case exec fails
> 
> 



From alet at unice.fr  Mon Feb 19 15:08:24 2001
From: alet at unice.fr (Jerome Alet)
Date: Mon, 19 Feb 2001 15:08:24 +0100
Subject: [ANNOUNCE] New version of the JAXML module
Message-ID: <3A9128D8.35D0EE22@unice.fr>

I'm very pleased to announce you the new 2.10 version of the JAXML
module.

With the JAXML module, generating XML documents has never been easier,
this module:

        * Takes care of the indentation for you: your XML documents are
          readable by humans.
        * Takes care of closing tags for you.
        * Can generate every tag you want in an incredibly easy way.
        * Offers two templating facilities (see below)
        * Can generate XHTML documents and be used in CGI scripts
because it
          takes care of the HTTP headers too.

Here's a sample:

--- CUT ---
import jaxml
x = jaxml.XML_document()
x.something(value=5).dummy("some dummy text", isntit="YES")
print x

will produce:

<?xml version="1.0" encoding="iso-8859-1"?>
<something value="5">
    <dummy isntit="YES">some dummy text</dummy>
</something>
--- CUT ---

The JAXML module is GPLed, you can download it freely from:

        http://cortex.unice.fr/~jerome/jaxml/

Changes:

        * A new and very powerful templating facility was added: there
is now
          two templating facilities in JAXML
        * Better code documentation.
        * When testing truth value on documents, returns false on empty
          documents and true on not empty ones.

The well commented test program and module's sources are your best
friends.

Comments are very welcome.

-- 
Jerome Alet - alet at unice.fr - http://cortex.unice.fr/~jerome
Fac de Medecine de Nice        http://wwwmed.unice.fr 
Tel: (+33) 4 93 37 76 30     Fax: (+33) 4 93 53 15 15
28 Avenue de Valombrose - 06107 NICE Cedex 2 - FRANCE


From gerhard.nospam at bigfoot.de  Mon Feb 19 16:42:38 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Mon, 19 Feb 2001 16:42:38 +0100
Subject: [ANNOUNCE] New version of the JAXML module
References: <3A9128D8.35D0EE22@unice.fr>
Message-ID: <slrn992gj2.j5.gerhard.nospam@lilith.hqd-internal>

On Mon, 19 Feb 2001 15:08:24 +0100, Jerome Alet <alet at unice.fr> wrote:
>I'm very pleased to announce you the new 2.10 version of the JAXML
>module.
>
>With the JAXML module, generating XML documents has never been easier,
>this module:
>
> [snip]
>
>The JAXML module is GPLed, you can download it freely from:
>
>        http://cortex.unice.fr/~jerome/jaxml/
>
> [snip]
>
>Comments are very welcome.

Could you please change the license to LGPL or another more liberal license?
This would make your module usable by a lot more people. Putting the GPL on
a library prevents it from being used in a commercial application.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de
web:    http://highqualdev.com


From alet at unice.fr  Tue Feb 20 10:00:59 2001
From: alet at unice.fr (Jerome Alet)
Date: Tue, 20 Feb 2001 10:00:59 +0100
Subject: [ANNOUNCE] New version of the JAXML module
References: <3A9128D8.35D0EE22@unice.fr> <slrn992gj2.j5.gerhard.nospam@lilith.hqd-internal>
Message-ID: <3A92324B.3F7CD22D@unice.fr>

Gerhard H?ring wrote:
> 
> On Mon, 19 Feb 2001 15:08:24 +0100, Jerome Alet <alet at unice.fr> wrote:
> >The JAXML module is GPLed, you can download it freely from:
> >
> >        http://cortex.unice.fr/~jerome/jaxml/

> Could you please change the license to LGPL or another more liberal license?
> This would make your module usable by a lot more people. Putting the GPL on
> a library prevents it from being used in a commercial application.

Several people have already asked the same thing...

The purpose of the GPL on this module is not to forbid you to use it in
a commercial application, but to forbid you to use it in a proprietary
application, which is very different.

If your commercial application is delivered under the GPL then there's
no problem: feel free to sell it, and sell my module if you want. But
you must also give the same rights to
the people you sell it.

Despite its very small size, I think this module is (or could be made)
powerful, and it took me some time to write it, so I don't want it to be
used in non-Free Software. I think (maybe I'm wrong) that its simplicity
of use compared to the official (and of course more complete) XML
modules is attractive and is an advantage given to the Free Software
world.

Since I want to give my copyright on this module to the FSF, their
policy might be different and they could decide to put it under the
LGPL, but now it is GPLed.

So, not to be rude, but your only options are:

	* use this module internally and don't redistribute your work
	* distribute your application under the GPL licence (best choice for
people)
	* don't use my module (probably your best choice for a proprietary
application)

bye,

-- 
Jerome Alet - alet at unice.fr - http://cortex.unice.fr/~jerome
Fac de Medecine de Nice        http://wwwmed.unice.fr 
Tel: (+33) 4 93 37 76 30     Fax: (+33) 4 93 53 15 15
28 Avenue de Valombrose - 06107 NICE Cedex 2 - FRANCE


From chris at voodooland.net  Tue Feb 20 10:20:57 2001
From: chris at voodooland.net (Chris Watson)
Date: Tue, 20 Feb 2001 03:20:57 -0600 (CST)
Subject: [ANNOUNCE] New version of the JAXML module
In-Reply-To: <3A92324B.3F7CD22D@unice.fr>
Message-ID: <Pine.BSF.4.21.0102200316150.65043-100000@open-systems.net>

> Several people have already asked the same thing...

	<AOL> Me too! </AOL>

> So, not to be rude, but your only options are:
> 
> 	* use this module internally and don't redistribute your work
> 	* distribute your application under the GPL licence (best choice for
> people)
> 	* don't use my module (probably your best choice for a proprietary
> application)

	All of which you are right about obviously. Any chance you will
reconsider perhaps maybe a dual license? GPL and BSDL ?

--
=============================================================================
-Chris Watson         (316) 326-3862 | FreeBSD Consultant, FreeBSD Geek 
Work:              scanner at jurai.net | Open Systems Inc., Wellington, Kansas
Home:  scanner at deceptively.shady.org | http://open-systems.net
=============================================================================
WINDOWS: "Where do you want to go today?"
LINUX: "Where do you want to go tomorrow?"
BSD: "Are you guys coming or what?"
=============================================================================
GNU = Gnu's Non-portable & Unstable
irc.openprojects.net #FreeBSD -Join the revolution!
ICQ: 20016186




From gerhard.nospam at bigfoot.de  Tue Feb 20 16:10:42 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Tue, 20 Feb 2001 16:10:42 +0100
Subject: [ANNOUNCE] New version of the JAXML module
References: <3A9128D8.35D0EE22@unice.fr> <slrn992gj2.j5.gerhard.nospam@lilith.hqd-internal> <3A92324B.3F7CD22D@unice.fr>
Message-ID: <slrn995336.2pm.gerhard.nospam@lilith.hqd-internal>

On Tue, 20 Feb 2001 10:00:59 +0100, Jerome Alet <alet at unice.fr> wrote:
>Gerhard H?ring wrote:
>> 
>> On Mon, 19 Feb 2001 15:08:24 +0100, Jerome Alet <alet at unice.fr> wrote:
>> >The JAXML module is GPLed, you can download it freely from:
>> >
>> >        http://cortex.unice.fr/~jerome/jaxml/
>
>> Could you please change the license to LGPL or another more liberal license?
>> This would make your module usable by a lot more people. Putting the GPL on
>> a library prevents it from being used in a commercial application.
>
>Several people have already asked the same thing...
>
>The purpose of the GPL on this module is not to forbid you to use it in
>a commercial application, but to forbid you to use it in a proprietary
>application, which is very different.

Yes, I know. Using a GPL-ed library in any of my programs restricts options
on what I can do with *my code*, though. I just like to have all the
options. I might troll and say the GPL is less free than the BSD license,
but let's not start this flamewar again.

>If your commercial application is delivered under the GPL then there's
>no problem: feel free to sell it, and sell my module if you want. But
>you must also give the same rights to
>the people you sell it.

Yes that's my point. If I use a GPL'd module the author forces me to GPL the
rest of my code. That's why using it restricts me. The LGPL would protect
*your* work, but wouldn't try to force *mine* under GPL, too. That's why I
prefer LGPL or BSD licensed libraries.

>Despite its very small size, I think this module is (or could be made)
>powerful, and it took me some time to write it, so I don't want it to be
>used in non-Free Software. I think (maybe I'm wrong) that its simplicity
>of use compared to the official (and of course more complete) XML
>modules is attractive and is an advantage given to the Free Software
>world.
>
>Since I want to give my copyright on this module to the FSF, their
>policy might be different and they could decide to put it under the
>LGPL, but now it is GPLed.
>
>So, not to be rude, but your only options are:
>
>	* use this module internally and don't redistribute your work
>	* distribute your application under the GPL licence (best choice for
>people)
>	* don't use my module (probably your best choice for a proprietary
>application)

Just to be clear: Of course I respect your licensing decision. You perfectly
know the implications of putting *libary* under the GPL. That's fine. I just
wanted to be sure.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de
web:    http://highqualdev.com


From gregor at mediasupervision.de  Tue Feb 20 16:54:15 2001
From: gregor at mediasupervision.de (Gregor Hoffleit)
Date: Tue, 20 Feb 2001 16:54:15 +0100
Subject: [ANNOUNCE] New version of the JAXML module
In-Reply-To: <slrn995336.2pm.gerhard.nospam@lilith.hqd-internal>; from gerhard.nospam@bigfoot.de on Tue, Feb 20, 2001 at 04:10:42PM +0100
References: <3A9128D8.35D0EE22@unice.fr> <slrn992gj2.j5.gerhard.nospam@lilith.hqd-internal> <3A92324B.3F7CD22D@unice.fr> <slrn995336.2pm.gerhard.nospam@lilith.hqd-internal>
Message-ID: <20010220165415.B16924@mediasupervision.de>

On Tue, Feb 20, 2001 at 04:10:42PM +0100, Gerhard H?ring wrote:
> Yes that's my point. If I use a GPL'd module the author forces me to GPL the
> rest of my code. That's why using it restricts me. The LGPL would protect
> *your* work, but wouldn't try to force *mine* under GPL, too. That's why I
> prefer LGPL or BSD licensed libraries.

If you prefer BSD, just release your code under a BSD-like license. BSD-like
licenses go well with GPL code. You're not forced to use GPL, it just has to
be a 'reasonably free' license.

    Gregor
    



From alet at unice.fr  Tue Feb 20 17:05:43 2001
From: alet at unice.fr (Jerome Alet)
Date: Tue, 20 Feb 2001 17:05:43 +0100
Subject: [ANNOUNCE] New version of the JAXML module
References: <3A9128D8.35D0EE22@unice.fr> <slrn992gj2.j5.gerhard.nospam@lilith.hqd-internal> <3A92324B.3F7CD22D@unice.fr> <slrn995336.2pm.gerhard.nospam@lilith.hqd-internal>
Message-ID: <3A9295D7.9B2231AC@unice.fr>

Gerhard H?ring wrote:

> options. I might troll and say the GPL is less free than the BSD license,
> but let's not start this flamewar again.

No, it would be a very bad idea (tm) !-))

> Yes that's my point. If I use a GPL'd module the author forces me to GPL the
> rest of my code. That's why using it restricts me. The LGPL would protect
> *your* work, but wouldn't try to force *mine* under GPL, too. That's why I
> prefer LGPL or BSD licensed libraries.

yes but the LGPL wouldn't encourage you to write free software, since
you'd be allowed
to use it in a non free application.

> Just to be clear: Of course I respect your licensing decision. You perfectly
> know the implications of putting *libary* under the GPL. That's fine. I just
> wanted to be sure.

There's no problem. As mentionned by Martin above, I'm open to
discussion ;-)

bye,

Jerome Alet


From loewis at informatik.hu-berlin.de  Tue Feb 20 14:06:41 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 20 Feb 2001 14:06:41 +0100
Subject: [ANNOUNCE] New version of the JAXML module
References: <mailman.982660943.25321.python-list@python.org>
Message-ID: <p6qvgq5ld0e.fsf@informatik.hu-berlin.de>

Chris Watson <chris at voodooland.net> writes:

> > So, not to be rude, but your only options are:
> > 
> > 	* use this module internally and don't redistribute your work
> > 	* distribute your application under the GPL licence (best choice for
> > people)
> > 	* don't use my module (probably your best choice for a proprietary
> > application)
> 
> 	All of which you are right about obviously. Any chance you will
> reconsider perhaps maybe a dual license? 

I don't know where this desire of talking authors into changing their
license comes from: he who writes the code writes the license.

If you really need his software in an application where the GPL is
unacceptable, maybe you should offer him a certain amount of money to
change the license. Just by posting to this group is unlikely to
achieve that effect after he made it clear that choice of license was
on purpose, not by mistake.

Regards,
Martin



From ransen_spam_me_not at nemo.it  Mon Feb 19 16:31:52 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Mon, 19 Feb 2001 15:31:52 GMT
Subject: Py_SetProgramName?
Message-ID: <3a913b2f.4245179@news.newsguy.com>

Embeddinga and Extending, I'm not sure of the reasons 
for calling :

   Py_SetProgramName

Is it for getting the command line arguments of 
a C program calling Python? Or for something to
do with the search path of modules?


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From aleaxit at yahoo.com  Mon Feb 19 17:15:54 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 19 Feb 2001 17:15:54 +0100
Subject: Py_SetProgramName?
References: <3a913b2f.4245179@news.newsguy.com>
Message-ID: <96rgs712drn@news1.newsguy.com>

"Owen F. Ransen" <ransen_spam_me_not at nemo.it> wrote in message
news:3a913b2f.4245179 at news.newsguy.com...
> Embeddinga and Extending, I'm not sure of the reasons
> for calling :
>
>    Py_SetProgramName
>
> Is it for getting the command line arguments of
> a C program calling Python? Or for something to
> do with the search path of modules?

The latter.  Quoting from "1.4, Embedding Python" in
the Python/C API Reference Manual:

"""
The embedding application can steer the search by
calling Py_SetProgramName(file) before calling
Py_Initialize(). Note that $PYTHONHOME still overrides
this and $PYTHONPATH is still inserted in front of the
standard path. An application that requires total
control has to provide its own implementation of
Py_GetPath(), Py_GetPrefix(), Py_GetExecPrefix(),
and Py_GetProgramFullPath() (all defined in
Modules/getpath.c).
"""

If your embedding application wants to set the
sys.argv list, it can do so, quite separately
from Py_SetProgramName, by calling
PySys_SetArgv(argc, argv) *after* it calls
Py_Initialize.


Alex





From ransen_spam_me_not at nemo.it  Tue Feb 20 07:04:29 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Tue, 20 Feb 2001 06:04:29 GMT
Subject: Py_SetProgramName?
References: <3a913b2f.4245179@news.newsguy.com> <96rgs712drn@news1.newsguy.com>
Message-ID: <3a920669.2025119@news.newsguy.com>

>Quoting from "1.4, Embedding Python" in
>the Python/C API Reference Manual:
>
>"""
>The embedding application can steer the search by
>calling Py_SetProgramName(file) before calling
>Py_Initialize(). Note that $PYTHONHOME still overrides
>this and $PYTHONPATH is still inserted in front of the
>standard path. An application that requires total
>control has to provide its own implementation of
>Py_GetPath(), Py_GetPrefix(), Py_GetExecPrefix(),
>and Py_GetProgramFullPath() (all defined in
>Modules/getpath.c).
>"""

I read that several times before posting my original
message. It still is not clear to me what or why I
should call the function.

Like many manuals it assumes that you understand EVERYTHING
else in the manual except the entry you are currently
reading!  ;)

Maybe someone can give me a practical example of when
I would want to use Py_SetProgramName(file)


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From zope at thewebsons.com  Mon Feb 19 17:14:30 2001
From: zope at thewebsons.com (Ben Ocean)
Date: Mon, 19 Feb 2001 08:14:30 -0800
Subject: Testing for Speed
Message-ID: <5.0.2.1.0.20010219081411.009f3c60@lefeberbulb.com>

Hi;
How do I determine a visitor's connection speed? Clock a round trip?
TIA,
BenO




From erno-news at erno.iki.fi  Mon Feb 19 17:59:54 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 19 Feb 2001 18:59:54 +0200
Subject: Testing for Speed
References: <mailman.982599381.15269.python-list@python.org>
Message-ID: <kuvgq6sj5h.fsf@lasipalatsi.fi>

In article <mailman.982599381.15269.python-list at python.org>, Ben Ocean
<zope at thewebsons.com> writes:

| Hi;
| How do I determine a visitor's connection speed? Clock a round trip?

i'd suggest a laser doppler velocity gauge, that way you don't
have to ask your visitor to make a whole round trip.

  -- erno


From emile at fenx.com  Mon Feb 19 18:08:18 2001
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 19 Feb 2001 09:08:18 -0800
Subject: Testing for Speed
References: <mailman.982599381.15269.python-list@python.org>
Message-ID: <96rk35$bh94k$1@ID-11957.news.dfncis.de>

One approach used on web sites that do this (see
http://home.cfl.rr.com/eaa/Bandwidth.htm )is to embed a
large amount of non-content in a web document and time the
transmission.  Others use client side javascript to time a
download from their site.  The ones that I've looked at all
seem to measure the speed of the narrowest point between the
client and server.  If you want to know the speed of the
connection at the client side, I think you'd want a client
side script that connects to many places and then take the
highest calculated rate.

HTH,

--

Emile van Sebille
emile at fenx.com
-------------------


"Ben Ocean" <zope at thewebsons.com> wrote in message
news:mailman.982599381.15269.python-list at python.org...
> Hi;
> How do I determine a visitor's connection speed? Clock a
round trip?
> TIA,
> BenO
>
>




From grante at visi.com  Tue Feb 20 19:47:10 2001
From: grante at visi.com (Grant Edwards)
Date: Tue, 20 Feb 2001 18:47:10 GMT
Subject: Testing for Speed
References: <mailman.982599381.15269.python-list@python.org>
Message-ID: <OYyk6.597$S7.49782@ruti.visi.com>

In article <mailman.982599381.15269.python-list at python.org>, Ben Ocean wrote:

>How do I determine a visitor's connection speed? Clock a round trip?

"Connection speed" is a rather vague term.  The only meaningful
results are going to be from measuring whatever it is you care
about. If you want to know how long it takes to download a 100K
file, then transfer a 100K file and measure it.  For file
transfers, you should have a pretty representative number by
measuring how much data you can transfer over a period of a
several seconds.

If you just want round-trip packet timing, doing a "ping" will
tell you that, but ping times don't correlate with data
transfer rates.  Ping measures how long the pipe is, data transfer
rate is determined by its diameter.

-- 
Grant Edwards                   grante             Yow!  I'm reporting for
                                  at               duty as a modern person. I
                               visi.com            want to do the Latin
                                                   Hustle now!


From gabriel_ambuehl at buz.ch  Mon Feb 19 18:18:17 2001
From: gabriel_ambuehl at buz.ch (Gabriel Ambuehl)
Date: Mon, 19 Feb 2001 18:18:17 +0100
Subject: Testing for Speed
In-Reply-To: <5.0.2.1.0.20010219081411.009f3c60@lefeberbulb.com>
References: <5.0.2.1.0.20010219081411.009f3c60@lefeberbulb.com>
Message-ID: <89448442806.20010219181817@buz.ch>

-----BEGIN PGP SIGNED MESSAGE-----

Hello Ben,

Monday, February 19, 2001, 5:14:30 PM, you wrote:

> Hi;
> How do I determine a visitor's connection speed? Clock a round trip?

You can't. pings might give you an impression but even those will only
make sense if there are only a few hops between you and the visitor.

I, for example, got 512kbit cable here. In Switzerland, I got ping
times
of under 20ms to really fast servers and can download with 50kbyte/s.
For fast servers in the USA, I have about 110ms ping times but still
I'm able to download with 50kbyte/s.

Conclusion: there's no such thing as accurate connection speed
detection. The ONLY thing you can do is round trip time measurement
but as pointed out above, this won't tell you much.


Best regards,
 Gabriel

-----BEGIN PGP SIGNATURE-----
Version: PGP 6.0.2i

iQEVAwUBOpFHUcZa2WpymlDxAQElbwgAolim4olzagCC+3eR9Ps93I0Y0BKCiBn1
H/FN1whb3QxmF/SC0v3ekl0OQAqy59nXpseo3iMQwKGT8lv+c1wISGSXjhhw/Ze+
T2xDTLTqE8HfOKGby+fcIBI9KxrMq1ZerExz5chu1vN2/InLZo7CM62y0v6M3wxW
SkIgUufw2ALAV1+Tv75UGlJjR06sMBAR0b7N1r9I/nCcxYMKKllVX25Ux2DoIvoW
4x0qpAGwEX5UgPwzce6sR/atI1QDSm6G0JmGzJnwC/oXMpLB8sqcwDEhPXw20GJc
Ev5cewR2hraEceSrh2C2vaMfTQHnggiJzfWKjq952Nuax9IV5T47Eg==
=xS8Z
-----END PGP SIGNATURE-----





From ransen_spam_me_not at nemo.it  Mon Feb 19 19:06:35 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Mon, 19 Feb 2001 18:06:35 GMT
Subject: Where to get Python20_d.lib for Win32 ?
Message-ID: <3a91435d.6339893@news.newsguy.com>

Hello All,

Where can I find python20_d.lib to download?

(I'd rather not rebuild the sources myself since
 I am already up to my neck in it understanding
 "embedding and extending")

TIA


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From trentm at ActiveState.com  Mon Feb 19 23:18:54 2001
From: trentm at ActiveState.com (Trent Mick)
Date: Mon, 19 Feb 2001 14:18:54 -0800
Subject: Where to get Python20_d.lib for Win32 ?
In-Reply-To: <3a91435d.6339893@news.newsguy.com>; from ransen_spam_me_not@nemo.it on Mon, Feb 19, 2001 at 06:06:35PM +0000
References: <3a91435d.6339893@news.newsguy.com>
Message-ID: <20010219141854.C31647@ActiveState.com>

On Mon, Feb 19, 2001 at 06:06:35PM +0000, Owen F. Ransen wrote:
> Hello All,
> 
> Where can I find python20_d.lib to download?
> 
> (I'd rather not rebuild the sources myself since
>  I am already up to my neck in it understanding
>  "embedding and extending")

One place is:

http://www.activestate.com/Products/ActivePython/Download.html

look for the "ActivePython-2.0.0.202.debug.tar.gz" link


Cheers,
Trent


-- 
Trent Mick
TrentM at ActiveState.com



From MarkH at ActiveState.com  Mon Feb 19 23:32:27 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Mon, 19 Feb 2001 22:32:27 GMT
Subject: Where to get Python20_d.lib for Win32 ?
References: <3a91435d.6339893@news.newsguy.com>
Message-ID: <3A919E21.4080903@ActiveState.com>

Owen F. Ransen wrote:

> Hello All,
> 
> Where can I find python20_d.lib to download?

You can find one at http://www.activestate.com/Products/ActivePython/Download.html

> (I'd rather not rebuild the sources myself since
>  I am already up to my neck in it understanding
>  "embedding and extending")

I strongly suggest you do!  It is not that painful and if you are going through extending and embedding pain having everything built locally makes stepping through the source code etc much simpler.

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From tim.one at home.com  Mon Feb 19 23:35:21 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 19 Feb 2001 17:35:21 -0500
Subject: Where to get Python20_d.lib for Win32 ?
In-Reply-To: <3a91435d.6339893@news.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEIJJAAA.tim.one@home.com>

[posted & mailed]

[Owen F. Ransen]
> Hello All,
>
> Where can I find python20_d.lib to download?

Go to

    https://sourceforge.net/project/showfiles.php?group_id=5470

and pick up Python-2.0-Debug.zip.  It contains all the 2.0 _d stuff
(including DLLs and .pyds and .exes -- the .libs alone won't do you much
good).

> (I'd rather not rebuild the sources myself since I am already
> up to my neck in it understanding "embedding and extending")

Then rebuilding from source yourself is a darned good exercise!  At heart,
Python is itself a C application and all the std C modules are "extensions"
to it.  Fiddling the source for Python itself can be very helpful in
figuring out mysteries in your own extensions.




From ransen_spam_me_not at nemo.it  Tue Feb 20 07:12:06 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Tue, 20 Feb 2001 06:12:06 GMT
Subject: Where to get Python20_d.lib for Win32 ?
References: <3a91435d.6339893@news.newsguy.com>
Message-ID: <3a930aa5.3109364@news.newsguy.com>

>Where can I find python20_d.lib to download?
Thanks all for the pointers :)


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From ransen_spam_me_not at nemo.it  Tue Feb 20 16:28:37 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Tue, 20 Feb 2001 15:28:37 GMT
Subject: Where to get Python20_d.lib for Win32 ?
References: <mailman.982622181.11732.python-list@python.org>
Message-ID: <3a928cf1.1181595@news.newsguy.com>

On Mon, 19 Feb 2001 17:35:21 -0500, "Tim Peters" <tim.one at home.com>
wrote:

>> Where can I find python20_d.lib to download?
>Go to
>    https://sourceforge.net/project/showfiles.php?group_id=5470
>and pick up Python-2.0-Debug.zip.
>Then rebuilding from source yourself is a darned good exercise!  

Where can I download the sources? That page does not seem to
have them, or maybe I am missing something?


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From zamurai at gmx.net  Mon Feb 19 19:15:29 2001
From: zamurai at gmx.net (Zamurai)
Date: Mon, 19 Feb 2001 19:15:29 +0100
Subject: Does an Tkinter widget exist, that is like JTabbedPane (a java-swing widget)???
Message-ID: <982606529.1988133305@news.t-online.de>



From kleistereimer at gmx.de  Mon Feb 19 19:16:45 2001
From: kleistereimer at gmx.de (kleistereimer at gmx.de)
Date: Mon, 19 Feb 2001 19:16:45 +0100 (MET)
Subject: docs/examples for building windows modules (dll's) from c++.  distutils ???
Message-ID: <11418.982606605@www10.gmx.net>

hi!

i'm a c++ professional, but i have converted to python now. (wx..)
my 2 problems are: 

1. i have lot's of hardware-dependend c++ classes, i need to extend
   python with.
2. i need to make a (binary) installer for linux and windows.
   (at best not releasing the sources, and including a python interpreter
   already. (i can do this with a modified py2exe already))

I suppose the distutils will (at least) help with both (?) problems.
(i like my software to be standart-conform)

1:
there are examples and docs to extend python with c-code,
but not for c++ classes.
so how to wrap c++ code to make it callable from python?
(c++ code means a complete class hierarchy in aprox 30 files
which contains one interface class and one iterator class on top,
which both needs to be accessible from python)

2:
next thing is how to compile the wrapped c++.

i found references to a 'compile.py' which generates vc6 project-files.
but it's not available anymore. (use distutils everyone tells me..)
and i wont use outdated tools..

the documentation of distutils is not very clear.
and the examples-pages are empty.

all i need is a minimal example where a c++ class
is interfaced to python, and compiled to a dll
using bc, or vc.
at best all driven by distutils to make it standart-conform.

distutils seem to encapsulate compilers, so
maybe i'm able to generate the dll using distutils?

or am i completely wrong?
please point me to some information! THANK YOU!!

-- 
Sent through GMX FreeMail - http://www.gmx.net




From thomas.heller at ion-tof.com  Mon Feb 19 19:25:21 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Mon, 19 Feb 2001 19:25:21 +0100
Subject: [Distutils] docs/examples for building windows modules (dll's) from c++.  distutils ???
References: <11418.982606605@www10.gmx.net>
Message-ID: <031301c09aa1$51a95d40$e000a8c0@thomasnotebook>

kleistereimer wrote:

> hi!
> 
> i'm a c++ professional, but i have converted to python now. (wx..)
> my 2 problems are: 
> 
> 1. i have lot's of hardware-dependend c++ classes, i need to extend
>    python with.
> 2. i need to make a (binary) installer for linux and windows.
>    (at best not releasing the sources, and including a python interpreter
>    already. (i can do this with a modified py2exe already))

As the author of py2exe I would certainly like to hear more about this!

> 
> I suppose the distutils will (at least) help with both (?) problems.
> (i like my software to be standart-conform)
> 
> 1:
> there are examples and docs to extend python with c-code,
> but not for c++ classes.
> so how to wrap c++ code to make it callable from python?
> (c++ code means a complete class hierarchy in aprox 30 files
> which contains one interface class and one iterator class on top,
> which both needs to be accessible from python)
> 
> 2:
> next thing is how to compile the wrapped c++.
> 


Take a look at SWIG: http://www.swig.org/

Distutils contains (in an early stage) support for swig.

Thomas




From qrczak at knm.org.pl  Mon Feb 19 19:29:43 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 19 Feb 2001 18:29:43 GMT
Subject: An alternative approach to bound methods
Message-ID: <slrn992pgm.i5n.qrczak@qrnik.zagroda>

Here is how I would handle bound methods. Similar to Python but
a bit different.

A disadvantage of current function-to-bound-method conversion is
that it checks the type of the object and does an adjustment in a
specific case, instead of working uniformly for all types. It breaks
when a method is implemented as some other object than a function
and introduces unnecessary complexity to the language definition.

Why to have unbound methods as a separate concept from functions? Why
objects of other types can't act as methods? Why I can store any
value as a data of a class *except* functions (which are converted
to unbound methods)? And I would like to have static methods.

The conversion happens during extraction of an attribute out of
a class, in case it's a function. IMHO it should happen during
translation of extract-from-instance to extract-from-class, for
all cases.

obj.x, when obj is an instance object, first searches for 'x' in its
__dict__. If not found, it should try obj.__class__.x, converting
it to a bound method by making a proxy object which, when called,
calls the function with obj argument prepended (and perhaps passes
other attribute accesses transparently, but it's not essential -
it could provide the real function as im_func only as well).

Nothing magical happens when attributes are extracted from a class
itself, no matter what their types are. There are no unbound methods,
just functions. You can define static methods by plain def m() in
a class, without the self parameter. You can call any method as a
static method, with any object as its first argument. Only when you
call a method through an instance, self is prepended. Any callable
object can behave as a method.

What we lose is the ability to access class attributes by getting an
attribute of an instance (in case they are not functions, because you
could not do it when they are functions anyway). It's not a problem
except compatibility. When scopes nest, you can just refer to these
names unqualified. You can also refer to them qualified by the class
name, which works as now from any place where the class is accessible.

What do you think?

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From drtimcouper at yahoo.co.uk  Tue Feb 20 16:55:08 2001
From: drtimcouper at yahoo.co.uk (=?iso-8859-1?q?Tim=20Couper?=)
Date: Tue, 20 Feb 2001 15:55:08 +0000 (GMT)
Subject: An alternative approach to bound methods
In-Reply-To: <E14UvH4-00054g-00@mail.python.org>
Message-ID: <20010220155508.4879.qmail@web6403.mail.yahoo.com>

Marcin wrote
_____
A disadvantage of current function-to-bound-method
conversion is that it checks the type of the object
and does an adjustment in a specific case, instead of
working uniformly for all types. It breaks when a
method is implemented as some other object than a
function and introduces unnecessary complexity to the
language definition.

Why to have unbound methods as a separate concept from
functions? Why objects of other types can't act as
methods? Why I can store any value as a data of a
class *except* functions (which are converted
to unbound methods)? And I would like to have static
methods.

The conversion happens during extraction of an
attribute out of a class, in case it's a function.
IMHO it should happen during translation of
extract-from-instance to extract-from-class, for all
cases.

obj.x, when obj is an instance object, first searches
for 'x' in its __dict__. If not found, it should try
obj.__class__.x, converting it to a bound method by
making a proxy object which, when called, calls the
function with obj argument prepended (and perhaps
passes other attribute accesses transparently, but
it's not essential - it could provide the real
function as im_func only as well).

Nothing magical happens when attributes are extracted
from a class itself, no matter what their types are.
There are no unbound methods, just functions. You can
define static methods by plain def m() in a class,
without the self parameter. You can call any method as
a static method, with any object as its first
argument. Only when you call a method through an
instance, self is prepended. Any callable object can
behave as a method.

What we lose is the ability to access class attributes
by getting an attribute of an instance (in case they
are not functions, because you could not do it when
they are functions anyway). It's not a problem except
compatibility. When scopes nest, you can just refer to
these names unqualified. You can also refer to them
qualified by the class name, which works as now from
any place where the class is accessible.

What do you think?

_________________

FWIW I think that this seems an excellent suggestion;
for one, I would find it a real asset to have such a
seemingly python-esque way of having class
functions/static methods available in this manner.
What is the procedure to bring this suggestion the the
powers-that-be?

Tim Couper 

____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From jeremy at alum.mit.edu  Tue Feb 20 18:11:16 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Tue, 20 Feb 2001 12:11:16 -0500 (EST)
Subject: An alternative approach to bound methods
In-Reply-To: <20010220155508.4879.qmail@web6403.mail.yahoo.com>
References: <E14UvH4-00054g-00@mail.python.org>
	<20010220155508.4879.qmail@web6403.mail.yahoo.com>
Message-ID: <14994.42292.696689.559105@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "TC" == Tim Couper writes:

  TC> What we lose is the ability to access class attributes by
  TC> getting an attribute of an instance (in case they are not
  TC> functions, because you could not do it when they are functions
  TC> anyway). It's not a problem except compatibility. When scopes
  TC> nest, you can just refer to these names unqualified. You can
  TC> also refer to them qualified by the class name, which works as
  TC> now from any place where the class is accessible.

  TC> What do you think?

I'm not sure I really followed this argument.  Perhaps some code
examples would help.

The ability to access class attributes via self is pretty fundamental
property of Python.  I don't expect it will go away.

Also note that nested scopes do not allow access to class attributes
as unqualified names.  When resolving free variables, class scopes are
skipped.  This decision was made because we wanted to enforce the rule
that all access to class and instance data should be via self.

  TC> _________________

  TC> FWIW I think that this seems an excellent suggestion; for one, I
  TC> would find it a real asset to have such a seemingly python-esque
  TC> way of having class functions/static methods available in this
  TC> manner.  What is the procedure to bring this suggestion the the
  TC> powers-that-be?

Write a PEP.
http://python.sourceforge.net/peps/

Jeremy



From qrczak at knm.org.pl  Wed Feb 21 23:43:20 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 21 Feb 2001 22:43:20 GMT
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org>
Message-ID: <slrn998h48.1te.qrczak@qrnik.zagroda>

Tue, 20 Feb 2001 12:11:16 -0500 (EST), Jeremy Hylton <jeremy at alum.mit.edu> pisze:

> The ability to access class attributes via self is pretty fundamental
> property of Python.  I don't expect it will go away.

It's a pity.

With my proposal it could be at most retained as a compatibility hack:
when translating the object attribute access to class attribute access,
check if it's callable, and do the self-prepending transformation
only if it is - because it would be meaningless otherwise anyway.

But it does not fit the spirit of the proposal at all, it's bad to
rely on absence of an interface, and thus it would be deprecated.
No way to have it as a preferred access. You can refer to these names
unqualified, or qualified by self.__class__, or qualified by the class
name (each way has a slightly different semantics wrt. inheritance
and name rebinding), but not with self. Unless you copy them to
instances in __init__ (class attributes already can't be assigned to
from methods, so it should work in all cases).

Current Python does rely on a negative constraint: class attributes
can be any types *except* functions. This is bad in principle, even
if mostly harmless in practice. It's not quite harmless today: we
don't have class methods because of this, even though it's obvious
how they would be written if they worked, and they can be simulated
by other callable objects. It's also ugly that the builtin function
type has special rights that no other type can emulate.

Sorry, this proposal does change a fundamental Python's rule.
I believe it would be a saner design ignoring the status quo.
But it's difficult to apply it to an existing language without
breaking compatibility.

I can specify the details, write a PEP, write open questions about
the details, write examples etc., but for all that to make sense,
I would have to hear positive feedback from a person which can decide
about changing such core things.

> Also note that nested scopes do not allow access to class attributes
> as unqualified names.  When resolving free variables, class scopes are
> skipped.  This decision was made because we wanted to enforce the rule
> that all access to class and instance data should be via self.

I would definitely change that and let all scopes nest, classes too.

A recursive function should be able to refer to itself in the same
way no matter if it's defined at the top level, in another function,
or in a class. The first two cases were unified in 2.1b2, but the
third has not been unified.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From aleaxit at yahoo.com  Thu Feb 22 11:14:05 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 11:14:05 +0100
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda>
Message-ID: <972opb0ngg@news2.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn998h48.1te.qrczak at qrnik.zagroda...
    [snip]
> I would definitely change that and let all scopes nest, classes too.
>
> A recursive function should be able to refer to itself in the same
> way no matter if it's defined at the top level, in another function,
> or in a class. The first two cases were unified in 2.1b2, but the
> third has not been unified.

I *STRONGLY* disagree, because in the third case, AND IN IT ONLY,
inheritance and overriding come into play!!!

class X:
    def foo(self, N):
        if N>23: return 1+2*self.foo(N/2-4)
        else: return N/7

_This_ recursively-defined function IS subject to overriding -- I
know this, and I see it very clearly, because of the "self.foo".

It would be just as clear and explicit that overring is NOT
supported if I used X.foo in the recursive call.  _Normally_
the potential for polymorphism and overriding IS desired, of
course, but it's important to be able to explicitly express
"...and here, strangely enough, NO overriding predicted".

But what should it mean if just "1+foo(N/2-4)" was coded, in
your "all scopes nest" ideal?  Would you thus facilitate the
EXCEPTIONAL case, in which X.foo is wanted, to the detriment
of the NORMAL case, in which self.foo is wanted?  Or have
some C++-like black magick, to have foo mean self.foo...?!

Thanks, but, NO thanks.

Nesting is for the birds!-)


Alex





From qrczak at knm.org.pl  Thu Feb 22 15:38:53 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Feb 2001 14:38:53 GMT
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com>
Message-ID: <slrn99a93s.vn8.qrczak@qrnik.zagroda>

Thu, 22 Feb 2001 11:14:05 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:

> I *STRONGLY* disagree, because in the third case, AND IN IT ONLY,
> inheritance and overriding come into play!!!

Sure. So depending on whether you want recursive call to refer to
the function itself or to its version overridden by inheritance,
you would write either
    foo
    X.foo
or
    self.__class__.foo
    self.foo # After initialization of self.foo in __init__

> _Normally_ the potential for polymorphism and overriding IS desired,

In recursive calls to class methods? Not at all. If the overridden
version calls me, it probably expects that I would do what I was
supposed to do in the superclass. No matter if I do this by calling
myself recursively or not.

Not to say that class methods would rarely be overridden at all -
they are not called in the context of an instance after all.

Remember that I am talking about class methods, which don't exist
now at all. They can be currently expressed as module functions which
are not subject to inheritance. Instance methods are normally called
through self as always!

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From aleaxit at yahoo.com  Thu Feb 22 17:30:24 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 17:30:24 +0100
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com> <slrn99a93s.vn8.qrczak@qrnik.zagroda>
Message-ID: <973er001d39@news2.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn99a93s.vn8.qrczak at qrnik.zagroda...
> Thu, 22 Feb 2001 11:14:05 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:
>
> > I *STRONGLY* disagree, because in the third case, AND IN IT ONLY,
> > inheritance and overriding come into play!!!
>
> Sure. So depending on whether you want recursive call to refer to
> the function itself or to its version overridden by inheritance,
> you would write either
>     foo
>     X.foo

> > _Normally_ the potential for polymorphism and overriding IS desired,
>
> In recursive calls to class methods? Not at all. If the overridden
> version calls me, it probably expects that I would do what I was
> supposed to do in the superclass. No matter if I do this by calling
> myself recursively or not.

Where do you read/infer the limitation to *class* methods?

> Remember that I am talking about class methods, which don't exist
> now at all. They can be currently expressed as module functions which
> are not subject to inheritance. Instance methods are normally called
> through self as always!

But how does a given function, whose def is inside a class, know
whether it's supposed to recurse through its (potential) overrider
(via a same-name method of the first argument) or by defeating the
override (via its bare name, or, equivalently if your proposal was
adopted, using itself as an attribute in its enclosing class)?

Recursion apart, what distinguishes, e.g., a method foo in class
Derived(Base), calling the version it overrides as Base.foo(self,x,y),
and a "class-method" bar calling the same-name class-method of
the base class as Base.bar(x,y,z)?  They're indistinguishable cases
now -- would they get distinguished in your proposal, and, if so, how?


Alex





From qrczak at knm.org.pl  Thu Feb 22 20:14:49 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Feb 2001 19:14:49 GMT
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com> <slrn99a93s.vn8.qrczak@qrnik.zagroda> <973er001d39@news2.newsguy.com>
Message-ID: <slrn99ap97.lm7.qrczak@qrnik.zagroda>

Thu, 22 Feb 2001 17:30:24 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:

> Where do you read/infer the limitation to *class* methods?

My proposal doesn't change how instance methods and instance attributes
can be used. Only class attributes change.

Or I didn't understand what you mean?

> But how does a given function, whose def is inside a class, know
> whether it's supposed to recurse through its (potential) overrider
> (via a same-name method of the first argument) or by defeating the
> override (via its bare name, or, equivalently if your proposal was
> adopted, using itself as an attribute in its enclosing class)?

Are you asking what should one choose when he is writing the class?
I don't know any case where it should recurse through its potential
overrider yet, but perhaps they exist. This question can't have a
general answer, similarly as questions like 'which class should I
make an instance of here'.

Are you asking about the semantics of a particular piece of code?
If so, which code? My proposal is unambiguous.

> Recursion apart, what distinguishes, e.g., a method foo in class
> Derived(Base), calling the version it overrides as Base.foo(self,x,y),
> and a "class-method" bar calling the same-name class-method of
> the base class as Base.bar(x,y,z)?

Technically nothing.

Ideologically foo is an instance method, and bar is a class method,
i.e. the first argument of foo is supposed to be an instance of a
subclass of Base (even Derived at this usage point).

Instance methods are special cases of class methods, where the
first argument is an object of the class we are in. They are not
distinguished by declaration: a method can play the role of both.

We have a shortcut for calling instance methods: instead of
    obj.__class__.method (obj, arg1, arg2)
you can also write
    obj.method (arg1, arg2)

It happens at the time of accessing methods from the instance, i.e.
    obj.method
itself means
    lambda *args, **kwargs: obj.__class__.method (obj, *args, **kwargs)
when type (obj) == types.InstanceType
         and not obj.__dict__.has_key ('method').

That's it.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From aleaxit at yahoo.com  Fri Feb 23 12:33:00 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 23 Feb 2001 12:33:00 +0100
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com> <slrn99a93s.vn8.qrczak@qrnik.zagroda> <973er001d39@news2.newsguy.com> <slrn99ap97.lm7.qrczak@qrnik.zagroda>
Message-ID: <975hpa05qm@news1.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn99ap97.lm7.qrczak at qrnik.zagroda...
> Thu, 22 Feb 2001 17:30:24 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:
>
> > Where do you read/infer the limitation to *class* methods?
>
> My proposal doesn't change how instance methods and instance attributes
> can be used. Only class attributes change.
>
> Or I didn't understand what you mean?

Let's say we have...:

class Aclass:
    def amethod(x, *args):
        # body snipped

is this 'amethod' a "class method" or an "instance method"?  Does
calling it via

    x=Aclass();x.amethod(y,z)

or via

    x=Aclass();Aclass.amethod(x,y,z)

change anything?


> > But how does a given function, whose def is inside a class, know
> > whether it's supposed to recurse through its (potential) overrider
> > (via a same-name method of the first argument) or by defeating the
> > override (via its bare name, or, equivalently if your proposal was
> > adopted, using itself as an attribute in its enclosing class)?
>
> Are you asking what should one choose when he is writing the class?
> I don't know any case where it should recurse through its potential
> overrider yet, but perhaps they exist. This question can't have a
> general answer, similarly as questions like 'which class should I
> make an instance of here'.
>
> Are you asking about the semantics of a particular piece of code?
> If so, which code? My proposal is unambiguous.

And it does unambiguously mean that mentioning 'amethod' within
the above amethod's code body will in fact refer to Aclass.amethod,
not self.amethod, then?

In which case, the key objection is one I already raised, that
this is designed to make the easiest/most natural expression
the one for the case one wants *most rarely*.  People will often
and erroneously make recursive calls to "amethod(a,b,c)" when
they should be going through self, because "going through self"
IS the normal, most frequently desired case.

If you mean other things, then I have already expressed most
of my other objections.


> > Recursion apart, what distinguishes, e.g., a method foo in class
> > Derived(Base), calling the version it overrides as Base.foo(self,x,y),
> > and a "class-method" bar calling the same-name class-method of
> > the base class as Base.bar(x,y,z)?
>
> Technically nothing.
>
> Ideologically foo is an instance method, and bar is a class method,
> i.e. the first argument of foo is supposed to be an instance of a
> subclass of Base (even Derived at this usage point).

And this is checked at runtime, and gives a very clear
error message if the user makes a mistake about it:

TypeError: unbound method must be called with class instance 1st argument

I don't want to lose this very helpful error, and error
message, pursuing some kind of "ideological" target.
Practicality beats purity.

If "class methods" must exist (and I, for one, feel quite
happy without them), then they'd have to be distinguished
"technically" (in a compiler-visible way) from instance
methods, to cover the pragmatical need for clear messages
when programming errors are made.

class Base:
    def foo(self, *args):
        print "Base.foo",args

class Derived(Base):
    def foo(self, *args):
        print "Derived.foo",args
        Base.foo(*args)

I don't want this error to become a silent, mysterious
one; we know have clear diagnostics for it, and losing
them would be a serious step backwards.


Alex






From qrczak at knm.org.pl  Sat Feb 24 15:03:18 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 24 Feb 2001 14:03:18 GMT
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com> <slrn99a93s.vn8.qrczak@qrnik.zagroda> <973er001d39@news2.newsguy.com> <slrn99ap97.lm7.qrczak@qrnik.zagroda> <975hpa05qm@news1.newsguy.com>
Message-ID: <slrn99ffp6.ep3.qrczak@qrnik.zagroda>

Fri, 23 Feb 2001 12:33:00 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:

> class Aclass:
>     def amethod(x, *args):
>         # body snipped
> 
> is this 'amethod' a "class method" or an "instance method"?

Both.

> Does calling it via
> 
>     x=Aclass();x.amethod(y,z)
> 
> or via
> 
>     x=Aclass();Aclass.amethod(x,y,z)
> 
> change anything?

No.

> > Are you asking about the semantics of a particular piece of code?
> > If so, which code? My proposal is unambiguous.
> 
> And it does unambiguously mean that mentioning 'amethod' within
> the above amethod's code body will in fact refer to Aclass.amethod,
> not self.amethod, then?

Yes.

> People will often and erroneously make recursive calls to
> "amethod(a,b,c)" when they should be going through self, because
> "going through self" IS the normal, most frequently desired case.

Both are equally easy. Why would anybody write
    amethod(self,b,c)
instead of
    self.amethod(b,c)
when he means the latter?

> And this is checked at runtime, and gives a very clear
> error message if the user makes a mistake about it:
> 
> TypeError: unbound method must be called with class instance 1st argument

It's not necessarily a mistake.

> class Derived(Base):
>     def foo(self, *args):
>         print "Derived.foo",args
>         Base.foo(*args)
> 
> I don't want this error to become a silent, mysterious one; we know
> have clear diagnostics for it, and losing them would be a serious
> step backwards.

I admit that this error would have a worse diagnostics (usually:
too few arguments, i.e. not that bad). The same can be said about
writing 'self' in method definitions: when one forgets it, the error
is equally mysterious.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From aleaxit at yahoo.com  Sat Feb 24 19:05:29 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 24 Feb 2001 19:05:29 +0100
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com> <slrn99a93s.vn8.qrczak@qrnik.zagroda> <973er001d39@news2.newsguy.com> <slrn99ap97.lm7.qrczak@qrnik.zagroda> <975hpa05qm@news1.newsguy.com> <slrn99ffp6.ep3.qrczak@qrnik.zagroda>
Message-ID: <978t7802ghb@news1.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn99ffp6.ep3.qrczak at qrnik.zagroda...
    [snip]
> > People will often and erroneously make recursive calls to
> > "amethod(a,b,c)" when they should be going through self, because
> > "going through self" IS the normal, most frequently desired case.
>
> Both are equally easy. Why would anybody write
>     amethod(self,b,c)
> instead of
>     self.amethod(b,c)
> when he means the latter?

The pragmatical reason for which people erroneously omit 'self'
(and try to call amethod(a,b,c), not amethod(self,a,b,c)) from
within the code for a method is no doubt that there are many
languages in which this "self." (or its equivalent "this->") can be
left implicit.  But we can't change those other languages (C++,
Java, ...), so, people who come from such languages and/or
use them primarily, and only use Python more rarely, _will_ no
doubt keep making this mistake frequently -- even though in
theory there should be no intrinsic reason why they should.

The key point is that we need to keep diagnosing this error
at runtime clearly and unambiguously, just like Python does
now -- this is and will remain an important pragmatical issue,
and pragmaticity trumps purity.


> > I don't want this error to become a silent, mysterious one; we know
> > have clear diagnostics for it, and losing them would be a serious
> > step backwards.
>
> I admit that this error would have a worse diagnostics (usually:
> too few arguments, i.e. not that bad).

Much worse if the method has at least one optional argument,
which is a VERY frequent case; then, we've pushed the error
into the dark realm of mysteriously erroneous runtime behavior

> The same can be said about
> writing 'self' in method definitions: when one forgets it, the error
> is equally mysterious.

My impression, which comes from working directly with a modest
number of Python newbies and helping many more through the
net, is that the error of omitting self in the _definition_ is much
rarer than that of omitting it in the _calls_ to methods within other
methods.  And when you make that mistake in one method out
of N, getting diagnostics on _every_ call to that method (and for
every attempted use of self within that method -- if one makes
TWO mistakes, omitting self in definition AND in use, then of
course it's not easy to diagnose both correctly & clearly -- but
the single-error case is more important) tends to pinpoint the
error pretty clearly.


Alex





From jpet at eskimo.com  Sun Feb 25 21:34:56 2001
From: jpet at eskimo.com (Jeff Petkau)
Date: Sun, 25 Feb 2001 20:34:56 GMT
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com> <slrn99a93s.vn8.qrczak@qrnik.zagroda> <973er001d39@news2.newsguy.com> <slrn99ap97.lm7.qrczak@qrnik.zagroda> <975hpa05qm@news1.newsguy.com> <slrn99ffp6.ep3.qrczak@qrnik.zagroda> <978t7802ghb@news1.newsguy.com>
Message-ID: <Q%dm6.948$zZ4.217355@paloalto-snr1.gtei.net>

Alex Martelli <aleaxit at yahoo.com> wrote in message
news:978t7802ghb at news1.newsguy.com...
> "Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
> news:slrn99ffp6.ep3.qrczak at qrnik.zagroda...
>     [snip]
> > > People will often and erroneously make recursive calls to
> > > "amethod(a,b,c)" when they should be going through self, because
> > > "going through self" IS the normal, most frequently desired case.
> >
> > Both are equally easy. Why would anybody write
> >     amethod(self,b,c)
> > instead of
> >     self.amethod(b,c)
> > when he means the latter?
>
> [...]
> The key point is that we need to keep diagnosing this error
> at runtime clearly and unambiguously, just like Python does
> now -- this is and will remain an important pragmatical issue,
> and pragmaticity trumps purity.
>
> > > I don't want this error to become a silent, mysterious one; we know
> > > have clear diagnostics for it, and losing them would be a serious
> > > step backwards.

Unbound methods don't get involved in this error message anyway--
the name lookup fails.

>>> class C:
        def spam(self):
            spam(self)

>>> c = C()
>>> c.spam()
Traceback (innermost last):
  File "<pyshell#5>", line 1, in ?
    c.spam()
  File "<pyshell#3>", line 3, in spam
    spam(self)
NameError: There is no variable named 'spam'
>>>

I can't really see any use whatsoever for unbound method
objects. The only effect they have is to make class
methods not work, by adding a typecheck that makes
unwarranted assumptions about what the first parameter
should be.

The semantics could also be a little simpler if they went
away. If you put something in a class, you get the same
object out. If you put something in an instance, you get
the same object out. If you put something in a class and
retrieve it through the instance, *then* it would add the
magic 'self' parameter.

--Jeff Petkau (jpet at eskimo.com)





From tim at 2wave.net  Wed Feb 21 18:17:03 2001
From: tim at 2wave.net (Tim Couper)
Date: Wed, 21 Feb 2001 17:17:03 -0000
Subject: An alternative approach to bound methods
In-Reply-To: <E14VExO-0004On-00@mail.python.org>
Message-ID: <NDBBKCJHKCHKEGIIANBKKENDCOAA.tim@2wave.net>

Jeremy Hylton wrote
> .... I'm not sure I really followed this argument.  Perhaps some code
> examples would help....
>

What attracted me to Marcin Kowalczy's original posting was that it appeared
to give a way of defining class functions, by merely extending the search
algorithm ("..obj.x, when obj is an instance object, first searches for 'x'
in its __dict__. If not found, it should try
obj.__class__.x ...")

Here's an example of the code I'd like to be able to run, extending a class
attributes example I was using:

class myclass:
    alist=['initial']

    def __init__(self,thelist=['the list']):
        self.__class__.alist=thelist

    def show(self):
        print self.__class__.alist

    """

    # I want to be able to declare a class function: (see unit test element
below)
    # so, as it's a class function, it doesn't need the self argument,
    # but is accessible through __class__. So I think it should take the
form
    #

    def static(args):
        #do something (only with class attributes)
        return SomeValue

    # or perhaps something like (but I think this is indistinguishable from
    # instance methods)

    def static(self,args):
        if not self:
           #do something (only with class attributes)
        else:
           # this is an instance call of this function
        return SomeValue
    """

if __name__=='__main__':

    print
    print 'Usual test'
    print '----------'
    print
    print 'string class attribute: ',myclass.astring

    a=myclass()
    print
    a.show()
    print

    b=myclass(['orig list'])
    a.show()
    b.show()

    print
    print 'Re-Assignment'
    print '-------------'

    a.__class__.alist=['new','list']

    a.show()
    b.show()

    print
    print 'Method'
    print '------'
    print

    # assuming that the static function returns something,
    # I'd like to write, in a completely analogous manner to the
    # class attribute:
    #
    # print myclass.static(args)
    # print a.__class__.static(args)
    # print b.__class__.static(args)






From drtimcouper at yahoo.co.uk  Fri Feb 23 12:02:45 2001
From: drtimcouper at yahoo.co.uk (=?iso-8859-1?q?Tim=20Couper?=)
Date: Fri, 23 Feb 2001 11:02:45 +0000 (GMT)
Subject: An alternative approach to bound methods
In-Reply-To: <E14W1co-0004F7-00@mail.python.org>
Message-ID: <20010223110245.26841.qmail@web6401.mail.yahoo.com>

Marcin

As I said before (IMHO) you are addressing something
important here, which seems to be self-evident to your
good self, but needs to be clearly understood by
others. Being a simple soul, I wonder if it might be
possible to restate the proposal again, maybe in a
top-down manner. Here's where I'm at, as they say.

Would I be right in saying that an objective (or is it
a consequence?) of the proposal is to permit
Pythonesque class methods as well as class attributes
in Python? If so, does the "this-is-what-I-would-like"
example which I gave in a previous email demonstrate
the requirement correctly?

This could then lead into a presentation again of the
changes that would be needed to make Python operate in
this way, and then the implications of such changes.
Again I think you see all these clearly (and they are
maybe obvious to you and others!)

It would appear that the changes covered by the
proposal are more far-reacing than the introduction of
class methods; it would be good to understand this
scope, the further benefits and downsides. Having the
class methods illustration to hand would I think be a
good stepping stone to what seems to be a more general
case.

If this is an inappropriate suggestion, please say so,
I realise it could be just me that is a bit slow :-)

Tim
  


> 
> > Where do you read/infer the limitation to *class*
> methods?
> 
> My proposal doesn't change how instance methods and
> instance attributes
> can be used. Only class attributes change.
> 
> Or I didn't understand what you mean?
> 
> > But how does a given function, whose def is inside
> a class, know
> > whether it's supposed to recurse through its
> (potential) overrider
> > (via a same-name method of the first argument) or
> by defeating the
> > override (via its bare name, or, equivalently if
> your proposal was
> > adopted, using itself as an attribute in its
> enclosing class)?
> 
> Are you asking what should one choose when he is
> writing the class?
> I don't know any case where it should recurse
> through its potential
> overrider yet, but perhaps they exist. This question
> can't have a
> general answer, similarly as questions like 'which
> class should I
> make an instance of here'.
> 
> Are you asking about the semantics of a particular
> piece of code?
> If so, which code? My proposal is unambiguous.
> 
> > Recursion apart, what distinguishes, e.g., a
> method foo in class
> > Derived(Base), calling the version it overrides as
> Base.foo(self,x,y),
> > and a "class-method" bar calling the same-name
> class-method of
> > the base class as Base.bar(x,y,z)?
> 
> Technically nothing.
> 
> Ideologically foo is an instance method, and bar is
> a class method,
> i.e. the first argument of foo is supposed to be an
> instance of a
> subclass of Base (even Derived at this usage point).
> 
> Instance methods are special cases of class methods,
> where the
> first argument is an object of the class we are in.
> They are not
> distinguished by declaration: a method can play the
> role of both.
> 
> We have a shortcut for calling instance methods:
> instead of
>     obj.__class__.method (obj, arg1, arg2)
> you can also write
>     obj.method (arg1, arg2)
> 
> It happens at the time of accessing methods from the
> instance, i.e.
>     obj.method
> itself means
>     lambda *args, **kwargs: obj.__class__.method
> (obj, *args, **kwargs)
> when type (obj) == types.InstanceType
>          and not obj.__dict__.has_key ('method').
> 
> That's it.
> 
> -- 
>  __("<  Marcin Kowalczyk * qrczak at knm.org.pl
> http://qrczak.ids.net.pl/
>  \__/
>   ^^                      SYGNATURA ZAST?PCZA
> QRCZAK
> 
> 
> > -- 
> http://mail.python.org/mailman/listinfo/python-list
> 


____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From qrczak at knm.org.pl  Sat Feb 24 15:17:53 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 24 Feb 2001 14:17:53 GMT
Subject: An alternative approach to bound methods
References: <mailman.982926184.31337.python-list@python.org>
Message-ID: <slrn99fgkh.ep3.qrczak@qrnik.zagroda>

Fri, 23 Feb 2001 11:02:45 +0000 (GMT), Tim Couper <drtimcouper at yahoo.co.uk> pisze:

> Being a simple soul, I wonder if it might be possible to restate
> the proposal again, maybe in a top-down manner.

I will write it precisely at some time.

> Would I be right in saying that an objective (or is it
> a consequence?) of the proposal is to permit Pythonesque
> class methods as well as class attributes in Python?

I would say that having class methods is a consequence. Class
attributes are already present - currently they can have any type
except the function type, i.e. except what would make them act as
class methods.

> If so, does the "this-is-what-I-would-like" example which I gave
> in a previous email demonstrate the requirement correctly?

I'm not sure if I understands its intent, but yes, it will work
(with static(args), not static(self,args)).

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From joconnor at cybermesa.com  Mon Feb 19 19:38:26 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 19 Feb 2001 18:38:26 +0000
Subject: Monetary package?
Message-ID: <3A916822.8E2B7573@cybermesa.com>

All,

	Are there any packages available that handle monetary units?

Take care,
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython


From cg at gaia.intranet.cdegroot.com  Tue Feb 20 00:17:34 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 20 Feb 2001 00:17:34 +0100
Subject: Monetary package?
References: <3A916822.8E2B7573@cybermesa.com>
Message-ID: <96s9ie$knl$1@gaia.intranet.cdegroot.com>

Jay O'Connor  <joconnor at cybermesa.com> said:
>	Are there any packages available that handle monetary units?
>
No, but there's a decent fixed point calculation package around (check
Parnassus - I think it's called FixedPoint.py), and the rest is relatively
simple.

-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From zamurai at gmx.net  Mon Feb 19 20:23:27 2001
From: zamurai at gmx.net (Zamurai)
Date: Mon, 19 Feb 2001 20:23:27 +0100
Subject: what is tkinter´s equivalent to wxpytohn´s wxNotebook?
Message-ID: <982610607.838946826@news.t-online.de>



From s2mdalle at titan.vcu.edu  Tue Feb 20 04:14:05 2001
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Tue, 20 Feb 2001 03:14:05 GMT
Subject: what is tkinter´s equivalent to wxpytohn´s wxNotebook?
References: <982610607.838946826@news.t-online.de>
Message-ID: <1ilk6.6048$pS4.2368620@typhoon2.ba-dsg.net>

Tkinter doesn't have one AFAIK, but Pmw does:


http://www.dscpl.com.au/pmw/doc/NoteBook.html

-- 
David Allen
http://opop.nols.com/
----------------------------------------
Diplomacy is the art of saying "nice doggy" until you can find a gun.


From pinard at iro.umontreal.ca  Mon Feb 19 20:32:43 2001
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=e7ois?= Pinard)
Date: 19 Feb 2001 14:32:43 -0500
Subject: quopri.encode bug in Python 1.5.2, maybe in 2.0 (?)
Message-ID: <oqsnlatqn8.fsf@lin2.sram.qc.ca>

Hi, people.

I do not have Python 2 handy here, and would like that some kind soul
forwards this report appropriately if the bug still exist in latest versions.

The problem is that `quopri.encode' does not properly handle files which do
not have a line terminator on the last line.  It produces the same result
as if the newline was there.

The proper action in this case is to append a `=' at end of the last produced
line of the `out' file, before the concluding newline, when there was no
end of line to terminate the original, non QP-ed file.  Hmph!  I guess my
explanations are not very clear.  Better a simple example, then! :-)

$ echo a | recode ../qp
a
$ echo -n a | recode ../qp
a=

These matters were clarified to me by Nathaniel Borenstein, the author
of the MIME standard.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard



From wilson.austin.aj at bhp.com.au  Mon Feb 19 20:36:03 2001
From: wilson.austin.aj at bhp.com.au (Austin Wilson)
Date: Tue, 20 Feb 2001 06:36:03 +1100
Subject: Problem with py2exe
Message-ID: <96rsie$dsc$1@gossamer.itmel.bhp.com.au>

Hi

I am trying to freeze a program using py2exe 0.2.2.  However,
I am getting the following output.  Can anyone help.

BTW:  I am using Python20 and PIL 1.1.1 on Win98.

Thanks
Austin

Resolving binary dependencies:
Traceback (most recent call last):
  File "buildwatcher.py", line 6, in ?
    scripts=["watcher.py"],
  File "c:\program files\python20\lib\distutils\core.py", line 138, in setup
    dist.run_commands()
  File "c:\program files\python20\lib\distutils\dist.py", line 829, in
run_commands
    self.run_command(cmd)
  File "c:\program files\python20\lib\distutils\dist.py", line 849, in
run_command
    cmd_obj.run()
  File "py2exe\py2exe.py", line 354, in run
    dlls, unfriendly_dlls = self.find_dependend_dlls(use_runw, dlls)
  File "py2exe\py2exe.py", line 464, in find_dependend_dlls
    alldlls, warnings = bin_depends(loadpath, images)
  File "py2exe\py2exe.py", line 674, in bin_depends
    for result in py2exe_util.depends(image, loadpath).items():
py2exe_util.bind_error: tcl83.dll






From thomas.heller at ion-tof.com  Mon Feb 19 21:17:29 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Mon, 19 Feb 2001 21:17:29 +0100
Subject: Problem with py2exe
References: <96rsie$dsc$1@gossamer.itmel.bhp.com.au>
Message-ID: <96rv0s$m7sd5$1@ID-59885.news.dfncis.de>

"Austin Wilson" <wilson.austin.aj at bhp.com.au> wrote in message news:96rsie$dsc$1 at gossamer.itmel.bhp.com.au...
> Hi
>
> I am trying to freeze a program using py2exe 0.2.2.  However,
> I am getting the following output.  Can anyone help.
>
> BTW:  I am using Python20 and PIL 1.1.1 on Win98.
>
> Thanks
> Austin
>
> Resolving binary dependencies:
> Traceback (most recent call last):
>   File "buildwatcher.py", line 6, in ?
>     scripts=["watcher.py"],
>   File "c:\program files\python20\lib\distutils\core.py", line 138, in setup
>     dist.run_commands()
>   File "c:\program files\python20\lib\distutils\dist.py", line 829, in
> run_commands
>     self.run_command(cmd)
>   File "c:\program files\python20\lib\distutils\dist.py", line 849, in
> run_command
>     cmd_obj.run()
>   File "py2exe\py2exe.py", line 354, in run
>     dlls, unfriendly_dlls = self.find_dependend_dlls(use_runw, dlls)
>   File "py2exe\py2exe.py", line 464, in find_dependend_dlls
>     alldlls, warnings = bin_depends(loadpath, images)
>   File "py2exe\py2exe.py", line 674, in bin_depends
>     for result in py2exe_util.depends(image, loadpath).items():
> py2exe_util.bind_error: tcl83.dll
>
It seems you are trying to freeze a Tkinter program, which
is not really supported up to now.
Does your python script run from the command line correctly?

Even if your build would succeed, the resulting executable
would probably fail to run like this, because the executable
would not find the tcl/tk installation:

  C:\sf\py2exe\tests\dist\_test_hanoi>_test_hanoi
  Traceback (most recent call last):
    File "<string>", line 154, in ?
  File "<string>", line 146, in main
    File "<string>", line 31, in __init__
  File "Tkinter.pyc", line 1482, in __init__
  TclError: Can't find a usable init.tcl in the following directories:
      {} C:/sf/py2exe/tests/dist/lib/tcl8.3 C:/sf/py2exe/tests/dist/lib/tcl8.3 C:/sf/py2exe/tests/lib/
  tcl8.3 C:/sf/py2exe/tests/lib/tcl8.3/library C:/sf/py2exe/tests/library C:/sf/py2exe/tests/../tcl8.3
  /library C:/sf/py2exe/../tcl8.3/library

  This probably means that Tcl wasn't installed properly.

Sorry,

Thomas




From corour01 at motorola.com  Mon Feb 19 20:44:27 2001
From: corour01 at motorola.com (O'Rourke Clodagh-corour01)
Date: Mon, 19 Feb 2001 19:44:27 -0000
Subject: Global Variables
Message-ID: <13EE655665F4D311AB4D0008C789498A01EE2C43@zei02exm02.cork.cig.mot.com>

Hi Everybody,

I am a beginner-ish at Python and I am working on a large project using it right now.

I am having a lot of problems with my variable scope.

Where exactly should global variables be declared?

I have written a class that reads the contects of an INFORMIX database into variables and arrays for manipulation.  
Several methods access the same arrays/variables.  I read the values in no problem but when I try to access them from another function aside from  the one they were initiated in they all contain PY_VARO ?  I s'ppose they cannot be seen by the function in which I am trying to use them so where should they be declared?

If anyone could offer my some much needed help I would be very grateful,

Thank You,
Clodagh O' Rourke
  



From aahz at panix.com  Mon Feb 19 20:57:51 2001
From: aahz at panix.com (Aahz Maruch)
Date: 19 Feb 2001 11:57:51 -0800
Subject: Global Variables
References: <mailman.982611925.2927.python-list@python.org>
Message-ID: <96rtrv$ok1$1@panix3.panix.com>

In article <mailman.982611925.2927.python-list at python.org>,
O'Rourke Clodagh-corour01 <corour01 at motorola.com> wrote:
>
>I am having a lot of problems with my variable scope.
>Where exactly should global variables be declared?

Don't.  Seriously.

>I have written a class that reads the contects of an INFORMIX database
>into variables and arrays for manipulation.  Several methods access
>the same arrays/variables.  I read the values in no problem but when I
>try to access them from another function aside from the one they were
>initiated in they all contain PY_VARO ? I s'ppose they cannot be seen
>by the function in which I am trying to use them so where should they
>be declared?

If you have a class, you should be setting the instance variables using
the "self." notation.  That will allow the variables to be accessed by
other methods of the class.  If you're using functions, make sure to
return the value ("value" being a number, string, tuple, list,
dictionary, or class instance) that you want to use, then pass that value
to any other functions that need to access it.  Use one of the mutable
values (list, dictionary, class) if the later functions need to modify
the value.

BTW, much of this would be clearer if you worked your way through the
entire tutorial: http://www.python.org/doc/current/tut/tut.html
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From sav at ulmen.mv.ru  Mon Feb 19 20:54:43 2001
From: sav at ulmen.mv.ru (Alexander Semenov)
Date: Mon, 19 Feb 2001 22:54:43 +0300
Subject: Multithread ZODB.
Message-ID: <96rtm6$2i2l$1@news1.simtel.ru>

Hi!

Can someone give me a python code snippet, or skeleton of multithread
program which uses ZODB? Multithread examples are missing from ZODB
guide. It says what I must create connection for each thread, but I don't
know
for what? Should I make root objects separately for each thread? I prefer to
have common object for all threads. Should each thread commit transactions
on common objects? Shoud I use synchronization objects for safe access to
ZODB objects? I can't use ZEO, my program for Windows.
Luke, where is your sources?

WBR,
Alexander Semenov.

PS: Sorry for my English, my native is Russian.





From amk at mira.erols.com  Tue Feb 20 02:13:52 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 20 Feb 2001 01:13:52 GMT
Subject: Multithread ZODB.
References: <96rtm6$2i2l$1@news1.simtel.ru>
Message-ID: <slrn993i1i.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>

On Mon, 19 Feb 2001 22:54:43 +0300, 
	Alexander Semenov <sav at ulmen.mv.ru> wrote:
>Can someone give me a python code snippet, or skeleton of multithread
>program which uses ZODB? Multithread examples are missing from ZODB
>guide. It says what I must create connection for each thread, but I don't

Look through the code in ZODB/*.py; much of it is commented to a
middling degree.  If you come up with small test programs, please send
me copies for incorporation; multithreaded ZODB programming isn't
covered because I haven't looked at it and, given that I'm cool on the
idea of threads, it's not a big priority just yet.

--amk



From cobrien at Radix.Net  Wed Feb 21 00:09:15 2001
From: cobrien at Radix.Net (Cary O'Brien)
Date: 20 Feb 2001 18:09:15 -0500
Subject: Multithread ZODB.
References: <96rtm6$2i2l$1@news1.simtel.ru> <slrn993i1i.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>
Message-ID: <96uter$b4t$1@saltmine.radix.net>

In article <slrn993i1i.hb.amk at 207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>,
A.M. Kuchling <akuchlin at mems-exchange.org> wrote:
>On Mon, 19 Feb 2001 22:54:43 +0300, 
>	Alexander Semenov <sav at ulmen.mv.ru> wrote:
>>Can someone give me a python code snippet, or skeleton of multithread
>>program which uses ZODB? Multithread examples are missing from ZODB
>>guide. It says what I must create connection for each thread, but I don't
>
>Look through the code in ZODB/*.py; much of it is commented to a
>middling degree.  If you come up with small test programs, please send
>me copies for incorporation; multithreaded ZODB programming isn't
>covered because I haven't looked at it and, given that I'm cool on the
>idea of threads, it's not a big priority just yet.
>

Isn't ZODB thread safe?  I mean Zope runs as multiple threads, doesn't
it?

-- cary (who spent the day working on a Corba + ZODB test program).




From amk at mira.erols.com  Wed Feb 21 03:21:14 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 21 Feb 2001 02:21:14 GMT
Subject: Multithread ZODB.
References: <96rtm6$2i2l$1@news1.simtel.ru> <slrn993i1i.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com> <96uter$b4t$1@saltmine.radix.net>
Message-ID: <slrn996abs.k4.amk@207-172-112-118.s118.tnt4.ann.va.dialup.rcn.com>

On 20 Feb 2001 18:09:15 -0500, Cary O'Brien <cobrien at Radix.Net> wrote:
>Isn't ZODB thread safe?  I mean Zope runs as multiple threads, doesn't
>it?

Yes it does, but I have no clear idea what the ground rules are for doing that,
and so can't really document them.

--amk



From db3l at fitlinxx.com  Fri Feb 23 03:06:28 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 22 Feb 2001 21:06:28 -0500
Subject: Multithread ZODB.
References: <96rtm6$2i2l$1@news1.simtel.ru> <slrn993i1i.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com> <96uter$b4t$1@saltmine.radix.net> <slrn996abs.k4.amk@207-172-112-118.s118.tnt4.ann.va.dialup.rcn.com>
Message-ID: <un1be5f17.fsf@ctwd0143.fitlinxx.com>

amk at mira.erols.com (A.M. Kuchling) writes:

> On 20 Feb 2001 18:09:15 -0500, Cary O'Brien <cobrien at Radix.Net> wrote:
> >Isn't ZODB thread safe?  I mean Zope runs as multiple threads, doesn't
> >it?
> 
> Yes it does, but I have no clear idea what the ground rules are for
> doing that, and so can't really document them.

It's actually more than thread-safe, it's designed to have the object
store run via multiple threads, and each thread sees its own copy of
the objects independently.

That way, if you have two users working on the web site simultaneous,
within a transaction, each sees an internally consistent view of the
object store, without being impacted by any changes from the other
user with that other transaction.  If the first user commits changes,
the second user will then see those changes in his/her reference to
the objects from the local connection - and receive a conflict error
if trying to commit changes to the same objects now changed.

If I read it correctly, the original question in this thread is
actually asking about a simpler scenario - where the application is
multi-threaded, but access to the object store need not be, and in
fact is preferred not to be.  That's what our application also wanted
- the object store represented a single state that was managed by a
multi-threaded application, but the application had it's own locking
and management of access to the data - it didn't represent independent
users.

>From my understanding, it's like ZODB solved the much harder problem,
but in doing so it made some assumptions (like per-thread connections)
that actually makes the application approach (not multi-user) more
difficult :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From db3l at fitlinxx.com  Fri Feb 23 03:02:01 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 22 Feb 2001 21:02:01 -0500
Subject: Multithread ZODB.
References: <96rtm6$2i2l$1@news1.simtel.ru>
Message-ID: <upuga5f8m.fsf@ctwd0143.fitlinxx.com>

"Alexander Semenov" <sav at ulmen.mv.ru> writes:

> Can someone give me a python code snippet, or skeleton of multithread
> program which uses ZODB? Multithread examples are missing from ZODB
> guide. It says what I must create connection for each thread, but I don't
> know
> for what? Should I make root objects separately for each thread? I prefer to
> have common object for all threads.

We had a similar goal and had actually written a multi-threaded
application using ZODB before it really sinking in that not only does
it support separate connections for each thread, but its current
implementation makes it hard to avoid (the default get_transaction()
is implicitly thread-specific).  We also just wanted a single set of
persistent objects, as we were maintaining our own controlled access
to them amongst the threads - e.g., a typical shared state object
setup that we just wanted to overlay on a persistent object store.

In the end, it turned out that as long as we were committing changes
right as they were being made, and in the thread that made them,
everything worked with a single set of objects.  We did have I think
two instances where it appeared that we were losing data that were
tracked back to not committing in the thread making the change - one
of the code paths had skipped the get_transaction().commit() call.  It
was actually the first of these cases that highlighted the
multi-connection dependency noted above.

At the time, I did get some e-mail feedback from Jim Fulton in this
regard (which I hope he doesn't mind my reproducing here - it's just
some general stuff) - in particular (">" is my note to which he was
responding):

  "> ZODB3 is documented as supporting multiple connections to a storage
   > which supports independent object caches for multiple threads.
   > However, although I've looked around at a bunch of documents, I can't
   > seem to find if this is required or simply available. 

   It isn't required, but it is by far the common usage pattern.

   Hm, I guess that the current implementation actually makes it
   hard to institute a different policy.

   > That is, if I'm
   > using ZODB in a multi-threaded Python application, must I create
   > separate connections for each thread, or can they share a single
   > connection.

   Multiple threads *can* share a single connection, but
   if you do this, you'll need to:

     - Perform whatever locking is required to serialize
       access to the shared objects,

     - You'll need to override get_transaction with a version
       that doesn't manage transactions by thread. "

We've been running our application for several months now without
actually have taken the step of replacing get_transaction() yet, as
long as we stick to the commit right away in the thread that made the
change before another thread goes to change the object again :-)

In terms of why it needs to be replaced (presumably to remove that
commit behavior restriction), Jim said:

  "The logic for the Persistent mix-in class calls get_transaction
   to get the current transaction when it needs to
   register that an object has changed.  The standard get_transaction
   manages transaction objects per thread.  In addition, the
   function is cached the first time it is called, to avoid
   global lookups each time. 

   You can replace get_transacton by installing a different version
   in builtins, but you will need to do so immediately after importing
   ZODB, to make sure your version gets cached."

We've been intending to test reworking it with separate connections,
but haven't had a chance yet.  The biggest wrinkle I imagined in that
is that we share object references among threads, and I wasn't sure
how the references from the different connections would work.

In our application, we serialize our access to the set of persistent
objects amongst all threads (effectively using a global lock on the
root object that is obtained whenever a thread is going to be playing
with state data), and any work a thread does on those objects is
committed before the lock is released.  So there's no chance of
collision with two threads trying to update/commit the same persistent
object at a time.  We also used ZODB strictly for persistence (no
versioning or transactions).

But within that structure, it's worked extremely well and has proven
very reliable and robust.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From man_at_moon at hotmail.com  Mon Feb 19 21:08:42 2001
From: man_at_moon at hotmail.com (David A.)
Date: Mon, 19 Feb 2001 20:08:42 -0000
Subject: A simple (newbie) question.
Message-ID: <9VRk6.116$DT7.616555@newsserver.ip.pt>

Hello to everyone,

    maybe I have already sent this question to the group, but I am not sure
(the first one never reached the group I think). Anyway, I am a newbie, so
probably my question is a realy simple for the most experienced python
users.

Well what I want to is this:
>>> a='python'
>>> b='print'
>>> c=b+a
>>> eval(c)
Traceback (innermost last):
  File "<pyshell#19>", line 1, in ?
    eval(c)
  File "<string>", line 1
    print"python"
        ^
SyntaxError: invalid syntax
>>>

    Of course, what I wanted was that the eval(c) prints "python"!
what am I doing wrong? Any help is apreciated.

Thx in advance David Asfaha.
--
"Quidquid latine dictum sit, altum viditur."
Whatever is said in Latin sounds profound.






From wolfgang.grafen at marconi.com  Wed Feb 21 17:35:38 2001
From: wolfgang.grafen at marconi.com (Wolfgang Grafen)
Date: Wed, 21 Feb 2001 17:35:38 +0100
Subject: A simple (newbie) question.
References: <9VRk6.116$DT7.616555@newsserver.ip.pt>
Message-ID: <3A93EE5A.A9F62C46@marconi.com>

Hi David,

following should work:

>>> a='"python"'
>>> b='print'
>>> c=b+' '+a
>>> exec(c)
python

regards

Wolfgang

"David A." wrote:
> 
> Hello to everyone,
> 
>     maybe I have already sent this question to the group, but I am not sure
> (the first one never reached the group I think). Anyway, I am a newbie, so
> probably my question is a realy simple for the most experienced python
> users.
> 
> Well what I want to is this:
> >>> a='python'
> >>> b='print'
> >>> c=b+a
> >>> eval(c)
> Traceback (innermost last):
>   File "<pyshell#19>", line 1, in ?
>     eval(c)
>   File "<string>", line 1
>     print"python"
>         ^
> SyntaxError: invalid syntax
> >>>
> 
>     Of course, what I wanted was that the eval(c) prints "python"!
> what am I doing wrong? Any help is apreciated.
> 
> Thx in advance David Asfaha.
> --
> "Quidquid latine dictum sit, altum viditur."
> Whatever is said in Latin sounds profound.


From dsh8290 at rit.edu  Wed Feb 21 18:10:06 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 21 Feb 2001 12:10:06 -0500
Subject: A simple (newbie) question.
In-Reply-To: <9VRk6.116$DT7.616555@newsserver.ip.pt>; from man_at_moon@hotmail.com on Mon, Feb 19, 2001 at 08:08:42PM +0000
References: <9VRk6.116$DT7.616555@newsserver.ip.pt>
Message-ID: <20010221121006.A23949@harmony.cs.rit.edu>

On Mon, Feb 19, 2001 at 08:08:42PM +0000, David A. wrote:
| Hello to everyone,
| 
|     maybe I have already sent this question to the group, but I am not sure
| (the first one never reached the group I think). Anyway, I am a newbie, so
| probably my question is a realy simple for the most experienced python
| users.
| 
| Well what I want to is this:
| >>> a='python'
| >>> b='print'
| >>> c=b+a
| >>> eval(c)
| Traceback (innermost last):
|   File "<pyshell#19>", line 1, in ?
|     eval(c)
|   File "<string>", line 1
|     print"python"
|         ^
| SyntaxError: invalid syntax

Try this to find the problem :

>>> print c
printpython


Once you check the data you are sending to eval you know that

printpython

is indeed invalid syntax.  Instead, try this:

a = '"python"'
b = 'print'
c = b + ' ' + a
eval( c )


HTH,
-D





From embed at geocities.com  Mon Feb 19 22:19:37 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 19 Feb 2001 16:19:37 -0500
Subject: How to make a custom type "true" or not, in Python 1.5.2
Message-ID: <z6gk6.96$TW.423@tor-nn1.netcom.ca>

I am writing a little event object, here is the PyTypeObject declaration:

static PyTypeObject win32eventtype = {
 PyObject_HEAD_INIT(&PyType_Type)
 0,   /*ob_size*/
 "win32event",   /*tp_name*/
 sizeof(win32eventobject), /*tp_basicsize*/
 0,   /*tp_itemsize*/
 /* methods */
 (destructor)win32event_dealloc, /*tp_dealloc*/
 0,   /*tp_print*/
 (getattrfunc)win32event_getattr, /*tp_getattr*/
 (setattrfunc)win32event_setattr, /*tp_setattr*/
 0,   /*tp_compare*/
 (reprfunc)win32event_repr,   /*tp_repr*/
 0,   /*tp_as_number*/
 0,   /*tp_as_sequence*/
 0,   /*tp_as_mapping*/
 0,   /*tp_hash*/
};

Now, how do I make it true or false, it appears that all C Type object
instances are True, because They Are A Thing, not Nothing, but what if you
want to make something that can be a Logically False Entity?  Do I have to
implement the sequence protocol, and implement len(x)?  It doesn't seem that
it calls my getattr function to check anything. It's just plain true.

This is using Python 1.5.2, on Win32.

Warren Postma





From mwh21 at cam.ac.uk  Mon Feb 19 22:29:42 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Feb 2001 21:29:42 +0000
Subject: How to make a custom type "true" or not, in Python 1.5.2
References: <z6gk6.96$TW.423@tor-nn1.netcom.ca>
Message-ID: <m3r90utl89.fsf@atrus.jesus.cam.ac.uk>

"Warren Postma" <embed at geocities.com> writes:

> I am writing a little event object, here is the PyTypeObject declaration:
[bobbit]
> Now, how do I make it true or false, it appears that all C Type object
> instances are True, because They Are A Thing, not Nothing, but what if you
> want to make something that can be a Logically False Entity?  Do I have to
> implement the sequence protocol, and implement len(x)?  It doesn't seem that
> it calls my getattr function to check anything. It's just plain true.

You want nb_nonzero in the number methods struct, from memory...  and
the nb_nonzero field from the tp_as_number field from checking the
source (! where does my brain keep all these bits of information?).

Cheers,
M.

-- 
  On the other hand,  the following areas are subject to boycott
  in reaction to the rampant impurity of design or execution, as
  determined after a period of study, in no particular order: 
    ...                              http://www.naggum.no/profile.html


From embed at geocities.com  Mon Feb 19 22:36:28 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 19 Feb 2001 16:36:28 -0500
Subject: How to make a custom type "true" or not, in Python 1.5.2
References: <z6gk6.96$TW.423@tor-nn1.netcom.ca> <m3r90utl89.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <nmgk6.100$TW.986@tor-nn1.netcom.ca>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3r90utl89.fsf at atrus.jesus.cam.ac.uk...
> "Warren Postma" <embed at geocities.com> writes:
>
> You want nb_nonzero in the number methods struct, from memory...  and
> the nb_nonzero field from the tp_as_number field from checking the
> source (! where does my brain keep all these bits of information?).
>

Alas it appears you can't just fill this in. I looked at MxDateTime as a
sample extension, and it says the first 10 items in the tp_as_number table
must be non-null or python will crash, thus you have to write a little stub
function. Kind of yucky. My objects have no possible numeric protocol.

So I tried my earlier hunch about if len(x) is 0, even if the rest of the
sequence protocol is undefined, it seems to work.  The weird thing here is
that len(x) doesn't mean anything, other than I suppose, the object is an
empty set if len(x) is zero.

That seemed to work.  Has anyone got any opinion on the Relative Goodness or
Badness of using Len(x) => 0 to impute Non-Thingy-ness upon some thing, when
implementing the whole tp_as_number protocol is not the Right Thing to Do?

;-)

Warren





From mwh21 at cam.ac.uk  Mon Feb 19 22:55:39 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Feb 2001 21:55:39 +0000
Subject: How to make a custom type "true" or not, in Python 1.5.2
References: <z6gk6.96$TW.423@tor-nn1.netcom.ca> <m3r90utl89.fsf@atrus.jesus.cam.ac.uk> <nmgk6.100$TW.986@tor-nn1.netcom.ca>
Message-ID: <m3lmr2tk10.fsf@atrus.jesus.cam.ac.uk>

"Warren Postma" <embed at geocities.com> writes:

> "Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
> news:m3r90utl89.fsf at atrus.jesus.cam.ac.uk...
> > "Warren Postma" <embed at geocities.com> writes:
> >
> > You want nb_nonzero in the number methods struct, from memory...  and
> > the nb_nonzero field from the tp_as_number field from checking the
> > source (! where does my brain keep all these bits of information?).
> >
> 
> Alas it appears you can't just fill this in. I looked at MxDateTime as a
> sample extension, and it says the first 10 items in the tp_as_number table
> must be non-null or python will crash, thus you have to write a little stub
> function. Kind of yucky. My objects have no possible numeric protocol.

Looking at Misc/HISTORY in the Python source implies that this hasn't
been the case since 1.5.2a1.  Do you need to support that far back?

> So I tried my earlier hunch about if len(x) is 0, even if the rest of the
> sequence protocol is undefined, it seems to work.  The weird thing here is
> that len(x) doesn't mean anything, other than I suppose, the object is an
> empty set if len(x) is zero.

I wouldn't be sure that this would work in the case that the
PyNumberMethods approach wouldn't.  I could extract 1.5.1 from cvs, I
suppose, but I'm not going to.  Actually, I just did, and I think

<your object> + 1

would probably crash (I think Python would try to call sq_concat
without checking it).

> That seemed to work.  Has anyone got any opinion on the Relative
> Goodness or Badness of using Len(x) => 0 to impute Non-Thingy-ness
> upon some thing, when implementing the whole tp_as_number protocol
> is not the Right Thing to Do?

Well, it too is a bit yucky.  If you can get away with demanding
1.5.2, go for the number methods.

Cheers,
M.

-- 
42. You can measure a programmer's perspective by noting his 
    attitude on the continuing vitality of FORTRAN.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From bkc at Murkworks.com  Mon Feb 19 22:41:08 2001
From: bkc at Murkworks.com (Brad Clements)
Date: Mon, 19 Feb 2001 16:41:08 -0500
Subject: Looking for RDS/ADO server-side XML components on Python
Message-ID: <sjgk6.6$R4.360@news.clarkson.edu>

I'd like to use Microsoft's RDS/ADO 2.5 components in an IE 5.5 client, but
connect to Zope or some other Python based "server".

I think MS has documented just enough of the XML schema for creating an
editable recordset in XML.. Was wondering if anyone has worked on this or
thought about it.

I don't want to do IIS.


--
Brad Clements,
bkc at murkworks.com




From ngreco at softhome.net  Mon Feb 19 22:46:46 2001
From: ngreco at softhome.net (Nahuel Greco)
Date: Mon, 19 Feb 2001 18:46:46 -0300
Subject: Python STL ?
Message-ID: <20010219184646.356a5e2f.ngreco@softhome.net>

There is something like the C++ STL for python (i dont mean CXX), or any project
aimed to this?



----------------------------------------------------------
Nahuel Greco                 Web Development - Open Source
http://www.codelarvs.com.ar  Game Programming - Research
Freelance coding / sysadmin  Networking. The answer is 42.
----------------------------------------------------------



From aahz at panix.com  Mon Feb 19 23:03:28 2001
From: aahz at panix.com (Aahz Maruch)
Date: 19 Feb 2001 14:03:28 -0800
Subject: Python STL ?
References: <mailman.982619303.32151.python-list@python.org>
Message-ID: <96s57g$246$1@panix2.panix.com>

In article <mailman.982619303.32151.python-list at python.org>,
Nahuel Greco  <ngreco at softhome.net> wrote:
>
>There is something like the C++ STL for python (i dont mean CXX), or
>any project aimed to this?

Not being familiar with STL, I might still be able to answer your
question if you were a little more specific about what STL features you
wanted.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From andrew at andrewcooke.free-online.co.uk  Mon Feb 19 23:27:07 2001
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Mon, 19 Feb 2001 22:27:07 +0000
Subject: Python STL ?
References: <mailman.982619303.32151.python-list@python.org>
Message-ID: <3A919DBB.77FF09CB@andrewcooke.free-online.co.uk>

There's no need for templates in a dynamically typed language and Python
already has arrays, strings, hash functions and generic "sequence"
operations.  There's discussion on iterators elsewhere in this group. 
So what do you want?

Andrew

Nahuel Greco wrote:
> 
> There is something like the C++ STL for python (i dont mean CXX), or any project
> aimed to this?
> 
> ----------------------------------------------------------
> Nahuel Greco                 Web Development - Open Source
> http://www.codelarvs.com.ar  Game Programming - Research
> Freelance coding / sysadmin  Networking. The answer is 42.
> ----------------------------------------------------------


From jwbnews at scandaroon.com  Tue Feb 20 06:14:41 2001
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Mon, 19 Feb 2001 21:14:41 -0800
Subject: Python STL ?
References: <mailman.982619303.32151.python-list@python.org> <3A919DBB.77FF09CB@andrewcooke.free-online.co.uk>
Message-ID: <jwbnews-C459C9.21144119022001@news.olympus.net>

All the joy of STL with none of the unending pain.  ;-)   --john


In article <3A919DBB.77FF09CB at andrewcooke.free-online.co.uk>, Andrew 
Cooke <andrew at andrewcooke.free-online.co.uk> wrote:

> There's no need for templates in a dynamically typed language and Python
> already has arrays, strings, hash functions and generic "sequence"
> operations.  There's discussion on iterators elsewhere in this group. 
> So what do you want?
> 
> Andrew
> 
> Nahuel Greco wrote:
> > 
> > There is something like the C++ STL for python (i dont mean CXX), or 
> > any project
> > aimed to this?
> > 
> > ----------------------------------------------------------
> > Nahuel Greco                 Web Development - Open Source
> > http://www.codelarvs.com.ar  Game Programming - Research
> > Freelance coding / sysadmin  Networking. The answer is 42.
> > ----------------------------------------------------------

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From jajvirta at cc.helsinki.fi  Tue Feb 20 08:34:39 2001
From: jajvirta at cc.helsinki.fi (Jarno J Virtanen)
Date: 20 Feb 2001 07:34:39 GMT
Subject: Python STL ?
References: <mailman.982619303.32151.python-list@python.org> <3A919DBB.77FF09CB@andrewcooke.free-online.co.uk> <jwbnews-C459C9.21144119022001@news.olympus.net>
Message-ID: <slrn9947gf.iu0.jajvirta@sirppi.helsinki.fi>

Mon, 19 Feb 2001 21:14:41 -0800 John W. Baxter wrote:
>All the joy of STL with none of the unending pain.  ;-)   --john

That's Python! :-)

sincerely yours, jarno



From marvind at Glue.umd.edu  Mon Feb 19 23:16:01 2001
From: marvind at Glue.umd.edu (Arvind Mani)
Date: Mon, 19 Feb 2001 17:16:01 -0500
Subject: python and multi-threaded C programs
Message-ID: <3A919B21.37B39900@Glue.umd.edu>

Hi,

I am new to python. I have a multi-threaded application in C. I would
like to create the threads in python and then use the thread routine
written in C. Is this possible?

I am doing this because, I had some problems with global variables that
are initialized in one thread but used in another. I dont need these
variables to be directly visible to the python interpreter. They are
required in functions that I call from python.


Thanks,
Arvind



From bryan at eevolved.com  Tue Feb 20 08:51:19 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Tue, 20 Feb 2001 07:51:19 GMT
Subject: python and multi-threaded C programs
References: <3A919B21.37B39900@Glue.umd.edu>
Message-ID: <Xlpk6.283774$JT5.9785293@news20.bellglobal.com>

Arvind Mani wrote:

> Hi,
> 
> I am new to python. I have a multi-threaded application in C. I would
> like to create the threads in python and then use the thread routine
> written in C. Is this possible?
> 
> I am doing this because, I had some problems with global variables that
> are initialized in one thread but used in another. I dont need these
> variables to be directly visible to the python interpreter. They are
> required in functions that I call from python.
> 
> 
> Thanks,
> Arvind
> 

Arvind, 

I noticed your post over on the SWIG list. I can't help but wonder just how 
valuable those thread routines in C are to you... You would save yourself a 
great headache (not to mention maintainability nightmares) if you ported 
your socket routines to python. It's not hard you know. Plus you can wield 
python's super-powers :)

Out of curiosity, are you making a distributed, decentralized file-sharing 
network? :)
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"Scientific beliefs are supported by evidence, and they get results. Myths 
and faiths are not and do not" -- Richard Dawkins



From marvind at Glue.umd.edu  Thu Feb 22 22:13:03 2001
From: marvind at Glue.umd.edu (Arvind Mani)
Date: Thu, 22 Feb 2001 16:13:03 -0500
Subject: python and multi-threaded C programs
References: <3A919B21.37B39900@Glue.umd.edu> <Xlpk6.283774$JT5.9785293@news20.bellglobal.com>
Message-ID: <3A9580DF.58FE8617@Glue.umd.edu>

hi,

I saw your reply just now.

Well I am implementing a key managment and distribution protocol for dynamic
groups. I have one thread waiting for messages from the initiator who controls
just the group membership but does not take part in key formation after the
initial round, one thread sending/receiving messages from other group members,
one thread for the rekey signal (USR1) so that they can form the secret key
again, one thread to receive fractional keys from other members and to send
its own contribution to the shared key and finally one thread to start the
python script. This is on the host side. The initiator code is simpler and has
only 2 threads - gui and the other to receive messages indicating membership
changes, round nos., etc. I have finished as much.

This application will sit on top of a dynamic routing protocol, say TORA. The
user interface will be in Tkinter.

I have completed the first part and also extended certain functions so that
they could be called from python. I got stuck with globals that are used in
other threads.

And I have semaphores to enforce certain oredering of events and mutexes to
protect globals. I think all the above are possible using Python.

I would be very grateful if you could tell me if using python to do the above
is straightforward. Can you point me to a tutorial on creating, cancelling
threads in python.
I did not need much effort for the first part and so I can redo the whole
thing.

Thanks,
Arvind




Bryan Mongeau wrote:

> Arvind Mani wrote:
>
> > Hi,
> >
> > I am new to python. I have a multi-threaded application in C. I would
> > like to create the threads in python and then use the thread routine
> > written in C. Is this possible?
> >
> > I am doing this because, I had some problems with global variables that
> > are initialized in one thread but used in another. I dont need these
> > variables to be directly visible to the python interpreter. They are
> > required in functions that I call from python.
> >
> >
> > Thanks,
> > Arvind
> >
>
> Arvind,
>
> I noticed your post over on the SWIG list. I can't help but wonder just how
> valuable those thread routines in C are to you... You would save yourself a
> great headache (not to mention maintainability nightmares) if you ported
> your socket routines to python. It's not hard you know. Plus you can wield
> python's super-powers :)
>
> Out of curiosity, are you making a distributed, decentralized file-sharing
> network? :)
> --
> <==================================>
> Bryan Mongeau
> Lead Developer, Director
> eEvolved Real-Time Technologies Inc.
> http://www.eevolved.com
> <==================================>
>
> "Scientific beliefs are supported by evidence, and they get results. Myths
> and faiths are not and do not" -- Richard Dawkins



From tjg at exceptionalminds.com  Tue Feb 20 00:33:19 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Mon, 19 Feb 2001 15:33:19 -0800
Subject: Numerous file parsers vs. XML
Message-ID: <20010219153319.G6222@trufflehunter.avalongroup.net>

Hi all,

I'm working on a project, and have been using a Python tool
called LParser. It works very well, and I have no complaints.
It is a tool for parsing an input file and returning a
dictionary of values.

I created a tool configuration file that LParser reads very
nicely and I can access all my configuration options for many
different sections through dictionary entries that are
themselves dictionaries (e.g., options['wltrakker']['application'])

I have recently been studying XML, and it looks like it would
be an excellent standardized tool for creating configuration
files.

Are there any XML modules available that would read and XML
configuration file for me and return a dictionary of the
contents?

Am I being silly and looking for a solution where I don't have
a problem?

As always, all input is greatly appreciated.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  34 days  3:40 hours ago<<



From shaleh at valinux.com  Tue Feb 20 00:43:49 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Mon, 19 Feb 2001 15:43:49 -0800
Subject: Numerous file parsers vs. XML
In-Reply-To: <20010219153319.G6222@trufflehunter.avalongroup.net>; from tjg@exceptionalminds.com on Mon, Feb 19, 2001 at 03:33:19PM -0800
References: <20010219153319.G6222@trufflehunter.avalongroup.net>
Message-ID: <20010219154349.B29288@valinux.com>

On Mon, Feb 19, 2001 at 03:33:19PM -0800, Timothy Grant wrote:
> I have recently been studying XML, and it looks like it would
> be an excellent standardized tool for creating configuration
> files.
> 
> Are there any XML modules available that would read and XML
> configuration file for me and return a dictionary of the
> contents?
> 
> Am I being silly and looking for a solution where I don't have
> a problem?
>

most of the tools out ther for XML make it easy for you to deal with the file.
So typically it is :

found_tag(foo)
foo_open()
foo_close()

so, <foo>Bar</foo> would trigger the above call. 



From loewis at informatik.hu-berlin.de  Tue Feb 20 14:04:04 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 20 Feb 2001 14:04:04 +0100
Subject: Numerous file parsers vs. XML
References: <mailman.982625482.24413.python-list@python.org>
Message-ID: <p6qy9v1ld4r.fsf@informatik.hu-berlin.de>

Timothy Grant <tjg at exceptionalminds.com> writes:

> Are there any XML modules available that would read and XML
> configuration file for me and return a dictionary of the
> contents?
> 
> Am I being silly and looking for a solution where I don't have
> a problem?

The easiest way may be to read the document into a DOM tree using
xml.dom.minidom of Python 2.0, then use the DOM API to find the bits
and pieces you are looking for.

Regards,
Martin


From tjg at exceptionalminds.com  Tue Feb 20 17:50:29 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Tue, 20 Feb 2001 08:50:29 -0800
Subject: Numerous file parsers vs. XML
In-Reply-To: <p6qy9v1ld4r.fsf@informatik.hu-berlin.de>; from loewis@informatik.hu-berlin.de on Tue, Feb 20, 2001 at 02:04:04PM +0100
References: <mailman.982625482.24413.python-list@python.org> <p6qy9v1ld4r.fsf@informatik.hu-berlin.de>
Message-ID: <20010220085029.M6222@trufflehunter.avalongroup.net>

Thanks Martin,

I'll take a look at it and see if it fits.

I'm still debating whether I'm looking for a solution where
there is no problem!


On Tue, Feb 20, 2001 at 02:04:04PM +0100, Martin von Loewis wrote:
> Timothy Grant <tjg at exceptionalminds.com> writes:
> 
> > Are there any XML modules available that would read and XML
> > configuration file for me and return a dictionary of the
> > contents?
> > 
> > Am I being silly and looking for a solution where I don't have
> > a problem?
> 
> The easiest way may be to read the document into a DOM tree using
> xml.dom.minidom of Python 2.0, then use the DOM API to find the bits
> and pieces you are looking for.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  34 days 21:02 hours ago<<



From wmiller at mediaone.net  Tue Feb 20 00:39:27 2001
From: wmiller at mediaone.net (Walter Miller)
Date: Mon, 19 Feb 2001 15:39:27 -0800
Subject: Screen capture window to *.bmp file
Message-ID: <L6ik6.96666$Wq1.39740056@nnrp5-w.sbc.net>

I'm trying to screen capture a window to a *.bmp file even if it is hidden
behind other windows.  After some research, it seems like the only way to do
this is to use the Win32 API messages: WM_PRINT or WM_PRINTCLIENT as
discribed in this article:  http://www.fengyuan.com/article/wmprint.html

Any ideas how to implement this in Python?

TIA,

Walter




From billings at eudoramail.com  Tue Feb 20 01:28:48 2001
From: billings at eudoramail.com (Scott Billings)
Date: Mon, 19 Feb 2001 18:28:48 -0600
Subject: Embedded Python (in HTML)
Message-ID: <96se0k$l4l$1@ins21.netins.net>

Is there any project out there, other than maybe PMZ, that is along the 
lines of ASP, PHP, HTML::Mason or EmbPerl for Python?

PMZ might not be so bad if it didn't seem like a CGI script that parses 
files, and then hands off the Python code to the Python interpreter.

And I have looked at such things as Zope (quite a bit overkill for what I 
want), Webware (also quite overkill)..... I would love to see basically PHP 
with a Python syntax. Does such a beast exist?
-- 
-Scott-


From ws-news at gmx.at  Tue Feb 20 17:02:15 2001
From: ws-news at gmx.at (Werner Schiendl)
Date: Tue, 20 Feb 2001 17:02:15 +0100
Subject: Embedded Python (in HTML)
References: <96se0k$l4l$1@ins21.netins.net>
Message-ID: <982685199.468853@newsmaster-04.atnet.at>

For windows you can have ASP pages with python scripts.
When using the ActivePython distribution from www.activestate.com, this
works right after you installed it.

hth
werner

Scott Billings <billings at eudoramail.com> wrote in message
news:96se0k$l4l$1 at ins21.netins.net...
> Is there any project out there, other than maybe PMZ, that is along the
> lines of ASP, PHP, HTML::Mason or EmbPerl for Python?
>
> PMZ might not be so bad if it didn't seem like a CGI script that parses
> files, and then hands off the Python code to the Python interpreter.
>
> And I have looked at such things as Zope (quite a bit overkill for what I
> want), Webware (also quite overkill)..... I would love to see basically
PHP
> with a Python syntax. Does such a beast exist?
> --
> -Scott-




From sciasbat at inorbit.com  Wed Feb 21 17:07:19 2001
From: sciasbat at inorbit.com (Fabio Forno)
Date: Wed, 21 Feb 2001 17:07:19 +0100
Subject: Embedded Python (in HTML)
References: <96se0k$l4l$1@ins21.netins.net>
Message-ID: <3A93E7B7.2C6C42CE@inorbit.com>

Scott Billings wrote:
> 
> Is there any project out there, other than maybe PMZ, that is along the
> lines of ASP, PHP, HTML::Mason or EmbPerl for Python?
> 

mod_snake may suit

ByE,
FF


From jmurray at agyinc.com  Tue Feb 20 01:44:45 2001
From: jmurray at agyinc.com (Joe Murray)
Date: Mon, 19 Feb 2001 16:44:45 -0800
Subject: postscript generation
Message-ID: <3A91BDFD.D1BA8527@agyinc.com>

Dear All,

I hope I am not missing some obvious module, but are there any more
Python modules for generating PostScript?  I've looked at PIL and PGPlot
(and cousin plplot) and even the 'yet-to-be-released' PythonPS module
(http://www.pugo.org/ps).

I'm looking for something like HTMLgen for PostScript.  Is the
PostScript Utility Library going to provide some of this functionality? 
I'm trying to combine the GNU Barcode library with some textual labeling
to implement a some sorta LIMS-type system.  So I really need access to
PostScript text layout tools.  (BTW, most of what I'm doing I could code
up quickly (quickly, in _python_ of course), but if there's a module,
why not use it...)

Thanks,

joe

-- 
Joseph Murray
Bioinformatics Specialist, AGY Therapeutics
290 Utah Avenue, South San Francisco, CA 94080
(650) 228-1146



From aahz at panix.com  Tue Feb 20 02:15:05 2001
From: aahz at panix.com (Aahz Maruch)
Date: 19 Feb 2001 17:15:05 -0800
Subject: postscript generation
References: <mailman.982630284.10805.python-list@python.org>
Message-ID: <96sgep$a0c$1@panix6.panix.com>

In article <mailman.982630284.10805.python-list at python.org>,
Joe Murray  <jmurray at agyinc.com> wrote:
>
>I hope I am not missing some obvious module, but are there any more
>Python modules for generating PostScript?  I've looked at PIL and PGPlot
>(and cousin plplot) and even the 'yet-to-be-released' PythonPS module
>(http://www.pugo.org/ps).

http://www.reportlab.com/
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"Laws that restrict speech are an abomination for a simple reason.
They're like a poison gas; they seem to be a good weapon to vanquish
your enemy, but the wind has a way of shifting."  -- Ira Glasser


From alet at unice.fr  Tue Feb 20 11:22:46 2001
From: alet at unice.fr (Jerome Alet)
Date: Tue, 20 Feb 2001 11:22:46 +0100
Subject: postscript generation
References: <mailman.982630284.10805.python-list@python.org> <96sgep$a0c$1@panix6.panix.com>
Message-ID: <3A924576.AEB4CCC0@unice.fr>

Aahz Maruch wrote:

> http://www.reportlab.com/

ReportLab produces only pdf, you then have to convert it to ps using
ghostscript (pdf2ps).

However it's a fantastic module ! Go and get it !

-- 
Jerome Alet - alet at unice.fr - http://cortex.unice.fr/~jerome
Fac de Medecine de Nice        http://wwwmed.unice.fr 
Tel: (+33) 4 93 37 76 30     Fax: (+33) 4 93 53 15 15
28 Avenue de Valombrose - 06107 NICE Cedex 2 - FRANCE


From clee at gnwy100.wuh.wustl.edu  Tue Feb 20 15:48:21 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 20 Feb 2001 08:48:21 -0600
Subject: postscript generation
References: <mailman.982630284.10805.python-list@python.org> <96sgep$a0c$1@panix6.panix.com> <3A924576.AEB4CCC0@unice.fr>
Message-ID: <7uzofhif62.fsf@gnwy100.wuh.wustl.edu>


If PDF is acceptible, I would recommend reportlab (http://www.reportlab.com).

If you're looking to generate postscript directly, Sping
(http://piddle.sourceforge.net) includes a postscript module that supports
either EPS or multipage PS generation.  It's not as complete as reportlab,
but it is still useful.  At worse, it would provide a base for anything you
wanted to build.  (Plus, I happen to know that the maintainer is in the
middle of writing it's documentation and would welcome any useful
additions. ;-)

-chris


From shaleh at valinux.com  Tue Feb 20 02:21:47 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Mon, 19 Feb 2001 17:21:47 -0800
Subject: a better getopt than getopt.py?
Message-ID: <20010219172147.F29288@valinux.com>

Writing my fisrt real python app that people will run from the command line
(most of my previous coding has been modules for other software to web stuff)
I decided to use getopt.  My setup looks like this:

LONG_OPTIONS = ['help', 'setup-lab', 'lab=']
SHORT_OPTIONS = 'hS'
    
try:
    optlist, args = getopt.getopt(sys.argv[1:], SHORT_OPTIONS,
                                  LONG_OPTIONS)
except getopt.error, e:
    print e
    print USAGE
    sys.exit(1)

for option, argument in optlist:
    if option in ('-h', '--help'):
        print USAGE
        sys.exit(0)
    elif option in ('-S', '--setup-lab'):
        print 'Want to set up the lab ...'
    elif option == 'lab':
        print 'Put lab in: ' + argument

this means I have my options defined twice and then once again in the USAGE
output.  Is there not a better way to handle arguments?



From jhauser at ifm.uni-kiel.de  Tue Feb 20 09:17:25 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 20 Feb 2001 09:17:25 +0100
Subject: a better getopt than getopt.py?
References: <mailman.982632084.17484.python-list@python.org>
Message-ID: <874rxpydii.fsf@lisboa.ifm.uni-kiel.de>

Search for CommandLineApp.py! It's a nice OO way to handle commandline
options and it generates the help/usage page automatically.

HTH,
__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From aleaxit at yahoo.com  Tue Feb 20 13:58:12 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 13:58:12 +0100
Subject: a better getopt than getopt.py?
References: <mailman.982632084.17484.python-list@python.org>
Message-ID: <96tpnt013kr@news2.newsguy.com>

"Sean 'Shaleh' Perry" <shaleh at valinux.com> wrote in message
news:mailman.982632084.17484.python-list at python.org...
> Writing my fisrt real python app that people will run from the command
line
> (most of my previous coding has been modules for other software to web
stuff)
> I decided to use getopt.  My setup looks like this:
>
> LONG_OPTIONS = ['help', 'setup-lab', 'lab=']
> SHORT_OPTIONS = 'hS'
>
> try:
>     optlist, args = getopt.getopt(sys.argv[1:], SHORT_OPTIONS,
>                                   LONG_OPTIONS)
> except getopt.error, e:
>     print e
>     print USAGE
>     sys.exit(1)
>
> for option, argument in optlist:
>     if option in ('-h', '--help'):
>         print USAGE
>         sys.exit(0)
>     elif option in ('-S', '--setup-lab'):
>         print 'Want to set up the lab ...'
>     elif option == 'lab':
>         print 'Put lab in: ' + argument
>
> this means I have my options defined twice and then once again in the
USAGE
> output.  Is there not a better way to handle arguments?

I think you basically want to group an option's short-form (if any),
its long-form (if any), and a unique-identifier for the option itself
(off which you can key an if/elif tree) into one 'chunk'.  You can
dress and package this up in several ways, of course, but, at the
root, it will be (before optimization...!-) something like:

Options = ( ('help',      'h',  'H'),
            ('setup-lab', 'S',  'S'),
            ('lab=',      None, 'L')
          )

def long(Options):
    return [long for long, short, unique in Options if not long is None]
def short(Options):
    return [short for long, short, unique in Options if not short is None]
def unique(option, Options):
    def short_to_unique(a_short, Options):
        for long, short, unique in Options:
            if a_short == short: return unique
        return None
    def long_to_unique(a_long, Options):
        for long, short, unique in Options:
            if a_long == long: return unique
        return None
    if option.startswith('--'):
        return long_to_unique(option[2:], Options)
    elif option.startswith('-'):
        return short_to_unique(option[1:], Options)
    else:
        return long_to_unique(option+'=', Options)

and your mainline code becomes, in turn:

try:
    optlist, args = getopt.getopt(sys.argv[1:],
        short(Options), long(Options))
except getopt.error, e:
    print e
    print USAGE
    sys.exit(1)

for option, argument in optlist:
    optionId = unique(option, Options)
    if optionId == 'H':
        print USAGE
        sys.exit(0)
    elif optionId == 'S':
        print 'Want to set up the lab ...'
    elif optionId == 'L':
        print 'Put lab in: ' + argument


Packaging this up as one nice support class or function
allowing a direct (and faster...) loop of the form:

allOpts, args = GetAllOpts(Options)
for optionId, argument in allOpts:
    if optionId == 'H':

etc etc, is left as an easy exercise for the reader:-).


Alex





From thomas at cbs.dtu.dk  Tue Feb 20 18:29:32 2001
From: thomas at cbs.dtu.dk (Thomas Sicheritz-Ponten)
Date: 20 Feb 2001 18:29:32 +0100
Subject: a better getopt than getopt.py?
References: <mailman.982632084.17484.python-list@python.org> <96tpnt013kr@news2.newsguy.com>
Message-ID: <y9vpugds1oj.fsf@delphinus.cbs.dtu.dk>

> > this means I have my options defined twice and then once again in the
> USAGE
> > output.  Is there not a better way to handle arguments?
> 

I agree - that is actually the ONLY part I miss from Tcl.  Tcl has this
really nice optparse.tcl package which parses the commandline arguments and
help documentation in one go ...

package require opt
::tcl::OptProc checkopts {
    {file -string      "FASTA file"}
    {-blast -string blastp  "blast program"}
    {-db -string "swiss" "quoted, space seperated list of blast databases"}
    {-flags -string {} "blast flags ( -flags \"\\-hspmax 5\")"}
    {-z "compress"}
    {-subdir "create subdir for each entry"}
    {-html "generate HTML summary"}
    {-grep /opt/gnu/bin/grep "location of GNU's grep"}
    {-report 1 "report \#n hit"}
    {-begin {} "begin with entry"}
    {-rev  "start with last entry"}
    {-switch  "switch order of db and seq ( for non blasts like hmmpfam)"}
} {
    foreach i [info locals] {
        if {$i=="args" || $i=="Args"} continue
        set ::data($i) [ set $i]
    }
}

-thomas

-- 
Sicheritz-Ponten Thomas, Ph.D  CBS, Department of Biotechnology
thomas at biopython.org           The Technical University of Denmark
CBS:  +45 45 252489            Building 208, DK-2800 Lyngby
Fax   +45 45 931585            http://www.cbs.dtu.dk/thomas

	De Chelonian Mobile ... The Turtle Moves ...


From cfelling at iae.nl  Tue Feb 20 14:17:54 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 20 Feb 2001 14:17:54 +0100
Subject: a better getopt than getopt.py?
References: <mailman.982632084.17484.python-list@python.org>
Message-ID: <96tqq2$2fh$1@animus.fel.iae.nl>

Sean 'Shaleh' Perry <shaleh at valinux.com> wrote:
> Writing my fisrt real python app that people will run from the command line
> (most of my previous coding has been modules for other software to web stuff)
> I decided to use getopt.  My setup looks like this:

...example code snipped

> this means I have my options defined twice and then once again in the USAGE
> output.  Is there not a better way to handle arguments?

Well you could make a class that hides all of this:) Sometime ago
(1999 me thinks) some people (including me) exchanged some ideas and
example code for this on c.l.py.  Though looking back at my own code,
I think I would do things differently now that I've more experience
with Python, but if you're interested I don't mind sending it.
-- 
groetjes, carel


From Roberto  Thu Feb 22 09:55:35 2001
From: Roberto (Roberto@lupex.net)
Date: Thu, 22 Feb 2001 08:55:35 GMT
Subject: a better getopt than getopt.py? [long]
References: <mailman.982632084.17484.python-list@python.org>
Message-ID: <slrn999l60.13a.Roberto@localhost.localdomain>

Sean 'Shaleh' Perry <shaleh at valinux.com> wrote:
>this means I have my options defined twice and then once again in the USAGE
>output.  Is there not a better way to handle arguments?

I have written the attached module as this issue bothers me too.
The code is not well tested (its very recent).

Here's a commented example from a program that uses it:

# -------------------- Assign default for command line args
#                      you can put them in any python dictionary
#                      ( globals() is the dictionary in this case )

low_uids_level = 0
root_check = 1
dup_UIDs = 1
dup_names = 1
verbose = 0
ignore_list = []
passwd_file = '/etc/passwd'

# -------------------- Create a command line parser

cmd = CmdLine2.CmdLine('Passwd file checks', # program title
		usage_msg='[options]', # usage message 
		                       # (default: "[options] args")
		context = globals()    # where to store option values
		                       # should be a dict-like object
				       # (default: an empty dict)
				       # accessible thru cmd.context
		)
cmd.add_help_option() # add a -h --help option
                      # to print the list of available options

cmd.add_set_option(                    # an option to set
  ('low_uids_level',string.atoi),      # context['low_uids_level']
  ['l:','low-uid='],                   # to the integer value
  "Check for UIDs < [argument]")       # of its argument
  
  # if the first argument of cmd_add_XXX_option() is a tuple,
  # the first element is the variable name/item key,
  # the second elemnt is a conversion function that is applied to the arg
  
cmd.add_toggle_option(                 # a boolean flag option, as in:
  'root_check',                        #   root_check = not root_check
  ['r','root'],
  "Enable/disable report of superusers")
  
cmd.add_check_option(                  # a boolean set option:
  'verbose',                           #   verbose = 1
  ['v','verbose'],
  "Print configution before processing")
  
cmd.add_append_option(                 # append arguments to ignore_list
  'ignore_list',
  ['i:','ignore='],
  "Add the specified user to the ignore list")

def option_handler(opt, arg):
    # do something sensible with opt and arg
    pass

cmd.add_action_option(                # custom option that
    option_handler,                   # execute option_handler(opt,arg)
    ['bingo'],
    "Prints lucky numbers for bingo")

cmd.getopt()                           # parse the command line

# the rest of the program

if verbose:
    print "I'll check", passwd_file,":"
    if low_uids_level > 0:
	print "  - for UIDs <", low_uid_level
    if root_check:
	print "  - for root users"
    if dup_UIDs:
	print "  - for duplicated UIDs"
    if dup_names:
	print "  - for duplicated usernames"
    print "Ignore list:", list2str(ignore_list)
    print

...

Here's the help message of the program (generated by add_help_option):

Passwd file checks

chk-passwd [options]

  -h --help             Print this help message
  -l: --low-uid=        Check for UIDs < [argument]
  -r --root             Enable/disable report of superusers
  -u --uids             En/dis report of duplicated UIDs
  -n --names            En/dis report of duplicated usernames
  -v --verbose          Print configution before processing
  -i: --ignore=         Add the specified user to the ignore list
  -p: --passwd=         Process the specified file (default: /etc/passwd)

(the complete script has a few more options than the example code above)

And here's the CmdLine2.py module:

# CmdLine2 - Command Line Processing v2.0
#
#
# 22-feb-2001 - Released as public domain code
#               by Roberto Lupi <Roberto at lupi.ws>

import sys
import getopt
import re
import os

debug = 0

short_option_re = re.compile('.:?$')

def RemoveLast(st, postfix):
    if st[-len(postfix):] == postfix:
        return st[:-len(postfix)]
    else:
        return st

def ListToStr(lst, sep=' '):
    return reduce( lambda a,b,sep=sep: str(a)+sep+str(b), lst)

######################################################################
## 
## Class Option
## 
######################################################################

class Option:
    def __init__(self, opts, help_msg):
        short_op = []
        long_op = []
        for opt in opts:
            if short_option_re.match(opt):
                short_op.append(opt)
            else:
                long_op.append(opt)
        self.short_op = short_op
        self.long_op = long_op
        self.help_msg = help_msg

    def fire(self, opt, arg):
        pass

    def short_options(self, arg_spec=None,op_prefix='-'):
        if self.short_op:
            if arg_spec:
                return map( lambda x,op_prefix=op_prefix: op_prefix + \
		  x, self.short_op )
            else:
                return map( lambda x,op_prefix=op_prefix: op_prefix + \
		  RemoveLast(x,':'), self.short_op )
        else:
            return []

    def long_options(self, arg_spec=None,op_prefix='--'):
        if self.long_op:
            if arg_spec:
                return map( lambda x,op_prefix=op_prefix: op_prefix + \
		  x, self.long_op )
            else:
                return map( lambda x,op_prefix=op_prefix: op_prefix + \
		  RemoveLast(x,'='), self.long_op )
        else:
            return []

    def all_options(self, arg_spec=None):
        return self.short_options(arg_spec) + self.long_options(arg_spec)

    def print_help(self):
        if self.help_msg:
            print "  %-20s  %s" % (ListToStr(self.all_options(1)), \
	      self.help_msg)

######################################################################
## 
## Class OptionAction
## 
######################################################################

class OptionAction(Option):
    def __init__(self, action, ops, help_msg):
        Option.__init__(self,ops,help_msg)
        self.action = action

    def fire(self, opt, arg):
        apply(self.action, (opt, arg))

######################################################################
## 
## Class OptionContextBased
## 
######################################################################
        
class OptionContextBased(Option):
    def __init__(self, action_spec, context, ops, help_msg):
        Option.__init__(self,ops,help_msg)
        self.action_spec = action_spec
        self.context = context

######################################################################
## 
## Class OptionSet
## 
######################################################################

class OptionSet(OptionContextBased):
    def fire(self, opt, arg):
        if type(self.action_spec) == type(""):
            vname = self.action_spec
            cvt = lambda x: x
        else:
            (vname, cvt) = self.action_spec
        self.context[vname] = cvt(arg)

######################################################################
## 
## Class OptionToggle
## 
######################################################################
        
class OptionToggle(OptionContextBased):
    def fire(self, opt, arg):
        vname = self.action_spec
        if self.context.has_key(vname):
            old_value = self.context[vname]
        else:
            old_value = 0
        self.context[vname] = not old_value

######################################################################
## 
## Class OptionCheck
## 
######################################################################

class OptionCheck(OptionContextBased):
    def fire(self, opt, arg):
        vname = self.action_spec
        self.context[vname] = 1

######################################################################
## 
## Class OptionAppend
## 
######################################################################

class OptionAppend(OptionContextBased):
    def fire(self, opt, arg):
        if type(self.action_spec) == type(""):
            vname = self.action_spec
            cvt = lambda x: x
        else:
            (vname, cvt) = self.action_spec
        if self.context.has_key(vname) and self.context[vname]:
            self.context[vname].append(cvt(arg))
        else:
            self.context[vname] = [ cvt(arg) ]

######################################################################
## 
## Class CmdLine
## 
######################################################################

class CmdLine:
    def __init__(self, program_title=None, usage_msg="[options] files", \
      context={}):
        self.program_title = program_title
        self.usage_msg = usage_msg
        self.options = []
        self.context = context

    def getopt(self, args=None):
        if self.program_title:
            print self.program_title
            print
        if args is None:
            args = sys.argv[1:]
        # build option map
        option_map = {}
        short_ops = ""
        long_ops = []
        for opt in self.options:
            if opt.short_op:
                short_ops = short_ops + \
		  ListToStr(opt.short_options(arg_spec=1,op_prefix=''),'')
            if opt.long_op:
                long_ops.extend(opt.long_options(arg_spec=1,op_prefix=''))
            for op in opt.all_options():
                option_map[op] = opt
        if debug:
            print "OPTION MAP:"
            for op_spec, Opt in option_map.items():
                print "  %10s -> %s" % (op_spec, str(Opt))
        # process cmd line
        try:
            opts, other_args = getopt.getopt(args, short_ops, long_ops)
            for opt, arg in opts:
                Op = option_map[opt]
                if debug:
                    print "DEBUG: Firing %s (opt=%s, arg=%s)" % \
		      (str(Op), opt, arg)
                Op.fire(opt, arg)
            return other_args
        except getopt.error, e:
            print str(e)
            sys.exit(1)

    def print_help(self, opt=None, arg=None):
        print "%s %s" % (os.path.basename(sys.argv[0]), self.usage_msg)
        print
        for opt in self.options:
            opt.print_help()
        print
        if opt: # invoked thru -h
            sys.exit(1)

    def add_option(self, Opt):
        self.options.append(Opt)

    def add_help_option(self):
        self.add_option(OptionAction(self.print_help,['h','help'], \
	  "Print this help message"))

    def add_action_option(self, action, ops, help_msg=None):
        self.add_option(OptionAction(action,ops,help_msg))

    def add_set_option(self, vname, ops, help_msg=None):
        self.add_option(OptionSet(vname,self.context,ops,help_msg))

    def add_toggle_option(self, vname, ops, help_msg=None):
        self.add_option(OptionToggle(vname,self.context,ops,help_msg))

    def add_check_option(self, vname, ops, help_msg=None):
        self.add_option(OptionCheck(vname,self.context,ops,help_msg))

    def add_append_option(self, vname, ops, help_msg=None):
        self.add_option(OptionAppend(vname,self.context,ops,help_msg))


-- 
Roberto Lupi


From tjg at exceptionalminds.com  Tue Feb 20 17:54:49 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Tue, 20 Feb 2001 08:54:49 -0800
Subject: a better getopt than getopt.py?
In-Reply-To: <20010219172147.F29288@valinux.com>; from shaleh@valinux.com on Mon, Feb 19, 2001 at 05:21:47PM -0800
References: <20010219172147.F29288@valinux.com>
Message-ID: <20010220085449.N6222@trufflehunter.avalongroup.net>

I recently went through a very similar experience with getopt,
though I was working on subclassing a class that handled
command line arguments (getopt proved to be completely
unsuitable for that task).  I turned to t module I found on
Parnassus called cmdline.py. It works beautifully, as expected,
and the implementation is much more readable than the similar
getopt version.

for some reason it doesn't show up in a Parnassus search for
some reason it doesn't show up in a search for cmdline.py, but
it does show up as one of about three hits in a search for
getopt.


On Mon, Feb 19, 2001 at 05:21:47PM -0800, Sean 'Shaleh' Perry wrote:
> Writing my fisrt real python app that people will run from the command line
> (most of my previous coding has been modules for other software to web stuff)
> I decided to use getopt.  My setup looks like this:
> 
> LONG_OPTIONS = ['help', 'setup-lab', 'lab=']
> SHORT_OPTIONS = 'hS'
>     
> try:
>     optlist, args = getopt.getopt(sys.argv[1:], SHORT_OPTIONS,
>                                   LONG_OPTIONS)
> except getopt.error, e:
>     print e
>     print USAGE
>     sys.exit(1)
> 
> for option, argument in optlist:
>     if option in ('-h', '--help'):
>         print USAGE
>         sys.exit(0)
>     elif option in ('-S', '--setup-lab'):
>         print 'Want to set up the lab ...'
>     elif option == 'lab':
>         print 'Put lab in: ' + argument
> 
> this means I have my options defined twice and then once again in the USAGE
> output.  Is there not a better way to handle arguments?
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  34 days 21:04 hours ago<<



From DavidLNonoSpam at raqia.com  Tue Feb 20 19:37:22 2001
From: DavidLNonoSpam at raqia.com (David Lees)
Date: Tue, 20 Feb 2001 13:37:22 -0500
Subject: a better getopt than getopt.py?
References: <20010219172147.F29288@valinux.com> <mailman.982688004.1785.python-list@python.org>
Message-ID: <3A92B962.17BA30ED@raqia.com>

Timothy,

Thanks for posting this info.  I tried getopt and was not crazy about
it.  This looks well documented and usefull.

David Lees

Timothy Grant wrote:
> 
> I recently went through a very similar experience with getopt,
> though I was working on subclassing a class that handled
> command line arguments (getopt proved to be completely
> unsuitable for that task).  I turned to t module I found on
> Parnassus called cmdline.py. It works beautifully, as expected,
> and the implementation is much more readable than the similar
> getopt version.
> 
> for some reason it doesn't show up in a Parnassus search for
> some reason it doesn't show up in a search for cmdline.py, but
> it does show up as one of about three hits in a search for
> getopt.
> 
> On Mon, Feb 19, 2001 at 05:21:47PM -0800, Sean 'Shaleh' Perry wrote:
> > Writing my fisrt real python app that people will run from the command line
> > (most of my previous coding has been modules for other software to web stuff)
> > I decided to use getopt.  My setup looks like this:
> >
> > LONG_OPTIONS = ['help', 'setup-lab', 'lab=']
> > SHORT_OPTIONS = 'hS'
> >
> > try:
> >     optlist, args = getopt.getopt(sys.argv[1:], SHORT_OPTIONS,
> >                                   LONG_OPTIONS)
> > except getopt.error, e:
> >     print e
> >     print USAGE
> >     sys.exit(1)
> >
> > for option, argument in optlist:
> >     if option in ('-h', '--help'):
> >         print USAGE
> >         sys.exit(0)
> >     elif option in ('-S', '--setup-lab'):
> >         print 'Want to set up the lab ...'
> >     elif option == 'lab':
> >         print 'Put lab in: ' + argument
> >
> > this means I have my options defined twice and then once again in the USAGE
> > output.  Is there not a better way to handle arguments?
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
> 
> --
> Stand Fast,
>     tjg.
> 
> Timothy Grant                         tjg at exceptionalminds.com
> Red Hat Certified Engineer            www.exceptionalminds.com
> Avalon Technology Group, Inc.                   (503) 246-3630
> >>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
> >>>>This machine was last rebooted:  34 days 21:04 hours ago<<


From jmarshal at mathworks.com  Tue Feb 20 02:25:22 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 20 Feb 2001 01:25:22 GMT
Subject: Proposal: allow '?' and '!' in identifiers
Message-ID: <96sh22$jq5$1@news.mathworks.com>

Anybody have any thoughts on allowing the characters '!' and '?' in
identifiers?  It's possible I'm missing something, but I think the
only ambiguity this introduces is in the case of something like the
following:

  a!=b

But whitespace easily disambiguates ("a! = b" vs "a != b").

The only snag I see is in the lexer.  For backward compatibility, "a!=b"
would need to be lexed as "a != b", but this may be unexpected, since the
lexer is currently greedy, trying to make tokens as large as possible
before moving on.

If the addition of '!' as a valid identifier character causes too many
waves, it would still be nice to see '?'.


From s2mdalle at titan.vcu.edu  Tue Feb 20 04:17:09 2001
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Tue, 20 Feb 2001 03:17:09 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com>
Message-ID: <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net>

In article <96sh22$jq5$1 at news.mathworks.com>, "Joshua Marshall"
<jmarshal at mathworks.com> wrote:

> Anybody have any thoughts on allowing the characters '!' and '?' in
> identifiers?  It's possible I'm missing something, but I think the only
> ambiguity this introduces is in the case of something like the following:
> 
>   a!=b
> 
> But whitespace easily disambiguates ("a! = b" vs "a != b").
> 
> The only snag I see is in the lexer.  For backward compatibility, "a!=b"
> would need to be lexed as "a != b", but this may be unexpected, since the
> lexer is currently greedy, trying to make tokens as large as possible before
> moving on.
> 
> If the addition of '!' as a valid identifier character causes too many
> waves, it would still be nice to see '?'.

Why?  How would this make the language more expressive?
How would it make your programming tasks easier?

Is introducing new and strange behavior (such as your
a!=b example) worth it?  Why?

-- 
David Allen
http://opop.nols.com/
----------------------------------------
Diplomacy is the art of saying "nice doggy" until you can find a gun.


From jmarshal at mathworks.com  Tue Feb 20 04:43:24 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 20 Feb 2001 03:43:24 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net>
Message-ID: <96sp4s$ni7$1@news.mathworks.com>

David Allen <s2mdalle at titan.vcu.edu> wrote:
...

>> If the addition of '!' as a valid identifier character causes too many
>> waves, it would still be nice to see '?'.

> Why?  How would this make the language more expressive?
> How would it make your programming tasks easier?

> Is introducing new and strange behavior (such as your
> a!=b example) worth it?  Why?

Introducing some strange behavior with '!' may very well not be worth
it.  However, while adding '?' as a legal identifier character does
not make the Python more expressive, it can make Python programs more
self-documenting.  The convention of ending predicate function names
with '?' is useful, in my opinion.


From max at alcyone.com  Tue Feb 20 05:32:52 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 19 Feb 2001 20:32:52 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com>
Message-ID: <3A91F374.580B1A10@alcyone.com>

Joshua Marshall wrote:

> Introducing some strange behavior with '!' may very well not be worth
> it.  However, while adding '?' as a legal identifier character does
> not make the Python more expressive, it can make Python programs more
> self-documenting.  The convention of ending predicate function names
> with '?' is useful, in my opinion.

But you don't need to literally have a ? at the end of the identifier to
get the message across if that is your desire.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ If love be good, from whence cometh my woe?
\__/ Chaucer
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From n8gray at caltech.edu.is.my.email.address  Tue Feb 20 07:21:35 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Mon, 19 Feb 2001 22:21:35 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com>
Message-ID: <96t2oi$4ce@gap.cco.caltech.edu>

Joshua Marshall wrote:

> David Allen <s2mdalle at titan.vcu.edu> wrote:
> ...
> 
> >> If the addition of '!' as a valid identifier character causes too many
> >> waves, it would still be nice to see '?'.
> 
> > Why?  How would this make the language more expressive?
> > How would it make your programming tasks easier?
> 
> > Is introducing new and strange behavior (such as your
> > a!=b example) worth it?  Why?
> 
> Introducing some strange behavior with '!' may very well not be worth
> it.  However, while adding '?' as a legal identifier character does
> not make the Python more expressive, it can make Python programs more
> self-documenting.  The convention of ending predicate function names
> with '?' is useful, in my opinion.

Hear hear!

(Or should that be "Hear hear?")

IMHO it's not worth adding "!", but if "?" doesn't break anything then it's 
a net gain.  I assert that it _would_ make Python more expressive to add an 
elegant way of indicating that a function returns only boolean values.

Unfortunately, I doubt you'll see much support for deviating from [a-zA-Z_] 
on this newsgroup.  Programmers of languages not descended from C (other 
than Python, of course) are in the minority, even among Pythonistas.  There 
may be some cause for hope, though, because among python-dev subscribers I 
suspect they're in the majority!

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From rc-bashar at gmx.net  Tue Feb 20 19:20:02 2001
From: rc-bashar at gmx.net (Christoph Horst)
Date: Tue, 20 Feb 2001 19:20:02 +0100
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu>
Message-ID: <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de>

Nathaniel Gray <n8gray at caltech.edu.is.my.email.address> wrote:
>IMHO it's not worth adding "!", but if "?" doesn't break anything then it's 
>a net gain.  I assert that it _would_ make Python more expressive to add an 
>elegant way of indicating that a function returns only boolean values.

What's wrong with prefixing the function name with 'is'?


From jmarshal at mathworks.com  Tue Feb 20 22:16:52 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 20 Feb 2001 21:16:52 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de>
Message-ID: <96ums4$qk2$1@news.mathworks.com>

Christoph Horst <rc-bashar at gmx.net> wrote:
> Nathaniel Gray <n8gray at caltech.edu.is.my.email.address> wrote:
>>IMHO it's not worth adding "!", but if "?" doesn't break anything then it's 
>>a net gain.  I assert that it _would_ make Python more expressive to add an 
>>elegant way of indicating that a function returns only boolean values.

> What's wrong with prefixing the function name with 'is'?

I wouldn't say anything is wrong with that approach.  But then I could
also ask "Why do we need underscores as valid identifier characters?
You can just capitalize the next letter."
The question mark is an unused character, and carries some
natural-language meaning.  It's useful.


From root at rainerdeyke.com  Wed Feb 21 04:34:58 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Wed, 21 Feb 2001 03:34:58 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <96ums4$qk2$1@news.mathworks.com>
Message-ID: <CHGk6.271777$ge4.94312785@news2.rdc2.tx.home.com>

"Joshua Marshall" <jmarshal at mathworks.com> wrote in message
news:96ums4$qk2$1 at news.mathworks.com...
> I wouldn't say anything is wrong with that approach.  But then I could
> also ask "Why do we need underscores as valid identifier characters?
> You can just capitalize the next letter."
> The question mark is an unused character, and carries some
> natural-language meaning.  It's useful.

This is mostly a question of aesthetics.  I like the way Python (and C/C++)
does it.  To me, the question mark looks like an operator; the underscore
does not.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From piet at cs.uu.nl  Wed Feb 21 15:39:18 2001
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 21 Feb 2001 15:39:18 +0100
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <96ums4$qk2$1@news.mathworks.com>
Message-ID: <uwvakgkx5.fsf@cs.uu.nl>

>>>>> Joshua Marshall <jmarshal at mathworks.com> (JM) writes:

JM> Christoph Horst <rc-bashar at gmx.net> wrote:
>> Nathaniel Gray <n8gray at caltech.edu.is.my.email.address> wrote:
>>> IMHO it's not worth adding "!", but if "?" doesn't break anything then it's 
>>> a net gain.  I assert that it _would_ make Python more expressive to add an 
>>> elegant way of indicating that a function returns only boolean values.

>> What's wrong with prefixing the function name with 'is'?

JM> I wouldn't say anything is wrong with that approach.  But then I could
JM> also ask "Why do we need underscores as valid identifier characters?
JM> You can just capitalize the next letter."
JM> The question mark is an unused character, and carries some
JM> natural-language meaning.  It's useful.

Please don't do it. There are more useful things to do with ? like the
ternary operator for conditional expressions.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From wolfson at midway.uchicago.edu  Wed Feb 21 22:36:35 2001
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Wed, 21 Feb 2001 21:36:35 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <96ums4$qk2$1@news.mathworks.com> <uwvakgkx5.fsf@cs.uu.nl>
Message-ID: <DxWk6.170$l25.2030@newsfeed.uchicago.edu>

In article <uwvakgkx5.fsf at cs.uu.nl>,  <piet at cs.uu.nl> wrote:
>>>>>> Joshua Marshall <jmarshal at mathworks.com> (JM) writes:
>
>JM> Christoph Horst <rc-bashar at gmx.net> wrote:
>>> Nathaniel Gray <n8gray at caltech.edu.is.my.email.address> wrote:
>>>> IMHO it's not worth adding "!", but if "?" doesn't break anything
>then it's 
>>>> a net gain.  I assert that it _would_ make Python more expressive
>to add an 
>>>> elegant way of indicating that a function returns only boolean values.
>
>>> What's wrong with prefixing the function name with 'is'?
>
>JM> I wouldn't say anything is wrong with that approach.  But then I could
>JM> also ask "Why do we need underscores as valid identifier characters?
>JM> You can just capitalize the next letter."
>JM> The question mark is an unused character, and carries some
>JM> natural-language meaning.  It's useful.
>
>Please don't do it. There are more useful things to do with ? like the
>ternary operator for conditional expressions.

Python isn't heading this way, but it's possible to allow a ? in
identifiers and still have it be an operator, by mandating that
operators have whitespace on both sides.



-- 
Barnabas T. Rumjuggler
For the world, I count it not an inn, but an hospital; and a place not to
live, but to die in.
 -- Thomas Browne, _Religio Medici_


From n8gray at caltech.edu.is.my.e-mail.address  Wed Feb 21 04:49:05 2001
From: n8gray at caltech.edu.is.my.e-mail.address (Nathan Gray)
Date: Tue, 20 Feb 2001 19:49:05 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de>
Message-ID: <3A933AB1.108@caltech.edu.is.my.e-mail.address>

Christoph Horst wrote:

> Nathaniel Gray <n8gray at caltech.edu.is.my.email.address> wrote:
> 
>> IMHO it's not worth adding "!", but if "?" doesn't break anything then it's 
>> a net gain.  I assert that it _would_ make Python more expressive to add an 
>> elegant way of indicating that a function returns only boolean values.
> 
> 
> What's wrong with prefixing the function name with 'is'?

Nothing.  I just find it pleasingly clear to read languages where 
predicates end with '?'.  It's simple and elegant.

Like I said, though, I'd be very surprised if this proposal ends up 
being accepted.

--
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
              Nathaniel Gray
    California Institute of Technology
      Computation and Neural Systems
      n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From n8gray at caltech.edu.is.my.email.address  Sun Feb 25 01:55:37 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Sat, 24 Feb 2001 16:55:37 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <3A933AB1.108@caltech.edu.is.my.e-mail.address> <m1r90qwkco.fsf@bertrand.carcosa.net>
Message-ID: <979lha$jg5@gap.cco.caltech.edu>

Jason F. McBrayer wrote:

> >>>>> "NG" == Nathan Gray <n8gray at caltech.edu.is.my.e-mail.address>
> >>>>> writes:
> 
> NG> Nothing.  I just find it pleasingly clear to read languages where
> NG> predicates end with '?'.  It's simple and elegant.
> 
> Why not be old-skool and end predicates with p?
> 

Why not be hardcore and allow both?

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From max at alcyone.com  Sun Feb 25 02:10:04 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 24 Feb 2001 17:10:04 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <3A933AB1.108@caltech.edu.is.my.e-mail.address> <m1r90qwkco.fsf@bertrand.carcosa.net> <979lha$jg5@gap.cco.caltech.edu>
Message-ID: <3A985B6C.C94048D1@alcyone.com>

Nathaniel Gray wrote:

> Why not be hardcore and allow both?

Because using a suffixed p to indicate predicates, if that is your
liking, is already supported by the language.  Allowing ? in identifiers
is not, and would require all sorts of weird parser logic to allow it in
identifiers but not break backward compatibility.  It is simply not
worth it.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ A physicist is an atom's way of knowing about atoms.
\__/ George Wald
    blackgirl international / http://www.blackgirl.org/
 The Internet resource for black women.


From grumble at mailcity.com  Sun Feb 25 17:36:31 2001
From: grumble at mailcity.com (jcm)
Date: 25 Feb 2001 16:36:31 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <3A933AB1.108@caltech.edu.is.my.e-mail.address> <m1r90qwkco.fsf@bertrand.carcosa.net> <979lha$jg5@gap.cco.caltech.edu> <3A985B6C.C94048D1@alcyone.com>
Message-ID: <97bcaf$noh$1@news.mathworks.com>

Erik Max Francis <max at alcyone.com> wrote:
> Nathaniel Gray wrote:

>> Why not be hardcore and allow both?

> Because using a suffixed p to indicate predicates, if that is your
> liking, is already supported by the language.  Allowing ? in identifiers
> is not, and would require all sorts of weird parser logic to allow it in
> identifiers but not break backward compatibility.  It is simply not
> worth it.

I understand that you dislike the idea of allowing '?' in identifiers.
But it would not necessarily require "all sorts of weird parser
logic".  I expect it's a simple change to the lexer.  I haven't looked
at the Python codebase, so I don't know for sure.  But I've written
enough lexers and parsers to know what should be easy or difficult.


From mwh21 at cam.ac.uk  Sun Feb 25 19:02:28 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 25 Feb 2001 18:02:28 +0000
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <3A933AB1.108@caltech.edu.is.my.e-mail.address> <m1r90qwkco.fsf@bertrand.carcosa.net> <979lha$jg5@gap.cco.caltech.edu> <3A985B6C.C94048D1@alcyone.com> <97bcaf$noh$1@news.mathworks.com>
Message-ID: <m3bsrqsksr.fsf@atrus.jesus.cam.ac.uk>

jcm <grumble at mailcity.com> writes:

> Erik Max Francis <max at alcyone.com> wrote:
> > Nathaniel Gray wrote:
> 
> >> Why not be hardcore and allow both?
> 
> > Because using a suffixed p to indicate predicates, if that is your
> > liking, is already supported by the language.  Allowing ? in identifiers
> > is not, and would require all sorts of weird parser logic to allow it in
> > identifiers but not break backward compatibility.  It is simply not
> > worth it.
> 
> I understand that you dislike the idea of allowing '?' in identifiers.
> But it would not necessarily require "all sorts of weird parser
> logic".  I expect it's a simple change to the lexer.  

It is.  It takes about a minute if you know where to look (which is in
Parser/tokenizer.c).  I did it once, but seem to have lost the patch.

Cheers,
M.

-- 
  ARTHUR:  Yes.  It was on display in the bottom of a locked filing
           cabinet stuck in a disused lavatory with a sign on the door
           saying "Beware of the Leopard".
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 1


From max at alcyone.com  Sun Feb 25 20:18:27 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 25 Feb 2001 11:18:27 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de> <3A933AB1.108@caltech.edu.is.my.e-mail.address> <m1r90qwkco.fsf@bertrand.carcosa.net> <979lha$jg5@gap.cco.caltech.edu> <3A985B6C.C94048D1@alcyone.com> <97bcaf$noh$1@news.mathworks.com>
Message-ID: <3A995A83.C20EF617@alcyone.com>

jcm wrote:

> I understand that you dislike the idea of allowing '?' in identifiers.
> But it would not necessarily require "all sorts of weird parser
> logic".  I expect it's a simple change to the lexer.  I haven't looked
> at the Python codebase, so I don't know for sure.  But I've written
> enough lexers and parsers to know what should be easy or difficult.

Sorry, I was comingling your requests for both '?' and '!' in
identifiers.  It was '!' that would have required ambiguity checks, not
'?'.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It is only to the individual that a soul is given.
\__/ Albert Einstein
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From grante at visi.com  Tue Feb 20 19:10:37 2001
From: grante at visi.com (Grant Edwards)
Date: Tue, 20 Feb 2001 18:10:37 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com>
Message-ID: <xqyk6.533$S7.49256@ruti.visi.com>

In article <96sp4s$ni7$1 at news.mathworks.com>, Joshua Marshall wrote:
>David Allen <s2mdalle at titan.vcu.edu> wrote:
>...
>
>>> If the addition of '!' as a valid identifier character causes too many
>>> waves, it would still be nice to see '?'.
>
>> Why?  How would this make the language more expressive?
>> How would it make your programming tasks easier?
>
>> Is introducing new and strange behavior (such as your
>> a!=b example) worth it?  Why?
>
>Introducing some strange behavior with '!' may very well not be worth
>it.  However, while adding '?' as a legal identifier character does
>not make the Python more expressive, it can make Python programs more
>self-documenting.  The convention of ending predicate function names
>with '?' is useful, in my opinion.

Definitely.  I always found that convention in Scheme to be
very expresive and compact.  The convention of ending
predicates with "!" when they modified internal state was also
useful though perhaps not as intuitive as "?".

-- 
Grant Edwards                   grante             Yow!  Now I am depressed...
                                  at               
                               visi.com            


From othello at javanet.com  Wed Feb 21 01:17:03 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 20 Feb 2001 19:17:03 -0500
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <xqyk6.533$S7.49256@ruti.visi.com>
Message-ID: <3A9308FF.D07FB37F@javanet.com>

>
> >Introducing some strange behavior with '!' may very well not be worth
> >it.  However, while adding '?' as a legal identifier character does
> >not make the Python more expressive, it can make Python programs more
> >self-documenting.  The convention of ending predicate function names
> >with '?' is useful, in my opinion.
>
> Definitely.  I always found that convention in Scheme to be
> very expresive and compact.  The convention of ending
> predicates with "!" when they modified internal state was also
> useful though perhaps not as intuitive as "?".
>

These naming conventions were useful to me iin Forth and Scheme;
however, they're ugly (IMHO).  The plain English forms are much
more expressive and flexible without carrying the risk of confusing
my optical lexer when I read code.

Count my vote:  -1

Raymond

"In theory, there is no difference between theory and practice.
 In practice, there is." -- Yogi Berra




From gregj at pobox.com  Sun Feb 25 20:10:16 2001
From: gregj at pobox.com (Gregory Jorgensen)
Date: Sun, 25 Feb 2001 19:10:16 GMT
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <xqyk6.533$S7.49256@ruti.visi.com> <3A9308FF.D07FB37F@javanet.com>
Message-ID: <sMcm6.959$aw5.1833@www.newsranger.com>

I vote against. Don't try to change Python into Scheme or Forth or whatever. If
you find another language more expressive or elegant than Python, use that
language. I find Python expressive and elegant because it doesn't have a lot of
chaff and historical baggage.

While I'm ranting... features of programming languages aren't intuitive. Some
people may deduce the syntax easier than others, but that is experience, not
intuition. My intuition tells me that any non-alphanumeric character in a symbol
name may mean something special to the compiler; why would I think ? relates to
its English usage when _ doesn't? In the context of reading and learning Python
neither ? nor _ say anything to my intuition.


In article <3A9308FF.D07FB37F at javanet.com>, Raymond Hettinger says...
>> Definitely.  I always found that convention in Scheme to be
>> very expresive and compact.  The convention of ending
>> predicates with "!" when they modified internal state was also
>> useful though perhaps not as intuitive as "?".
>>
>
>These naming conventions were useful to me iin Forth and Scheme;
>however, they're ugly (IMHO).  The plain English forms are much
>more expressive and flexible without carrying the risk of confusing
>my optical lexer when I read code.
>
>Count my vote:  -1
>
>Raymond

Greg Jorgensen
Deschooling Society
gregj at pobox.com


From max at alcyone.com  Tue Feb 20 05:31:58 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 19 Feb 2001 20:31:58 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com>
Message-ID: <3A91F33E.C779FF1@alcyone.com>

Joshua Marshall wrote:

> Anybody have any thoughts on allowing the characters '!' and '?' in
> identifiers?  It's possible I'm missing something, but I think the
> only ambiguity this introduces is in the case of something like the
> following:
> 
>   a!=b
> 
> But whitespace easily disambiguates ("a! = b" vs "a != b").

But missing or adding a seemingly harmless space means that you've got a
completely different symbol.  Bad news.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I don't like principles.  I prefer prejudices.
\__/ Oscar Wilde
    Alcyone Systems / http://www.alcyone.com/
 Alcyone Systems, San Jose, California.


From lsloan at bigfoot.com  Tue Feb 20 02:28:41 2001
From: lsloan at bigfoot.com (Lance E Sloan)
Date: Tue, 20 Feb 2001 01:28:41 -0000
Subject: multiple parameters to append()?
Message-ID: <96sh89+rk7v@eGroups.com>

I downloaded an old-ish Python module, Julian, and
my Python 2.0 complained about this line from its
Zone.py file (line 61):

  self.ttinfos.append(up.unpack_int(), ord(data[6*i + 4]), 
ord(data[6*i + 5]))

So I changed it to:

  self.ttinfos.append((up.unpack_int(), ord(data[6*i + 4]), 
ord(data[6*i + 5])))

Basically, I just added another set of parens to make the
arguments all part of a tuple that is appended instead.
I don't think this is correct, though.  Although Python doesn't
complain any more, the module doesn't work exactly as
described.

I saw in the Python documentation that this multi-parameter
usage of append() is no longer allowed and has been
deprecated for some time.  How exactly did the old usage
of append() work?  I don't think I can correct the problem in
this module unless I know that.






From amk at mira.erols.com  Tue Feb 20 02:43:26 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 20 Feb 2001 01:43:26 GMT
Subject: multiple parameters to append()?
References: <mailman.982632563.19686.python-list@python.org>
Message-ID: <slrn993jp0.t1.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>

On Tue, 20 Feb 2001 01:28:41 -0000, 
	Lance E Sloan <lsloan at bigfoot.com> wrote:
>deprecated for some time.  How exactly did the old usage
>of append() work?  I don't think I can correct the problem in
>this module unless I know that.

Exactly as you expected:

Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> L=[1,2]
>>> L.append(3,4)
>>> L
[1, 2, (3, 4)]
>>>

Perhaps the module is broken for some other reason; good luck with
your debugging.

--amk


From ruediger.maehl at dzsh.de  Wed Feb 21 12:35:07 2001
From: ruediger.maehl at dzsh.de (Ruediger Maehl)
Date: Wed, 21 Feb 2001 12:35:07 +0100
Subject: multiple parameters to append()?
References: <mailman.982632563.19686.python-list@python.org>
Message-ID: <newscache$osu39g$l2k$1@www.dzsh.de>

Lance E Sloan schrieb in Nachricht ...

>  self.ttinfos.append(up.unpack_int(), ord(data[6*i + 4]),
>ord(data[6*i + 5]))
>
how about:

self.ttinfos = self.ttinfos + [up.unpack_int(), ord(data[6*i + 4]),
ord(data[6*i + 5])]

HTH
R?diger




From shaleh at valinux.com  Tue Feb 20 02:44:06 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Mon, 19 Feb 2001 17:44:06 -0800
Subject: multiple parameters to append()?
In-Reply-To: <96sh89+rk7v@eGroups.com>; from lsloan@bigfoot.com on Tue, Feb 20, 2001 at 01:28:41AM -0000
References: <96sh89+rk7v@eGroups.com>
Message-ID: <20010219174406.H29288@valinux.com>

On Tue, Feb 20, 2001 at 01:28:41AM -0000, Lance E Sloan wrote:
> I downloaded an old-ish Python module, Julian, and
> my Python 2.0 complained about this line from its
> Zone.py file (line 61):
>

what you changed seems valid.  Where did you find the module?  Would like to
inspect it myself. 



From lsloan at yahoo.com  Tue Feb 20 15:41:40 2001
From: lsloan at yahoo.com (Lance E Sloan)
Date: Tue, 20 Feb 2001 06:41:40 -0800 (PST)
Subject: multiple parameters to append()?
In-Reply-To: <20010219174406.H29288@valinux.com>
Message-ID: <20010220144140.83710.qmail@web9504.mail.yahoo.com>

--- Sean 'Shaleh' Perry <shaleh at valinux.com> wrote:
> what you changed seems valid.  Where did you find the module?  Would
> like to inspect it myself. 

It's Ted Horst's Julian module, which I found on Parnassus.  You can
get it directly from <URL:http://www.lib.uchicago.edu/keith/software/Julian.tar.gz>.

=====
Lance E Sloan
lsloan at bigfoot.com


__________________________________________________
Do You Yahoo!?
Get personalized email addresses from Yahoo! Mail - only $35 
a year!  http://personal.mail.yahoo.com/



From dsavitsk at e-coli.net  Tue Feb 20 03:42:23 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Mon, 19 Feb 2001 20:42:23 -0600
Subject: Creating Python Data Server
Message-ID: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net>

This is sort of Python related, that is, if I write it, it will be in
Python.  I am really wondering if this exists, or is even practical ...

I am trying to set up client/server system like a database server with
stored procedures and what not where the procedures are written in python
(or any other language), live on the server, and that isn't tied to any
particular database. So the user would press a button on the interface
sending a signal to the server which would instantiate whatever objects were
necessary on the server, process the data, and then return the data to the
client.

i though about using apache/mod_python, but i want the user to feel like
they are running things on their computer, that is, like all of the
processes are local and the client and server are the same.  Thus http
doesn't seem to be what i want.

I'll write the interface in VB or Delphi, but I want all of the processing
to take place on the big fast server.  A Telnet-like system would be great,
but the win32 style interface with all of the prerequisite buttons and
progress bars is essential. Also, I would rather use UNIX/MySQL on the
server, but i don't want the users to know that.

Does this sound like a reasonable project? and is there anything similar or
does anyone have any better ideas?

thanks,
doug





From cg at gaia.intranet.cdegroot.com  Tue Feb 20 09:18:49 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 20 Feb 2001 09:18:49 +0100
Subject: Creating Python Data Server
References: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net>
Message-ID: <96t999$qrl$1@gaia.intranet.cdegroot.com>

dsavitsk <dsavitsk at e-coli.net> said:
>This is sort of Python related, that is, if I write it, it will be in
>Python.  I am really wondering if this exists, or is even practical ...
>
Well, it seems that PostgreSQL has the concept of pluggable back-end
languages although only Tcl and Perl have been implemented yet. For
this sort of projects, I'd recommend PostgreSQL over MySQL: PG is a
real RDBMS and is object-relational (so you can do subclassing etcetera
in the database instead of in some O/R mapping layer). MySQL is fun
for storing user accounts and cookies behind websites, but not for
serious data management.

-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From stephen_purcell at yahoo.com  Tue Feb 20 10:19:28 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 20 Feb 2001 10:19:28 +0100
Subject: Creating Python Data Server
In-Reply-To: <96t999$qrl$1@gaia.intranet.cdegroot.com>; from cg@gaia.intranet.cdegroot.com on Tue, Feb 20, 2001 at 09:18:49AM +0100
References: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net> <96t999$qrl$1@gaia.intranet.cdegroot.com>
Message-ID: <20010220101928.A14400@freedom.puma-ag.com>

Cees de Groot wrote:
> dsavitsk <dsavitsk at e-coli.net> said:
> >This is sort of Python related, that is, if I write it, it will be in
> >Python.  I am really wondering if this exists, or is even practical ...
> >
> Well, it seems that PostgreSQL has the concept of pluggable back-end
> languages although only Tcl and Perl have been implemented yet. For
> this sort of projects, I'd recommend PostgreSQL over MySQL: PG is a
> real RDBMS and is object-relational (so you can do subclassing etcetera
> in the database instead of in some O/R mapping layer). MySQL is fun
> for storing user accounts and cookies behind websites, but not for
> serious data management.

Agreed. PostGreSQL is the best database I've ever used, and it keeps getting
better. It manages many hundreds of queries per second on my laptop...
and it doesn't take a DBA to install it. (Hint: rpm)

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From aleaxit at yahoo.com  Tue Feb 20 11:39:43 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 11:39:43 +0100
Subject: Creating Python Data Server
References: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net>
Message-ID: <96thhf0rc1@news2.newsguy.com>

"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:jQkk6.117367$Ch.20710042 at newsrump.sjc.telocity.net...
    [snip]
> i though about using apache/mod_python, but i want the user to feel like
> they are running things on their computer, that is, like all of the
> processes are local and the client and server are the same.  Thus http
> doesn't seem to be what i want.

Why not?  It's an excellent protocol for your needs, and you'll find
lots of infrastructure already in place for it.  To give your user
the illusion they're running things locally, you may just need to
build a custom *client* -- which you can do _after_ you've got the
basic functionality running.  Easy-as-pie.

> I'll write the interface in VB or Delphi, but I want all of the processing
> to take place on the big fast server.  A Telnet-like system would be
great,

Nothing will stop you from writing the ('a'?) custom-client in VB or
Delphi at some point.  But HTTP is a better protocol than Telnet for
your needs -- and even if it weren't, just the infrastructure issues
might justify it anyway (think firewalls and proxies, etc, etc).


Alex





From joseamartin at venmex.com  Tue Feb 20 13:18:17 2001
From: joseamartin at venmex.com (Jose Martin)
Date: Tue, 20 Feb 2001 08:18:17 -0400
Subject: Creating Python Data Server
References: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net>
Message-ID: <t94orrqlvcjae7@corp.supernews.com>

Do you Know Gadfy, a SQL Client Server Database writen in python ?
Look at  the vault of parnasus.
Works fine.

dsavitsk <dsavitsk at e-coli.net> escribi? en el mensaje de noticias
jQkk6.117367$Ch.20710042 at newsrump.sjc.telocity.net...
> This is sort of Python related, that is, if I write it, it will be in
> Python.  I am really wondering if this exists, or is even practical ...
>
> I am trying to set up client/server system like a database server with
> stored procedures and what not where the procedures are written in python
> (or any other language), live on the server, and that isn't tied to any
> particular database. So the user would press a button on the interface
> sending a signal to the server which would instantiate whatever objects
were
> necessary on the server, process the data, and then return the data to the
> client.
>
> i though about using apache/mod_python, but i want the user to feel like
> they are running things on their computer, that is, like all of the
> processes are local and the client and server are the same.  Thus http
> doesn't seem to be what i want.
>
> I'll write the interface in VB or Delphi, but I want all of the processing
> to take place on the big fast server.  A Telnet-like system would be
great,
> but the win32 style interface with all of the prerequisite buttons and
> progress bars is essential. Also, I would rather use UNIX/MySQL on the
> server, but i don't want the users to know that.
>
> Does this sound like a reasonable project? and is there anything similar
or
> does anyone have any better ideas?
>
> thanks,
> doug
>
>
>




From pbrian at demon.net  Tue Feb 20 13:46:49 2001
From: pbrian at demon.net (Paul Brian)
Date: Tue, 20 Feb 2001 12:46:49 -0000
Subject: Creating Python Data Server
References: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net>
Message-ID: <982673034.5300.0.nnrp-10.c1c3e154@news.demon.co.uk>

May I say that unless there is an extremely strong reason for doing anymore
than presentation like work on the client that you do use apache/mod_pyhton
(or PSP or PHP anything else).

In general, the rule of thumb that puts data and business logic on central
servers and presentation on web servers or client is a good one. And
breaking that rule in the first stages of design does not sound good.

If you want the look and feel of a intelligent client rather than bare web
pages however, perhaps the following solutions might make sense:

-pickle python dictionary objects that represent the results table, squirt
as data through http, unpickle at the client end.
-Alternatively use apache to create results tables that it delviers as plain
text (XML?)

Then create a client that does what you want, (staying with Python) using
PyQT and QTdesigner (almost as easy as VB), or JPython and Swing (if you
know this - I have not yet found a neat way of WYSIWYG'ing Swing and still
being able to use jPython with it - any ideas anyone?

However may I suggest that you concentrate on the functionality that your
users desire first, and put it into plain boring web pages using
apache/mod_python / PHP/JSP. That way you
a) avoid the big management headaches of maintaining uptodate versions on
all clients (I am asuming this is a inhouse business problem)
b) Avoid big investment in gui design before functionality is all fixed
c) can change functionality faster
d) surface product quicker
e) can add in authentication and authorisation far more easly....
f) can throw it away if it is just completely not wanted.

good luck


"dsavitsk" <dsavitsk at e-coli.net> wrote in message
news:jQkk6.117367$Ch.20710042 at newsrump.sjc.telocity.net...
> This is sort of Python related, that is, if I write it, it will be in
> Python.  I am really wondering if this exists, or is even practical ...
>
> I am trying to set up client/server system like a database server with
> stored procedures and what not where the procedures are written in python
> (or any other language), live on the server, and that isn't tied to any
> particular database. So the user would press a button on the interface
> sending a signal to the server which would instantiate whatever objects
were
> necessary on the server, process the data, and then return the data to the
> client.
>
> i though about using apache/mod_python, but i want the user to feel like
> they are running things on their computer, that is, like all of the
> processes are local and the client and server are the same.  Thus http
> doesn't seem to be what i want.
>
> I'll write the interface in VB or Delphi, but I want all of the processing
> to take place on the big fast server.  A Telnet-like system would be
great,
> but the win32 style interface with all of the prerequisite buttons and
> progress bars is essential. Also, I would rather use UNIX/MySQL on the
> server, but i don't want the users to know that.
>
> Does this sound like a reasonable project? and is there anything similar
or
> does anyone have any better ideas?
>
> thanks,
> doug
>
>
>




From dsavitsk at e-coli.net  Wed Feb 21 00:11:18 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 20 Feb 2001 17:11:18 -0600
Subject: Creating Python Data Server
References: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net> <982673034.5300.0.nnrp-10.c1c3e154@news.demon.co.uk>
Message-ID: <qQCk6.120734$Ch.21260050@newsrump.sjc.telocity.net>

"Paul Brian" <pbrian at demon.net> wrote in message
news:982673034.5300.0.nnrp-10.c1c3e154 at news.demon.co.uk...
> May I say that unless there is an extremely strong reason for doing
anymore
> than presentation like work on the client that you do use
apache/mod_pyhton
> (or PSP or PHP anything else).
>
> In general, the rule of thumb that puts data and business logic on central
> servers and presentation on web servers or client is a good one. And
> breaking that rule in the first stages of design does not sound good.

I'm not meaning to break it, in fact, i am trying to use what i have to
extend it as much as i can.  I can use MySQL or Postgres for some stuff, but
i have to produce results in MS Access tables or i'll get in trouble. thus
making access look like a db server would be beneficial.  I have argued for
a change for 2 years to no avail.

> If you want the look and feel of a intelligent client rather than bare web
> pages however, perhaps the following solutions might make sense:
>
> -pickle python dictionary objects that represent the results table, squirt
> as data through http, unpickle at the client end.
> -Alternatively use apache to create results tables that it delviers as
plain
> text (XML?)

The biggest problem, and it may not really be a big one, is that some
commands take a long time to run.  It is a test scoring program, and
generating the statistics can take a while (obviously i'll write this stuff
in something that isn't as slow as python).  Thus I need the server to
return buffered data so that the client can update progress settings.
assuming apache/mod_python/mod_otherstuff can do this then it shouldn't be
too tough.

>
> Then create a client that does what you want, (staying with Python) using
> PyQT and QTdesigner (almost as easy as VB), or JPython and Swing (if you
> know this - I have not yet found a neat way of WYSIWYG'ing Swing and still
> being able to use jPython with it - any ideas anyone?
>
> However may I suggest that you concentrate on the functionality that your
> users desire first, and put it into plain boring web pages using
> apache/mod_python / PHP/JSP. That way you



> a) avoid the big management headaches of maintaining uptodate versions on
> all clients (I am asuming this is a inhouse business problem)
> b) Avoid big investment in gui design before functionality is all fixed
> c) can change functionality faster
> d) surface product quicker

This project is in essence rewriting a VB program from a while back.  There
are certain necessary additions/problems with the original that make it
easier to start over.  I do have the luxury of not needing to produce
incremental versions, and especially not needing to install on more than 1
computer until it is working.  Also, i have found that if I make VB
interfaces that are dynamic and configure themselves with registry/text file
settings it makes updates easier as only major overhalls need
reinstallation.

Thanks for all of the suggestions.

doug




From db3l at fitlinxx.com  Fri Feb 23 03:35:41 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 22 Feb 2001 21:35:41 -0500
Subject: Creating Python Data Server
References: <jQkk6.117367$Ch.20710042@newsrump.sjc.telocity.net>
Message-ID: <uk86i5doi.fsf@ctwd0143.fitlinxx.com>

"dsavitsk" <dsavitsk at e-coli.net> writes:

> I am trying to set up client/server system like a database server with
> stored procedures and what not where the procedures are written in python
> (or any other language), live on the server, and that isn't tied to any
> particular database. So the user would press a button on the interface
> sending a signal to the server which would instantiate whatever objects were
> necessary on the server, process the data, and then return the data to the
> client.

It sounds like a good fit for a straight forward RPC (remote procedure
call) or remote component based sort of setup.

Have you considered doing your client in Python as well (you could use
wxPython or Tkinter for a GUI)?  If so, you should definitely look at
XML-RPC.  It's an awfully simple (yet powerful) mechanism for making
remote calls, and absolutely dirt simple if both ends are in Python.
Support also exists for other clients (including a COM client that
could be used from VB).

See http://www.xmlrpc.org and http://www.pythonware.com/products/xmlrpc
for some more info.

Or, if you definitely want to use VB/Delphi for the client, you could
consider implementing your server procedures as COM objects in Python
(with the win32all package), and letting VB access them as it would
other system components - using DCOM in this case for a remote
invocation.

You mentioned in a later post about wanting to have status on the
client over the course of execution - that's really going to come into
play when you define what the remote calls are and/or your component
interface.  You could have one call to trigger the overall procedure,
but just return immediately after starting it, with a handle that you
could use later to check back.  Or you could even multi-thread your
client, and let the server issue its own callback remote calls during
course of execution to the client.

> i though about using apache/mod_python, but i want the user to feel like
> they are running things on their computer, that is, like all of the
> processes are local and the client and server are the same.  Thus http
> doesn't seem to be what i want.

Don't mistake HTTP (the protocol) with clients such as browsers that
may use the protocol.  While it's got warts, HTTP is often a fine
choice as the basic transport mechanism between clients and servers,
particularly as it has a big advantage in getting through firewalls or
other security precautions that are often locked down against anything
besides "web" access.  Besides, at its root, it's just a TCP stream
with some header information, which is fine.

XML-RPC defaults to using HTTP as its own transport, but its all
beneath the covers, so your client/presentation module can do anything
you want.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From s2mdalle at titan.vcu.edu  Tue Feb 20 03:51:16 2001
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Tue, 20 Feb 2001 02:51:16 GMT
Subject: Tkinter Menu.add_checkbutton()
Message-ID: <EYkk6.4853$lx.2398287@typhoon2.ba-dsg.net>

Just a quick question:

When I add a checkbutton to a menu:

v = BooleanVar()
v.set(FALSE)

foomenu.add_checkbutton(label="Option Foo",
                        variable=v,
                        showcolor='#00FF00',
                        command=some_callback)

The checkbutton always displays as 'off' regardless
of the value of 'v'.  See, depending on the value
of 'v', I'd like to have it start as on if v is
true.

I've also tried invoke()'ing this item, since in
the Tkinter class reference, that says it's the same
thing as the user clicking on it, but that doesn't
work either.

Any suggestions on how to make this start up as
checked?  The default is unchecked.
-- 
David Allen
http://opop.nols.com/
----------------------------------------
A language that doesn't affect the way you think about programming is 
not worth knowing.


From matt at mondoinfo.com  Tue Feb 20 04:57:27 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Tue, 20 Feb 2001 03:57:27 GMT
Subject: Tkinter Menu.add_checkbutton()
References: <EYkk6.4853$lx.2398287@typhoon2.ba-dsg.net>
Message-ID: <slrn993qp7.t3.matt@happy-hour.mondoinfo.com>

On Tue, 20 Feb 2001 02:51:16 GMT, David Allen <s2mdalle at titan.vcu.edu>
wrote:

>Just a quick question:
>
>When I add a checkbutton to a menu:

[. . .]

>The checkbutton always displays as 'off' regardless of the value of
>'v'.  See, depending on the value of 'v', I'd like to have it start
>as on if v is true.

I'm not sure what's going on with your code, but setting a checkbox to
on at the start works for me.  One thing that I was confused by is
that the default action is to toggle the state of the checkbox: it's
not necessary to do that in a callback. But that's probably not the
problem that you're having. I'll append a minimal example that seems
to work for me.

BTW, Fredrik Lundh's excellent An Introduction to Tkinter is very
useful. I almost always have a copy open when I'm doing Tkinter
programming. It's at:

http://www.pythonware.com/library/an-introduction-to-tkinter.htm

And I find Pmw's menus easier to deal with than the native Tkinter
kind. Pmw is at:

http://pmw.sourceforge.net/

Regards
Matt


import sys
from Tkinter import *

class mainWin:
  def __init__(self):
    self.root=Tk()
    self.toggleVar=IntVar()
    self.toggleVar.set(1)
    self.createWidgets()
    return None

  def createWidgets(self):
    menuBar=Menu(self.root)
    fileMenu=Menu(menuBar, tearoff=0)
    fileMenu.add_checkbutton(label="Toggle",variable=self.toggleVar)
    fileMenu.add_command(label="Quit", command=self.quitCB)
    menuBar.add_cascade(label="File",menu=fileMenu)
    self.root.config(menu=menuBar)
    return None

  def quitCB(self):
    sys.exit(0)

def main():
  mainWin()
  mainloop()

if __name__== '__main__':
  main()


From m_hussain at trillium.com  Tue Feb 20 06:15:23 2001
From: m_hussain at trillium.com (Hussain, Mohammed)
Date: Mon, 19 Feb 2001 21:15:23 -0800
Subject: Info requested
Message-ID: <D1C0BF20D4AFD411AB98009027AE9988C3FE9B@FMSMSX40>

Hi,
I am a new user of Python and would like to know whether there is any useful
website which gives a brief but detailed description of Python.
Thanks in advance.

Thanks and Regards,
Mohd Azaz Hussain





From pplumlee at omnigon.com  Tue Feb 20 06:52:15 2001
From: pplumlee at omnigon.com (Phlip)
Date: 20 Feb 2001 05:52:15 GMT
Subject: Info requested
References: <mailman.982646242.5470.python-list@python.org>
Message-ID: <3A920590.EFDB3EE9@omnigon.com>

"Hussain, Mohammed" wrote:
> 
> Hi,
> I am a new user of Python and would like to know whether there is any useful
> website which gives a brief but detailed description of Python.

Well, the Python community is extremely cryptic and obscure, so I'm
certain there are no links that say "tutorial" on this site:

	http://www.python.org

Or even this one:

	http://www.google.com

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  Founding member of NuGWa -
        Nudists for Global Warming  --


From eq3pvl at eq.uc.pt  Tue Feb 20 14:05:33 2001
From: eq3pvl at eq.uc.pt (Pedro Vale Lima)
Date: Tue, 20 Feb 2001 13:05:33 +0000
Subject: Info requested
References: <mailman.982646242.5470.python-list@python.org>
Message-ID: <3A926B9D.6BEF3A75@eq.uc.pt>

"Hussain, Mohammed" wrote:

> Hi,
> I am a new user of Python and would like to know whether there is any useful
> website which gives a brief but detailed description of Python.
> Thanks in advance.
>
> Thanks and Regards,
> Mohd Azaz Hussain

www.python.org

and also

diveintopython.org

regards
pedro



From grante at visi.com  Tue Feb 20 20:02:03 2001
From: grante at visi.com (Grant Edwards)
Date: Tue, 20 Feb 2001 19:02:03 GMT
Subject: Info requested
References: <mailman.982646242.5470.python-list@python.org>
Message-ID: <Lazk6.598$S7.50229@ruti.visi.com>

In article <mailman.982646242.5470.python-list at python.org>, Hussain, Mohammed wrote:

>I am a new user of Python and would like to know whether there is any useful
>website which gives a brief but detailed description of Python.
>Thanks in advance.

The computing world has become so web-centric that it
absolutely stuns me when somebody finds comp.lang.python but
can't find www.python.org. When looking for info on XYZ, I
assume that everybody naturally starts with
www.XYZ.[org|com|net].

Anyway, the web pages you're looking for is

  http://www.python.org/doc/Summary.html
  http://www.python.org/doc/current/tut/tut.html

-- 
Grant Edwards                   grante             Yow!  Imagine--a WORLD
                                  at               without POODLES...
                               visi.com            


From AshKapur at worldnet.att.net  Tue Feb 20 07:48:05 2001
From: AshKapur at worldnet.att.net (Ash Kapur)
Date: Tue, 20 Feb 2001 06:48:05 GMT
Subject: Static DLL
Message-ID: <Fqok6.5506$TD1.439939@bgtnsc05-news.ops.worldnet.att.net>

Hi,

I have an Win32 app that already loads a static dll (e.g. links with the
.lib, so
autoloads the dll when it starts). Using Dev Studio on 98.

I am trying to embed python20 into this app, so that I can use it to call
some
of the functions in the dll to test them.  The embedding went smoothly, I
can pass
command lines and file names to python and it executes them.

I used SWIG to create wrappers for some of the dll functions I want to call,
that also
went quite smoothly.

The problem is when I do an "import  dllname" it seems to load a new copy of
the dll,
(or at least there is a second set of static data at different places, and
breakpoints etc
do not work when the code is called from python).

Does anyone know of any way I can get python to know of my dll without doing
an
import, so that the already-loaded dll is used ? It would be a pain to have
to go back
to my existing app and change the architecture, as the dll exports lots of
functions.

Thanks in advance

Ash Kapur






From fredrik at pythonware.com  Tue Feb 20 08:45:41 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 20 Feb 2001 07:45:41 GMT
Subject: Safe to change SRE config to allow more than 100 groups?
References: <48b39tsc9eivp3fqeqe6pd5hn3fj3h148a@4ax.com>
Message-ID: <Fgpk6.14586$Qb7.2327195@newsb.telia.net>

Aaron Freeman wrote:
> I have an egregiously long regex, and I'm getting "AssertionError:
> sorry, but this version only supports 100 named groups" from SRE.
>
> After looking at SRE_Compile.py, I found where this is hardcoded to
> 100 (line 367), but but couldn't figure out why this was the arbitrary
> limitation.
>
> I tried changing it to 1000, and everything seemed to work fine

I doubt you ran it with 1000 named groups...  if you do,
you'll overrun buffers inside the _sre engine.  look in the
C sources for details.

> I also changed the maxcashe size in sre.py to 1000

if you have that many individual patterns, you're probably better
off compiling them in your script.

Cheers /F




From macky at edsamail.com  Tue Feb 20 09:33:53 2001
From: macky at edsamail.com (Macky)
Date: Tue, 20 Feb 2001 16:33:53 +0800
Subject: subsribe
Message-ID: <000b01c09b17$db7e1440$5764a8c0@edsamail.com>

----------------------------------------------------------------------
The information contained in this message (including any attachments)
is confidential and intended solely for the attention and use of the
named addressee(s). It must not be copied, distributed nor disclosed
to any person. If you are not the intended recipient, please delete
it from your system and notify sender immediately. Any disclosure,
copying or distribution thereof or any action taken or omitted to be
taken in reliance thereon is prohibited and may be unlawful.
----------------------------------------------------------------------




From macky at edsamail.com  Tue Feb 20 10:34:07 2001
From: macky at edsamail.com (Macky)
Date: Tue, 20 Feb 2001 17:34:07 +0800
Subject: Python for AOLServer
Message-ID: <003a01c09b20$45a872e0$5764a8c0@edsamail.com>

Im installing Python for AOLServer."PyWX" and encountered this error in

# make all
c++ -c
s_pywx.cc -DAOLSERVER_VERSION_3_2 -I. -I/usr/local/include/python2.0  -g -O2
 -o ns_pywx.o
In file included from ns_pywx.cc:64:
ns_pywx.hh:36: Python.h: No such file or directory
In file included from ns_pywx.cc:65:
PyWX_buffer.hh:68: Python.h: No such file or directory
PyWX_buffer.hh:69: compile.h: No such file or directory
PyWX_buffer.hh:70: eval.h: No such file or directory
PyWX_buffer.hh:71: structmember.h: No such file or directory
make: *** [ns_pywx.o] Error 1

AOLServer 3.2
INSTALLED COMPONENTS.
    SWIG 1.1
    Python 2.0


any suggestion and/or comments will be highly appreciated..





----------------------------------------------------------------------
The information contained in this message (including any attachments)
is confidential and intended solely for the attention and use of the
named addressee(s). It must not be copied, distributed nor disclosed
to any person. If you are not the intended recipient, please delete
it from your system and notify sender immediately. Any disclosure,
copying or distribution thereof or any action taken or omitted to be
taken in reliance thereon is prohibited and may be unlawful.
----------------------------------------------------------------------




From fredrik at pythonware.com  Tue Feb 20 10:58:26 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 20 Feb 2001 09:58:26 GMT
Subject: Python for AOLServer
References: <mailman.982661843.28561.python-list@python.org>
Message-ID: <6drk6.14593$Qb7.2329730@newsb.telia.net>

Macky wrote:
> Im installing Python for AOLServer."PyWX" and encountered this error in
>
> # make all
> c++ -c
> s_pywx.cc -DAOLSERVER_VERSION_3_2 -I. -I/usr/local/include/python2.0  -g -O2
>  -o ns_pywx.o
> In file included from ns_pywx.cc:64:
> ns_pywx.hh:36: Python.h: No such file or directory
> In file included from ns_pywx.cc:65:
> PyWX_buffer.hh:68: Python.h: No such file or directory
> PyWX_buffer.hh:69: compile.h: No such file or directory
> PyWX_buffer.hh:70: eval.h: No such file or directory
> PyWX_buffer.hh:71: structmember.h: No such file or directory
> make: *** [ns_pywx.o] Error 1
>
> AOLServer 3.2
> INSTALLED COMPONENTS.
>     SWIG 1.1
>     Python 2.0
>
> any suggestion and/or comments will be highly appreciated..

$ ls /usr/local/include/python2.0
$ more /usr/local/include/python2.0/Python.h

Cheers /F




From mhagger at alum.mit.edu  Tue Feb 27 05:15:23 2001
From: mhagger at alum.mit.edu (Michael Haggerty)
Date: Tue, 27 Feb 2001 04:15:23 GMT
Subject: Python for AOLServer
References: <mailman.982661843.28561.python-list@python.org>
Message-ID: <m28zms7odg.fsf@freak.kaiserty.com>

"Macky" <macky at edsamail.com> writes:

> Im installing Python for AOLServer."PyWX" and encountered this error in
> 
> # make all
> c++ -c
> s_pywx.cc -DAOLSERVER_VERSION_3_2 -I. -I/usr/local/include/python2.0  -g -O2
>  -o ns_pywx.o
> In file included from ns_pywx.cc:64:
> ns_pywx.hh:36: Python.h: No such file or directory

You might have to edit the Makefile (the autoconfiguration of PyWX is
not so good.)  Perhaps the -I include file is incorrect.  Make sure
that you include the directory that includes the Python include files
like Python.h.

If the Python include files are not present on your system at all, it
is possible that you don't have all of Python installed.  Maybe there
is a "python-devel" package or some such that you need to install.

Michael

-- 
Michael Haggerty
mhagger at alum.mit.edu


From macky at edsamail.com  Tue Feb 20 10:48:41 2001
From: macky at edsamail.com (Macky)
Date: Tue, 20 Feb 2001 17:48:41 +0800
Subject: Newsgroup
Message-ID: <005401c09b22$4eb8c860$5764a8c0@edsamail.com>

Im using OUTLOOK EXPRESS to read news group.. and its not the easy to use...
If im going to add newsgroup on OUTLOOK it alays asks for         News
(NNTP) Server?


----------------------------------------------------------------------
The information contained in this message (including any attachments)
is confidential and intended solely for the attention and use of the
named addressee(s). It must not be copied, distributed nor disclosed
to any person. If you are not the intended recipient, please delete
it from your system and notify sender immediately. Any disclosure,
copying or distribution thereof or any action taken or omitted to be
taken in reliance thereon is prohibited and may be unlawful.
----------------------------------------------------------------------




From ujen at nino.ru  Tue Feb 20 10:57:36 2001
From: ujen at nino.ru (ujen)
Date: 20 Feb 2001 09:57:36 GMT
Subject: Font size in IDLE
Message-ID: <01c09b23$99332380$3201a8c0@ws9201>

Hi there.

How to change font size in IDLE (Python 1.5.2 Linux version )



Thanks

Eugene Tsymbalov, Russia


From bryan at eevolved.com  Tue Feb 20 13:49:30 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Tue, 20 Feb 2001 12:49:30 GMT
Subject: Font size in IDLE
References: <01c09b23$99332380$3201a8c0@ws9201>
Message-ID: <uJtk6.284555$JT5.9794163@news20.bellglobal.com>

ujen wrote:

> Hi there.
> 
> How to change font size in IDLE (Python 1.5.2 Linux version )
> 
> 
> 
> Thanks
> 
> Eugene Tsymbalov, Russia

grep through the source code looking for the word "font". Once you find it, 
change it.
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"Reality is merely an illusion, albeit a very persistent one."-- Einstein



From glenfant at equod.com.nospam  Tue Feb 20 15:12:47 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Tue, 20 Feb 2001 15:12:47 +0100
Subject: Font size in IDLE
References: <01c09b23$99332380$3201a8c0@ws9201> <uJtk6.284555$JT5.9794163@news20.bellglobal.com>
Message-ID: <96ttlh$8rp$1@reader1.imaginet.fr>

Yes... the default window font for idle is ugly and not legible with Linux

Edit EditorWindow.py, search for this:
-----
...
        if sys.platform[:3] == 'win':
            text['font'] = ("lucida console", 8)
...
-----
And change it against this:
-----
...
#        if sys.platform[:3] == 'win': # Works with others than Windows
        text['font'] = ("lucidatypewriter", 12) # pretty with a 100 pix/inch
display
...
-----
Test your new font setting for IDLE as "root" to have this effective (users
have no write access to /usr/lib... so they use the "old" EditorWindow.pyc).

That's all!

"Bryan Mongeau" <bryan at eevolved.com> a ?crit dans le message news:
uJtk6.284555$JT5.9794163 at news20.bellglobal.com...
> ujen wrote:
>
> > Hi there.
> >
> > How to change font size in IDLE (Python 1.5.2 Linux version )
> >
> >
> >
> > Thanks
> >
> > Eugene Tsymbalov, Russia
>
> grep through the source code looking for the word "font". Once you find
it,
> change it.
> --
> <==================================>
> Bryan Mongeau
> Lead Developer, Director
> eEvolved Real-Time Technologies Inc.
> http://www.eevolved.com
> <==================================>
>
> "Reality is merely an illusion, albeit a very persistent one."-- Einstein
>



From gregj at pobox.com  Wed Feb 21 08:57:50 2001
From: gregj at pobox.com (Greg Jorgensen)
Date: Wed, 21 Feb 2001 07:57:50 GMT
Subject: Font size in IDLE
References: <01c09b23$99332380$3201a8c0@ws9201>
Message-ID: <2yKk6.452424$U46.13427774@news1.sttls1.wa.home.com>

"ujen" <ujen at nino.ru> wrote in message
news:01c09b23$99332380$3201a8c0 at ws9201...
>
> How to change font size in IDLE (Python 1.5.2 Linux version )

Look in the file config.txt in the directory IDLE is in.

--
Greg Jorgensen
PDXperts
Portland, Oregon, USA
gregj at pobox.com




From lorenzo at sancho.ccd.uniroma2.it  Tue Feb 20 11:10:15 2001
From: lorenzo at sancho.ccd.uniroma2.it (lorenzo at sancho.ccd.uniroma2.it)
Date: Tue, 20 Feb 2001 10:10:15 -0000
Subject: Drew's sslserv support into 2.1?
Message-ID: <96tfq7+jjhk@eGroups.com>

On Sep 22, 2000 Drew Csillag posted a patch adding an sslserver
object to the socket module. Since it could be (very) useful, even
though we all know about the shortcomings of such basic support
when confronted with a complete support such as that which could
come from a real SSL module, I hope the patch would not be forgot,
and get in before 2.1 beta/freeze.

A nice time to all of the readers.

Lorenzo




From bryan at eevolved.com  Tue Feb 20 11:48:54 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Tue, 20 Feb 2001 10:48:54 GMT
Subject: Drew's sslserv support into 2.1?
References: <mailman.982663930.4003.python-list@python.org>
Message-ID: <qYrk6.284085$JT5.9787899@news20.bellglobal.com>

lorenzo at sancho.ccd.uniroma2.it wrote:

> On Sep 22, 2000 Drew Csillag posted a patch adding an sslserver
> object to the socket module. Since it could be (very) useful, even
> though we all know about the shortcomings of such basic support
> when confronted with a complete support such as that which could
> come from a real SSL module, I hope the patch would not be forgot,
> and get in before 2.1 beta/freeze.
> 
> A nice time to all of the readers.
> 
> Lorenzo
> 
> 

Hi Lorenzo,

I am unfamiliar with the patch you mention. However, if it is anything like 
Brian Gallew's patch for 2.0, it is based on OpenSSL/SSLeay. Granted the 
functionality is nice, but do all users want to include such a large 
external library?  Weigh the price of the code bloat vs. the amount of 
people that will actually use it and I'm sure you'll conclude its status as 
a patch is justified. Additionally, SSL socket support is already a trivial 
matter with extensions like amkCrypto and M2Crypto.  Not to mention export 
restrictions should it be incorporated into the code.

However, I would like to take a look at it. Do you have a link?

Regards,
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"Scientific beliefs are supported by evidence, and they get results. Myths 
and faiths are not and do not" -- Richard Dawkins



From drew.csillag at starmedia.net  Tue Feb 20 18:09:25 2001
From: drew.csillag at starmedia.net (Drew Csillag)
Date: Tue, 20 Feb 2001 12:09:25 -0500
Subject: Drew's sslserv support into 2.1?
In-Reply-To: <qYrk6.284085$JT5.9787899@news20.bellglobal.com>; from bryan@eevolved.com on Tue, Feb 20, 2001 at 10:48:54AM +0000
References: <mailman.982663930.4003.python-list@python.org> <qYrk6.284085$JT5.9787899@news20.bellglobal.com>
Message-ID: <20010220120925.D911@naxos.skunk.org>

It has been *rightly* rejected by the BDFL for the following reasons:

   * M2Crypto does a *much* better job
   * If your server key is encrypted (it should be), it pops up a password
     prompt every time a connection is sslserv()'d which, in a word, sucks.

M2Crypto's Homepage is http://mars.post1.com/home/ngps/m2/.  Check it out.

Drew


On Tue, Feb 20, 2001 at 10:48:54AM +0000, Bryan Mongeau wrote:
> lorenzo at sancho.ccd.uniroma2.it wrote:
> 
> > On Sep 22, 2000 Drew Csillag posted a patch adding an sslserver
> > object to the socket module. Since it could be (very) useful, even
> > though we all know about the shortcomings of such basic support
> > when confronted with a complete support such as that which could
> > come from a real SSL module, I hope the patch would not be forgot,
> > and get in before 2.1 beta/freeze.
> > 
> > A nice time to all of the readers.
> > 
> > Lorenzo
> > 
> > 
> 
> Hi Lorenzo,
> 
> I am unfamiliar with the patch you mention. However, if it is anything like 
> Brian Gallew's patch for 2.0, it is based on OpenSSL/SSLeay. Granted the 
> functionality is nice, but do all users want to include such a large 
> external library?  Weigh the price of the code bloat vs. the amount of 
> people that will actually use it and I'm sure you'll conclude its status as 
> a patch is justified. Additionally, SSL socket support is already a trivial 
> matter with extensions like amkCrypto and M2Crypto.  Not to mention export 
> restrictions should it be incorporated into the code.
> 
> However, I would like to take a look at it. Do you have a link?
> 
> Regards,
> -- 
> <==================================>
> Bryan Mongeau
> Lead Developer, Director
> eEvolved Real-Time Technologies Inc.
> http://www.eevolved.com
> <==================================>
> 
> "Scientific beliefs are supported by evidence, and they get results. Myths 
> and faiths are not and do not" -- Richard Dawkins
> 

-- 
Drew Csillag                                              



From akuchlin at mems-exchange.org  Tue Feb 20 18:07:16 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 20 Feb 2001 12:07:16 -0500
Subject: Drew's sslserv support into 2.1?
References: <mailman.982663930.4003.python-list@python.org>
Message-ID: <3dg0h99tbv.fsf@ute.cnri.reston.va.us>

lorenzo at sancho.ccd.uniroma2.it writes:
> come from a real SSL module, I hope the patch would not be forgot,
> and get in before 2.1 beta/freeze.

This refers to patch #101647, which was marked 'Rejected'.
http://sourceforge.net/patch/?func=detailpatch&patch_id=101647&group_id=5470

Installing and using M2Crypto is a better idea than using the poor
support in the socket module.

--amk


From aleaxit at yahoo.com  Tue Feb 20 11:27:11 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 11:27:11 +0100
Subject: ANOVA and logistic regression?
Message-ID: <96tgpt0qj8@news2.newsguy.com>

Can anybody recommend C-coded Python extensions for doing
ANOVA, and logistic regression, on largish data sets, in a
reasonably speedy way?  Pure-Python solutions also welcome,
but I suspect they might be a bit slow.  I need to run this
stuff with Python 2.0 on a Win32 box (and have no problem
building extensions from sources, although I'd rather not
have to tweak/port the sources if possible -- but it won't
kill me if I do have to do that), but it would be nice if
the solution was cross-platform and redistributable (GPL,
LGPL, OpenSource, etc, preferred, but free-as-in-free-beer
is the key issue here:-).

Answers most welcome on comp.lang.python, or directly in
my mailbox (if anybody asks, I'll summarize any answers
received in mail -- I realize this marks me as a NetNews
dinosaur, but this used to be the correct protocol once:-).

Advthanksance...


Alex





From reuter at Uni-Hohenheim.DE  Tue Feb 20 14:59:42 2001
From: reuter at Uni-Hohenheim.DE (Daniel Reuter)
Date: Tue, 20 Feb 2001 14:59:42 +0100 (MET)
Subject: ANOVA and logistic regression?
In-Reply-To: <96tgpt0qj8@news2.newsguy.com>
Message-ID: <Pine.SOL.3.94.1010220145423.7897A-100000@ultra5>

Hello there,

I have a stats module, which can do at least ANOVA, found it at the
Vaults of Parnassus (check the mathematics section), but it's plain
python, and I don't know how well it behaves on large datasets. You might
try it out.

Regards,
Daniel

On Tue, 20 Feb 2001, Alex Martelli wrote:

> Can anybody recommend C-coded Python extensions for doing
> ANOVA, and logistic regression, on largish data sets, in a
> reasonably speedy way?  Pure-Python solutions also welcome,
> but I suspect they might be a bit slow.  I need to run this
> stuff with Python 2.0 on a Win32 box (and have no problem
> building extensions from sources, although I'd rather not
> have to tweak/port the sources if possible -- but it won't
> kill me if I do have to do that), but it would be nice if
> the solution was cross-platform and redistributable (GPL,
> LGPL, OpenSource, etc, preferred, but free-as-in-free-beer
> is the key issue here:-).
>  
> Advthanksance...
> 
> 
> Alex





From janne at oops.nnets.fi  Tue Feb 20 23:56:43 2001
From: janne at oops.nnets.fi (Janne Sinkkonen)
Date: 21 Feb 2001 00:56:43 +0200
Subject: ANOVA and logistic regression?
References: <96tgpt0qj8@news2.newsguy.com>
Message-ID: <m3vgq5otec.fsf@kinos.nnets.fi>

"Alex Martelli" <aleaxit at yahoo.com> writes:

> Can anybody recommend C-coded Python extensions for doing
> ANOVA, and logistic regression, on largish data sets, in a
> reasonably speedy way?

I have a "fast enough" Numeric Python implementation of basic (and
regularized, i.e. a gaussian prior for the parameters) log-linear
regression.  Only one binary depedent variable.

-- 
Janne


From thomas at cintra.no  Tue Feb 20 12:08:59 2001
From: thomas at cintra.no (Thomas Weholt)
Date: Tue, 20 Feb 2001 12:08:59 +0100
Subject: Any experience with 4Suite Server??
Message-ID: <gfsk6.19$t21.165@news3.oke.nextra.no>

Hi,

I need to store alot of xml-documents in some sort of structure that
implements automatic indexing of the contents of the documents, fast
searches and some way of producing reports, preferrably using XSLT to
transform the stored XML. Looked at 4Suite Server but I don't get a fully
understanding of how it works, or if it could fit my needs.

I'm looking for views on this package ( 4Suite Server ), even if they don't
relate to my specific problem.

Thomas




From rich_somerfield at tertio.com  Tue Feb 20 12:15:59 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Tue, 20 Feb 2001 11:15:59 -0000
Subject: Problems with Tkinter / Python - displaying JPEG.
Message-ID: <vlsk6.1473$MN.14971@news2-win.server.ntlworld.com>

Hi,

I had Python 1.5 and Pil 0.3 as part of my Linux (SuSE6.4) installation, and
was trying to execute a Python Script that loaded and displayed images.  [I
use the Python1.5 distribution from PythonWare under windows, but wanted to
do development under Linux instead.]  However I encountered a problem when I
was trying to display a JPEG file (i have indicated the relavent line).

thebackgroundimage = thebackgroundimage.resize((canvaswidth, canvasheight))
******* thebackgroundimage = ImageTk.PhotoImage(thebackgroundimage) ********
thebackground = arena.create_image(0,0,image=thebackgroundimage)
middlex = canvaswidth
middley = canvasheight
locationx = middlex / 2
locationy = middley / 2
arena.move(thebackground, locationx, locationy)

The interpreter indicated that the parameter "PyPhotoImage" was not set and
a call was returning the wrong type (string / int).

I thought this might be a problem with an earlier version of PIL / Python /
Tkinter that didnt support JPEG files. So I downloaded Python2.0 and PIL
1.1.1.  Compiled them (after messing with the Makefiles), checked that I had
got the JPEG libraries (libjpeg).

However I get exactly the same error message.  Any ideas where the problem
is coming from?  Pil / Tkinter / Tk?

Regards
Rich




From mrq at for.mat.bham.ac.uk  Tue Feb 20 12:36:01 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Tue, 20 Feb 2001 11:36:01 +0000
Subject: Calling a sequence of commands with a sequence of Tkinter buttons
Message-ID: <Pine.GSO.4.21.0102201050570.2787-100000@babbage>

I'm trying to create a number of buttons each of which has a similar sort
of command attached to it.  Obviously the following doesn't work since I'm
attaching a called function to the button rather than the function to
call.  Can anyone tell me the correct way to do it?  (Equally obviously, I
could create a different function for each button, but this seems lengthy
for a large number of buttons).

Thanks
Martyn

MY ATTEMPT:

from Tkinter import *

def f(i):	# Typically I'll have a more complicated function
	print i	# but this is just for illustration

root = Tk()

buttons = []	# initialize list to contain the button widgets

for i in range(20):
	buttons.append(Button(root,text="Button "+`i`,command=f(i)))
	buttons[i].pack()
	# This doesn't work - it just prints 1,2,..,20 down the screen
	# immediately, since f(i) is called at the creation of the button.
	

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick



From stephen_purcell at yahoo.com  Tue Feb 20 14:10:59 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 20 Feb 2001 14:10:59 +0100
Subject: Calling a sequence of commands with a sequence of Tkinter buttons
In-Reply-To: <Pine.GSO.4.21.0102201050570.2787-100000@babbage>; from mrq@for.mat.bham.ac.uk on Tue, Feb 20, 2001 at 11:36:01AM +0000
References: <Pine.GSO.4.21.0102201050570.2787-100000@babbage>
Message-ID: <20010220141058.A16164@freedom.puma-ag.com>

Martyn Quick wrote:
> MY ATTEMPT:
> 
> from Tkinter import *
> 
> ...
> 
> for i in range(20):
> 	buttons.append(Button(root,text="Button "+`i`,command=f(i)))
> 	buttons[i].pack()
> 	# This doesn't work - it just prints 1,2,..,20 down the screen
> 	# immediately, since f(i) is called at the creation of the button.

Right, you found the problem; f() is called immediately. Instead, give
Tkinter a new function that takes no parameters and that can be called later:

	for i in range(20):
	  def f_i(j=i):
	    f(j)
	  buttons.append(Button(root,text="Button "+`i`,command=f_i))
	  buttons[i].pack()


Of, if you prefer:

	for i in range(20):
	  buttons.append(Button(root,text="Button "+`i`,
				command=lambda j=i, f(j)))
	  buttons[i].pack()


It would perhaps be neater to solve this with a callable class instance than
with a default function argument:

	class PrintNumberAction:
	   def __init__(self, i):
	      self.i = i
	   def __call__(self):
	      print i  # or whatever

then:

	for i in range(20):
	   action = PrintNumberAction(i)
	   buttons.append(Button(root,text="Button "+`i`,command=action))
	   buttons[i].pack()

Hope that helps.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From rosemann at imkt.uni-hannover.de  Wed Feb 21 09:39:40 2001
From: rosemann at imkt.uni-hannover.de (Harald Rosemann)
Date: Wed, 21 Feb 2001 09:39:40 +0100
Subject: Calling a sequence of commands with a sequence of Tkinter buttons
References: <Pine.GSO.4.21.0102201050570.2787-100000@babbage>
Message-ID: <3A937ECC.E2EF40E7@imkt.uni-hannover.de>

Martyn Quick wrote:
> 
> I'm trying to create a number of buttons each of which has a similar sort
> of command attached to it.  Obviously the following doesn't work since I'm
> attaching a called function to the button rather than the function to
> call.  Can anyone tell me the correct way to do it?  (Equally obviously, I
> could create a different function for each button, but this seems lengthy
> for a large number of buttons).
> 
> Thanks
> Martyn
> 
> MY ATTEMPT:
> 
> from Tkinter import *
> 
> def f(i):       # Typically I'll have a more complicated function
>         print i # but this is just for illustration
> 
> root = Tk()
> 
> buttons = []    # initialize list to contain the button widgets
> 
> for i in range(20):
>         buttons.append(Button(root,text="Button "+`i`,command=f(i)))
>         buttons[i].pack()
>         # This doesn't work - it just prints 1,2,..,20 down the screen
>         # immediately, since f(i) is called at the creation of the button.
> 
> 
> --------------------------------------------------------
> Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
> University of Birmingham, Edgbaston, Birmingham, UK.
> http://www.mat.bham.ac.uk/M.R.Quick





You may use a "lambda expression". 
Something like the following will work.



def show_ind(i, event=None):
    print i
...

my_tst[i] = Tkinter.Frame(zzz, ....)
my_tst[i].pack()
....

my_but[i] = Tkinter.Button(my_tst[i], text = ".....", state = status)
my_but[i].pack(side = Tkinter.LEFT)
my_but[i].bind("<Button-1>", (lambda event, ind=i: show_ind(ind,
event)))


With best regards,
Harald Rosemann

-- 
Prof.Dr.-Ing. Harald Rosemann
am Institut fuer Maschinenelemente, 
   Konstruktionstechnik und Tribologie
Universitaet Hannover


From Achim.Gaedke at zpr.uni-koeln.de  Tue Feb 20 13:25:28 2001
From: Achim.Gaedke at zpr.uni-koeln.de (Achim Gaedke)
Date: Tue, 20 Feb 2001 13:25:28 +0100
Subject: Problem with Tkinter.PhotoImage
Message-ID: <3A926237.C9BD74FE@zpr.uni-koeln.de>

Hello everybody!

I've tested some features of the Tkinter.Canvas and found an odd
behaviour.

Here is the example that works:

import Tkinter

r=Tkinter.Tk()
c=Tkinter.Canvas(r)
p=Tkinter.PhotoImage(file='linuxorg.gif')
i=c.create_image(0,0,anchor=Tkinter.NW,image=p)
c.pack()

r.mainloop()

And now a simple example that does not show the image and does not
generate runtime errors.

import Tkinter

r=Tkinter.Tk()
c=Tkinter.Canvas(r)
i=c.create_image(0,0,anchor=Tkinter.NW,
                 image=Tkinter.PhotoImage(file='linuxorg.gif'))
c.pack()

r.mainloop()

This example is the simplest. Here is a more elaborate case (my original
problem):

import Tkinter

class picture:

    def __init__(self,canvas):
        self.p=Tkinter.PhotoImage(file='linuxorg.gif')
        self.i=c.create_image(0,0,anchor=Tkinter.NW,
                              image=self.p)


r=Tkinter.Tk()
c=Tkinter.Canvas(r)
picture(c)
c.pack()

r.mainloop()

This doesn't work. Just change the line

picture(c)

to

p=picture(c).

this reveals the reason of the strange behaviour. There is no reference
left on Tkinter.PhotoImage, so it is destructed.
But I do not think, that it should be, just because Phython is an "easy
going" language with reference counters and should keep one reference,
although this is held by the canvas-item image.

Does someone have the same problem with this behaviour?

Yours

Achim




From javalist at 21cn.com  Tue Feb 20 14:05:11 2001
From: javalist at 21cn.com (javalist)
Date: Tue, 20 Feb 2001 21:05:11 +0800
Subject: How did python handle forward declaration?
Message-ID: <1251166452389.20010220210511@21cn.com>

Hello python-list,

        in c/c++ we can use class xxx or struct xxx to forward declared what we really implement
        later,how python work this out?

Best regards,
 javalist                          mailto:javalist at 21cn.com





From stephen_purcell at yahoo.com  Tue Feb 20 14:16:34 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 20 Feb 2001 14:16:34 +0100
Subject: How did python handle forward declaration?
In-Reply-To: <1251166452389.20010220210511@21cn.com>; from javalist@21cn.com on Tue, Feb 20, 2001 at 09:05:11PM +0800
References: <1251166452389.20010220210511@21cn.com>
Message-ID: <20010220141634.A16302@freedom.puma-ag.com>

javalist wrote:
> 
>         in c/c++ we can use class xxx or struct xxx to forward declared what we really implement
>         later,how python work this out?
> 

Answer 1: There is no way to do this in Python

Answer 2: Write good documentation for your Python code; it has the same effect


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From ben.hutchings at roundpoint.com  Tue Feb 20 23:16:15 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 20 Feb 2001 14:16:15 -0800
Subject: How did python handle forward declaration?
References: <mailman.982674622.10844.python-list@python.org>
Message-ID: <uk86l2e6o.fsf@roundpoint.com>

javalist <javalist at 21cn.com> writes:

> Hello python-list,
> 
>         in c/c++ we can use class xxx or struct xxx to forward
>         declared what we really implement later,how python work this
>         out?

C and C++ compilers try to resolve all identifiers and type names at
the point they are used, at compile time.  This is a problem if you
need to define two classes/structures that refer to each other.  For
some purposes they can accept a declaration that promises there will
be a definition somewhere else (later, or in another translation
unit), and that's how you deal with that need.

Python is a much more dynamic language, which postpones almost all
name resolution to run-time.  The external things (classes, modules,
etc) that a piece of code relies on just need to be defined at the
time it gets to run.  When you define a class, none of the code in its
member functions runs, so you can define two classes/structures that
refer to each other without any need for forward-declarations.

You might have problems in Python if you want to define mutually-
dependent modules, but then a mutually-dependent set of modules
is effectively just one module and should be defined as such.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From costas at springmail.com  Tue Feb 20 14:09:28 2001
From: costas at springmail.com (Costas Menico)
Date: Tue, 20 Feb 2001 13:09:28 GMT
Subject: Tkinter does not work with PythonWin?
Message-ID: <3a926b7b.2462791@News.CIS.DFN.DE>

The documentation says that Tkinter does not work with PythonWin (a
great IDE for Python). Does anyone have any info about an alternative
I/F?


Thanks

costas


From chrisa at ASPATECH.COM.BR  Tue Feb 20 14:18:12 2001
From: chrisa at ASPATECH.COM.BR (Chris Richard Adams)
Date: Tue, 20 Feb 2001 10:18:12 -0300
Subject: Where is CASE?????
Message-ID: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR>

Someone please tell me why I cannot find a CASE statement in the Python
documentation index...

There is a CASE isn't there???



From stephen_purcell at yahoo.com  Tue Feb 20 14:48:36 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 20 Feb 2001 14:48:36 +0100
Subject: Where is CASE?????
In-Reply-To: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR>; from chrisa@ASPATECH.COM.BR on Tue, Feb 20, 2001 at 10:18:12AM -0300
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR>
Message-ID: <20010220144836.A16568@freedom.puma-ag.com>

Chris Richard Adams wrote:
> Someone please tell me why I cannot find a CASE statement in the Python
> documentation index...

Because there isn't one.

Use if/elif/else instead.

It doesn't sting as much as you might expect.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From chrisw at nipltd.com  Tue Feb 20 15:18:03 2001
From: chrisw at nipltd.com (Chris Withers)
Date: Tue, 20 Feb 2001 14:18:03 +0000
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com>
Message-ID: <3A927C9B.29C38751@nipltd.com>

Steve Purcell wrote:
> 
> Chris Richard Adams wrote:
> > Someone please tell me why I cannot find a CASE statement in the Python
> > documentation index...
> 
> Because there isn't one.
> 
> Use if/elif/else instead.
> 
> It doesn't sting as much as you might expect.

Oh but it does ;-)

cheers,

Chris



From dsh8290 at rit.edu  Tue Feb 20 16:05:49 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 20 Feb 2001 10:05:49 -0500
Subject: Where is CASE?????
In-Reply-To: <3A927C9B.29C38751@nipltd.com>; from chrisw@nipltd.com on Tue, Feb 20, 2001 at 02:18:03PM +0000
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com> <3A927C9B.29C38751@nipltd.com>
Message-ID: <20010220100549.C18013@harmony.cs.rit.edu>

On Tue, Feb 20, 2001 at 02:18:03PM +0000, Chris Withers wrote:
| Steve Purcell wrote:
| > 
| > Chris Richard Adams wrote:
| > > Someone please tell me why I cannot find a CASE statement in the Python
| > > documentation index...
| > 
| > Because there isn't one.
| > 
| > Use if/elif/else instead.
| > 
| > It doesn't sting as much as you might expect.
| 
| Oh but it does ;-)

Why does it?  CASE is little more than a special if-else ladder for
integral types only.  If you are simply trying to get a value, use a
dictionary lookup instead.

-D




From phlip_cpp at my-deja.com  Tue Feb 20 16:16:08 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 20 Feb 2001 15:16:08 GMT
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com> <3A927C9B.29C38751@nipltd.com> <mailman.982681590.5974.python-list@python.org>
Message-ID: <96u1no$dii@dispatch.concentric.net>

Proclaimed D-Man from the mountaintops:

> Why does it?  CASE is little more than a special if-else ladder for
> integral types only.  If you are simply trying to get a value, use a
> dictionary lookup instead.

If you are simply trying to run a jump table, use a dictionary where the 
stored values are functions. Modules written after the putative 'switch' 
statement could enrole their functions in it without changing its code.

Mr. Manners reminds the Original Gentle Poster that many things 
traditionally coded as 'switch' statements (typically those that repeat 
often in code) ought to go thru the refactor Replace Conditional with 
Polymorphism, per the book /Refactoring/.

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  It's a small Web, after all...  --


From max at alcyone.com  Tue Feb 20 18:41:17 2001
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 20 Feb 2001 09:41:17 -0800
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com> <mailman.982678705.26827.python-list@python.org>
Message-ID: <3A92AC3D.B91FDCB1@alcyone.com>

Chris Withers wrote:

> Steve Purcell wrote:
>
> > Chris Richard Adams wrote:
> >
> > > Someone please tell me why I cannot find a CASE statement in the
> > > Python
> > > documentation index...
> >
> > Because there isn't one.
> >
> > Use if/elif/else instead.
> >
> > It doesn't sting as much as you might expect.
> 
> Oh but it does ;-)

Not particularly.  I come from C and C++, which have switch...case
statements, and it isn't a big deal for them to be missing.  Case
statements are realy nothing more than glorified repeated if...else
if... statements, and really don't have any more power (in fact, in C
and C++, case statements have less).

If you're really doing something where there would potentially be many,
many if...else if... clauses, then a better approach in Python would be
to use a dictionary whose keys are the case statements and whose values
are the data to retrieve or the function to call.  This is syntactically
a lot cleaner and allows for in-code modification of the transition
table.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Some mistakes we must carry with us.
\__/ Speaker-to-Animals
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.


From larry at smith-house.org  Tue Feb 27 19:39:50 2001
From: larry at smith-house.org (Larry Smith)
Date: Tue, 27 Feb 2001 13:39:50 -0500
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com> <mailman.982678705.26827.python-list@python.org> <3A92AC3D.B91FDCB1@alcyone.com>
Message-ID: <3A9BF476.EF2DE06@smith-house.org>

Erik Max Francis wrote:

> Not particularly.  I come from C and C++, which have switch...case
> statements, and it isn't a big deal for them to be missing.  Case
> statements are realy nothing more than glorified repeated if...else
> if... statements, and really don't have any more power (in fact, in C
> and C++, case statements have less).

Case statements can be modelled with if-else chains
but they are not the same internally.  Case statements
are usually compiled as jump tables, rather than as a
series of if-elses.  This is hugely faster than if-else
on average.

However, in Guido's defense, I will also point out the
primary use of case statements is to select alternatives
based on a type flag - which is not needed in an object
oriented language.

-- 
 .-.    .-. .---. .---. .-..-. | Do not mistake  my cynicism
 | |__ / | \| |-< | |-<  >  /  | for dispair for deep inside
 `----'`-^-'`-'`-'`-'`-' `-'   | me is an  optimist  forever
  http://www.smith-house.org/  | asking, "Why not?"


From Norman_Shelley-RRDN60 at email.sps.mot.com  Tue Feb 27 19:55:14 2001
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Tue, 27 Feb 2001 11:55:14 -0700
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com> <mailman.982678705.26827.python-list@python.org> <3A92AC3D.B91FDCB1@alcyone.com> <3A9BF476.EF2DE06@smith-house.org>
Message-ID: <3A9BF812.D5DA7F91@email.sps.mot.com>

This argument is similiar to Bertrand Meyer's argument for not having a
CASE statement in early versions of Eiffel.  He finally "saw the light"
and added in the CASE statement.  I believe he finally saw the need to
efficiently be able to handle character and string data (real world data
that has not yet been objectized).



Larry Smith wrote:

> ...
>
> However, in Guido's defense, I will also point out the
> primary use of case statements is to select alternatives
> based on a type flag - which is not needed in an object
> oriented language.
>
> --
>  .-.    .-. .---. .---. .-..-. | Do not mistake  my cynicism
>  | |__ / | \| |-< | |-<  >  /  | for dispair for deep inside
>  `----'`-^-'`-'`-'`-'`-' `-'   | me is an  optimist  forever
>   http://www.smith-house.org/  | asking, "Why not?"



From joconnor at cybermesa.com  Tue Feb 27 19:59:49 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 27 Feb 2001 11:59:49 -0700
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com> <mailman.982678705.26827.python-list@python.org> <3A92AC3D.B91FDCB1@alcyone.com> <3A9BF476.EF2DE06@smith-house.org>
Message-ID: <3A9BF925.600AFF65@cybermesa.com>


Larry Smith wrote:

>
> However, in Guido's defense, I will also point out the
> primary use of case statements is to select alternatives
> based on a type flag - which is not needed in an object
> oriented language.

LOL!  Most of my background is in Smalltalk, which also does not have CASE
statements and I'm so used to programming without them that, up until I
read this post, it really didn't occur to me whether Python had them or not

Take care,
Jay O'Connor
http://www.cybermesa.com/~joconnor

Python Language Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython



From db3l at fitlinxx.com  Wed Feb 28 02:33:50 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 20:33:50 -0500
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <20010220144836.A16568@freedom.puma-ag.com> <mailman.982678705.26827.python-list@python.org> <3A92AC3D.B91FDCB1@alcyone.com> <3A9BF476.EF2DE06@smith-house.org>
Message-ID: <ur90jmw01.fsf@ctwd0143.fitlinxx.com>

Larry Smith <larry at smith-house.org> writes:

> Case statements can be modelled with if-else chains
> but they are not the same internally.  Case statements
> are usually compiled as jump tables, rather than as a
> series of if-elses.  This is hugely faster than if-else
> on average.

It's also what often restricts the case statements to integral values
within some range, and the compilation can be less efficient if your
values are sparsely spread over a very large range (in some cases the
compiler may even end up with the same code as an if-else chain).

But in such a case (lookups), dictionaries in Python are very
efficient for lookups and can offer similar performance advantages,
which still being generic in terms of what is being looked up.

> However, in Guido's defense, I will also point out the
> primary use of case statements is to select alternatives
> based on a type flag - which is not needed in an object
> oriented language.

Well, either that, or keeping the list of alternatives in the
dictionary keyed by the type flag is a nice fit.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From emile at fenx.com  Tue Feb 20 15:20:17 2001
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 20 Feb 2001 06:20:17 -0800
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <mailman.982676904.20104.python-list@python.org>
Message-ID: <96tuh8$mocjg$1@ID-11957.news.dfncis.de>

Or you may consider wrapping it in a dictionary:

Type "copyright", "credits" or "license" for more
information.
IDLE 0.6 -- press F1 for help
>>> def caseA(*args):
 print "CaseA ", args


>>> def caseB(*args):
 print "CaseB ", args


>>> def caseOther(*args):
 print "CaseOther ", args


>>> switch = {'A':caseA, 'B':caseB}
>>> switch.get('A',caseOther)
<function caseA at 0103153C>
>>> switch.get('A',caseOther)(1,2,3)
CaseA  (1, 2, 3)
>>> switch.get('B',caseOther)(1,2,3)
CaseB  (1, 2, 3)
>>> switch.get('C',caseOther)(1,2,3)
CaseOther  (1, 2, 3)
>>>

--

Emile van Sebille
emile at fenx.com
-------------------


"Steve Purcell" <stephen_purcell at yahoo.com> wrote in message
news:mailman.982676904.20104.python-list at python.org...
> Chris Richard Adams wrote:
> > Someone please tell me why I cannot find a CASE
statement in the Python
> > documentation index...
>
> Because there isn't one.
>
> Use if/elif/else instead.
>
> It doesn't sting as much as you might expect.
>
> -Steve
>
> --
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Get servlets at http://pyserv.sourceforge.net/
> "Even snakes are afraid of snakes." -- Steven Wright
>




From hgg9140 at cola.ca.boeing.com  Tue Feb 20 15:56:37 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Tue, 20 Feb 2001 14:56:37 GMT
Subject: Where is CASE?????
References: <mailman.982674921.12321.python-list@python.org>
Message-ID: <xqxd7cd4d3u.fsf@cola.ca.boeing.com>

No, but do you really need it?  First, if you do a performance
profiling, you will find if...elif...elif... is not so bad.  But it
isn't pretty either.  So consider using a map:

color_function={'red':do_red,'green':do_green,'blue':do_blue}
#get some case value in domain red,green,blue, then:
    ftn=color_function[value]
    ftn()

Or even:
    color_function[value]()

Further, this approach lends itself to dynamically defining and adding
cases, which can't be done with statically compiled CASE statements.
Thus registering new colors and their actions is a piece of cake.

"Chris Richard Adams" <chrisa at ASPATECH.COM.BR> writes:

> Someone please tell me why I cannot find a CASE statement in the Python
> documentation index...
> 
> There is a CASE isn't there???
> 

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From duncan at rcp.co.uk  Tue Feb 20 16:00:48 2001
From: duncan at rcp.co.uk (Duncan Booth)
Date: Tue, 20 Feb 2001 15:00:48 +0000 (UTC)
Subject: Where is CASE?????
References: <mailman.982674921.12321.python-list@python.org>
Message-ID: <Xns904E985DEFF7Bduncanrcpcouk@127.0.0.1>

"Chris Richard Adams" <chrisa at ASPATECH.COM.BR> wrote in 
<mailman.982674921.12321.python-list at python.org>:

>Someone please tell me why I cannot find a CASE statement in the Python
>documentation index...
>
>There is a CASE isn't there???
>

While there are undoubtedly occasions where a CASE statement would be 
useful, they are actually much rarer in Python than you might expect. A 
large proportion of the instances where you might use CASE can be written 
better as a dictionary or an attribute lookup.

For example in most languages if you were to try to handle errors from 
opening a web URL, you might reasonably expect to find a case statement 
switching on the different error codes you could receive. If you look at the 
urllib module you will see that the Pythonic way to do this is to have class 
methods with names such as http_error_401, http_error_301, and 
http_error_default. When an error code is returned, the appropriate method 
is called (or the default method if there is no better match) by using 
getattr to look up the method by name. This technique allows the end user to 
add their own handler for any error number, even those the original class 
knows nothing about.

Another common use for CASE in some languages is when you have created a 
structure and put a type field in it. You want to process the object 
according to the type field. In any object oriented language (not just 
Python), this situation is better handled by including a method in the 
object that does the appropriate action for the object's type.

If you post a message outlining the situation where you think you need a 
CASE statement you might get suggestions of better (or just different) ways 
to solve your particular problem.


From Lance.Ellinghaus at eConnections.com  Tue Feb 27 20:26:04 2001
From: Lance.Ellinghaus at eConnections.com (Lance.Ellinghaus at eConnections.com)
Date: Tue, 27 Feb 2001 13:26:04 -0600
Subject: Where is CASE?????
Message-ID: <OF62DEBF3D.D34C6DCE-ON86256A00.006A3A26@econnections.com>

This can be done using a dictionary and functions pretty easily. At least
that is how I have always done it. For example...

def func_a():
     pass

def func_b():
     pass

def func_c():
     pass

func_dict = {
     'A': func_a,
     'B': func_b,
     'C': func_c
     }

try:
     apply(func_dict[test_value], ())
except:
     print '%c is not a valid choice' % test_value


This works for any type of value in 'test_value' and is pretty fast
executing.


--
Lance Ellinghaus
Phone: 940-466-7251
Pager: 800-226-4526 or lellinghaus at bellsouthips.com
Sr. Architect - eConnections, Inc.



                                                                                                                                      
                    Norman Shelley                                                                                                    
                    <Norman_Shelley-RRDN60 at email.sp        To:     python-list at python.org                                             
                    s.mot.com>                             cc:                                                                        
                    Sent by:                               Subject:     Re: Where is CASE?????                                        
                    python-list-admin at python.org                                                                                      
                                                                                                                                      
                                                                                                                                      
                    02/27/01 12:55 PM                                                                                                 
                                                                                                                                      
                                                                                                                                      




This argument is similiar to Bertrand Meyer's argument for not having a
CASE statement in early versions of Eiffel.  He finally "saw the light"
and added in the CASE statement.  I believe he finally saw the need to
efficiently be able to handle character and string data (real world data
that has not yet been objectized).



Larry Smith wrote:

> ...
>
> However, in Guido's defense, I will also point out the
> primary use of case statements is to select alternatives
> based on a type flag - which is not needed in an object
> oriented language.
>
> --
>  .-.    .-. .---. .---. .-..-. | Do not mistake  my cynicism
>  | |__ / | \| |-< | |-<  >  /  | for dispair for deep inside
>  `----'`-^-'`-'`-'`-'`-' `-'   | me is an  optimist  forever
>   http://www.smith-house.org/  | asking, "Why not?"

--
http://mail.python.org/mailman/listinfo/python-list







From ohyvarin at cc.helsinki.fi  Tue Feb 20 15:25:26 2001
From: ohyvarin at cc.helsinki.fi (Otto T Hyvarinen)
Date: 20 Feb 2001 14:25:26 GMT
Subject: Python training in Finland?
Message-ID: <slrn994vil.vjq.ohyvarin@sirppi.helsinki.fi>

We starting to use more Python in our department.  So we are looking for
someone to teach us something :-)

I am aware that Secret Labs/Pythonware is relatively near and they are
willing to come and train us in English, but some of us would like to be
teached in Finnish.  (So why am I asking this in an English-language
group? Outoa??)

You may reply to this post or use directly my work email:
otto.hyvarinen at fmi.fi

Thank you,
Otto Hyv?rinen



From janne at oops.nnets.fi  Tue Feb 20 23:58:11 2001
From: janne at oops.nnets.fi (Janne Sinkkonen)
Date: 21 Feb 2001 00:58:11 +0200
Subject: Python training in Finland?
References: <slrn994vil.vjq.ohyvarin@sirppi.helsinki.fi>
Message-ID: <m3r90totbw.fsf@kinos.nnets.fi>

ohyvarin at cc.helsinki.fi (Otto T Hyvarinen) writes:

> I am aware that Secret Labs/Pythonware is relatively near and they
> are willing to come and train us in English, but some of us would
> like to be teached in Finnish.  (So why am I asking this in an
> English-language group? Outoa??)

Varsin outoa. What kind of teaching do you want?

-- 
Janne


From ohyvarin at cc.helsinki.fi  Wed Feb 21 12:22:17 2001
From: ohyvarin at cc.helsinki.fi (Otto T Hyvarinen)
Date: 21 Feb 2001 11:22:17 GMT
Subject: Python training in Finland?
References: <slrn994vil.vjq.ohyvarin@sirppi.helsinki.fi> <m3r90totbw.fsf@kinos.nnets.fi>
Message-ID: <slrn997979.dck.ohyvarin@sirppi.helsinki.fi>

In article <m3r90totbw.fsf at kinos.nnets.fi>, Janne Sinkkonen wrote:
>ohyvarin at cc.helsinki.fi (Otto T Hyvarinen) writes:
>
>> I am aware that Secret Labs/Pythonware is relatively near and they
>> are willing to come and train us in English, but some of us would
>> like to be teached in Finnish.  (So why am I asking this in an
>> English-language group? Outoa??)
>
>Varsin outoa. What kind of teaching do you want?
>

Some kind of quick introduction to Python and its standard libraries. 
Then something like http://www.python9.org/p9-tutorialsten.html , but
maybe not so much about parallel programming.

If you or your company can deliver that, then maybe should switch to
email?

Otto Hyv?rinen




From kmo at stekt60.oulu.fi  Wed Feb 28 10:57:03 2001
From: kmo at stekt60.oulu.fi (Kimmo Kansanen)
Date: 28 Feb 2001 09:57:03 GMT
Subject: Python training in Finland?
References: <slrn994vil.vjq.ohyvarin@sirppi.helsinki.fi> <m3r90totbw.fsf@kinos.nnets.fi> <slrn997979.dck.ohyvarin@sirppi.helsinki.fi>
Message-ID: <97ii1f$htc$2@ousrvr3.oulu.fi>

Otto T Hyvarinen <ohyvarin at cc.helsinki.fi> wrote:
: In article <m3r90totbw.fsf at kinos.nnets.fi>, Janne Sinkkonen wrote:
:>ohyvarin at cc.helsinki.fi (Otto T Hyvarinen) writes:
:>
:>> I am aware that Secret Labs/Pythonware is relatively near and they
:>> are willing to come and train us in English, but some of us would
:>> like to be teached in Finnish.  (So why am I asking this in an
:>> English-language group? Outoa??)
:>
:>Varsin outoa. What kind of teaching do you want?
:>

: Some kind of quick introduction to Python and its standard libraries. 
: Then something like http://www.python9.org/p9-tutorialsten.html , but
: maybe not so much about parallel programming.

You maybe have considered this allready, but do you think it is wise
to try and avoid english in learning python, when all available documentation
for the language (which you will be using later on in your work) is in english?

In other words: do you really think it is possible to avoid english in this 
case, or are you just trying to make the learning curve less steep?

Lykky? pyttyyn.

Kimmo
-- 
# Smile. It makes people wonder what you're thinking about. #


From jkraska1 at san.rr.com  Sat Feb 24 06:25:20 2001
From: jkraska1 at san.rr.com (Courageous)
Date: Sat, 24 Feb 2001 05:25:20 GMT
Subject: Python training in Finland?
References: <slrn994vil.vjq.ohyvarin@sirppi.helsinki.fi>
Message-ID: <olhe9tk22q9u1b5l7srrg2tcqumje1kb7r@4ax.com>

On 20 Feb 2001 14:25:26 GMT, ohyvarin at cc.helsinki.fi (Otto T Hyvarinen) wrote:

>We starting to use more Python in our department.  So we are looking for
>someone to teach us something :-)

I suggest that you use the inability of a software person to
learn Python without training as a good indicator that you
need to let them go. Anyone who can't pick up Python on
their own is beyond redemption.

`;--)

C//



From xyzmats at laplaza.org  Tue Feb 27 18:17:04 2001
From: xyzmats at laplaza.org (Mats Wichmann)
Date: Tue, 27 Feb 2001 17:17:04 GMT
Subject: Python training in Finland?
References: <slrn994vil.vjq.ohyvarin@sirppi.helsinki.fi> <olhe9tk22q9u1b5l7srrg2tcqumje1kb7r@4ax.com>
Message-ID: <3a9bdf95.8760106@news.laplaza.org>

On Sat, 24 Feb 2001 05:25:20 GMT, Courageous <jkraska1 at san.rr.com>
wrote:

>On 20 Feb 2001 14:25:26 GMT, ohyvarin at cc.helsinki.fi (Otto T Hyvarinen) wrote:
>
>>We starting to use more Python in our department.  So we are looking for
>>someone to teach us something :-)
>
>I suggest that you use the inability of a software person to
>learn Python without training as a good indicator that you
>need to let them go. Anyone who can't pick up Python on
>their own is beyond redemption.

That's a pretty harsh statement.  Yes, Python is "easy to learn".  I'm
not convinced that it's easy for /everyone/ to become quickly
productive in Python, however.  I hear from folks who have a project
to start, and they don't want to wait while the participants learn
Python at various different rates to varying degrees of effectiveness.
There's room for training programs, although they're clearly not for
everyone - this newsgroup being a testament to that!

(Otto: sorry I can't offer to help, my parents reserved Finnish as the
"private" language in the house, I grew up speaking Swedish instead)


From tgw at email.msn.com  Tue Feb 20 15:31:18 2001
From: tgw at email.msn.com (Thom Wysong)
Date: Tue, 20 Feb 2001 09:31:18 -0500
Subject: newbie - winsock error 10054 - connection reset by peer
Message-ID: <eHhinn0mAHA.268@cpmsnbbsa07>

I'm attempting to run a simple "Fetch Data via HTTP" example on p.283 of the
book "Python Programming on Win32". It is ending in a "10054, winsock
error". I don't know what's causing this or how to fix it. I'm hoping
someone can help.

I'm running Win98 on a Gateway laptop, connected to the Internet with MSN.

First, I typed in the following, which went fine...
-------------------------
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>>
>>> import httplib
>>> http=httplib.HTTP('www.python.org')
>>>
>>> http.putrequest('GET', '/index.html')
>>> http.putrequest('Accept', 'text/html')
>>> http.putrequest('Accept', 'text/plain')
>>> http.endheaders()
>>>
-------------------------

Next, the screen is supposed to look like this....
-------------------------
>>> errcode, errmsg, headers = http.getreply()
>>> print errcode, errmsg, headers
200  OK  <mimetools.Message instance at 1073680>
>>>
-------------------------

However, what actually happens is this...
-------------------------
>>> errcode, errmsg, headers = http.getreply()
Traceback (innermost last):
  File "<pyshell#9>", line 1, in ?
    errcode, errmsg, headers = http.getreply()
  File "C:\Program Files\Python\Lib\httplib.py", line 121, in getreply
    line = self.file.readline()
  File "C:\Program Files\Python\Lib\plat-win\socket.py", line 117, in
readline
    new = self._sock.recv(self._rbufsize)
error: (10054, 'winsock error')
>>>
-------------------------

The winsock error 10054 shows up for no apparent reason. The page at
http://support.microsoft.com/support/kb/articles/Q150/5/37.asp states that
winsock error 10054 means "Connection reset by peer".

Does anyone have any idea what causes this or how to fix it?

Thanks,
tgw






From fredrik at pythonware.com  Wed Feb 21 10:30:25 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 09:30:25 GMT
Subject: newbie - winsock error 10054 - connection reset by peer
References: <eHhinn0mAHA.268@cpmsnbbsa07>
Message-ID: <RULk6.14734$Qb7.2359810@newsb.telia.net>

Thom Wysong wrote:
> The winsock error 10054 shows up for no apparent reason.

where "no apparent reason" means that either python.org
or someone on the way to python.org doesn't want to talk
to your machine.

maybe you're using a proxy?

httplib doesn't work with proxies. use urllib instead:

    import urllib, os
    os.environ["http_proxy"] = "yourproxy:port"
    file = urllib.urlopen("http://www.python.org")

(if you're using IE and Python 2.0, you can leave out the
os.environ line)

Cheers /F




From lreilly at cs.strath.ac.uk  Tue Feb 20 17:20:03 2001
From: lreilly at cs.strath.ac.uk (Lee Reilly CS1997)
Date: Tue, 20 Feb 2001 16:20:03 +0000
Subject: retrieving file data from an external server - where to start?
Message-ID: <3A929933.AA4F81@cs.strath.ac.uk>

hi there,

i'm using python with zope and am trying to do the following with python
method:

retrieveFileInfo(matric){

  username = getUsername(matric)
  filename = username + ".gif"
  server = "http://blahblahblah/yaddieyaddie/"

  # some code to determine whether or not the file exists
  # where do i start
  # e.g if (retrieve.(server+filename) != none):
  #        do something
}

e.g. given the matric value '9728430' i find that the corresponding user
has username == 'lreilly'. if this person has their image stored on the
external server then it will be called 'lreilly.gif'.

i don't need to actually retrieve the file from python - only determine
whether or not it exists or not. i.e. whether the http request reports a
404 or maybe even if the image size == 0 bytes.

could someone possible be kind enough to give me an idea where to start?
a url? a snippet of code?

thanks very much in advance,

lee


From lreilly at cs.strath.ac.uk  Tue Feb 20 17:46:11 2001
From: lreilly at cs.strath.ac.uk (Lee Reilly CS1997)
Date: Tue, 20 Feb 2001 16:46:11 +0000
Subject: additional info
References: <3A929933.AA4F81@cs.strath.ac.uk>
Message-ID: <3A929F53.9E9D0140@cs.strath.ac.uk>

i'm trying to do this using python methods, which only support commonly
used modules e.g. string, math, etc. i had a look at the httplib module,
which seems like the ideal solution but this isn't supported in python
methods.

is the use of this module the only solution to the problem?

nb: i can use the module in an *external* python method but this is a
last resort.

thanks for the bandwidth,

lee


From flognat at flognat.myip.org  Tue Feb 20 20:42:40 2001
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Tue, 20 Feb 2001 19:42:40 GMT
Subject: additional info
References: <3A929933.AA4F81@cs.strath.ac.uk> <3A929F53.9E9D0140@cs.strath.ac.uk>
Message-ID: <m33dd9b0qj.fsf@master.athome>

Lets see.. You want to do this (see if you can fetch a page) doing as
few imports as possible.. Or even no imports??

I think you have got the fact 'python methods' wrong.. 

Well first of all, to be able to get data over the internet, you have
to import the socket module, it can be compiled into the python
interpreter, or can be as pyd(?) file.. (ehm to be honest, some of the
new Solaris versions have some kind of /dev/ device that you can open
and write some data to in a normal way, and read the answer from the
other end of the internet)

Ok so we go for sockets?? Then it is easy for you to hack a small
non-generic functionality to get certain webpages, you kan hardcode
strings and so on..

Write to the socket ('''HEAD [WEBPAGE] HTTP/1.1\n\r[some headers]''')
read from the socket, and see if you return a 404 or 200.. 

        /Andy


/ Lee Reilly CS1997 <lreilly at cs.strath.ac.uk> wrote:
| i'm trying to do this using python methods, which only support commonly
| used modules e.g. string, math, etc. i had a look at the httplib module,
| which seems like the ideal solution but this isn't supported in python
| methods.
| 
| is the use of this module the only solution to the problem?
| 
| nb: i can use the module in an *external* python method but this is a
| last resort.
| 
| thanks for the bandwidth,
| 
| lee


From embed at geocities.com  Tue Feb 20 17:21:45 2001
From: embed at geocities.com (Warren Postma)
Date: Tue, 20 Feb 2001 11:21:45 -0500
Subject: Better Threading, starting with Events
Message-ID: <jRwk6.270$TW.2706@tor-nn1.netcom.ca>

Here is my little 'event' module __doc__ string:

event module 1.0

  Provides an efficient way for threads to go to sleep while waiting for
  work to do, freeing up the python interpreter to run the other threads,
  and yet waking up immediately upon signalling the worker thread

  Uses CreateEvent/SetEvent/WaitForMultipleObjects Win32 functions.

  Usage Sample:

     import event
     sd = event.new('Shutdown')
     sd.MyAttribute = 1 # stores something in event dictionary
     work = event.new('DoSomeWork',1) # 'one-shot'
     work .work_to_do = func # pass the function to the thread
     ret = event.select( (sd,work), 5000 ) # wait 5 seconds for signal
     if ret:  # Got signal?
         print repr(ret)

------------------------------------------------------------
                Here are the module methods:
------------------------------------------------------------

event.new( 'name', manualResetFlag=1, initialStateFlag=0)
   -> Creates and returns a new event object
        manualResetFlag = does event reset automatically when received
        (default=1)
        initialStateFlag = event signalled or not signalled initially
        (default=0)

 e = event.new( 'ASampleEvent')

 Using a WinEvent:
        e.name       name of event <string-attribute>
        e.wait()     wait indefinitely (until python shutdown) for event
        e.wait(500)  wait up to 500 milliseconds (0.5 seconds) for event
        e.state()    query state (non blocking) returns 0 or 1
        e.set()      set signalled state
        e.reset()    reset to non-signalled state
        e.handle()   get internal handle value (as integer)

To wait for several events in parallel, use event.select( (e1,e2,..),
timeout )

------------------------------------------------------------

event.select( <event tuple>, timeout=None)
   -> Waits for a single event or any of a group of events,
    returns None if the timeout occurs.
      Note that all waits are interruptable by shutdown,
      which will cause an exception to be raised.

  Example:
     ev1 = event.new('DoSomething')
     ev2 = event.new('DoSomethingElse')
     ret=event.select( (ev1,ev2), 5000 ) # wait up to 5000 millisec.
     if ret:
          print 'got event', ret.name  # which is it? DoSomething, or
DoSomethingElse
     else: # none?
          print 'timeout'


--------------------------- End ----------------------------

Now, for a discussion on my Evil Plan:

Anyone interested in native C Type for Event Signalling? I have written a
1.0, or 0.1, that works, it just ain't portable off of Win32. I'll make it
portable if there is some agreement on the basic idea behind what I'm doing.

I'd like feedback on whether my Evil Plan calls for a PEP.  Before I go
PEP'ing, I'd like to get some initial comments on other people who have
implemented worker threads. Is anyone else dissatisfied with threading.py?

My first problem with the existing facilities of threading.py is its
existence. I don't like that the Primitives are not Primitive, the interface
is great, the implementation is in Python. Urgh!  Great for Jython/JPython.
Not so great for CPython guys like me.

Why should Locks, Events, and Signals and Semaphores and so on all live in a
py file?  Their efficiency is therefore sub-par, and it will negatively
affect the performance of all apps that use it.  I also feel that a
consistent API and underlying implementation ought to permeate downwards
into the core of CPython.  Anyone with me? At that stage, you can see my
Evil Plan is beginning to get into swing.

But back to reality for a minute. The first thing I needed was extremely
efficient Event Objects, which could make a thread be "sleepy", waking up
periodically every 5 seconds to do some housekeeping, then going back to
sleep, but to be instantly interruptable at any time and immediately start
doing whatever work is required. They let go of the python interpreter lock,
and don't make the interpreter do any work, running functions in
threading.py while the thread ought to be idle, etcetera.

Also, my eventual evil plan involves thread-by-thread signalling in ways not
currently supported by the current frame objects, and the code in ceval.c.
It should also involve making various blocking functions "interruptable".
For example, time.sleep, which calls the C function floatsleep in
timemodule.c:
      Py_BEGIN_ALLOW_THREADS
      Sleep((unsigned long)millisecs);
      Py_END_ALLOW_THREADS

This should be, IMHO:
      Py_BEGIN_ALLOW_THREADS
      ret=WaitForSingleObject( hSystemShutdown, INIFINITE);
      Py_END_ALLOW_THREADS
      if (ret != WAIT_TIMEOUT)
            return NULL; // system is shutting down.

I am basically doing my R&D here anyways to get threading maximally
efficient on my embedded systems.  But I am more than willing, excited even,
to try my hand at proposing a PEP.   Not for my little event module, but for
a whole bunch of threading primitives, in C. Perhaps the CPython version of
threading.py could be adapted to use them.

The PEP would probably be for an expanded  "thread" module, and a revised
threading system that improves "pervasive threading" inside CPython.

Anyone? Anyone? Bueller?

Warren




From rwklee at home.com  Wed Feb 21 03:23:26 2001
From: rwklee at home.com (Rick Lee)
Date: Wed, 21 Feb 2001 02:23:26 GMT
Subject: Better Threading, starting with Events
References: <jRwk6.270$TW.2706@tor-nn1.netcom.ca>
Message-ID: <3A93269A.8AFFC7BB@home.com>

I like the idea of being able to wait for several events at once.   That's
something I can't do with queues in the Queue module.  So my wish list would
extend to queues as well.  Another thing on my wish list is to be able to join
to several threads at once.  And then finally, to be able to select from any
combination of events, queues, joins, and maybe anything else that blocks.

Maybe you also want to keep on eye on compatibility with Stackless.

- Rick Lee

Warren Postma wrote:

> Here is my little 'event' module __doc__ string:
>
> event module 1.0
>
>   Provides an efficient way for threads to go to sleep while waiting for
>   work to do, freeing up the python interpreter to run the other threads,
>   and yet waking up immediately upon signalling the worker thread
>
>   Uses CreateEvent/SetEvent/WaitForMultipleObjects Win32 functions.
>
>   Usage Sample:
>
>      import event
>      sd = event.new('Shutdown')
>      sd.MyAttribute = 1 # stores something in event dictionary
>      work = event.new('DoSomeWork',1) # 'one-shot'
>      work .work_to_do = func # pass the function to the thread
>      ret = event.select( (sd,work), 5000 ) # wait 5 seconds for signal
>      if ret:  # Got signal?
>          print repr(ret)
>
> ------------------------------------------------------------
>                 Here are the module methods:
> ------------------------------------------------------------
>
> event.new( 'name', manualResetFlag=1, initialStateFlag=0)
>    -> Creates and returns a new event object
>         manualResetFlag = does event reset automatically when received
>         (default=1)
>         initialStateFlag = event signalled or not signalled initially
>         (default=0)
>
>  e = event.new( 'ASampleEvent')
>
>  Using a WinEvent:
>         e.name       name of event <string-attribute>
>         e.wait()     wait indefinitely (until python shutdown) for event
>         e.wait(500)  wait up to 500 milliseconds (0.5 seconds) for event
>         e.state()    query state (non blocking) returns 0 or 1
>         e.set()      set signalled state
>         e.reset()    reset to non-signalled state
>         e.handle()   get internal handle value (as integer)
>
> To wait for several events in parallel, use event.select( (e1,e2,..),
> timeout )
>
> ------------------------------------------------------------
>
> event.select( <event tuple>, timeout=None)
>    -> Waits for a single event or any of a group of events,
>     returns None if the timeout occurs.
>       Note that all waits are interruptable by shutdown,
>       which will cause an exception to be raised.
>
>   Example:
>      ev1 = event.new('DoSomething')
>      ev2 = event.new('DoSomethingElse')
>      ret=event.select( (ev1,ev2), 5000 ) # wait up to 5000 millisec.
>      if ret:
>           print 'got event', ret.name  # which is it? DoSomething, or
> DoSomethingElse
>      else: # none?
>           print 'timeout'
>
> --------------------------- End ----------------------------
>
> Now, for a discussion on my Evil Plan:
>
> Anyone interested in native C Type for Event Signalling? I have written a
> 1.0, or 0.1, that works, it just ain't portable off of Win32. I'll make it
> portable if there is some agreement on the basic idea behind what I'm doing.
>
> I'd like feedback on whether my Evil Plan calls for a PEP.  Before I go
> PEP'ing, I'd like to get some initial comments on other people who have
> implemented worker threads. Is anyone else dissatisfied with threading.py?
>
> My first problem with the existing facilities of threading.py is its
> existence. I don't like that the Primitives are not Primitive, the interface
> is great, the implementation is in Python. Urgh!  Great for Jython/JPython.
> Not so great for CPython guys like me.
>
> Why should Locks, Events, and Signals and Semaphores and so on all live in a
> py file?  Their efficiency is therefore sub-par, and it will negatively
> affect the performance of all apps that use it.  I also feel that a
> consistent API and underlying implementation ought to permeate downwards
> into the core of CPython.  Anyone with me? At that stage, you can see my
> Evil Plan is beginning to get into swing.
>
> But back to reality for a minute. The first thing I needed was extremely
> efficient Event Objects, which could make a thread be "sleepy", waking up
> periodically every 5 seconds to do some housekeeping, then going back to
> sleep, but to be instantly interruptable at any time and immediately start
> doing whatever work is required. They let go of the python interpreter lock,
> and don't make the interpreter do any work, running functions in
> threading.py while the thread ought to be idle, etcetera.
>
> Also, my eventual evil plan involves thread-by-thread signalling in ways not
> currently supported by the current frame objects, and the code in ceval.c.
> It should also involve making various blocking functions "interruptable".
> For example, time.sleep, which calls the C function floatsleep in
> timemodule.c:
>       Py_BEGIN_ALLOW_THREADS
>       Sleep((unsigned long)millisecs);
>       Py_END_ALLOW_THREADS
>
> This should be, IMHO:
>       Py_BEGIN_ALLOW_THREADS
>       ret=WaitForSingleObject( hSystemShutdown, INIFINITE);
>       Py_END_ALLOW_THREADS
>       if (ret != WAIT_TIMEOUT)
>             return NULL; // system is shutting down.
>
> I am basically doing my R&D here anyways to get threading maximally
> efficient on my embedded systems.  But I am more than willing, excited even,
> to try my hand at proposing a PEP.   Not for my little event module, but for
> a whole bunch of threading primitives, in C. Perhaps the CPython version of
> threading.py could be adapted to use them.
>
> The PEP would probably be for an expanded  "thread" module, and a revised
> threading system that improves "pervasive threading" inside CPython.
>
> Anyone? Anyone? Bueller?
>
> Warren



From gmcm at hypernet.com  Sat Feb 24 03:13:59 2001
From: gmcm at hypernet.com (Gordon McMillan)
Date: 24 Feb 2001 02:13:59 GMT
Subject: Better Threading, starting with Events
References: <jRwk6.270$TW.2706@tor-nn1.netcom.ca>
Message-ID: <9051D40E0gmcmhypernetcom@199.171.54.154>

embed at geocities.com (Warren Postma) wrote in
<jRwk6.270$TW.2706 at tor-nn1.netcom.ca>: 

[huge snip]

>Anyone interested in native C Type for Event Signalling? I have written
>a 1.0, or 0.1, that works, it just ain't portable off of Win32. I'll
>make it portable if there is some agreement on the basic idea behind
>what I'm doing. 

If you want to work with native Win32 events, locks etc. you can use Mark 
Hammond's Win32event stuff. But before you go making cross-platform 
promises, you might want to look carefully at threadmodule.c and friends.

Believe it or not, it took a lot of effort to come up with a set of Python 
primitives which behave reasonably cross-platform. What's efficient on one 
platform is likely to a dog on another.

- Gordon


From akuchlin at mems-exchange.org  Tue Feb 20 18:11:59 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: Tue, 20 Feb 2001 12:11:59 -0500
Subject: What of Siphon?
Message-ID: <E14VGKV-0007Ia-00@ute.cnri.reston.va.us>

Suchandra Thapa's Siphon code is now available.

http://sourceforge.net/project/showfiles.php?group_id=20509&release_id=24001

Comments?  Let's not have the issue go to sleep again!

--amk



From phlip_cpp at my-deja.com  Tue Feb 20 19:07:31 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: Tue, 20 Feb 2001 10:07:31 -0800
Subject: What of Siphon?
References: <mailman.982689203.3370.python-list@python.org>
Message-ID: <Npyk6.96825$Wq1.39953628@nnrp5-w.sbc.net>

Andrew Kuchling wrote:

> Suchandra Thapa's Siphon code is now available.
> 
> 
http://sourceforge.net/project/showfiles.php?group_id=20509&release_id=24001
> 
> Comments?  Let's not have the issue go to sleep again!

What does it do? Or is that a secret?

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  All sensors report Patti having a very good time  --



From andrew at andrewcooke.free-online.co.uk  Tue Feb 20 20:41:33 2001
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Tue, 20 Feb 2001 19:41:33 +0000
Subject: What of Siphon?
References: <mailman.982689203.3370.python-list@python.org> <Npyk6.96825$Wq1.39953628@nnrp5-w.sbc.net>
Message-ID: <3A92C86D.282751E3@andrewcooke.free-online.co.uk>

My ISP seems to have decided the rest of the world doesn't exist (apart
from usenet, which makes it a very odd world that remains), but isn't
this connected with the CPAN-for-Python project?  The name rings a bell.

May be completely wrong...
Andrew

Phlip wrote:
> 
> Andrew Kuchling wrote:
> 
> > Suchandra Thapa's Siphon code is now available.
> >
> >
> http://sourceforge.net/project/showfiles.php?group_id=20509&release_id=24001
> >
> > Comments?  Let's not have the issue go to sleep again!
> 
> What does it do? Or is that a secret?
> 
> --
>   Phlip                          phlip_cpp at my-deja.com
> ============ http://c2.com/cgi/wiki?PhlIp ============
>   --  All sensors report Patti having a very good time  --


From amk at mira.erols.com  Wed Feb 21 01:53:00 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 21 Feb 2001 00:53:00 GMT
Subject: What of Siphon?
References: <mailman.982689203.3370.python-list@python.org> <Npyk6.96825$Wq1.39953628@nnrp5-w.sbc.net> <3A92C86D.282751E3@andrewcooke.free-online.co.uk>
Message-ID: <slrn99656f.h0.amk@207-172-36-141.s141.tnt6.ann.va.dialup.rcn.com>

On Tue, 20 Feb 2001 19:41:33 +0000, 
	Andrew Cooke <andrew at andrewcooke.free-online.co.uk> wrote:
>My ISP seems to have decided the rest of the world doesn't exist (apart
>from usenet, which makes it a very odd world that remains), but isn't
>this connected with the CPAN-for-Python project?  The name rings a bell.

Correct.  Discussion on this topic died down immediately after a brief
flurry of activity (as usual).

--amk



From doughellmann at bigfoot.com  Wed Feb 21 02:49:29 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Wed, 21 Feb 2001 01:49:29 GMT
Subject: What of Siphon?
References: <mailman.982689203.3370.python-list@python.org> <Npyk6.96825$Wq1.39953628@nnrp5-w.sbc.net> <3A92C86D.282751E3@andrewcooke.free-online.co.uk> <slrn99656f.h0.amk@207-172-36-141.s141.tnt6.ann.va.dialup.rcn.com>
Message-ID: <J8Fk6.72357$m5.4255033@news1.rdc1.ga.home.com>

Some of the conversation was going to move to the SourceForge forum associated with
the project, but there wasn't much there either.  Suchandra Thapa, the
project creator, was having some system troubles and then got busy with
work this week and things have died down a bit.

Doug

In article
<slrn99656f.h0.amk at 207-172-36-141.s141.tnt6.ann.va.dialup.rcn.com>,
amk at mira.erols.com  wrote:

> On Tue, 20 Feb 2001 19:41:33 +0000, 
> 	Andrew Cooke <andrew at andrewcooke.free-online.co.uk> wrote:
>>My ISP seems to have decided the rest of the world doesn't exist (apart
>>from usenet, which makes it a very odd world that remains), but isn't
>>this connected with the CPAN-for-Python project?  The name rings a bell.
> 
> Correct.  Discussion on this topic died down immediately after a brief
> flurry of activity (as usual).
> 
> --amk
>


From ssthapa at classes.cs.uchicago.edu  Wed Feb 21 02:12:28 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Wed, 21 Feb 2001 01:12:28 GMT
Subject: What of Siphon?
References: <mailman.982689203.3370.python-list@python.org> <Npyk6.96825$Wq1.39953628@nnrp5-w.sbc.net> <3A92C86D.282751E3@andrewcooke.free-online.co.uk>
Message-ID: <slrn995obe.4nc.ssthapa@ntcs-ip45.uchicago.edu>

Andrew Cooke <andrew at andrewcooke.free-online.co.uk> wrote:
>
>My ISP seems to have decided the rest of the world doesn't exist (apart
>from usenet, which makes it a very odd world that remains), but isn't
>this connected with the CPAN-for-Python project?  The name rings a bell.
>
>May be completely wrong...
>Andrew

  Yes, it is.  


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From mal at lemburg.com  Tue Feb 20 19:26:41 2001
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 20 Feb 2001 19:26:41 +0100
Subject: [Distutils] What of Siphon?
References: <E14VGKV-0007Ia-00@ute.cnri.reston.va.us>
Message-ID: <3A92B6E1.3AD4789C@lemburg.com>

Andrew Kuchling wrote:
> 
> Suchandra Thapa's Siphon code is now available.
> 
> http://sourceforge.net/project/showfiles.php?group_id=20509&release_id=24001
> 
> Comments?  Let's not have the issue go to sleep again!

Hmm, no docs, no summary, no home page... 
what is this Siphon thingie ? ;-) 

-- 
Marc-Andre Lemburg
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Pages:                           http://www.lemburg.com/python/



From ssthapa at classes.cs.uchicago.edu  Wed Feb 21 02:12:29 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Wed, 21 Feb 2001 01:12:29 GMT
Subject: [Distutils] What of Siphon?
References: <E14VGKV-0007Ia-00@ute.cnri.reston.va.us> <mailman.982693646.8017.python-list@python.org>
Message-ID: <slrn995ohs.4nc.ssthapa@ntcs-ip45.uchicago.edu>

M.-A. Lemburg <mal at lemburg.com> wrote:
>Andrew Kuchling wrote:
>> 
>> Suchandra Thapa's Siphon code is now available.
>> 
>> http://sourceforge.net/project/showfiles.php?group_id=20509&release_id=24001
>> 
>> Comments?  Let's not have the issue go to sleep again!
>
>Hmm, no docs, no summary, no home page... 
>what is this Siphon thingie ? ;-) 

  It's an early project to implement a CPAN module for python.  I'm not really
familiar with sourceforge and lack the time to put all the supporting 
documentation on the web page but I'm still working on the code.  Hopefully
by the end of this week I'll have the time to flesh out the project details 
on sourceforge and more importantly add more code and better documentation 
for the code.

-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From wmiller at mediaone.net  Tue Feb 20 18:35:27 2001
From: wmiller at mediaone.net (Walter Miller)
Date: Tue, 20 Feb 2001 09:35:27 -0800
Subject: How to get device context for com object?
Message-ID: <NVxk6.15$uj7.15036@news.pacbell.net>

Anyone know how to get a device context for a com object?  For example:

import win32com.client

ie = win32com.client.Dispatch('InternetExplorer.Application.1')
ie.Visible = 1
ie.Navigate('http://www.yahoo.com')

hDCSrc = ie.GetDC()   # Thought this might be the way but it's not an option




From aleaxit at yahoo.com  Tue Feb 20 21:10:12 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 21:10:12 +0100
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net>
Message-ID: <96ujcc12r6k@news1.newsguy.com>

"Walter Miller" <wmiller at mediaone.net> wrote in message
news:NVxk6.15$uj7.15036 at news.pacbell.net...
> Anyone know how to get a device context for a com object?  For example:

There is no general way: the COM object might have/use any
number of windows, from 0 upwards, at any given time, and/or
other devices, so it's not clear at all what 'a device context for
the object' might mean (what does it mean to get a device
context for a process, or a thread, or a DLL, or...?).

Some specific COM interface for which it makes sense will
implement a GetDC method, of course (offhand, though, only
IDirectDrawSurface3 comes to mind -- which doesn't help
you much with an internet explorer application object!-)

Other will let you get a handle to a window object (again
only one comes to mind, IOleWindow, but that one has
several derivatives, and is used heavily in the in-place
activation protocol, so it _may_ be possible, with a little
bit of effort, to get *that* one for internet explorer -- if
you activate it in-place within a container of yours, or use
it as the container for your own activex control's in-place
activation).


Alex





From wmiller at mediaone.net  Tue Feb 20 23:03:22 2001
From: wmiller at mediaone.net (Walter Miller)
Date: Tue, 20 Feb 2001 14:03:22 -0800
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com>
Message-ID: <YQBk6.80$3e2.62458@news.pacbell.net>

The "InternetExplorer.Application.1" com object has a HWND property from
which you can get the window handle.  It seem reasonable to expect that one
could get a device context based on a valid window handle.

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:96ujcc12r6k at news1.newsguy.com...
> "Walter Miller" <wmiller at mediaone.net> wrote in message
> news:NVxk6.15$uj7.15036 at news.pacbell.net...
> > Anyone know how to get a device context for a com object?  For example:
>
> There is no general way: the COM object might have/use any
> number of windows, from 0 upwards, at any given time, and/or
> other devices, so it's not clear at all what 'a device context for
> the object' might mean (what does it mean to get a device
> context for a process, or a thread, or a DLL, or...?).
>
> Some specific COM interface for which it makes sense will
> implement a GetDC method, of course (offhand, though, only
> IDirectDrawSurface3 comes to mind -- which doesn't help
> you much with an internet explorer application object!-)
>
> Other will let you get a handle to a window object (again
> only one comes to mind, IOleWindow, but that one has
> several derivatives, and is used heavily in the in-place
> activation protocol, so it _may_ be possible, with a little
> bit of effort, to get *that* one for internet explorer -- if
> you activate it in-place within a container of yours, or use
> it as the container for your own activex control's in-place
> activation).
>
>
> Alex
>
>
>
>




From aleaxit at yahoo.com  Tue Feb 20 23:44:51 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 23:44:51 +0100
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com> <YQBk6.80$3e2.62458@news.pacbell.net>
Message-ID: <96us1u0dj8@news1.newsguy.com>

"Walter Miller" <wmiller at mediaone.net> wrote in message
news:YQBk6.80$3e2.62458 at news.pacbell.net...
> The "InternetExplorer.Application.1" com object has a HWND property from
> which you can get the window handle.  It seem reasonable to expect that
one
> could get a device context based on a valid window handle.

If you have a valid window handle, you can indeed get a device
context for that window -- that's got nothing to do with COM (or
however else you got your HWND), but rather is the job of the APIs
GetWindowDC (if you want the DC for the _entire_ window, i.e.,
including the non-client areas), GetDC (if you want the DC for
the client-area only), GetDCEx (if you want peculiar clipping, a
cached context, and/or intact DC attributes... THAT one I've never
used yet!-), ... -- just call the appropriate one (at least the first
two, I believe, are exposed by modules in the win32all package).


Alex





From wmiller at mediaone.net  Wed Feb 21 00:56:25 2001
From: wmiller at mediaone.net (Walter Miller)
Date: Tue, 20 Feb 2001 15:56:25 -0800
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com> <YQBk6.80$3e2.62458@news.pacbell.net> <96us1u0dj8@news1.newsguy.com>
Message-ID: <XuDk6.96953$Wq1.40024851@nnrp5-w.sbc.net>

That's exaclty what I'm trying to do but don't know how through python - get
the device context based on a window handle.  How do I call the GetDC api
through python based on a given window handle?  Sorry, this probably should
have been my original question.




From aleaxit at yahoo.com  Wed Feb 21 10:29:35 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 10:29:35 +0100
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com> <YQBk6.80$3e2.62458@news.pacbell.net> <96us1u0dj8@news1.newsguy.com> <XuDk6.96953$Wq1.40024851@nnrp5-w.sbc.net>
Message-ID: <9701q102nt4@news1.newsguy.com>

"Walter Miller" <wmiller at mediaone.net> wrote in message
news:XuDk6.96953$Wq1.40024851 at nnrp5-w.sbc.net...
> That's exaclty what I'm trying to do but don't know how through python -
get
> the device context based on a window handle.  How do I call the GetDC api
> through python based on a given window handle?  Sorry, this probably
should
> have been my original question.

You need to have installed win32all (or the ActiveState build
of Python for Win32, that comes with win32all).  Then,

import win32gui
hdc = win32gui.GetDC(hwnd)

will set hdc to the (int) handle for the device context
corresponding to (the client areas of) the window whose
handle is in (int) hwnd.  Remember to also call
win32gui.ReleaseDC(hwnd, hdc) when you're done (you
might want to use a try/finally...).

(Similarly, you may use win32gui.GetWindowDC to get the
hdc for the whole window including non-client areas, &c).


Alex








From wmiller at mediaone.net  Wed Feb 21 23:48:35 2001
From: wmiller at mediaone.net (Walter Miller)
Date: Wed, 21 Feb 2001 14:48:35 -0800
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com> <YQBk6.80$3e2.62458@news.pacbell.net> <96us1u0dj8@news1.newsguy.com> <XuDk6.96953$Wq1.40024851@nnrp5-w.sbc.net> <9701q102nt4@news1.newsguy.com>
Message-ID: <rBXk6.268$FG.70985@news.pacbell.net>

That's great!  Yes, it works but I'm also trying to call
CreateCompatibleDC() and it only works with the device context object not
the (int) handle for the device context.  Is there an api for creating a
device context object based on its own handle?

The ultimate plan is to try an replicate the visual basic example of copying
the screenshot of the client portion of the window and saving it as a bitmap
file.  I have an activex control that does this already but that's too
easy - I'm _actually_ trying to learn all the details of using Python with
win32.

> import win32gui
> hdc = win32gui.GetDC(hwnd)
>
> will set hdc to the (int) handle for the device context
> corresponding to (the client areas of) the window whose
> handle is in (int) hwnd.




From aleaxit at yahoo.com  Thu Feb 22 10:28:13 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 10:28:13 +0100
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com> <YQBk6.80$3e2.62458@news.pacbell.net> <96us1u0dj8@news1.newsguy.com> <XuDk6.96953$Wq1.40024851@nnrp5-w.sbc.net> <9701q102nt4@news1.newsguy.com> <rBXk6.268$FG.70985@news.pacbell.net>
Message-ID: <972m3d0keq@news2.newsguy.com>

"Walter Miller" <wmiller at mediaone.net> wrote in message
news:rBXk6.268$FG.70985 at news.pacbell.net...
> That's great!  Yes, it works but I'm also trying to call
> CreateCompatibleDC() and it only works with the device context object not
> the (int) handle for the device context.  Is there an api for creating a
> device context object based on its own handle?

win32ui.CreateDCFromHandle.


Alex





From ransen_spam_me_not at nemo.it  Tue Feb 20 19:11:23 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Tue, 20 Feb 2001 18:11:23 GMT
Subject: Py_Exit in a C/C++ program
Message-ID: <3a9298dc.4233599@news.newsguy.com>

Till my Python books arrive... ;)

When embedding Python in a C/C++ program I suppose 
that Py_Exit was created as an simple way of
exiting your C (C++) program and at the same time
cleaning up the Python environment?



--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From ransen_spam_me_not at nemo.it  Tue Feb 20 19:11:36 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Tue, 20 Feb 2001 18:11:36 GMT
Subject: When embedding with no natural console...?
Message-ID: <3a969ec3.5744237@news.newsguy.com>

When embedding with no natural console (I mean that
the application does not have a console output, it
is pure GUI) what suggestions are there for printing?

I can think of some, but someone with more experience
may be able to tell me which is best or easiest or
most dangerous...

*) Create some sort of console in the GUI app.
   Maybe with AllocConsole WriteConsole and 
   FreeConsole...
*) Ridirect output to a log file...
*) Do not print anything...
*) Other...

Since my final app will be aimed at beginning programmers
I think some sort of console output would be nice.


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From aleaxit at yahoo.com  Tue Feb 20 21:00:49 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 20 Feb 2001 21:00:49 +0100
Subject: When embedding with no natural console...?
References: <3a969ec3.5744237@news.newsguy.com>
Message-ID: <96ujca02r6k@news1.newsguy.com>

"Owen F. Ransen" <ransen_spam_me_not at nemo.it> wrote in message
news:3a969ec3.5744237 at news.newsguy.com...
> When embedding with no natural console (I mean that
> the application does not have a console output, it
> is pure GUI) what suggestions are there for printing?

Set the 'stdout' attribute of module 'sys' (PySys_SetObject
may work well for that, but it seems to be undocumented
so I'm not sure it's best; you might do it in Python itself,
executing a small piece of Python source from within your
embedding application) to a "suitable" object -- one that
implements a "suitable" write method, accepting a (buffer)
Python object (and lets Python set and get an attribute
named 'softspace').  It's simplest to do it in Python, with
a small Python class that just directs the argument of that
write method to a module-level function in your module.
(You may want to do the same for the 'stderr' attribute).

Your module-level function may do whatever you see fit --
emit the output with Win32 API "OutputDebugString", use
the event-logging system functions of NT, add it to a real
on-disk file, scroll it on a listbox or other within-a-GUI sort
of "console-alike", etc, etc.

> *) Create some sort of console in the GUI app.
>    Maybe with AllocConsole WriteConsole and
>    FreeConsole...

That's one possibilities, but real consoles don't work all
that well on Win95/Win98 and don't offer many advantages
(wrt a scrolling text window that you can embed/dock/etc
within your GUI) for text-output.

> *) Ridirect output to a log file...
> *) Do not print anything...
> *) Other...

It might be best to let the user choose (by configuration
options, menus, preferences, ...) among these (zero or
more thereof).

> Since my final app will be aimed at beginning programmers
> I think some sort of console output would be nice.

Scrolling text-output (with a long-enough 'history' that is
scrollable backwards, ability to save to a textfile, etc) should
be good enough for that.


Alex





From ransen_spam_me_not at nemo.it  Wed Feb 21 17:25:58 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Wed, 21 Feb 2001 16:25:58 GMT
Subject: When embedding with no natural console...?
References: <3a969ec3.5744237@news.newsguy.com> <96ujca02r6k@news1.newsguy.com>
Message-ID: <3a965a5c.2664472@news.newsguy.com>

>> Since my final app will be aimed at beginning programmers
>> I think some sort of console output would be nice.
>Scrolling text-output (with a long-enough 'history' that is
>scrollable backwards, ability to save to a textfile, etc) should
>be good enough for that.
>Alex

Good point.


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From dsavitsk at e-coli.net  Wed Feb 21 00:56:11 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 20 Feb 2001 17:56:11 -0600
Subject: When embedding with no natural console...?
References: <3a969ec3.5744237@news.newsguy.com>
Message-ID: <wuDk6.120741$Ch.21281841@newsrump.sjc.telocity.net>

I needed a similar thing (and was tired of my app covering the win32
console), so i put a (tkinter) text widget and a scroll widget at the bottom
of my interface. One advantage i found was the ability to control the color
of the output (a novelty on windows it would seem) so errors are red, data
is blue, instructions are white, etc.  There is a picture of it in "action"
at the bottom of this page www.e-coli.net/software/ecpse/inst.html

doug


"Owen F. Ransen" <ransen_spam_me_not at nemo.it> wrote in message
news:3a969ec3.5744237 at news.newsguy.com...
> When embedding with no natural console (I mean that
> the application does not have a console output, it
> is pure GUI) what suggestions are there for printing?
>
> I can think of some, but someone with more experience
> may be able to tell me which is best or easiest or
> most dangerous...
>
> *) Create some sort of console in the GUI app.
>    Maybe with AllocConsole WriteConsole and
>    FreeConsole...
> *) Ridirect output to a log file...
> *) Do not print anything...
> *) Other...
>
> Since my final app will be aimed at beginning programmers
> I think some sort of console output would be nice.
>
>
> --
> Owen F. Ransen
> http://www.ransen.com/
> Home of Gliftic & Repligator Image Generators




From ransen_spam_me_not at nemo.it  Wed Feb 21 17:26:12 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Wed, 21 Feb 2001 16:26:12 GMT
Subject: When embedding with no natural console...?
References: <3a969ec3.5744237@news.newsguy.com> <wuDk6.120741$Ch.21281841@newsrump.sjc.telocity.net>
Message-ID: <3a975a88.2708336@news.newsguy.com>

On Tue, 20 Feb 2001 17:56:11 -0600, "dsavitsk" <dsavitsk at e-coli.net>
wrote:

>I needed a similar thing (and was tired of my app covering the win32
>console), 
>doug

Good point.


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From jim at interet.com  Wed Feb 21 14:23:09 2001
From: jim at interet.com (James C. Ahlstrom)
Date: Wed, 21 Feb 2001 08:23:09 -0500
Subject: When embedding with no natural console...?
References: <3a969ec3.5744237@news.newsguy.com>
Message-ID: <3A93C13D.E4A7F08D@interet.com>

"Owen F. Ransen" wrote:
> 
> When embedding with no natural console (I mean that
> the application does not have a console output, it
> is pure GUI) what suggestions are there for printing?

You need to replace sys.stdout and sys.stderr with an object which
has a "write" method.  Then you need to make "write" do something.
My "write" saves the text, and displays it in a MessageBox when
the system is idle.  But you could write it to a different window
for example.

Remember that output is lost if something goes wrong before you
can replace sys.stdout/err.

JimA


From ransen_spam_me_not at nemo.it  Wed Feb 21 17:59:22 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Wed, 21 Feb 2001 16:59:22 GMT
Subject: When embedding with no natural console...?
References: <3a969ec3.5744237@news.newsguy.com>
Message-ID: <3a94f040.1756349@news.newsguy.com>

On Tue, 20 Feb 2001 18:11:36 GMT, ransen_spam_me_not at nemo.it (Owen F.
Ransen) wrote:

>When embedding with no natural console (I mean that
>the application does not have a console output, it
>is pure GUI) what suggestions are there for printing?

Thanks all for your suggestions, I think that in my
case a window embedded somewhere in the app is the
best solution...


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From stevemak at softhome.net  Tue Feb 20 19:53:03 2001
From: stevemak at softhome.net (Steve Mak)
Date: Tue, 20 Feb 2001 18:53:03 GMT
Subject: help with re.split()
Message-ID: <3A92BF37.63637CA0@softhome.net>

Hi guys,

    How do I use the re.split() function so it splits a line of text,
keeping only the word. ie: it excludes any symbols, spaces, etc. I tried
p=re.split('[. ]+', line), but some spaces are being kept.

if the line is Hello, how are you?

I need to split it so I get:
Hello
how
are
you



From cfelling at iae.nl  Wed Feb 21 00:31:24 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 21 Feb 2001 00:31:24 +0100
Subject: help with re.split()
References: <3A92BF37.63637CA0@softhome.net>
Message-ID: <96uuoc$dhs$1@animus.fel.iae.nl>

Steve Mak <stevemak at softhome.net> wrote:
> Hi guys,

>     How do I use the re.split() function so it splits a line of text,
> keeping only the word. ie: it excludes any symbols, spaces, etc. I tried
> p=re.split('[. ]+', line), but some spaces are being kept.

Maybe some space are taps in disguish?
The approach you toke has it's pitfalls, like in

>>> import re
>>> re.split(r'\W+', 'Hello, how are you?')
['Hello', 'how', 'are', 'you', '']

Notice this empty string at the end of the split-list.
That's because after the matched "?" there is an empty string
(there always is:).  The same would happen with a leading match.

Maybe finding all words in that list is simpler, like (\w+ matches
all alphanumeric strings.):

>>> import re
>>> re.findall(r'\w+', ' Hello, how are you? ')
['Hello', 'how', 'are', 'you']
>>> re.findall(r'\w+', ' Hello, 2 you 2! ')
['Hello', '2', 'you', '2']
-- 
groetjes, carel


From stevemak at softhome.net  Wed Feb 21 00:52:21 2001
From: stevemak at softhome.net (Steve Mak)
Date: Tue, 20 Feb 2001 23:52:21 GMT
Subject: ATTN: Carel Fellinger
References: <3A92BF37.63637CA0@softhome.net> <96uuoc$dhs$1@animus.fel.iae.nl>
Message-ID: <3A930560.F84586D8@softhome.net>

Thanks for your help!!!!!

Carel Fellinger wrote:

> Steve Mak <stevemak at softhome.net> wrote:
> > Hi guys,
>
> >     How do I use the re.split() function so it splits a line of text,
> > keeping only the word. ie: it excludes any symbols, spaces, etc. I tried
> > p=re.split('[. ]+', line), but some spaces are being kept.
>
> Maybe some space are taps in disguish?
> The approach you toke has it's pitfalls, like in
>
> >>> import re
> >>> re.split(r'\W+', 'Hello, how are you?')
> ['Hello', 'how', 'are', 'you', '']
>
> Notice this empty string at the end of the split-list.
> That's because after the matched "?" there is an empty string
> (there always is:).  The same would happen with a leading match.
>
> Maybe finding all words in that list is simpler, like (\w+ matches
> all alphanumeric strings.):
>
> >>> import re
> >>> re.findall(r'\w+', ' Hello, how are you? ')
> ['Hello', 'how', 'are', 'you']
> >>> re.findall(r'\w+', ' Hello, 2 you 2! ')
> ['Hello', '2', 'you', '2']
> --
> groetjes, carel



From ben.hutchings at roundpoint.com  Wed Feb 21 01:16:29 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 20 Feb 2001 16:16:29 -0800
Subject: help with re.split()
References: <3A92BF37.63637CA0@softhome.net>
Message-ID: <uelws3n6q.fsf@roundpoint.com>

Steve Mak <stevemak at softhome.net> writes:

> Hi guys,
> 
>     How do I use the re.split() function so it splits a line of text,
> keeping only the word. ie: it excludes any symbols, spaces, etc. I tried
> p=re.split('[. ]+', line), but some spaces are being kept.

If the line has a leading and/or trailing section that matches the
pattern, then the result list will begin and/or end with an empty
string, but no spaces will be kept.  If you then join the listed
strings back together with spaces then you will get leading and/or
trailing space.

For example:
    >>> p=re.split(' ', ' string with leading and trailing space ')
    >>> p
    ['', 'string', 'with', 'leading', 'and', 'trailing', 'space', '']
    >>> string.join(p, ' ')
    ' string with leading and trailing space '

Perhaps this is what you are seeing?

> if the line is Hello, how are you?
> 
> I need to split it so I get:
> Hello
> how
> are
> you

Then you must include ',' and '?' in the character class.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From tchur at optushome.com.au  Tue Feb 20 20:41:43 2001
From: tchur at optushome.com.au (Tim Churches)
Date: Wed, 21 Feb 2001 06:41:43 +1100
Subject: Optimisation problem - in C?
Message-ID: <3A92C877.C67AAD8@optushome.com.au>

After a bit of profiling I have worked out exactly which part of a data
crunching class on which I am working needs to be optimised - maybe in
C? The abstracted function which needs to run faster is as follows - not
surprisingly it is itself a loop which lives within two other loops -
hence it is called rather frequently:

############
def makevector(sourcelist,elementlist):
	resultvector = []
	for element in elementlist:
		resultvector.append(sourcelist[element])
	return resultvector
###########

My test data has about 300,000 elements in sourcelist and between about
10,000 and 70,000 elements in elementlist, but larger lists are
envisaged, hence the earnest desire for maximum speed.

Questions:
Can this function (which essentially does the fairly generic operation
of "fetch the elements of sequence a using the element indexes contained
in sequence b") be optimised without resorting to a C module?

If not, is anyone willing to hold my hand (it's clean, but rather
sweaty) while I fumble with trying to replace the function with a C
module? (Note: after years of unremitting effort, I have finally
mastered all aspects of hello_world.c)

Tim Churches
Sydney, Australia



From johann at physics.berkeley.edu  Tue Feb 20 21:09:33 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 20 Feb 2001 12:09:33 -0800
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org>
Message-ID: <mt8zn1xgjm.fsf@astron.berkeley.edu>

Tim Churches writes:

> Questions:
> Can this function (which essentially does the fairly generic operation
> of "fetch the elements of sequence a using the element indexes contained
> in sequence b") be optimised without resorting to a C module?

The Numeric module has a "take" function, which does exactly what you
want.

You will have to convert your lists into numeric arrays of python
objects, but that shouldn't be difficult.

--J

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From aleaxit at yahoo.com  Wed Feb 21 00:17:19 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 00:17:19 +0100
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org>
Message-ID: <96uts50ilg@news1.newsguy.com>

"Tim Churches" <tchur at optushome.com.au> wrote in message
news:mailman.982698204.12938.python-list at python.org...
    [snip]
> def makevector(sourcelist,elementlist):
> resultvector = []
> for element in elementlist:
> resultvector.append(sourcelist[element])
> return resultvector
    [snip]
> Questions:
> Can this function (which essentially does the fairly generic operation
> of "fetch the elements of sequence a using the element indexes contained
> in sequence b") be optimised without resorting to a C module?

You can easily make it a *little* bit faster (though using Numeric,
as others suggested, would be better still):

def makevector(sourcelist, elementlist):
    resultvector = [None] * len(elementlist)
    for i in range(len(elementlist)):
        resultvector[i] = sourcelist[elementlist[i]]
    return resultvector

this is a bit faster by avoiding re-allocations of resultvector.


Alex





From ben.hutchings at roundpoint.com  Wed Feb 21 01:07:45 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 20 Feb 2001 16:07:45 -0800
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org>
Message-ID: <uhf1o3nla.fsf@roundpoint.com>

Tim Churches <tchur at optushome.com.au> writes:
<snip>
> def makevector(sourcelist,elementlist):
> 	resultvector = []
> 	for element in elementlist:
> 		resultvector.append(sourcelist[element])
> 	return resultvector
<snip>
> My test data has about 300,000 elements in sourcelist and between about
> 10,000 and 70,000 elements in elementlist, but larger lists are
> envisaged, hence the earnest desire for maximum speed.
> 
> Questions:
> Can this function (which essentially does the fairly generic operation
> of "fetch the elements of sequence a using the element indexes contained
> in sequence b") be optimised without resorting to a C module?
<snip>

I think it can.  Since a list's contents are contiguous in memory (in
order to support random access), appending to it will cause it to
reallocate and move all its contents periodically.  If I remember
correctly, Python's lists increase their memory allocations by
constant amounts rather than by a constant factor, so this loop will
run in quadratic time, i.e. proportional to the square of the length
of elementlist!

Each of the following definitions should be equivalent to the above,
and I think at least one will run faster:

    def makevector(sourcelist,elementlist):
        resultvector = [None] * len(elementlist)
        for i in range(len(elementlist)):
            resultvector[i] = sourcelist[elementlist[i]]
        return resultvector

    def makevector(sourcelist,elementlist):
        return [sourcelist[element] for element in elementlist]

    def makevector(sourcelist,elementlist):
        return map(lambda i,l=sourcelist: l[i], elementlist)

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From djc at object-craft.com.au  Wed Feb 21 04:27:39 2001
From: djc at object-craft.com.au (Dave Cole)
Date: 21 Feb 2001 14:27:39 +1100
Subject: Optimisation problem - in C?
In-Reply-To: Tim Churches's message of "Wed, 21 Feb 2001 06:41:43 +1100"
References: <mailman.982698204.12938.python-list@python.org>
Message-ID: <m366i4lnpw.fsf@vole.object-craft.com.au>

>>>>> "Tim" == Tim Churches <tchur at optushome.com.au> writes:

Tim> def makevector(sourcelist,elementlist):
Tim>         resultvector = []
Tim>         for element in elementlist:
Tim>                 resultvector.append(sourcelist[element])
Tim>         return resultvector

Because I am trying to avoid writing documentation for some code that
I want to give away, I decided to whip up some code for you to try.

Extract the two following files and grab a copy of Makefile.pre.in
from any other extension module - the CSV module from

        http://www.object-craft.com.au/projects/csv/

is where I copied mine from.

Place all of the files in a directory then type:

% make -f Makefile.pre.in boot
% make
% python
Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian GNU/Linux)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import timbo
>>> a = ['a','b','c','d','e']
>>> b = [2,3,4]
>>> timbo.makevector(a, b)
['c', 'd', 'e']
>>> 

Tell me if it goes faster (or if it even works :-)

Because I am a bit lazy, I have assumed that elementlist is a sequence
of integers.

- Dave

- - 8< - Setup.in - - - - - - - - - - - - - - - - - - - - - - - - - -
# Setup file to build the timbo module

*shared*
timbo timbo.c
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - 8< - timbo.c - - - - - - - - - - - - - - - - - - - - - - - - - - -
#include "Python.h"

static char makevector__doc__[] =
"Works just like the following - only faster (I hope)\n"
"\n"
"def makevector(sourcelist, elementlist):\n"
"	 resultvector = []\n"
"	 for element in elementlist:\n"
"		 resultvector.append(sourcelist[element])\n"
"	 return resultvector";

static PyObject *makevector(PyObject *module, PyObject *args)
{
    PyObject *sourcelist, *elementlist;
    PyObject *resultlist = NULL;
    int len, i;

    if (!PyArg_ParseTuple(args, "OO", &sourcelist, &elementlist))
	return NULL;

    len = PySequence_Length(elementlist);
    if (PyErr_Occurred())
	return NULL;

    resultlist = PyList_New(len);
    if (resultlist == NULL)
	return NULL;

    for (i = 0; i < len; i++) {
	PyObject *item;
	int num;

	if ((item = PySequence_GetItem(elementlist, i)) == NULL)
	    goto error;
	num = PyInt_AsLong(item);
	if (PyErr_Occurred()) {
	    Py_DECREF(item);
	    goto error;
	}
	if ((item = PySequence_GetItem(sourcelist, num)) == NULL)
	    goto error;
	PyList_SetItem(resultlist, i, item);
    }

    return resultlist;

error:
    Py_DECREF(resultlist);
    return NULL;
}

/* List of methods defined in the module */
static struct PyMethodDef methods[] = {
    { "makevector", (PyCFunction)makevector, METH_VARARGS, makevector__doc__ },
    { NULL, (PyCFunction)NULL, 0, NULL } /* sentinel */
};

static char module__doc__[] = 
"Tim's optimisations.";

void inittimbo(void)
{
    Py_InitModule4("timbo", methods, module__doc__,
		   (PyObject*)NULL, PYTHON_API_VERSION);
    if (PyErr_Occurred())
	Py_FatalError("can't initialize module timbo");
}
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

-- 
http://www.object-craft.com.au




From djc at object-craft.com.au  Wed Feb 21 05:54:29 2001
From: djc at object-craft.com.au (Dave Cole)
Date: 21 Feb 2001 15:54:29 +1100
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org>
Message-ID: <m3y9v03abe.fsf@vole.object-craft.com.au>

Forgot to followup in the newsgroup...

>>>>> "Tim" == Tim Churches <tchur at optushome.com.au> writes:

Tim> def makevector(sourcelist,elementlist):
Tim>         resultvector = []
Tim>         for element in elementlist:
Tim>                 resultvector.append(sourcelist[element])
Tim>         return resultvector

Because I am trying to avoid writing documentation for some code that
I want to give away, I decided to whip up some code for you to try.

Extract the two following files and grab a copy of Makefile.pre.in
from any other extension module - the CSV module from

        http://www.object-craft.com.au/projects/csv/

is where I copied mine from.

Place all of the files in a directory then type:

% make -f Makefile.pre.in boot
% make
% python
Python 1.5.2 (#0, Apr  3 2000, 14:46:48)  [GCC 2.95.2 20000313 (Debian GNU/Linux)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import timbo
>>> a = ['a','b','c','d','e']
>>> b = [2,3,4]
>>> timbo.makevector(a, b)
['c', 'd', 'e']
>>> 

Tell me if it goes faster (or if it even works :-)

Because I am a bit lazy, I have assumed that elementlist is a sequence
of integers.

- Dave

- - 8< - Setup.in - - - - - - - - - - - - - - - - - - - - - - - - - -
# Setup file to build the timbo module

*shared*
timbo timbo.c
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

- - 8< - timbo.c - - - - - - - - - - - - - - - - - - - - - - - - - - -
#include "Python.h"

static char makevector__doc__[] =
"Works just like the following - only faster (I hope)\n"
"\n"
"def makevector(sourcelist, elementlist):\n"
"	 resultvector = []\n"
"	 for element in elementlist:\n"
"		 resultvector.append(sourcelist[element])\n"
"	 return resultvector";

static PyObject *makevector(PyObject *module, PyObject *args)
{
    PyObject *sourcelist, *elementlist;
    PyObject *resultlist = NULL;
    int len, i;

    if (!PyArg_ParseTuple(args, "OO", &sourcelist, &elementlist))
	return NULL;

    len = PySequence_Length(elementlist);
    if (PyErr_Occurred())
	return NULL;

    resultlist = PyList_New(len);
    if (resultlist == NULL)
	return NULL;

    for (i = 0; i < len; i++) {
	PyObject *item;
	int num;

	if ((item = PySequence_GetItem(elementlist, i)) == NULL)
	    goto error;
	num = PyInt_AsLong(item);
	if (PyErr_Occurred()) {
	    Py_DECREF(item);
	    goto error;
	}
	if ((item = PySequence_GetItem(sourcelist, num)) == NULL)
	    goto error;
	PyList_SetItem(resultlist, i, item);
    }

    return resultlist;

error:
    Py_DECREF(resultlist);
    return NULL;
}

/* List of methods defined in the module */
static struct PyMethodDef methods[] = {
    { "makevector", (PyCFunction)makevector, METH_VARARGS, makevector__doc__ },
    { NULL, (PyCFunction)NULL, 0, NULL } /* sentinel */
};

static char module__doc__[] = 
"Tim's optimisations.";

void inittimbo(void)
{
    Py_InitModule4("timbo", methods, module__doc__,
		   (PyObject*)NULL, PYTHON_API_VERSION);
    if (PyErr_Occurred())
	Py_FatalError("can't initialize module timbo");
}
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

-- 
http://www.object-craft.com.au


From scarblac at pino.selwerd.nl  Wed Feb 21 14:02:34 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 21 Feb 2001 13:02:34 GMT
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org>
Message-ID: <slrn997f03.h0.scarblac@pino.selwerd.nl>

Tim Churches <tchur at optushome.com.au> wrote in comp.lang.python:
> After a bit of profiling I have worked out exactly which part of a data
> crunching class on which I am working needs to be optimised - maybe in
> C? The abstracted function which needs to run faster is as follows - not
> surprisingly it is itself a loop which lives within two other loops -
> hence it is called rather frequently:
> 
> ############
> def makevector(sourcelist,elementlist):
> 	resultvector = []
> 	for element in elementlist:
> 		resultvector.append(sourcelist[element])
> 	return resultvector
> ###########
> 
> My test data has about 300,000 elements in sourcelist and between about
> 10,000 and 70,000 elements in elementlist, but larger lists are
> envisaged, hence the earnest desire for maximum speed.

Yes, I can see that this would be slow.

It's too bad that lists don't have a .get() method (so x.get(3) would be
equal to x[3]), because resultvector=map(sourcelist.get, elementlist)
would be fastest (all done in C). As it is, you have to fake that with
resultvector=map(lambda i, l=sourcelist: l[i], elementlist), and the
lambda calls mean that this is much slower than your version

resultvector = [sourcelist[element] for element in elementlist]
is Python 2.0, and it at least is somewhat faster than your version (0.290s
for your version, 0.250 for the list comprehension on my machine with
len(sourcelist)=300000 and len(elementlist)=70000.

Wait, one of the main problems of your first version is that the list has to
grow. What if we create a list of the same length as elementlist first, say
by copying it?

def makevector(sourcelist, elementlist):
   resultvector = elementlist[:]
   for i in range(len(elementlist)):
      resultvector[i] = sourcelist[elementlist[i]]
   return resultvector
   
This one runs in 0.170 seconds on my computer, on average.

I'm out of further ideas. Too bad there's no syntax construct for 'for i in
range(len(...))' yet.

> Can this function (which essentially does the fairly generic operation
> of "fetch the elements of sequence a using the element indexes contained
> in sequence b") be optimised without resorting to a C module?

By about 40%, but for more you'll need to do it in C. I don't know how much
of a gain you can expect there.

How do you use this list? If you only use part of the data later you might
choose not to compute the whole thing at all but put a lazy wrapper around it.

class Wrapper:
   def __init__(self, sourcelist, elementlist):
      self.sl = sourcelist
      self.el = elementlist
   def __getitem__(self, i):
      return self.sl[self.el[i]]

resultlist = Wrapper(sourcelist, elementlist)

This would be faster if you use only a small part of the whole list.

> If not, is anyone willing to hold my hand (it's clean, but rather
> sweaty) while I fumble with trying to replace the function with a C
> module? (Note: after years of unremitting effort, I have finally
> mastered all aspects of hello_world.c)

I'll pass this one on, haven't needed to do it myself yet...

-- 
Remco Gerlich


From sebastian.dauss at web.de  Wed Feb 21 15:11:24 2001
From: sebastian.dauss at web.de (Sebastian Dauss)
Date: Wed, 21 Feb 2001 15:11:24 +0100
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org> <slrn997f03.h0.scarblac@pino.selwerd.nl>
Message-ID: <3a93c91b$1@netnews.web.de>

> It's too bad that lists don't have a .get() method (so x.get(3) would be
> equal to x[3]), because resultvector=map(sourcelist.get, elementlist)
> would be fastest (all done in C).

Oh, in fact there is a list get()-method. It is hidden within the
'operator'-module ;). The only disadvantage is that it expects the sequence
to fetch the item from in its first argument, so you have to make a list of
references to the sourcelist in order to apply the map()-function. If memory
consumption is not an issue, you can try to do the most work at C-speed by
using:

   import operator
   def makevector_operator(sourcelist,elementlist):
       return map(operator.getitem, [sourcelist]*len(elementlist),
elementlist)



Let's summarize and profile the different approaches (300000 items in
sourcelist, 70000 in elementlist with random indices):

   Runtime UsrTim SysTim Count  TimerName
  0.517912   0.50   0.01     1: makevector
  0.361448   0.33   0.01     1: makevector_listc
  0.222591   0.22   0.00     1: makevector_noappend
  0.221596   0.22   0.00     1: makevector_lambda
  0.183666   0.19   0.00     1: makevector_operator

def makevector(sourcelist,elementlist):
    resultvector = []
    for element in elementlist:
        resultvector.append(sourcelist[element])
    return resultvector

def makevector_noappend(sourcelist, elementlist):
    resultvector = elementlist[:]
    for i in range(len(elementlist)):
       resultvector[i] = sourcelist[elementlist[i]]
    return resultvector

def makevector_lambda(sourcelist,elementlist):
    return map(lambda i, l=sourcelist: l[i], elementlist)

def makevector_listc(sourcelist,elementlist):
    return [sourcelist[element] for element in elementlist]

def makevector_operator(sourcelist,elementlist):
    return map(operator.getitem, [sourcelist]*len(elementlist), elementlist)



Of course, now I'm very interested in the performance of a C-implementation
;).
Maybe it would help to have something like a "constant sequence" in C to
avoid generating the reference-list. Then, you could write:

listreferences = xReferenceList(sourcelist, len(elementlist))
return map(operator.getitem, listreferences, elementlist)


Greetings,
        Sebastian.








From djc at object-craft.com.au  Wed Feb 21 15:16:26 2001
From: djc at object-craft.com.au (Dave Cole)
Date: 22 Feb 2001 01:16:26 +1100
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org> <slrn997f03.h0.scarblac@pino.selwerd.nl> <3a93c91b$1@netnews.web.de>
Message-ID: <m3snl8nmth.fsf@vole.object-craft.com.au>

>>>>> "sd" == Sebastian Dauss <sebastian.dauss at web.de> writes:

sd> Of course, now I'm very interested in the performance of a
sd> C-implementation ;).  Maybe it would help to have something like a
sd> "constant sequence" in C to avoid generating the
sd> reference-list. Then, you could write:

I wrote a C version - in my extremely primitive testing it seemed to
be around 6 times faster.  It would make more sense to test all of the
techniques on the same data though.

- Dave

-- 
http://www.object-craft.com.au


From sebastian.dauss at web.de  Wed Feb 21 15:47:42 2001
From: sebastian.dauss at web.de (Sebastian Dauss)
Date: Wed, 21 Feb 2001 15:47:42 +0100
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org> <slrn997f03.h0.scarblac@pino.selwerd.nl> <3a93c91b$1@netnews.web.de>
Message-ID: <3a93d19f$1@netnews.web.de>

Sorry for replying to my own message, but exactly the function you need does
already exist! It is included within the "mxTools" package:
    http://www.lemburg.com/files/python/mxTools.html
and named "extract".

So, what you really want (I hope so! ;) ) is:

  import mxTools
  def makevector_mxTools(sourcelist,elementlist):
      return mxTools.extract(sourcelist, elementlist)

And for performance considerations, take this:

   Runtime UsrTim SysTim Count  TimerName
  0.474198   0.47   0.00     1: makevector
  0.347249   0.34   0.00     1: makevector_listc
  0.214850   0.21   0.00     1: makevector_noappend
  0.212961   0.21   0.01     1: makevector_lambda
  0.180423   0.18   0.00     1: makevector_operator
  0.060965   0.06   0.00     1: makevector_mxTools

So, using mxTools results in a runtime that is 3 times faster than the
(probably) best "map-magic" in pure python.


Hope that really helped ;),
    Sebastian.






From barry at digicool.com  Tue Feb 20 21:22:32 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Tue, 20 Feb 2001 15:22:32 -0500
Subject: Update to PEP 232
Message-ID: <14994.53768.767065.272158@anthem.wooz.org>

After some internal discussions amongst the Pythonlabbers, we've had
to make some updates to PEP 232, Function Attributes.  Attached is the
complete current PEP draft, also available at

    http://python.sourceforge.net/peps/pep-0232.html

The PEP has been moved back to Draft status, but will be Accepted and
Finalized for Python 2.1.  It will also be propagated forward for
Python 2.2 for the next step in implementation.

-Barry

-------------------- snip snip --------------------
PEP: 232
Title: Function Attributes
Version: $Revision: 1.6 $
Author: barry at digicool.com (Barry A. Warsaw)
Status: Draft
Type: Standards Track
Created: 02-Dec-2000
Python-Version: 2.1 / 2.2
Post-History: 20-Feb-2001


Introduction

    This PEP describes an extension to Python, adding attribute
    dictionaries to functions and methods.  This PEP tracks the status
    and ownership of this feature.  It contains a description of the
    feature and outlines changes necessary to support the feature.
    This PEP summarizes discussions held in mailing list forums, and
    provides URLs for further information, where appropriate.  The CVS
    revision history of this file contains the definitive historical
    record.


Background

    Functions already have a number of attributes, some of which are
    writable, e.g. func_doc, a.k.a. func.__doc__.  func_doc has the
    interesting property that there is special syntax in function (and
    method) definitions for implicitly setting the attribute.  This
    convenience has been exploited over and over again, overloading
    docstrings with additional semantics.

    For example, John Aycock has written a system where docstrings are
    used to define parsing rules[1].  Zope's ZPublisher ORB[2] uses
    docstrings to signal "publishable" methods, i.e. methods that can
    be called through the web.  And Tim Peters has developed a system
    called doctest[3], where docstrings actually contain unit tests.

    The problem with this approach is that the overloaded semantics
    may conflict with each other.  For example, if we wanted to add a
    doctest unit test to a Zope method that should not be publishable
    through the web.


Proposal

    This proposal adds a new dictionary to function objects, called
    func_dict (a.k.a. __dict__).  This dictionary can be set and get
    using ordinary attribute set and get syntax.

    Methods also gain `getter' syntax, and they currently access the
    attribute through the dictionary of the underlying function
    object.  It is not possible to set attributes on bound or unbound
    methods, except by doing so explicitly on the underlying function
    object.  See the `Future Directions' discussion below for
    approaches in subsequent versions of Python.

    A function object's __dict__ can also be set, but only to a
    dictionary object (i.e. setting __dict__ to UserDict raises a
    TypeError). 


Examples

    Here are some examples of what you can do with this feature.

        def a():
	    pass

	a.publish = 1
	a.unittest = '''...'''

	if a.publish:
	    print a()

	if hasattr(a, 'unittest'):
	    testframework.execute(a.unittest)

	class C:
	    def a(self):
	        'just a docstring'
            a.publish = 1

	c = C()
	if c.a.publish:
            publish(c.a())


Other Uses

    Paul Prescod enumerated a bunch of other uses:

    http://mail.python.org/pipermail/python-dev/2000-April/003364.html


Future Directions

    - A previous version of this PEP (and the accompanying
      implementation) allowed for both setter and getter of attributes
      on unbound methods, and only getter on bound methods.  A number
      of problems were discovered with this policy.

      Because method attributes were stored in the underlying
      function, this caused several potentially surprising results:

      class C:
          def a(self): pass

      c1 = C()
      c2 = C()
      c1.a.publish = 1
      # c2.a.publish would now be == 1 also!

      Because a change to `a' bound c1 also caused a change to `a'
      bound to c2, setting of attributes on bound methods was
      disallowed.  However, even allowing setting of attributes on
      unbound methods has its ambiguities:

      class D(C): pass
      class E(C): pass

      D.a.publish = 1
      # E.a.publish would now be == 1 also!

      For this reason, the current PEP disallows setting attributes on
      either bound or unbound methods, but does allow for getting
      attributes on either -- both return the attribute value on the
      underlying function object.

      The proposal for Python 2.2 is to implement setting (bound or
      unbound) method attributes by setting attributes on the instance
      or class, using special naming conventions.  I.e.

      class C:
          def a(self): pass

      C.a.publish = 1
      C.__a_publish__ == 1 # true

      c = C()
      c.a.publish = 2
      c.__a_publish__ == 2 # true

      d = C()
      d.__a_publish__ == 1 # true

      Here, a lookup on the instance would look to the instance's
      dictionary first, followed by a lookup on the class's
      dictionary, and finally a lookup on the function object's
      dictionary.

    - Currently, Python supports function attributes only on Python
      functions (i.e. those that are written in Python, not those that
      are built-in).  Should it be worthwhile, a separate patch can be
      crafted that will add function attributes to built-ins.

    - __doc__ is the only function attribute that currently has
      syntactic support for conveniently setting.  It may be
      worthwhile to eventually enhance the language for supporting
      easy function attribute setting.  Here are some syntaxes
      suggested by PEP reviewers:

      def a {
          'publish' : 1,
          'unittest': '''...''',
          }
          (args):
          # ...

      def a(args):
          """The usual docstring."""
          {'publish' : 1,
           'unittest': '''...''',
           # etc.
           }

      It isn't currently clear if special syntax is necessary or
      desirable.


Dissenting Opinion

    When this was discussed on the python-dev mailing list in April
    2000, a number of dissenting opinions were voiced.  For
    completeness, the discussion thread starts here:

    http://mail.python.org/pipermail/python-dev/2000-April/003361.html

    The dissenting arguments appear to fall under the following
    categories:

    - no clear purpose (what does it buy you?)
    - other ways to do it (e.g. mappings as class attributes)
    - useless until syntactic support is included

    Countering some of these arguments is the observation that with
    vanilla Python 2.0, __doc__ can in fact be set to any type of
    object, so some semblance of writable function attributes are
    already feasible.  But that approach is yet another corruption of
    __doc__.

    And while it is of course possible to add mappings to class
    objects (or in the case of function attributes, to the function's
    module), it is more difficult and less obvious how to extract the
    attribute values for inspection.

    Finally, it may be desirable to add syntactic support, much the
    same way that __doc__ syntactic support exists.  This can be
    considered separately from the ability to actually set and get
    function attributes.


Reference Implementation

    The reference implementation is available on SourceForge as a
    patch against the Python CVS tree (patch #103123).  This patch
    doesn't include the regrtest module and output file.  Those are
    available upon request.

   http://sourceforge.net/patch/?func=detailpatch&patch_id=103123&group_id=5470

    This patch has been applied and will become part of Python 2.1.


References

    [1] Aycock, "Compiling Little Languages in Python",
    http://www.foretec.com/python/workshops/1998-11/proceedings/papers/aycock-little/aycock-little.html

    [2] http://classic.zope.org:8080/Documentation/Reference/ORB

    [3] ftp://ftp.python.org/pub/python/contrib-09-Dec-1999/System/doctest.py
    

Copyright

    This document has been placed in the Public Domain.



Local Variables:
mode: indented-text
indent-tabs-mode: nil
End:



From tony at lsl.co.uk  Wed Feb 21 11:06:34 2001
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Wed, 21 Feb 2001 10:06:34 -0000
Subject: Update to PEP 232
In-Reply-To: <14994.53768.767065.272158@anthem.wooz.org>
Message-ID: <000901c09bed$f861d750$f05aa8c0@lslp7o.int.lsl.co.uk>

Small pedantry (there's another sort?)

I note that:

>     - __doc__ is the only function attribute that currently has
>       syntactic support for conveniently setting.  It may be
>       worthwhile to eventually enhance the language for supporting
>       easy function attribute setting.  Here are some syntaxes
>       suggested by PEP reviewers:
[...elided to save space!...]
>       It isn't currently clear if special syntax is necessary or
>       desirable.

has not been changed since the last version of the PEP. I suggest that
it be updated in two ways:

1. Clarify the final statement - I seem to have the impression (sorry,
can't find a message to back it up) that either the BDFL or Tim Peters
is very against anything other than the "simple" #f.a = 1# sort of
thing - unless I'm mischannelling (?) again.

2. Reference the thread/idea a little while back that ended with #def
f(a,b) having (publish=1)# - it's certainly no *worse* than the
proposals in the PEP! (Michael Hudson got as far as a patch, I think).

Tibs

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
then-again-i-confuse-easily<wink>-ly y'rs  - tim
That's true -- I usually feel confused after reading one of your
posts. - Aahz
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From guido at digicool.com  Thu Feb 22 01:12:11 2001
From: guido at digicool.com (Guido van Rossum)
Date: Wed, 21 Feb 2001 19:12:11 -0500
Subject: [Python-Dev] RE: Update to PEP 232
In-Reply-To: Your message of "Wed, 21 Feb 2001 10:06:34 GMT."
             <000901c09bed$f861d750$f05aa8c0@lslp7o.int.lsl.co.uk> 
References: <000901c09bed$f861d750$f05aa8c0@lslp7o.int.lsl.co.uk> 
Message-ID: <200102220012.TAA12047@cj20424-a.reston1.va.home.com>

> Small pedantry (there's another sort?)
> 
> I note that:
> 
> >     - __doc__ is the only function attribute that currently has
> >       syntactic support for conveniently setting.  It may be
> >       worthwhile to eventually enhance the language for supporting
> >       easy function attribute setting.  Here are some syntaxes
> >       suggested by PEP reviewers:
> [...elided to save space!...]
> >       It isn't currently clear if special syntax is necessary or
> >       desirable.
> 
> has not been changed since the last version of the PEP. I suggest that
> it be updated in two ways:
> 
> 1. Clarify the final statement - I seem to have the impression (sorry,
> can't find a message to back it up) that either the BDFL or Tim Peters
> is very against anything other than the "simple" #f.a = 1# sort of
> thing - unless I'm mischannelling (?) again.

Agreed.

> 2. Reference the thread/idea a little while back that ended with #def
> f(a,b) having (publish=1)# - it's certainly no *worse* than the
> proposals in the PEP! (Michael Hudson got as far as a patch, I think).

Sure, reference it.  It will never be added while I'm in charge
though.

--Guido van Rossum (home page: http://www.python.org/~guido/)



From tony at lsl.co.uk  Thu Feb 22 10:02:51 2001
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Thu, 22 Feb 2001 09:02:51 -0000
Subject: [Python-Dev] RE: Update to PEP 232
In-Reply-To: <200102220012.TAA12047@cj20424-a.reston1.va.home.com>
Message-ID: <001b01c09cae$3c3fa360$f05aa8c0@lslp7o.int.lsl.co.uk>

Guido responded to my points thus:
> > 1. Clarify the final statement - I seem to have the
> > impression (sorry, can't find a message to back it up)
> > that either the BDFL or Tim Peters is very against
> > anything other than the "simple" #f.a = 1# sort of
> > thing - unless I'm mischannelling (?) again.
>
> Agreed.

That's a relief - I obviously had "heard" right!

> > 2. Reference the thread/idea a little while back that ended
> > with #def > f(a,b) having (publish=1)# ...
>
> Sure, reference it.  It will never be added while I'm in charge
> though.

Well, I'd kind of assumed that, given my "memory" of the first point.
But of the schemes that won't be adopted, that's the one *I* preferred.

(my own sense of "locality" means that I would prefer to be placing
function attributes near the declaration of the function, especially
given my penchant for long docstrings which move the end of the function
off-screen. But then I haven't *used* them yet, and I assume this sort
of point has been taken into account. And anyway I definitely prefer
your sense of language design to mine).

Keep on trying not to get run over by buses, and thanks again for the
neat language,

Tibs

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
"Bounce with the bunny. Strut with the duck.
 Spin with the chickens now - CLUCK CLUCK CLUCK!"
BARNYARD DANCE! by Sandra Boynton
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From embed at geocities.com  Thu Feb 22 19:00:10 2001
From: embed at geocities.com (Warren Postma)
Date: Thu, 22 Feb 2001 13:00:10 -0500
Subject: [Python-Dev] RE: Update to PEP 232
References: <mailman.982834926.6065.python-list@python.org>
Message-ID: <rucl6.815$TW.6205@tor-nn1.netcom.ca>

Has the following been considered as a 2.1+ PEP 232 improvement:

def myfunc(param1,param2),{ dict here }:
        "doc here, or in dict"

The above would also allow functions to share a dictionary:

group_dict = { a:1 }
def myfunc1(param1,param2),group_dict:
def myfunc2(param1,param2),group_dict:

The  #def ...# crap looks horrible to me. Suddenly we have Comments Doing
Things.
Ugh!

Warren




From barry at digicool.com  Mon Feb 26 18:57:09 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Mon, 26 Feb 2001 12:57:09 -0500
Subject: [Python-Dev] RE: Update to PEP 232
References: <14994.53768.767065.272158@anthem.wooz.org>
	<000901c09bed$f861d750$f05aa8c0@lslp7o.int.lsl.co.uk>
Message-ID: <15002.39157.936988.699980@anthem.wooz.org>

>>>>> "TJI" == Tony J Ibbs <tony at lsl.co.uk> writes:

    TJI> 1. Clarify the final statement - I seem to have the
    TJI> impression (sorry, can't find a message to back it up) that
    TJI> either the BDFL or Tim Peters is very against anything other
    TJI> than the "simple" #f.a = 1# sort of thing - unless I'm
    TJI> mischannelling (?) again.

>From the next draft of the PEP:

"The BDFL is currently against any such special syntactic support for
setting arbitrary function attributes."

    TJI> 2. Reference the thread/idea a little while back that ended
    TJI> with #def f(a,b) having (publish=1)# - it's certainly no
    TJI> *worse* than the proposals in the PEP! (Michael Hudson got as
    TJI> far as a patch, I think).

Any possibility that you have a url for this reference?

-Barry



From syver at NOSPAMcyberwatcher.com  Tue Feb 20 22:07:10 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Tue, 20 Feb 2001 22:07:10 +0100
Subject: pythoncom.Missing vs. pythoncom.Empty
Message-ID: <96umec$pks$1@troll.powertech.no>

Hi, what's the deal with pythoncom.Missing/Empty?

I've found while working with the MSMQ (MessageQueue) through COM on Win32
that default arguments don't work unless I change the definitions of
defaultNamedOptArg to pythoncom.Empty.

I am aware that the generated source states that there might be a reason to
try this, but why, and what is the difference between them? It would also be
very good to be able to specify in the gencache statement what to use for
the various values so that one wouldn't have to use custom edited genpy
files.




From MarkH at ActiveState.com  Tue Feb 20 23:23:16 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 20 Feb 2001 22:23:16 GMT
Subject: pythoncom.Missing vs. pythoncom.Empty
References: <96umec$pks$1@troll.powertech.no>
Message-ID: <3A92ED79.101@ActiveState.com>

Syver Enstad wrote:

> Hi, what's the deal with pythoncom.Missing/Empty?
> 
> I've found while working with the MSMQ (MessageQueue) through COM on Win32
> that default arguments don't work unless I change the definitions of
> defaultNamedOptArg to pythoncom.Empty.
> 
> I am aware that the generated source states that there might be a reason to
> try this, but why, and what is the difference between them? It would also be
> very good to be able to specify in the gencache statement what to use for
> the various values so that one wouldn't have to use custom edited genpy
> files.

The different is in the way the args are presented.  "Empty" means the 
arg is presented with a type of VT_EMPTY.  Missing means the arg is not 
even presented to the object.

I have heard this reported before.  It is "Missing" simply due to Office 
95!!  I should change this and see who screams - my guess is noone any more!

Mark.
-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From mkx at excite.com  Tue Feb 20 23:34:22 2001
From: mkx at excite.com (mkx at excite.com)
Date: Tue, 20 Feb 2001 17:34:22 -0500
Subject: ZLIB decompressing only portion of data
Message-ID: <s1s59tojmev7822ndpmuhmg4d5gepipq39@4ax.com>

I am experimenting with the ZLIB module to decompress some .ZLB files
I have acquired. The script below properly decompresses only the first
segment of the compressed file. For example, a 10KB .ZLB comressed
file only returns a 546 character string from the example below.

What have I done wrong?


###
import os, zlib

fd=os.open('d:/temp/zlb/snippets.zlb', os.O_RDONLY | os.O_BINARY )

ch=os.read(fd,1)
while ch:
    zlb=zlb+ch
    ch=os.read(fd,1)
os.close(fd)

print len(zlb)

str = zlib.decompress(zlb)
print str
print len(str)





From mkx at excite.com  Tue Feb 20 23:52:23 2001
From: mkx at excite.com (mkx at excite.com)
Date: Tue, 20 Feb 2001 17:52:23 -0500
Subject: ZLIB decompressing only portion of data
References: <s1s59tojmev7822ndpmuhmg4d5gepipq39@4ax.com>
Message-ID: <h5t59tsb1bgg6j7ids19vnljabs0mme9ds@4ax.com>

Of course, there is a better way to implement my test script (see
below), but I still get the same results. The input sample is 9,859
bytes long, where the string returned by the decompress method is only
546 bytes in length.


###

import os, zlib

fd=os.open('d:/temp/zlb/snippets.zlb', os.O_RDONLY | os.O_BINARY )
dco=zlib.decompressobj(15)

str=dco.decompress(os.read(fd,16384))

print len(str)
#print str



From amk at mira.erols.com  Wed Feb 21 01:51:16 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 21 Feb 2001 00:51:16 GMT
Subject: ZLIB decompressing only portion of data
References: <s1s59tojmev7822ndpmuhmg4d5gepipq39@4ax.com> <h5t59tsb1bgg6j7ids19vnljabs0mme9ds@4ax.com>
Message-ID: <slrn996537.h0.amk@207-172-36-141.s141.tnt6.ann.va.dialup.rcn.com>

On Tue, 20 Feb 2001 17:52:23 -0500, mkx at excite.com <mkx at excite.com> wrote:
>str=dco.decompress(os.read(fd,16384))

zlib often buffers data internally; try adding 'str = str + dco.flush()'
and see if you get more output.  

--amk


From fredrik at pythonware.com  Wed Feb 21 07:23:14 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 06:23:14 GMT
Subject: ZLIB decompressing only portion of data
References: <s1s59tojmev7822ndpmuhmg4d5gepipq39@4ax.com> <h5t59tsb1bgg6j7ids19vnljabs0mme9ds@4ax.com>
Message-ID: <m9Jk6.14719$Qb7.2357570@newsb.telia.net>

mkx at excite.com wrote:
> Of course, there is a better way to implement my test script (see
> below), but I still get the same results. The input sample is 9,859
> bytes long, where the string returned by the decompress method is only
> 546 bytes in length.

what's wrong with using "open" like everyone else?

> import os, zlib
>
> fd=os.open('d:/temp/zlb/snippets.zlb', os.O_RDONLY | os.O_BINARY )

file = open("d:/temp/zlb/snippets.zlb", "rb")

> dco=zlib.decompressobj(15)

dco = zlib.decompressobj() # 15 is default

> str=dco.decompress(os.read(fd,16384))

str = dco.decompress(file.read(16384))
str = str + dco.flush() # get rest of data

> print len(str)
> #print str

for more info on decompress/flush, see the docs (under
"decompression objects"):

http://www.python.org/doc/current/lib/module-zlib.html

(and of course, if the files are smaller than a couple of megabytes
uncompressed, it's better to use zlib.decompress...)

Cheers /F




From mkx at excite.com  Wed Feb 21 17:03:06 2001
From: mkx at excite.com (mkx at excite.com)
Date: Wed, 21 Feb 2001 11:03:06 -0500
Subject: ZLIB decompressing only portion of data
References: <s1s59tojmev7822ndpmuhmg4d5gepipq39@4ax.com> <h5t59tsb1bgg6j7ids19vnljabs0mme9ds@4ax.com> <m9Jk6.14719$Qb7.2357570@newsb.telia.net>
Message-ID: <71p79tk1c3448i50k9cdjc9nlj0dl898o3@4ax.com>

On Wed, 21 Feb 2001 06:23:14 GMT, "Fredrik Lundh"
<fredrik at pythonware.com> wrote:

>mkx at excite.com wrote:
>> Of course, there is a better way to implement my test script (see
>> below), but I still get the same results. The input sample is 9,859
>> bytes long, where the string returned by the decompress method is only
>> 546 bytes in length.
>
>what's wrong with using "open" like everyone else?

Nothing! Python is new to me, and sometimes I take the long way
around. In this case, I tried "open", but in the process of
determining that the file needed to be opened as Binary, I found my
self at os.open, instead of open(,"rb"). 

<...>

>str = dco.decompress(file.read(16384))
>str = str + dco.flush() # get rest of data

Ah...Through trial and error, I was only successful (partially) by
using dco.unused_data, but then I had to recreate the decompress
object with each iteration when .unusued_data was present.

It appears that the .decompress method is only decompressing the first
"segment" of the data. At this point I am unsure what delimits the
segments.

>for more info on decompress/flush, see the docs (under
>"decompression objects"):
>
>http://www.python.org/doc/current/lib/module-zlib.html
>
>(and of course, if the files are smaller than a couple of megabytes
>uncompressed, it's better to use zlib.decompress...)

That is where I started, but then I ran in to the problem with only
the first segment of the data being uncompressed. Using the decompress
object is the only thing I have working so far.



From mkx at excite.com  Wed Feb 21 17:24:26 2001
From: mkx at excite.com (mkx at excite.com)
Date: Wed, 21 Feb 2001 11:24:26 -0500
Subject: ZLIB decompressing only portion of data
References: <s1s59tojmev7822ndpmuhmg4d5gepipq39@4ax.com> <h5t59tsb1bgg6j7ids19vnljabs0mme9ds@4ax.com> <m9Jk6.14719$Qb7.2357570@newsb.telia.net>
Message-ID: <3lq79t4b000o9vvndsvjm9lllfon3bdvod@4ax.com>

On Wed, 21 Feb 2001 06:23:14 GMT, "Fredrik Lundh"
<fredrik at pythonware.com> wrote:

>str = dco.decompress(file.read(16384))
>str = str + dco.flush() # get rest of data

This methodology still only returns the first segment of uncompressed
data.

So far, only the following works, returning about 12K of uncompressed
data, before returning an Error -3 while decompressing: unknown
compression method.

###

import zlib

fd=open('d:/temp/zlb/snippets.zlb', 'rb' )
of=open('d:/temp/zlb/snippets.dat','w')

dco = zlib.decompressobj()
str = dco.decompress(fd.read(16384))
of.write(str)

while dco.unused_data:
    ustr = dco.unused_data
    dco=zlib.decompressobj()	# Having to recreate the dco object
				# each time, otherwise further data 
				# not returned
    str=dco.decompress(ustr)
    of.write(str)




From paul at fxtech.com  Wed Feb 21 00:04:21 2001
From: paul at fxtech.com (Paul Miller)
Date: Tue, 20 Feb 2001 17:04:21 -0600
Subject: docs on using "new" PyTypeObject entries?
Message-ID: <3A92F7F5.F930DC55@fxtech.com>

Where can I find some documentation on how to use the tp_traverse,
tp_clear, tp_getattro, etc slots in the PyTypeObject? I just discovered
them and don't know what they do.


From paul at fxtech.com  Wed Feb 21 00:25:25 2001
From: paul at fxtech.com (Paul Miller)
Date: Tue, 20 Feb 2001 17:25:25 -0600
Subject: give an extension type a __dict__?
Message-ID: <3A92FCE5.9CD17BE5@fxtech.com>

Is it possible to give an extension type a __dict__ to store
type-specific constants? Assume we create a Point object in C:

PyTypeObject point_type_info = 
{
	PyObject_HEAD_INIT(&PyType_Type)
	0,					// ob_size
	"Point",				// tp_name
	sizeof(PointType),			// tp_basicsize
	0,					// tp_itemsize

	etc...

In Python, I can create one of these:

	from point import Point
	p = Point()

But since my Point type *acts* like I class, I'd like to scope some
constants into its namespace, so I can do this:

	p = Point.ZERO

I realize I can put constants at the module scope in the module's
dictionary, but if I have many extension types, the global namespace
would be polluted.

Is this even possible?


From root at rainerdeyke.com  Wed Feb 21 04:49:13 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Wed, 21 Feb 2001 03:49:13 GMT
Subject: give an extension type a __dict__?
References: <3A92FCE5.9CD17BE5@fxtech.com>
Message-ID: <ZUGk6.271907$ge4.94357411@news2.rdc2.tx.home.com>

"Paul Miller" <paul at fxtech.com> wrote in message
news:3A92FCE5.9CD17BE5 at fxtech.com...
> Is it possible to give an extension type a __dict__ to store
> type-specific constants? Assume we create a Point object in C:
>
> PyTypeObject point_type_info =
> {
> PyObject_HEAD_INIT(&PyType_Type)
> 0, // ob_size
> "Point", // tp_name
> sizeof(PointType), // tp_basicsize
> 0, // tp_itemsize
>
> etc...
>
> In Python, I can create one of these:
>
> from point import Point
> p = Point()
>
> But since my Point type *acts* like I class, I'd like to scope some
> constants into its namespace, so I can do this:

You should realize that the name 'Point' within Python does not refer to a
type, but a function (type 'builtin_function_or_method').  You can easily
(or maybe not so easily) modify the extension module so that 'Point' is
instead an object that acts like a function but also has additional
properties, such as the ability to hold attributes.  This is totally
different from modifying the actual type 'Point'.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From tjg at exceptionalminds.com  Wed Feb 21 00:42:48 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Tue, 20 Feb 2001 15:42:48 -0800
Subject: Case insensitive re.sub()
Message-ID: <20010220154248.L15226@trufflehunter.avalongroup.net>

Hi all,

I am working on a little project where it would be really nice
to be able to do a case insensitive re.sub(). It appears that
re.sub() does not allow a flags argument. Is it possible, or do
I have to roll my own?

Thanks.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  35 days  3:54 hours ago<<



From aahz at panix.com  Wed Feb 21 01:13:44 2001
From: aahz at panix.com (Aahz Maruch)
Date: 20 Feb 2001 16:13:44 -0800
Subject: Case insensitive re.sub()
References: <mailman.982712484.28722.python-list@python.org>
Message-ID: <96v17o$h53$1@panix6.panix.com>

In article <mailman.982712484.28722.python-list at python.org>,
Timothy Grant  <tjg at exceptionalminds.com> wrote:
>
>I am working on a little project where it would be really nice
>to be able to do a case insensitive re.sub(). It appears that
>re.sub() does not allow a flags argument. Is it possible, or do
>I have to roll my own?

import re
regex = re.compile('foo', re.I)
print regex.sub('bar','FOOBAR')
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From tjg at exceptionalminds.com  Wed Feb 21 02:09:35 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Tue, 20 Feb 2001 17:09:35 -0800
Subject: Case insensitive re.sub()
In-Reply-To: <96v17o$h53$1@panix6.panix.com>; from aahz@panix.com on Tue, Feb 20, 2001 at 04:13:44PM -0800
References: <mailman.982712484.28722.python-list@python.org> <96v17o$h53$1@panix6.panix.com>
Message-ID: <20010220170935.S15226@trufflehunter.avalongroup.net>

Thanks Aazh.

That did the trick beautifully. 



On Tue, Feb 20, 2001 at 04:13:44PM -0800, Aahz Maruch wrote:
> In article <mailman.982712484.28722.python-list at python.org>,
> Timothy Grant  <tjg at exceptionalminds.com> wrote:
> >
> >I am working on a little project where it would be really nice
> >to be able to do a case insensitive re.sub(). It appears that
> >re.sub() does not allow a flags argument. Is it possible, or do
> >I have to roll my own?
> 
> import re
> regex = re.compile('foo', re.I)
> print regex.sub('bar','FOOBAR')

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  35 days  5:20 hours ago<<



From aleaxit at yahoo.com  Wed Feb 21 10:34:59 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 10:34:59 +0100
Subject: Case insensitive re.sub()
References: <mailman.982712484.28722.python-list@python.org>
Message-ID: <97024702ogj@news1.newsguy.com>

"Timothy Grant" <tjg at exceptionalminds.com> wrote in message
news:mailman.982712484.28722.python-list at python.org...
> Hi all,
>
> I am working on a little project where it would be really nice
> to be able to do a case insensitive re.sub(). It appears that
> re.sub() does not allow a flags argument. Is it possible, or do
> I have to roll my own?

Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import re
>>> cre=re.compile('pip',re.IGNORECASE)
>>> cre.sub("A pip and a PIPpop", 'plup')
'plup'
>>> cre.sub('plup', "A pip and a PIPpop")
'A plup and a pluppop'
>>>


Alex





From fredrik at pythonware.com  Wed Feb 21 11:29:27 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 10:29:27 GMT
Subject: Case insensitive re.sub()
References: <mailman.982712484.28722.python-list@python.org>
Message-ID: <bMMk6.14747$Qb7.2360418@newsb.telia.net>

Timothy Grant wrote:
> I am working on a little project where it would be really nice
> to be able to do a case insensitive re.sub(). It appears that
> re.sub() does not allow a flags argument. Is it possible, or do
> I have to roll my own?

add the flags to the pattern instead: "(?i)pattern" makes
"pattern" case-insensitive.

(note that there are issues with uppercase characters in case-
insensitive patterns in 2.0; to be on the safe side, only use
lowercase characters in case-insensitive patterns.  or use 2.1)

Cheers /F




From aahz at panix.com  Wed Feb 21 16:00:39 2001
From: aahz at panix.com (Aahz Maruch)
Date: 21 Feb 2001 07:00:39 -0800
Subject: Case insensitive re.sub()
References: <mailman.982712484.28722.python-list@python.org> <bMMk6.14747$Qb7.2360418@newsb.telia.net>
Message-ID: <970l6n$1bb$1@panix3.panix.com>

In article <bMMk6.14747$Qb7.2360418 at newsb.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
>
>(note that there are issues with uppercase characters in case-
>insensitive patterns in 2.0; to be on the safe side, only use
>lowercase characters in case-insensitive patterns.  or use 2.1)

Or install the new sre package you just released?
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From insanc at cc.gatech.edu  Wed Feb 21 00:55:54 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 20 Feb 2001 23:55:54 GMT
Subject: locking files
Message-ID: <96v06a$l16$1@solaria.cc.gatech.edu>

is there anyway to lock a file so that two different programs cannot access
the file at the same time. ie, my python module is editting file foo, and
at the same time user x (or program x) tries to read from it. is there 
anyway that the python module can prevent x from accessing the file for
a given amount of time?

-- 
---
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From insanc at cc.gatech.edu  Wed Feb 21 15:48:34 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 21 Feb 2001 14:48:34 GMT
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu>
Message-ID: <970kg2$cn0$1@solaria.cc.gatech.edu>

Joseph Holland King <insanc at cc.gatech.edu> wrote:

: is there anyway to lock a file so that two different programs cannot access
: the file at the same time. ie, my python module is editting file foo, and
: at the same time user x (or program x) tries to read from it. is there 
: anyway that the python module can prevent x from accessing the file for
: a given amount of time?

to clarify: i am on a unix system, i have tried to use flock however when i
locked the file i was still able to read, write and copy the file, none of 
which i want to happen while the file is locked. while using the flock 
method i was using LOCK_EX. also the third party program is not going to be
controlled, written, or have anything to do with mine. am i just not using
flock properly or is there something else that might do the job? thank you.

-- 
---
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From dnew at san.rr.com  Wed Feb 21 19:18:34 2001
From: dnew at san.rr.com (Darren New)
Date: Wed, 21 Feb 2001 18:18:34 GMT
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu>
Message-ID: <3A94067B.AB98CE53@san.rr.com>

Joseph Holland King wrote:
> flock properly or is there something else that might do the job? thank you.

On some UNIXes, there's a bit you set on the file (like, group sticky bit or
something) that means "No, really, lock the @*$& file already!"  You may
need to set this. Try

chmod g-x my_file
chmod g+l my_file

and see if the locking continues to fail. 

-- 
Darren New / Senior MTS & Free Radical / Invisible Worlds Inc.
San Diego, CA, USA (PST).  Cryptokeys on demand.
                 Ignorance can be cured. Naivety cures itself.


From donn at u.washington.edu  Wed Feb 21 19:40:57 2001
From: donn at u.washington.edu (Donn Cave)
Date: 21 Feb 2001 18:40:57 GMT
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu>
Message-ID: <97123p$ojg$1@nntp6.u.washington.edu>

Quoth Joseph Holland King <insanc at cc.gatech.edu>:
| Joseph Holland King <insanc at cc.gatech.edu> wrote:
| : is there anyway to lock a file so that two different programs cannot access
| : the file at the same time. ie, my python module is editting file foo, and
| : at the same time user x (or program x) tries to read from it. is there 
| : anyway that the python module can prevent x from accessing the file for
| : a given amount of time?
|
| to clarify: i am on a unix system, i have tried to use flock however when i
| locked the file i was still able to read, write and copy the file, none of 
| which i want to happen while the file is locked. while using the flock 
| method i was using LOCK_EX. also the third party program is not going to be
| controlled, written, or have anything to do with mine. am i just not using
| flock properly or is there something else that might do the job? thank you.

Basically, that's how file locking works.  It's advisory, for the benefit
of software components that are designed to work together.

But there is a variation that you can find supported on some filesystems,
where in combination with an obscure permission mode on the file the lock
becomes mandatory and effective against any and all I/O.  The permission
trick is setgid & not group execute, i.e., (oldperm | 02000) & ~010.  It
works for me on Digital UNIX local (AdvFS) filesystem.  I didn't try on
NFS, but from the documentation I doubt that it will work there.  It does
not work on NetBSD 1.5 local filesystem.  Where it does work, you must
use fcntl() or lockf() to obtain the lock, not a genuine flock().

While I'm here, let me add a little note about flock(), for the benefit
of those whose UNIX platforms don't have the deluxe man pages that you
get on NetBSD for example.  There are two kinds of locks, one obtained
by Berkeley flock() and the other by POSIX 1003.1 fcntl().  The lockf()
function is a different, X/Open interface that interoperates with fcntl().
Many modern platforms do not support a genuine flock().  Some of them
provide an flock() function that is actually implemented with fcntl(),
which means that the semantics of the lock are per fcntl() and not
flock().  Python follows this trend and does the same thing if there's
no flock(2), so it's up to the programmer to find out what kind of lock
fcntl.flock() really gets, if the difference is important.

	Donn Cave, donn at u.washington.edu


From grante at visi.com  Thu Feb 22 00:44:02 2001
From: grante at visi.com (Grant Edwards)
Date: Wed, 21 Feb 2001 23:44:02 GMT
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu> <97123p$ojg$1@nntp6.u.washington.edu>
Message-ID: <slrn998km2.cbp.grante@isis.visi.com>

In article <97123p$ojg$1 at nntp6.u.washington.edu>, Donn Cave wrote:

>| to clarify: i am on a unix system, i have tried to use flock however when i
>| locked the file i was still able to read, write and copy the file, none of 
>| which i want to happen while the file is locked. while using the flock 
>| method i was using LOCK_EX. also the third party program is not going to be
>| controlled, written, or have anything to do with mine. am i just not using
>| flock properly or is there something else that might do the job? thank you.
>
>Basically, that's how file locking works.  

IOW, it doesn't.  ;)

Really -- file locking under Unix is a royal mess and always has been.
Throw NFS into the picture and it's pretty much hopeless.

>It's advisory, for the benefit of software components that are designed to
>work together.

And then it's rarely portable, and there are about seven people on the
planet that understand what syscalls work on what systems and how to use
them.

I am not among those seven.

Here's a quick hack: When you want to "lock" the file, rename it.  I think
that on most Unices and filesystems a rename is an atomic operation (not
sure, about that).  I don't know how you're going to tell if the other
program is using the file.

>But there is a variation that you can find supported on some filesystems,
>where in combination with an obscure permission mode on the file the lock
>becomes mandatory and effective against any and all I/O.  The permission
[...]
 flock(2), so it's up to the programmer to find out what kind of lock
>fcntl.flock() really gets, if the difference is important.

Donn may be among the seven.  ;)

-- 
Grant Edwards                   grante             Yow!  I hope the
                                  at               "Eurythmics" practice birth
                               visi.com            control...


From chris.gonnerman at usa.net  Fri Feb 23 04:55:00 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 22 Feb 2001 21:55:00 -0600
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu> <97123p$ojg$1@nntp6.u.washington.edu> <slrn998km2.cbp.grante@isis.visi.com>
Message-ID: <00d301c09d4c$6ade4200$a100000a@local>

----- Original Message -----
From: "Grant Edwards" <grante at visi.com>
Subject: Re: locking files
> Here's a quick hack: When you want to "lock" the file, rename it.  I think
> that on most Unices and filesystems a rename is an atomic operation (not
> sure, about that).  I don't know how you're going to tell if the other
> program is using the file.

Renaming on Windows may be atomic, but on Unix/Linux/BSD etc.
it is the combination of link() and unlink() system calls, so that for a
moment the file has two names.  Also, on Unixoid OS's, if process A
has a file open and process B renames it, process A still has the same file
(inode) open and doesn't know about the rename... EXCEPT if the
file is on a remote NFS file system, in which case process A will either
start getting errors or have "undefined behavior" when it tries to use the
"open" file.  On Windoze, you may be prevented from renaming the file
if it is open in another process, or you may succeed and cause the other
process to have errors.

In other words, renaming is not necessarily a solution.






From erno-news at erno.iki.fi  Fri Feb 23 05:18:24 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 23 Feb 2001 06:18:24 +0200
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu> <97123p$ojg$1@nntp6.u.washington.edu> <slrn998km2.cbp.grante@isis.visi.com> <mailman.982900205.13479.python-list@python.org>
Message-ID: <ku8zmyrq0f.fsf@lasipalatsi.fi>

In article <mailman.982900205.13479.python-list at python.org>, "Chris
Gonnerman" <chris.gonnerman at usa.net> writes:

| on Unix/Linux/BSD etc.  it is the combination of link() and unlink()
| system calls, so that for a moment the file has two names.

at least on linux rename is a syscall, and the man page seems
to say that the operation will be atomic if the destination
path doesn't exist.

| Also, on Unixoid OS's, if process A
| has a file open and process B renames it, process A still has the same file
| (inode) open and doesn't know about the rename... EXCEPT if the
| file is on a remote NFS file system, in which case process A will either
| start getting errors or have "undefined behavior" when it tries to use the
| "open" file.  On Windoze, you may be prevented from renaming the file
| if it is open in another process, or you may succeed and cause the other
| process to have errors.

well, nfs doesn't pretend to offer unix filesystem semantics anyway,
usually it's not necessary to worry about how things work with it.

  -- erno


From chris.gonnerman at usa.net  Fri Feb 23 05:47:19 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 22 Feb 2001 22:47:19 -0600
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu> <97123p$ojg$1@nntp6.u.washington.edu> <slrn998km2.cbp.grante@isis.visi.com> <mailman.982900205.13479.python-list@python.org> <ku8zmyrq0f.fsf@lasipalatsi.fi>
Message-ID: <017201c09d53$b5463ee0$a100000a@local>

----- Original Message -----
From: "Erno Kuusela" <erno-news at erno.iki.fi>
Subject: Re: locking files
> at least on linux rename is a syscall, and the man page seems
> to say that the operation will be atomic if the destination
> path doesn't exist.

Huh.  You learn something new every day.

> well, nfs doesn't pretend to offer unix filesystem semantics anyway,
> usually it's not necessary to worry about how things work with it.

Usually.  But if you plan to use renaming to sub for locking, you need to
know this.





From clee at gnwy100.wuh.wustl.edu  Wed Feb 21 20:27:45 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 21 Feb 2001 13:27:45 -0600
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu>
Message-ID: <7ulmqzu98u.fsf@gnwy100.wuh.wustl.edu>

Joseph Holland King <insanc at cc.gatech.edu> wrote:
: is there anyway to lock a file so that two different programs cannot access
: the file at the same time. ie, my python module is editting file foo, and
: at the same time user x (or program x) tries to read from it. is there 
: anyway that the python module can prevent x from accessing the file for
: a given amount of time?

: to clarify: i am on a unix system, i have tried to use flock however when i
: locked the file i was still able to read, write and copy the file, none of 
: which i want to happen while the file is locked. while using the flock 
: method i was using LOCK_EX. also the third party program is not going to be
: controlled, written, or have anything to do with mine. am i just not using
: flock properly or is there something else that might do the job? thank you.

flock() is an advisory lock, if you don't have cooperation from the other
processes, and if your system supports the SysV interface you can enable
mandatory locks.  "man lockf" or "man fcntl" and search for information on SysV
mandatory locking on your system.

Files are marked as candidates for mandatory file locking by setting the
file mode with SGID but without the "group execute" bits, (chmod g+s; chmod
g-x).  Enabling mandatory locking is system dependent, but you need to use
fnctl/lockf, flock is always advisory.

Under linux, I can do it like this:
[as root]
# enable file locking by remounting /home file system w/ mand flag
mount -o remount -o mand /home   

[as normal user]
> chmod g+s tmp; chmod g-x tmp
> python
>> import fcntl, FCNTL
>> fp = open("tmp", "rw")
>> fd = fp.fileno()
>> fcntl.lockf(fd, FCNTL.F_LOCK)  # secures the lock, prevents
                                  # reading/writing 




From clee at gnwy100.wuh.wustl.edu  Wed Feb 21 20:42:36 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 21 Feb 2001 13:42:36 -0600
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu> <7ulmqzu98u.fsf@gnwy100.wuh.wustl.edu>
Message-ID: <7uitm3u8k3.fsf@gnwy100.wuh.wustl.edu>

>>>>> "Christopher" == Christopher Lee <clee at gnwy100.wuh.wustl.edu> writes:

    >> chmod g+s tmp; chmod g-x tmp python
    >>> import fcntl, FCNTL fp = open("tmp", "rw") fd = fp.fileno()
    >>> fcntl.lockf(fd, FCNTL.F_LOCK) # secures the lock, prevents

Oops, FCNTL.F_LOCK is wrong here (although when I typed it in the example
worked by coincidence).  Looking at the python source of fcntlmodule.c,
it's actually asking for fcntl's LOCK_EX, LOCK_SH, LOCK_UN, or LOCK_NB.

-chris


#ifndef LOCK_SH
#define LOCK_SH		1	/* shared lock */
#define LOCK_EX		2	/* exclusive lock */
#define LOCK_NB		4	/* don't block when locking */
#define LOCK_UN		8	/* unlock */


From insanc at cc.gatech.edu  Wed Feb 21 21:50:55 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 21 Feb 2001 20:50:55 GMT
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu> <7ulmqzu98u.fsf@gnwy100.wuh.wustl.edu>
Message-ID: <9719nf$lmf$1@solaria.cc.gatech.edu>

Christopher Lee <clee at gnwy100.wuh.wustl.edu> wrote:
: [as normal user]
:> chmod g+s tmp; chmod g-x tmp
:> python
:>> import fcntl, FCNTL
:>> fp = open("tmp", "rw")
:>> fd = fp.fileno()
:>> fcntl.lockf(fd, FCNTL.F_LOCK)  # secures the lock, prevents
:                                   # reading/writing 

following this i tried this code:

import FCNTL, fcntl

int_temp = 999999
fd = open("ftbl", "wr")
fd1 = fd.fileno()
fcntl.lockf(fd1, FCNTL.LOCK_EX)
while(int_temp > 0):
        int_temp = int_temp - 1
fcntl.lockf(fd1, FCNTL.LOCK_UN)
fd.close()

and then after did the following:
{gypsy:gte743n:116} cat mbox  > ftbl
{gypsy:gte743n:117} python testlock.py
{gypsy:gte743n:118} more ftbl

as you can see i cated my mailbox into ftbl and the ran the script the script
however empties the file. is there something in my code that is causing
this or is there something i am not doing? thank you.

-- 
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From clee at gnwy100.wuh.wustl.edu  Wed Feb 21 22:34:34 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 21 Feb 2001 15:34:34 -0600
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu> <7ulmqzu98u.fsf@gnwy100.wuh.wustl.edu> <9719nf$lmf$1@solaria.cc.gatech.edu>
Message-ID: <7ubsrvu3dh.fsf@gnwy100.wuh.wustl.edu>

chris> fp = open("tmp", "rw")

oops, indicision on my part.  see below

> fd = open("ftbl", "wr")

The "w" truncates the file.  You probably want either "r" or "r+" or "a+"
see the python library reference on open or the C fopen call.

try

fd = open("ftbl", "r+")

-chris


From ron_clarke at hp.com  Wed Feb 21 01:04:24 2001
From: ron_clarke at hp.com (Ron Clarke)
Date: Tue, 20 Feb 2001 16:04:24 -0800
Subject: Object type from a Python extension
Message-ID: <96v476$72k$1@news.vcd.hp.com>

I am creating a shared library (Win32 DLL and Linux '.so') using C++. I have
it all wired up and working from both environments, but I cannot understand
how to get Python scripts to recognize an object created in the C++ code.

For example:

class something( )
{
    //some variables
    //some public methods
};

something* ps = new something()
//...do stuff with ps object

//Now when I pass the ps object back to Python using Py_BuildValue(...), the
Python script knows about the object, but does not know it is of type
'something'. I'm obviously missing some connection, but have been through
all the extending and embedding documentation, along with a few Python
books, and don't see it.

Any ideas?

Thanks.




From rwklee at home.com  Wed Feb 21 01:05:27 2001
From: rwklee at home.com (Rick Lee)
Date: Wed, 21 Feb 2001 00:05:27 GMT
Subject: Is there an easier-to-use module than ftplib?
Message-ID: <3A930645.761A82C8@home.com>

Is there an easier-to-use module than ftplib in the standard library,
for both uploading and downloading files?  I can't use urllib because it
does not support uploading, it seems; and I am hoping I don't have to
learn FTP itself, which is required somewhat by ftplib.

- Rick Lee



From othello at javanet.com  Wed Feb 21 01:36:37 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 20 Feb 2001 19:36:37 -0500
Subject: Is there an easier-to-use module than ftplib?
References: <3A930645.761A82C8@home.com>
Message-ID: <3A930D95.2DC3B98C@javanet.com>

Rick Lee wrote:

> Is there an easier-to-use module than ftplib in the standard library,
> for both uploading and downloading files?  I can't use urllib because it
> does not support uploading, it seems; and I am hoping I don't have to
> learn FTP itself, which is required somewhat by ftplib.
>
> - Rick Lee

I felt frustrated with ftplib.py at first because the documentation
pre-supposes that you already know the ftp commands.

I made the following quick notes which show an examples that cover
95% of what you might want to do with ftp:

from ftplib import *
f = FTP('ftp.javanet.com','username','password')
f.retrlines('LIST')  # or f.dir() to list the directory
f.cwd('public_html')         # change directory

r = open('random.htm','w')
f.retrlines('retr random.htm',r.write)         # download a text file
r.close()

r = open('quoter.pyw','rb')                    # upload a binary file
f.storbinary('STOR quoter.pyw',r,512)
r.close()

f.rename('quoter.pyw','q.pyw')              # rename a file
f.delete('q.pyw')                                # delete a file

f.quit()                                        # terminate the session


Hope these note help,

Raymond





From rwklee at home.com  Wed Feb 21 03:09:02 2001
From: rwklee at home.com (Rick Lee)
Date: Wed, 21 Feb 2001 02:09:02 GMT
Subject: Is there an easier-to-use module than ftplib?
References: <3A930645.761A82C8@home.com> <3A930D95.2DC3B98C@javanet.com>
Message-ID: <3A932336.71661B16@home.com>

Thanks!  That looks simple enough.  Do I have to worry about active vs.
passive mode?

Raymond Hettinger wrote:

> Rick Lee wrote:
>
> > Is there an easier-to-use module than ftplib in the standard library,
> > for both uploading and downloading files?  I can't use urllib because it
> > does not support uploading, it seems; and I am hoping I don't have to
> > learn FTP itself, which is required somewhat by ftplib.
> >
> > - Rick Lee
>
> I felt frustrated with ftplib.py at first because the documentation
> pre-supposes that you already know the ftp commands.
>
> I made the following quick notes which show an examples that cover
> 95% of what you might want to do with ftp:
>
> from ftplib import *
> f = FTP('ftp.javanet.com','username','password')
> f.retrlines('LIST')  # or f.dir() to list the directory
> f.cwd('public_html')         # change directory
>
> r = open('random.htm','w')
> f.retrlines('retr random.htm',r.write)         # download a text file
> r.close()
>
> r = open('quoter.pyw','rb')                    # upload a binary file
> f.storbinary('STOR quoter.pyw',r,512)
> r.close()
>
> f.rename('quoter.pyw','q.pyw')              # rename a file
> f.delete('q.pyw')                                # delete a file
>
> f.quit()                                        # terminate the session
>
> Hope these note help,
>
> Raymond



From sholden at holdenweb.com  Wed Feb 28 19:27:02 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 28 Feb 2001 13:27:02 -0500
Subject: Is there an easier-to-use module than ftplib?
References: <3A930645.761A82C8@home.com> <3A930D95.2DC3B98C@javanet.com> <3A932336.71661B16@home.com>
Message-ID: <rtbn6.3331$jv4.240927@e420r-atl2.usenetserver.com>

"Rick Lee" <rwklee at home.com> wrote in message
news:3A932336.71661B16 at home.com...
> Thanks!  That looks simple enough.  Do I have to worry about active vs.
> passive mode?
>
You shouldn't: most servers will work either way. Note that the ftplib
software is (AFAICR) dure to change its default behavior to passive with the
2.1 release.

regards
 Steve





From stevemak at softhome.net  Wed Feb 21 01:22:10 2001
From: stevemak at softhome.net (Steve Mak)
Date: Wed, 21 Feb 2001 00:22:10 GMT
Subject: Carel Fellinger: one more question
Message-ID: <3A930C5E.4869355C@softhome.net>

I need to output some data to an output file. I am using:

outp = open("output.txt","w")
outp.write(data)
outp.close()

my problem is i need to have some data on separate lines. right now
everything is saved as 1 long line of data. how do I specify a carriage
return to the output file?




From adam at deprince.net  Tue Feb 20 02:57:27 2001
From: adam at deprince.net (Adam DePrince)
Date: Mon, 19 Feb 2001 20:57:27 -0500
Subject: Carel Fellinger: one more question
References: <3A930C5E.4869355C@softhome.net>
Message-ID: <3A91CF07.1DA27722@deprince.net>

Steve Mak wrote:
> 
> I need to output some data to an output file. I am using:
> 
> outp = open("output.txt","w")
> outp.write(data)
> outp.close()
> 
> my problem is i need to have some data on separate lines. right now
> everything is saved as 1 long line of data. how do I specify a carriage
> return to the output file?

Assuming that data is a sequence, each member representing the contents
of a line
(as opposed to the one-big-blob implied about) you can do this:

outp = open("output.txt","w")
outp.write(string.join(data, "\n"))
outp.close()

If you are using Python 2.0 of better you can say:

outp = open("output.txt","w")
for line in data:
	print >> outp, line
outp.close()

Or, if you lack a version <2.0, but not courage, you can:

outp = open("output.txt", "w" )
old_stdout = sys.stdout
sys.stdout = outp
for line in data:
	print line
sys.stdout = old_stdout
outp.close()

Just make sure your other threads don't disagree with your mistreatment
of stdout ... 


To see where I ripped off my explanation:
http://www.python.org/2.0/new-python.html

Search for "9.1" or "Minor Language Changes."  


Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From fredrik at pythonware.com  Wed Feb 21 07:42:28 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 06:42:28 GMT
Subject: Carel Fellinger: one more question
References: <3A930C5E.4869355C@softhome.net>
Message-ID: <orJk6.14720$Qb7.2357564@newsb.telia.net>

Steve Mak wrote:
> I need to output some data to an output file. I am using:
>
> outp = open("output.txt","w")
> outp.write(data)
> outp.close()
>
> my problem is i need to have some data on separate lines. right now
> everything is saved as 1 long line of data. how do I specify a carriage
> return to the output file?

outp.write("\n")

Cheers /F




From phd at phd.russ.ru  Wed Feb 21 02:03:49 2001
From: phd at phd.russ.ru (Oleg Broytmann)
Date: 20 Feb 2001 19:03:49 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Feb 20)
Message-ID: <D0D42E6D25BC1A27.425AE269661E6091.0F4BD16934D5BD2A@lp.airnews.net>

    Michael Hudson continues Python-dev Summaries after AMK. Thanks, Mike!
        http://lwn.net/daily/pydev-20010215.php3
    
    Pete Shinners announces new version (0.9) of pygame
        http://groups.google.com/groups?q=ANN:+pygame-0.9&hl=en&lr=&safe=off&rnum=1&ic=1
    
    LAMP (Linux, Apache, MySql, and Python) in action
        http://linuxpr.com/releases/3326.html
    
    Suchandra S Thapa moves Siphon project to SourceForge
        http://groups.google.com/groups?q=Suchandra+S+Thapa+Re:+CPAN+functionality+for+python&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Mark Hammond announces new version of win32com that allows you to implement
    any COM interface decribed in a type library, either IDispatch based or not
        http://groups.google.com/groups?q=Mark+Hammond+COM+support+for+vtable+based+interfaces&hl=en&lr=&safe=off&rnum=1&ic=1
    
    ActiveState invites nominations for scripting community heroes
        http://www.activestate.com/Corporate/Communications/Releases/Awards

    Moop project (MUD, Object Oriented, Pythonic) is now hosted at SourceForge
        http://moop.sourceforge.net/
    
    /F, as usual, shows an excellent snippet of code. This time it is
    GlobDirectoryWalker for doing recursive globbing
        http://groups.google.com/groups?q=Fredrik+Lundh+Re:+recursing+glob&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Dustin Boswell explains how to compile https support into Python2.0
        http://groups.google.com/groups?q=Dustin+Boswell+how+to+compile+https+support+into+Python2.0&hl=en&lr=&safe=off&rnum=1&ic=1
    
    David Mertz: Book Roundup
        http://gnosis.cx/cgi/txt2html.cgi?source=http://gnosis.cx/publish/programming/charming_python_12.txt
    
    Fredrik Lundh points to TryType (freetype) support for PIL
        http://groups.google.com/groups?q=Fredrik+Lundh+PIL+Fonts&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Carel Fellinger show how to handle overloading in OO manner
        http://groups.google.com/groups?q=Carel+Fellinger+Re:+Can+methods+functions+overloaded&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Steven D. Arnold announces updated version of Pure Python Encryption with
    fixed IDEA implementation
        http://groups.google.com/groups?q=Steven+D.+Arnold+ANN:+Pure+Python+Encryption+%5BUPDATE%5D&hl=en&lr=&safe=off&rnum=1&ic=1


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org
      Notice especially the master FAQ
          http://www.python.org/doc/FAQ.html

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
	  http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

      Michael Hudson continues Andrew Kuchling's marvelous tradition
      of summarizing action on the python-dev mailing list twice a
      month.
	  http://starship.python.net/crew/mwh/summaries/
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collect Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
	  http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From costas at meezon.com  Wed Feb 21 02:28:00 2001
From: costas at meezon.com (Costas Menico)
Date: Wed, 21 Feb 2001 01:28:00 GMT
Subject: Tkinter importing error
Message-ID: <3a93190a.462510@News.CIS.DFN.DE>

Does anyone know why I get the following message when I try to import
Tkinter on Windows 98?

>>> import Tkinter
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "d:\python20\lib\lib-tk\Tkinter.py", line 34, in ?
    import FixTk # Attempt to configure Tcl/Tk without requiring PATH
  File "d:\python20\lib\lib-tk\FixTk.py", line 1, in ?
    import sys, os, _tkinter
ImportError: DLL load failed: One of the library files needed to run
this application cannot be found.
>>> 

Where is Tk supposed to be installed? The Python I use is
ActiveState's version.  Thanks

Costas


From monpetit at orgio.net  Wed Feb 21 03:04:23 2001
From: monpetit at orgio.net (Monpetit)
Date: Wed, 21 Feb 2001 11:04:23 +0900
Subject: Tkinter importing error
References: <3a93190a.462510@News.CIS.DFN.DE>
Message-ID: <96v72n$p82$1@news.hitel.net>

Active Python does not include Tk module by default.
You have to install Tcl/Tk 8.3.2. Visit http://www.scriptics.com/ and
download Tcl/Tk.

"Costas Menico" <costas at meezon.com> wrote in message
news:3a93190a.462510 at News.CIS.DFN.DE...
> Does anyone know why I get the following message when I try to import
> Tkinter on Windows 98?
>
> >>> import Tkinter
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
>   File "d:\python20\lib\lib-tk\Tkinter.py", line 34, in ?
>     import FixTk # Attempt to configure Tcl/Tk without requiring PATH
>   File "d:\python20\lib\lib-tk\FixTk.py", line 1, in ?
>     import sys, os, _tkinter
> ImportError: DLL load failed: One of the library files needed to run
> this application cannot be found.
> >>>
>
> Where is Tk supposed to be installed? The Python I use is
> ActiveState's version.  Thanks
>
> Costas




From costas at meezon.com  Wed Feb 21 04:17:02 2001
From: costas at meezon.com (Costas Menico)
Date: Wed, 21 Feb 2001 03:17:02 GMT
Subject: Tkinter importing error
References: <3a93190a.462510@News.CIS.DFN.DE> <96v72n$p82$1@news.hitel.net>
Message-ID: <3a933279.997835@News.CIS.DFN.DE>

I did all that. I check the DOS path and it looks right. I copied the
Tcl/Tk DLLs and LIBs all over the place. Rebooted and still get the
same message.

Are there any instructions as to the exact installation?

Thanks

"Monpetit" <monpetit at orgio.net> wrote:

>
>Active Python does not include Tk module by default.
>You have to install Tcl/Tk 8.3.2. Visit http://www.scriptics.com/ and
>download Tcl/Tk.
>
>"Costas Menico" <costas at meezon.com> wrote in message
>news:3a93190a.462510 at News.CIS.DFN.DE...
>> Does anyone know why I get the following message when I try to import
>> Tkinter on Windows 98?
>>
>> >>> import Tkinter
>> Traceback (innermost last):
>>   File "<interactive input>", line 1, in ?
>>   File "d:\python20\lib\lib-tk\Tkinter.py", line 34, in ?
>>     import FixTk # Attempt to configure Tcl/Tk without requiring PATH
>>   File "d:\python20\lib\lib-tk\FixTk.py", line 1, in ?
>>     import sys, os, _tkinter
>> ImportError: DLL load failed: One of the library files needed to run
>> this application cannot be found.
>> >>>
>>
>> Where is Tk supposed to be installed? The Python I use is
>> ActiveState's version.  Thanks
>>
>> Costas
>
>



From costas at meezon.com  Wed Feb 21 05:00:39 2001
From: costas at meezon.com (Costas Menico)
Date: Wed, 21 Feb 2001 04:00:39 GMT
Subject: Tkinter importing error
References: <3a933279.997835@News.CIS.DFN.DE>
Message-ID: <3a933cfd.3690751@News.CIS.DFN.DE>

Found the problem. Tkinter only works with Tcl/Tk 8.3. 

Now for a more serious problem. Why doesnt it work with PythonWin???
PythinWin is such a great IDE. Its a shame it's not supported.

Costas


From marco at atmosp.physics.utoronto.ca  Wed Feb 21 02:46:09 2001
From: marco at atmosp.physics.utoronto.ca (marco at atmosp.physics.utoronto.ca)
Date: 21 Feb 2001 01:46:09 GMT
Subject: select.select example?
Message-ID: <96v6l1$3n3$1@news.netmar.com>

Hi all,

A few days ago I posted a little problem I had regarding
the blocking of UNIX pipes. Basically, when I ran:

(po, pi, pe) = popen2.popen3('some command')

and then:

po_out = po.read()
pe_err = pe.read()

it happened sometimes (in an upredictable manner) that
po.read() got stuck and my program stopped on its tracks
(the order in which I read po and pe does make a difference,
but while reading po.read() first works 99% of the time,
I execute this many times and that 1% kills me).

Two suggestions came up. One was using fcntl which seems to
work, e.g.:

>>> import popen2, fcntl, FCNTL
>>> (po, pi, pe) = popen2.popen3('some command')
>>>
>>> # set read pipe to NONBLOCK, so we won't block
>>> # on read() when there aren't any data available at the time
>>>
>>> fcntl.fcntl(po.fileno(), FCNTL.F_SETFL, FCNTL.O_NONBLOCK)
>>> fcntl.fcntl(pe.fileno(), FCNTL.F_SETFL, FCNTL.O_NONBLOCK)
>>>
>>> # for child -- close above opened file descriptors on exec()
>>> fcntl.fcntl(pi.fileno(), FCNTL.F_SETFD, 1)
>>> fcntl.fcntl(po.fileno(), FCNTL.F_SETFD, 1)
>>> fcntl.fcntl(pe.fileno(), FCNTL.F_SETFD, 1)
>>>
>>> pout = po.read()
>>> perr = pe.read()

The other option was using the select module. If I understand correctly,
select.select will retrieve the input, output and error (exceptional
condition?) without me having to worry which to read first. If none
is available a timeout can also be specified. If this is the case
it's probably what I want, but unfortunately the documentation is
very scarce. From what I can gather:

>>> import popen2, select
>>> (po, pi, pe) = popen2.popen3('some command')
>>> (spo, spi, spe) = select.select([po], [pi], [pe], 10)
>>> spo[0].read()

seems to work, although according to the Python 1.5.2 Library
Reference the order of the lists should be:
select.select(input, output, error, 10)
which doesn't seem to be working for me. Furthermore, if
'some command' worked OK I would expect spe[0].read() to
be '' but instead I get the error:
IOError: [Errno 9] Bad file descriptor

On the other hand, when 'some command' *does* raise an error, then
spo[0].read() gives me '' (which is expected), but spe[0].read()
gives the error:
IndexError: list index out of range
instead of std_err.

So, I'm confused...
If anyone can provide me with a simple example using select.select
on a UNIX pipe I would greatly appreciate it! (And, of course, any
other solutions are greatly welcomed :)

Cheers!

 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse at newsone.net


From donn at oz.net  Fri Feb 23 07:49:04 2001
From: donn at oz.net (Donn Cave)
Date: 23 Feb 2001 06:49:04 GMT
Subject: select.select example?
References: <96v6l1$3n3$1@news.netmar.com>
Message-ID: <975150$1jo$0@216.39.151.169>

Quoth marco at atmosp.physics.utoronto.ca:
| A few days ago I posted a little problem I had regarding
| the blocking of UNIX pipes. Basically, when I ran:
|
| (po, pi, pe) = popen2.popen3('some command')
|
| and then:
|
| po_out = po.read()
| pe_err = pe.read()
...

| The other option was using the select module. If I understand correctly,
| select.select will retrieve the input, output and error (exceptional
| condition?) without me having to worry which to read first. If none
| is available a timeout can also be specified. If this is the case
| it's probably what I want, but unfortunately the documentation is
| very scarce. From what I can gather:
|
| >>> import popen2, select
| >>> (po, pi, pe) = popen2.popen3('some command')
| >>> (spo, spi, spe) = select.select([po], [pi], [pe], 10)
| >>> spo[0].read()
|
| seems to work, although according to the Python 1.5.2 Library
| Reference the order of the lists should be:
| select.select(input, output, error, 10)
| which doesn't seem to be working for me.

Meaning in this context:
   input:  file descriptors to check for readability
   output: file descriptors to check for writability
   error:  file descriptors to check for I/O errors (not error I/O!)

So you were close, but it should really have been
   rds, wrs, ers = select.select([po, pe], [pi], [pi, po, pe], 10)

You could eventually be interested in whether your pipes are
ready to have more data written to them (and in fact you had
better learn to be interested in that if you want reliable
general purpose 2 way pipe communications between processes.
But by then, you will have weaned yourself off of file objects
too.)  For now, since you're interested only in readability,
you may leave the second set empty.  The errors that you
might get from the third set are actually fairly exotic, but
it doesn't hurt to check.

| ... Furthermore, if
| 'some command' worked OK I would expect spe[0].read() to
| be '' but instead I get the error:
| IOError: [Errno 9] Bad file descriptor

Can't say I see how that happened - certainly, anything
select() reports in its third, error set will not likely
be a valid file descriptor for more I/O, after the error
that got it into that set.  But I don't know right off
hand what that error would have been.

| On the other hand, when 'some command' *does* raise an error, then
| spo[0].read() gives me '' (which is expected), but spe[0].read()
| gives the error:
| IndexError: list index out of range
| instead of std_err.

That's fine, one doesn't expect I/O errors.

About file objects and select(), please bear in mind that the
file object is a C stdio buffer.  Output can be buffered and
not written to the pipe.  Input can be buffered and be ready
to read but not visible to select().  Reads can block until
the amount requested appears in the pipe, which is usually
not convenient.  You can ameliorate some of this brain damage
by turning off the buffering, but it's even better to just
get the file descriptor back out of the file object and use
it for everything.

    po, pi, pe = popen2.popen3('some command')
    pod = po.fileno()
    pid = pi.fileno()
    ped = pi.fileno()

    while 1:
        if data:
            ofs = [pid]
        else:
            ofs = []
        ifd, ofd, efd = select.select([pod, ped], ofs, [pod, ped, pid], 10)
        if efd:
            print >> sys.stderr, "error on units", efd
            break
        if pod in ifd:
            output = os.read(pod, 16384)
            if not output:
                print >> sys.stderr, "EOF from process"
                break
            else:
                print "process output", repr(output)
        if ped in ifd:
            errors = os.read(ped, 16384)
            if not errors:
                print >> sys.stderr, "EOF from process"
                break
            else:
                print "process errors", repr(errors)
        if data and pid in ofs:
            n = os.write(pid, data)
            data = data[n:]

I haven't checked to see if that actually works, but it looks OK to me.

	Donn Cave, donn at oz.net


From fimafeng at yahoo.com  Wed Feb 21 03:30:19 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Wed, 21 Feb 2001 02:30:19 GMT
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
Message-ID: <3A93280C.2080802@yahoo.com>

Hello I've been using BeOpen python on Windows for a while now. I have 
converted one of my home systems over to RedHat 7.0. I installed the 
actual 2.0 rpm just fine. However..when I tried to install 
Tkinter-2.0-8.0-3.i386.rpm I get the following errors


"Tkinter-2.0-8.0-3 requires libtcl8.0.so"
"Tkinter-2.0-8.0-3 requires libtk8.0.so"

I thought that TCL/TK was installed
I show the following RPMs as beeing installed...
tksysv 1.3-2
tcl 8.3.1-46
tcl 8.2.0-46
tk  8.3.1-46
tkinter 1.5.2-27
BeOpenPython 2.0
tcllib.0.4-46

What am I missing?

Also...How do I access the IDLE in Linux??

Many thanks,
Chris



From bowman at montana.com  Wed Feb 21 03:53:46 2001
From: bowman at montana.com (bowman)
Date: Tue, 20 Feb 2001 19:53:46 -0700
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com>
Message-ID: <28Gk6.499$c7.3553@newsfeed.slurp.net>

Chris Nelson <fimafeng at yahoo.com> wrote in message
news:3A93280C.2080802 at yahoo.com...
> "Tkinter-2.0-8.0-3 requires libtcl8.0.so"
> "Tkinter-2.0-8.0-3 requires libtk8.0.so"
> I show the following RPMs as beeing installed...
> tcl 8.3.1-46
> tcl 8.2.0-46
> tk  8.3.1-46

welcome to rpm hell. tcl8.0 is not tcl8.3. unless you have some
dependencies, you might remove the 8.3 rpms and install 8.0. I don't know if
the problems were resolved, but there were some issues with 8.3 that caused
people building packages that required tcl/tk to revert back to 8.0. rpm's
are fine things, but they imply your machine has all the bits and pieces
that the build machine had, which is not always the case.





From fimafeng at yahoo.com  Wed Feb 21 05:21:42 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Wed, 21 Feb 2001 04:21:42 GMT
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net>
Message-ID: <3A93422B.3040605@yahoo.com>

bowman wrote:

> Chris Nelson <fimafeng at yahoo.com> wrote in message
> 
> news:3A93280C.2080802 at yahoo.com...
> 
> 
>> "Tkinter-2.0-8.0-3 requires libtcl8.0.so"
> 
>> "Tkinter-2.0-8.0-3 requires libtk8.0.so"
> 
>> I show the following RPMs as beeing installed...
> 
>> tcl 8.3.1-46
> 
>> tcl 8.2.0-46
> 
>> tk  8.3.1-46
> 
> 
> 
> welcome to rpm hell. tcl8.0 is not tcl8.3. unless you have some
> 
> dependencies, you might remove the 8.3 rpms and install 8.0. I don't know if
> 
> the problems were resolved, but there were some issues with 8.3 that caused
> 
> people building packages that required tcl/tk to revert back to 8.0. rpm's
> 
> are fine things, but they imply your machine has all the bits and pieces
> 
> that the build machine had, which is not always the case.

I removed the 8.3 libraries (even though it said there were 
dependencies..installed the 8.0.5 series of tcl and tk....and TKinter 
installed all right. I was goint to then upgrade the 8.0's back to 8.3 
but I got the warning about those libraries mentioned above so I decided 
to wait.

BTW.. How do I start the IDLE in Linux???

Thanks,
Chris



From adam at deprince.net  Tue Feb 20 06:32:21 2001
From: adam at deprince.net (Adam DePrince)
Date: Tue, 20 Feb 2001 00:32:21 -0500
Subject: Off topic: Re: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net> <3A93422B.3040605@yahoo.com>
Message-ID: <3A920165.57957986@deprince.net>

Chris Nelson wrote:
> 
> bowman wrote:
> 
> > Chris Nelson <fimafeng at yahoo.com> wrote in message
> >
> > news:3A93280C.2080802 at yahoo.com...
> >
> >
> >> "Tkinter-2.0-8.0-3 requires libtcl8.0.so"
> >
> >> "Tkinter-2.0-8.0-3 requires libtk8.0.so"
> >
> >> I show the following RPMs as beeing installed...
> >
> >> tcl 8.3.1-46
> >
> >> tcl 8.2.0-46
> >> tk  8.3.1-46
> > welcome to rpm hell. tcl8.0 is not tcl8.3. unless you have some
> > dependencies, you might remove the 8.3 rpms and install 8.0. I don't know if
> > the problems were resolved, but there were some issues with 8.3 that caused
> > people building packages that required tcl/tk to revert back to 8.0. rpm's
> > are fine things, but they imply your machine has all the bits and pieces
> > that the build machine had, which is not always the case.
> I removed the 8.3 libraries (even though it said there were
> dependencies..installed the 8.0.5 series of tcl and tk....and TKinter
> installed all right. I was goint to then upgrade the 8.0's back to 8.3
> but I got the warning about those libraries mentioned above so I decided
> to wait.
> BTW.. How do I start the IDLE in Linux???

At risk of going off topic for this group, but you mentioned that you
use Redhat 7.0
I belive IDLE is an optional part of the RedHat install (well, its on my
machine, I run RH7, and it lives in /usr/bin, not the place my installed
goodies go, so I guess it comes w/ RH7).  If you have the space, select
the "everything" option during your install. 

Adam DePrince
Starmedia Network Inc.
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From fimafeng at yahoo.com  Wed Feb 21 11:16:19 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Wed, 21 Feb 2001 10:16:19 GMT
Subject: Off topic: Re: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net> <3A93422B.3040605@yahoo.com> <3A920165.57957986@deprince.net>
Message-ID: <3A939548.5070700@yahoo.com>

Adam DePrince wrote:

> Chris Nelson wrote:
> 
> 
>> bowman wrote:
> 
>> 
> 
>> 
>>> Chris Nelson <fimafeng at yahoo.com> wrote in message
> 
>>> 
> 
>>> news:3A93280C.2080802 at yahoo.com...
> 
>>> 
> 
>>> 
> 
>>> 
>>>> "Tkinter-2.0-8.0-3 requires libtcl8.0.so"
> 
>>> 
>>>> "Tkinter-2.0-8.0-3 requires libtk8.0.so"
> 
>>> 
>>>> I show the following RPMs as beeing installed...
> 
>>> 
>>>> tcl 8.3.1-46
> 
>>> 
>>>> tcl 8.2.0-46
> 
>>>> tk  8.3.1-46
> 
>>> 
>>> welcome to rpm hell. tcl8.0 is not tcl8.3. unless you have some
> 
>>> dependencies, you might remove the 8.3 rpms and install 8.0. I don't know if
> 
>>> the problems were resolved, but there were some issues with 8.3 that caused
> 
>>> people building packages that required tcl/tk to revert back to 8.0. rpm's
> 
>>> are fine things, but they imply your machine has all the bits and pieces
> 
>>> that the build machine had, which is not always the case.
> 
>> 
>> I removed the 8.3 libraries (even though it said there were
> 
>> dependencies..installed the 8.0.5 series of tcl and tk....and TKinter
> 
>> installed all right. I was goint to then upgrade the 8.0's back to 8.3
> 
>> but I got the warning about those libraries mentioned above so I decided
> 
>> to wait.
> 
>> BTW.. How do I start the IDLE in Linux???
> 
> 
> 
> At risk of going off topic for this group, but you mentioned that you
> 
> use Redhat 7.0
> 
> I belive IDLE is an optional part of the RedHat install (well, its on my
> 
> machine, I run RH7, and it lives in /usr/bin, not the place my installed
> 
> goodies go, so I guess it comes w/ RH7).  If you have the space, select
> 
> the "everything" option during your install. 
> 
> 
> 
> Adam DePrince
> 
> Starmedia Network Inc.
> 
> zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')

Does the BeOsPython distrivution contain an IDLE for 2.0? I just 
installed it so I was curious how to access it if it were included



From noselasd at frisurf.no  Wed Feb 21 10:13:21 2001
From: noselasd at frisurf.no (Nils O. Selåsdal)
Date: Wed, 21 Feb 2001 10:13:21 +0100
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net> <3A93422B.3040605@yahoo.com>
Message-ID: <AELk6.3963$X_2.96782@news1.oke.nextra.no>

"Chris Nelson" <fimafeng at yahoo.com> wrote in message
news:3A93422B.3040605 at yahoo.com...
> bowman wrote:
>
> BTW.. How do I start the IDLE in Linux???
Contained in some rpm .. python-tools or py-tools..
think its on the 2. cd of RH7. install it, and type idle





From john at dhh.gt.org  Wed Feb 21 06:00:05 2001
From: john at dhh.gt.org (John Hasler)
Date: Wed, 21 Feb 2001 05:00:05 GMT
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net>
Message-ID: <877l2ky6ju.fsf@toncho.dhh.gt.org>

bowman writes:
> unless you have some dependencies, you might remove the 8.3 rpms and
> install 8.0.

Why do you think he needs to remove 8.3?  tcl8.0, tcl8.2, and tcl8.3 all
co-exist peacefully on my Debian system.

> rpm's are fine things, but they imply your machine has all the bits and
> pieces that the build machine had, which is not always the case.

That's what dependencies are for (and you can have multiple versions of a
given library installed at the same time).
-- 
John Hasler
john at dhh.gt.org (John Hasler)
Dancing Horse Hill
Elmwood, WI


From ptb at oboe.it.uc3m.es  Wed Feb 21 12:46:15 2001
From: ptb at oboe.it.uc3m.es (Peter T. Breuer)
Date: Wed, 21 Feb 2001 12:46:15 +0100
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net> <877l2ky6ju.fsf@toncho.dhh.gt.org>
Message-ID: <7q9079.1hj.ln@news.it.uc3m.es>

In comp.os.linux.setup John Hasler <john at dhh.gt.org> wrote:
> bowman writes:
>> unless you have some dependencies, you might remove the 8.3 rpms and
>> install 8.0.

> Why do you think he needs to remove 8.3?  tcl8.0, tcl8.2, and tcl8.3 all
> co-exist peacefully on my Debian system.

Yes, well, to nitpick, he didn't say "need" and he appears to know what
he is talking about in general :-). I daresay he just wants to reduce
clutter in the case that these installs are really not necessary for
the system, thus simplifying whatever problem the OP is facing and
aiding diagnosis.

>> rpm's are fine things, but they imply your machine has all the bits and
>> pieces that the build machine had, which is not always the case.

> That's what dependencies are for (and you can have multiple versions of a
> given library installed at the same time).

Yes.

Peter


From fimafeng at yahoo.com  Wed Feb 21 14:05:59 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Wed, 21 Feb 2001 13:05:59 GMT
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net> <877l2ky6ju.fsf@toncho.dhh.gt.org> <7q9079.1hj.ln@news.it.uc3m.es>
Message-ID: <3A93BD0C.3000307@yahoo.com>

Peter T. Breuer wrote:

> In comp.os.linux.setup John Hasler <john at dhh.gt.org> wrote:
> 
> 
>> bowman writes:
> 
>> 
>>> unless you have some dependencies, you might remove the 8.3 rpms and
> 
>>> install 8.0.
> 
>> 
> 
>> Why do you think he needs to remove 8.3?  tcl8.0, tcl8.2, and tcl8.3 all
> 
>> co-exist peacefully on my Debian system.
> 
> 
> 
> Yes, well, to nitpick, he didn't say "need" and he appears to know what
> 
> he is talking about in general :-). I daresay he just wants to reduce
> 
> clutter in the case that these installs are really not necessary for
> 
> the system, thus simplifying whatever problem the OP is facing and
> 
> aiding diagnosis.
> 
> 
> 
> 
>>> rpm's are fine things, but they imply your machine has all the bits and
> 
>>> pieces that the build machine had, which is not always the case.
> 
>> 
> 
>> That's what dependencies are for (and you can have multiple versions of a
> 
>> given library installed at the same time).
> 
> 
> 
> Yes.
> 
> 
> 
> Peter


Ultimately hat I'm looking for is some version of Python 2.0 
(ActiveState or BeOs) to run with IDLE on RH 7.0

I'm back down to the 8.0 version of tcl & tk and 1.52 of tkinter

Does anyone have this working? Should I start a new thread?? w/this as 
topic?

any help is appreciated...Chris



From bsass at freenet.edmonton.ab.ca  Thu Feb 22 22:26:06 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Thu, 22 Feb 2001 14:26:06 -0700
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
In-Reply-To: <3A93BD0C.3000307@yahoo.com>
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net>
 <877l2ky6ju.fsf@toncho.dhh.gt.org> <7q9079.1hj.ln@news.it.uc3m.es>
 <3A93BD0C.3000307@yahoo.com>
Message-ID: <Pine.LNX.4.33.0102221401030.20294-100000@bms>


On Wed, 21 Feb 2001, Chris Nelson wrote:
> Ultimately hat I'm looking for is some version of Python 2.0
> (ActiveState or BeOs) to run with IDLE on RH 7.0
>
> I'm back down to the 8.0 version of tcl & tk and 1.52 of tkinter

The ActiveState idle wants tk/tcl-8.3, if you can get it to co-exist
with 8.0 you can run both 1.5.2 and as2.0's idle (that is what you
are after or the state you are in, right?).

To get 2.0 to live with 1.5.2 you can(must?) change all the *.py
executables in the py2 distribution to use the python2.0 executable.
A tweaked .../Tools/scripts/fixps.py program helps automate the task.


- Bruce



From bowman at montana.com  Wed Feb 21 15:32:04 2001
From: bowman at montana.com (bowman)
Date: Wed, 21 Feb 2001 07:32:04 -0700
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net> <877l2ky6ju.fsf@toncho.dhh.gt.org>
Message-ID: <AmQk6.162$yY.1257@newsfeed.slurp.net>

John Hasler <john at dhh.gt.org> wrote in message
news:877l2ky6ju.fsf at toncho.dhh.gt.org...
>
> Why do you think he needs to remove 8.3?  tcl8.0, tcl8.2, and tcl8.3 all
> co-exist peacefully on my Debian system.

because I'm lazy. because the OP seemed to be new to Linux, and maintaining
several version levels can require bit of tweaking. because I've been around
the barn with tcl/tk, Expect, and friends that won't build unless they can
find the right version and its hidden headers, etc. because I am not sure if
the OP could rpm 8.0 onto his system without forcing the install.

I've also been around the barn with earlier RH releases that came with
broken Python and Perl packages, and 7.0 had its share of alpha and beta
builds dressed up as stable packages.

Many people enjoy the sysadmin aspects of Linux, but as the OS gains
popularity, more users just want to get the job done, without a lot of
complications and surprises.





From tgagne at ix.netcom.com  Wed Feb 21 04:38:24 2001
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Tue, 20 Feb 2001 22:38:24 -0500
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za>
Message-ID: <3A933830.7F98D118@ix.netcom.com>

I just had an interesting conversation today with my brother who is learning
programming from scratch.  He's learning Smalltalk, and was watching me debug a
program I'm working on and we visited the subject of reuse.

One of the things I learned recently (over in comp.lang.python, I think) was
that it is as important to reuse idioms as it is to reuse code.  By designing
your classes to work within the idioms of the language you're using your
program ultimately reuses more, and becomes more reusable (and more easily
factored) itself.

For instance, I could create all kinds of classes with their own protocol for
doing something, or I could endow them with behaviors that make them great
collection citizens.  In Smalltalk, that would be thinking of how using do:,
select:, collect:, detect:, and inject:into: on collection containing my
objects would be either easier or more useful.

Long-time Smalltalkers, and apparently Python programmers, may have already had
this drilled into their heads.  I don't remember reading anything about
designing software to work well within the idiom of any language in any of the
books I've read.  Maybe others have?

x wrote:

> Hi,
>
> I have been reading up on the OO environment, focusing on the use of UML to
> speed up creation of objects, or even to create superior designs of objects.
>
> Its great. Everyone is creating their reusable objects, no efforts are being
> wasted, these huge libraries are being built....are they?
>
> Does everyone create their own libraries, with a little sharing when
> requested, or is there some huge repository of objects available to all and
> sundry where the benefits of mankind's OO efforts are being stored nice and
> neat... you know, like the Human Genome project?
>
> I'm curious to know.
>
> Thanks
>
> Michael
> michael.wijtenburg at iafrica.com

--
.tom





From phlip_cpp at my-deja.com  Wed Feb 21 06:39:41 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 21 Feb 2001 05:39:41 GMT
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com>
Message-ID: <96vkat$m8d@dispatch.concentric.net>

> x wrote:
> 
>> Hi,
>>
>> I have been reading up on the OO environment, focusing on the use of UML
>> to speed up creation of objects, or even to create superior designs of
>> objects.

You might think you must write UML first, then copy it to code. Whether you 
do this or not isn't as important as whether your code can always get 
better. For example, you could use UML to illustrate its current state, 
upgrade the code, and then use UML to demonstrate to colleagues who re-use 
the code what it now looks like & how it changed. Projects where any of the 
classes never change go downhill and become mired in their legacy 
expectations. Don't do that.

You can't change code without unit tests. They are more important than UML, 
therefor write them before you write code. Then as you write more code, 
"refactor" it to improve its design. (Python's typeless code & pristine 
syntax makes refactoring as easy as falling off a log.)

Long term, for large projects, making sure some "design document" matches 
the current state of the code has been discovered again and again to be a 
pure waste of time. Either the UML locks in a bad design, or the design 
gets better and the UML document does not reflect this.

"Well designed code" is not the same as "code with a lot of pictures 
describing it."

>> Its great. Everyone is creating their reusable objects, no efforts are
>> being wasted, these huge libraries are being built....are they?

No, UML won't do this alone. To write a re-usable library that others can 
use, you should develop it (just as iteratively as one-shot code) with at 
least three client applications all at the same time. That which all three 
find themselves doing belongs inside the library, and "nice to haves" don't.

>> Does everyone create their own libraries, with a little sharing when
>> requested, or is there some huge repository of objects available to all
>> and sundry where the benefits of mankind's OO efforts are being stored
>> nice and neat... you know, like the Human Genome project?

Easy. For every feature you must add, you write test code first and get it 
to fail. Then add the feature until the test code passes. Version this test 
code with the project - don't throw it away - and create a "super_test" 
that calls all the tests in a project. 

All this hacking might just make you think such a project will lead to poor 
design.

Refactor the code each time you see something break the rule 
OnceAndOnlyOnce. That means any concept, no matter how profound or trivial, 
must only appear in source once. As you refactor, re-usable modules grow 
out of the primordial ooze of functionality. And many of these re-usables 
might just appear where you could never have expected them to with all the 
up-front UML in the world.

Now why are unit tests crucial to re-use at all levels of a project? If 
every single feature you write has a unit test, then every feature already 
has two clients, not just one. It has a production client that sends it 
real messages, and it also has a test client sending it test messages. 
Test-first code is all already re-used once. This makes re-using it again 
blazingly easy.

        http://c2.com/cgi/wiki?CodeUnitTestFirst
        http://c2.com/cgi/wiki?RefactorMercilessly

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  How does Bugs Bunny do it? How does he know when he
      wakes up in the morning to put in his pocket 3 sticks
      of dynamite, a physician costume, and a bicycle pump?  --


From ctavares at develop.com  Wed Feb 21 07:30:58 2001
From: ctavares at develop.com (Chris Tavares)
Date: Wed, 21 Feb 2001 06:30:58 GMT
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com>
Message-ID: <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net>

"Thomas Gagne" <tgagne at ix.netcom.com> wrote in message
news:3A933830.7F98D118 at ix.netcom.com...
[... snip ...]
>
> Long-time Smalltalkers, and apparently Python programmers, may have
already had
> this drilled into their heads.  I don't remember reading anything about
> designing software to work well within the idiom of any language in any of
the
> books I've read.  Maybe others have?
>

Actually, this is in many ways the whole idea behind the STL in C++. Write
code to common idioms and then you can easily reuse different pieces
together.

While the world is totally different (static vs. dynamic typing), the
concepts are in many ways the same. Studying the ideas behind the STL are
well worth doing for everyone I would think.

-Chris





From cmh at bDistributed.com  Thu Feb 22 15:13:59 2001
From: cmh at bDistributed.com (Chris Hanson)
Date: Thu, 22 Feb 2001 08:13:59 -0600
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net>
Message-ID: <220220010813598042%cmh@bDistributed.com>

In article <CgJk6.515$v77.66568 at newsread2.prod.itd.earthlink.net>,
Chris Tavares <ctavares at develop.com> wrote:

> While the world is totally different (static vs. dynamic typing), the
> concepts are in many ways the same. Studying the ideas behind the STL are
> well worth doing for everyone I would think.

It's a great idea for learning what *not* to do.

The STL is horribly designed.  There's no hierarchy, no real
interchangeability in the container classes.  And they force you to do
all sorts of extra work (some of which just isn't feasible) to be able
to store your objects in containers.

STL was designed by number crunchers with the broken assumption that
programs are designed from the algorithm out and that efficiency is the
absolute most important goal.  Software in the real world is designed
from the business requirements out; this usually starts with a human
interface and a set of user requirements, and efficiency is only
measured compared to human use rather than the processor's maximum
throughput.

I've actually had STL proponents tell me it's a *good* thing that
std::vector<T> doesn't have a push_front() method because if you try to
do that your code won't compile and you'll know you should be using a
"more efficient" container class!  What if I *know* it's an expensive
operation, I know I need to do it occasionally, and I need the speed
that std::vector<T> will otherwise give me?  "Subclass std::vector<T>
then!"  (This was the real answer they gave.)  If I have to subclass a
standard container class to get simple functionality, it's *not*
well-designed.

Study the Smalltalk frameworks and Apple's Cocoa frameworks (formerly
the NeXT frameworks) if you want to learn good object-oriented design. 
<http://developer.apple.com/techpubs/macosx/Cocoa/CocoaTopics.html>

  -- Chris

-- 
Chris Hanson <cmh at bDistributed.com>
bDistributed.com: Making business distributed.
http://bDistributed.com/


From root at rainerdeyke.com  Thu Feb 22 16:59:15 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Thu, 22 Feb 2001 15:59:15 GMT
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com>
Message-ID: <nHal6.285701$ge4.97547638@news2.rdc2.tx.home.com>

"Chris Hanson" <cmh at bDistributed.com> wrote in message
news:220220010813598042%cmh at bDistributed.com...
> I've actually had STL proponents tell me it's a *good* thing that
> std::vector<T> doesn't have a push_front() method because if you try to
> do that your code won't compile and you'll know you should be using a
> "more efficient" container class!  What if I *know* it's an expensive
> operation, I know I need to do it occasionally, and I need the speed
> that std::vector<T> will otherwise give me?  "Subclass std::vector<T>
> then!"  (This was the real answer they gave.)  If I have to subclass a
> standard container class to get simple functionality, it's *not*
> well-designed.

The correct answer, of course, is to use the 'insert' method.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From aleaxit at yahoo.com  Thu Feb 22 18:10:21 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 18:10:21 +0100
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <nHal6.285701$ge4.97547638@news2.rdc2.tx.home.com>
Message-ID: <973hl801g60@news2.newsguy.com>

"Rainer Deyke" <root at rainerdeyke.com> wrote in message
news:nHal6.285701$ge4.97547638 at news2.rdc2.tx.home.com...
> "Chris Hanson" <cmh at bDistributed.com> wrote in message
> news:220220010813598042%cmh at bDistributed.com...
> > I've actually had STL proponents tell me it's a *good* thing that
> > std::vector<T> doesn't have a push_front() method because if you try to
> > do that your code won't compile and you'll know you should be using a
> > "more efficient" container class!  What if I *know* it's an expensive
> > operation, I know I need to do it occasionally, and I need the speed
> > that std::vector<T> will otherwise give me?  "Subclass std::vector<T>
> > then!"  (This was the real answer they gave.)  If I have to subclass a
> > standard container class to get simple functionality, it's *not*
> > well-designed.
>
> The correct answer, of course, is to use the 'insert' method.

I'd rather do a push_back and a rotate because, in my typical
use, N/2 swaps would be far less costly than N-1 copies.  But,
sure, "vec.insert(vec.begin(),x);" IS rather more concise than
"vec.push_back(x); std::rotate(vec.begin(),vec.end()-1,vec.end();"
(and unless you've specialized the swap template, so that the
swaps ARE fast, performance will be equivalent; however, note
that several specialized implementations of std::swap exist,
and that's what std::rotate is supposed to be using).


Alex







From aleaxit at yahoo.com  Thu Feb 22 17:56:06 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 17:56:06 +0100
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com>
Message-ID: <973gb401epe@news2.newsguy.com>

"Chris Hanson" <cmh at bDistributed.com> wrote in message
news:220220010813598042%cmh at bDistributed.com...
    [snip]
> STL was designed by number crunchers with the broken assumption that
> programs are designed from the algorithm out and that efficiency is the
> absolute most important goal.

If C++ is being used, then efficiency *HAD BETTER* be a paramount
goal: otherwise, the programming-language choice was utterly absurd.

> Software in the real world is designed

...in utterly absurd ways, very often, sure -- such as, using C++
for stuff that it's simply ridiculous to do in C++, as it does not
really need its runtime efficiency (and thus, should absolutely *NOT*
be paying the huge programmer-productivity costs connected with
C++ memory management &c).

But the STL designers were quite right not to take this into
account -- their key goal was to make C++ better *for the uses
to which it SHOULD be put* (which in turn implies, cases where
runtime efficiency is important) rather than for uses to which
it should NOT be put (although it often IS:-).


> I've actually had STL proponents tell me it's a *good* thing that
> std::vector<T> doesn't have a push_front() method because if you try to
> do that your code won't compile and you'll know you should be using a
> "more efficient" container class!

Yep.

> What if I *know* it's an expensive
> operation, I know I need to do it occasionally, and I need the speed
> that std::vector<T> will otherwise give me?

You don't make believe a METHOD exists, but rather call a FUNCTION
for the purpose.  E.g.,

template <class X>
inline void push_front(std::vector<X>& vec, const X& x)
{
    vec.push_back(x);
    std::rotate(vec.begin(), vec.end()-1, vec.end());
}

> "Subclass std::vector<T>
> then!"  (This was the real answer they gave.)  If I have to subclass a
> standard container class to get simple functionality, it's *not*
> well-designed.

The STL designers are not *class-obsessed* (the way you seem to
be, and, apparently, those 'STL proponents' you asked -- the
very *idea* of subclassing STL classes is laughable, so, they're
proponents of what they don't really understand).  Lots of
important functionality is packaged up as _function_ templates,
and is meant to be called with function syntax.  The 'algorithms'
thus packaged are _orthogonal_ to the container classes.

You don't have to subclass anything to 'get simple functionality':
you call functions, when that is more appropriate than calling
methods (funny enough, Python is based on this concept, too:-).

Apparently, you refuse to consider that calling a function can
be THE way to access functionality; and thus, to consider "well
designed" a class if it has methods to duplicate any "simple
functionality" that is in fact quite accessible with slightly
different syntax-sugar via function-calls.

Well, I have news for you -- those classes you seem to think
are "well designed" are what many of us would think of as "fat".

*MY* idea of a well-designed class, **for a language that has
free-standing (templated or dynamically typed) functions** as
well as classes, and thinking of foundational library classes,
is one which exposes a LEAN (but functionally complete) set of
methods, allowing a separate set of "convenience" functions to
do their job neatly.  Not 10 classes with 100 methods each,
as 10 * 100 == 1000; rather, 10 classes with 10 methods each,
and the other 90 things factored out as functions -- 190 is
a much smaller number than 1000, so the overall design is
simpler and leaner.


Alex





From ben.hutchings at roundpoint.com  Fri Feb 23 03:01:35 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 22 Feb 2001 18:01:35 -0800
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com>
Message-ID: <ud7cakvi8.fsf@roundpoint.com>

Chris Hanson <cmh at bDistributed.com> writes:

> In article <CgJk6.515$v77.66568 at newsread2.prod.itd.earthlink.net>,
> Chris Tavares <ctavares at develop.com> wrote:
> 
> > While the world is totally different (static vs. dynamic typing), the
> > concepts are in many ways the same. Studying the ideas behind the STL are
> > well worth doing for everyone I would think.
> 
> It's a great idea for learning what *not* to do.
> 
> The STL is horribly designed.  There's no hierarchy,

I think you mean it's not an OO design, and that's true.

There *are* common interfaces (random lookup, list-ops, front-ops,
stack-ops, iteration), but those interfaces aren't explicit so you
can't put off your choice of implementation to run-time - unless you
use wrappers.

> no real interchangeability in the container classes.

Some algorithms require different interfaces from others.  You can
interchange containers as long as you keep a clear idea of which
interfaces your algorithm(s) need and pick the right containers.  The
most common one is iteration, and almost all containers support that.

> And they force you to do all sorts of extra work (some of which just
> isn't feasible) to be able to store your objects in containers.

Storing polymorphic objects certainly seems to be too difficult, since
they have to be stored by reference and there is no ref-counting
smart-pointer class in the standard library (but there is
boost::shared_ptr).

> STL was designed by number crunchers with the broken assumption that
> programs are designed from the algorithm out and that efficiency is the
> absolute most important goal.  Software in the real world is designed
> from the business requirements out; this usually starts with a human
> interface and a set of user requirements, and efficiency is only
> measured compared to human use rather than the processor's maximum
> throughput.
<snip>

STL is not meant to do everything *for* the C++ programmer, but it is
meant to be used *by* every C++ programmer.  That's why it has to be
efficient and flexible, and not too abstract.  You can always write
wrapper classes on top of it that are more abstract but less flexible
or less efficient, but you could never write wrapper classes to make
an abstract container library more efficient!

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From topmind at technologist.com  Sun Feb 25 21:52:26 2001
From: topmind at technologist.com (Topmind)
Date: Sun, 25 Feb 2001 20:52:26 GMT
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com>
Message-ID: <MPG.150312ceb68f1ada98969f@news.earthlink.net>

>
> Some algorithms require different interfaces from others.  You can
> interchange containers as long as you keep a clear idea of which
> interfaces your algorithm(s) need and pick the right containers.  The
> most common one is iteration, and almost all containers support that.

This is a sore point with me. To me it seems silly (artificial)
to split collection handling into categories (stacks, sets, trees, etc.).
If needs change from one to another or a combo, then you are hosed
with a non-fitting interface.

IMO the features should be viewed as independent as possible 
(HAS-A, not IS-A), not as coupled to some taxonomy of collections.

That is unnecessary coupling.

Shape up!

-tmind-


From patrickdlogan at home.com  Sun Feb 25 22:20:04 2001
From: patrickdlogan at home.com (Patrick Logan)
Date: Sun, 25 Feb 2001 21:20:04 GMT
Subject: Collection interfaces (Was: New to OO concepts - re-usability)
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net>
Message-ID: <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com>

"Topmind" <topmind at technologist.com> wrote
in message news:MPG.150312ceb68f1ada98969f at news.earthlink.net...
>
> This is a sore point with me. To me it seems silly (artificial)
> to split collection handling into categories (stacks, sets, trees, etc.).
> If needs change from one to another or a combo, then you are hosed
> with a non-fitting interface.

Yes, and so OO language provide an advantage here. Collection classes
in these languages are accessible through common interfaces. You don't
care how they are implemented, you only care about the interface.

> Shape up!

Yeah!

--
Patrick Logan
mailto:patrickdlogan at home.com





From topmind at technologist.com  Mon Feb 26 02:32:51 2001
From: topmind at technologist.com (Topmind)
Date: Mon, 26 Feb 2001 01:32:51 GMT
Subject: Collection interfaces (Was: New to OO concepts - re-usability)
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com>
Message-ID: <MPG.1503548620e632919896a4@news.earthlink.net>

> "Topmind" <topmind at technologist.com> wrote
> in message news:MPG.150312ceb68f1ada98969f at news.earthlink.net...
> >
> > This is a sore point with me. To me it seems silly (artificial)
> > to split collection handling into categories (stacks, sets, trees, etc.).
> > If needs change from one to another or a combo, then you are hosed
> > with a non-fitting interface.
> 
> Yes, and so OO language provide an advantage here. Collection classes
> in these languages are accessible through common interfaces. You don't
> care how they are implemented, you only care about the interface.
> 


You mean something like SQL or ODBC?

And, this crap that only OO can do 
implimentation-hiding interfaces is
pure propaganda.

(Note that I am talking about using
the interfaces, not making DB
engines or whatnot.)


> > Shape up!
> 
> Yeah!
> 
> --
> Patrick Logan
> mailto:patrickdlogan at home.com
> 

-tmind-


From patrickdlogan at home.com  Mon Feb 26 03:57:48 2001
From: patrickdlogan at home.com (Patrick Logan)
Date: Mon, 26 Feb 2001 02:57:48 GMT
Subject: Collection interfaces (Was: New to OO concepts - re-usability)
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net>
Message-ID: <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com>

"Topmind" <topmind at technologist.com> wrote
in message news:MPG.1503548620e632919896a4 at news.earthlink.net...
> > >
> > > This is a sore point with me. To me it seems silly (artificial)
> > > to split collection handling into categories (stacks, sets, trees,
etc.).
> > > If needs change from one to another or a combo, then you are hosed
> > > with a non-fitting interface.
> >
> > Yes, and so OO language provide an advantage here. Collection classes
> > in these languages are accessible through common interfaces. You don't
> > care how they are implemented, you only care about the interface.
>
> You mean something like SQL or ODBC?

No, I mean collections, as with, say, the Smalltalk collection classes.

> And, this crap that only OO can do implimentation-hiding interfaces is
> pure propaganda.

It is possible to hide the implementation behind any procedure. But
it requires some kind of OO mechanism to be polymorphic. That
is what OO is, by definition.

--
Patrick Logan
mailto:patrickdlogan at home.com





From topmind at technologist.com  Mon Feb 26 07:03:54 2001
From: topmind at technologist.com (Topmind)
Date: Mon, 26 Feb 2001 06:03:54 GMT
Subject: Collection interfaces (Was: New to OO concepts - re-usability)
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com>
Message-ID: <MPG.150394195f7f04a19896a7@news.earthlink.net>

> 
> "Topmind" <topmind at technologist.com> wrote
> in message news:MPG.1503548620e632919896a4 at news.earthlink.net...
> > > >
> > > > This is a sore point with me. To me it seems silly (artificial)
> > > > to split collection handling into categories (stacks, sets, trees,
> etc.).
> > > > If needs change from one to another or a combo, then you are hosed
> > > > with a non-fitting interface.
> > >
> > > Yes, and so OO language provide an advantage here. Collection classes
> > > in these languages are accessible through common interfaces. You don't
> > > care how they are implemented, you only care about the interface.
> >
> > You mean something like SQL or ODBC?
> 
> No, I mean collections, as with, say, the Smalltalk collection classes.
> 

Smalltalk uses collection taxonomies, which I frown on.
Collection needs change, morph, and grow. If you tie your
application to a specific "type", then the chances of
getting screwed are high IME.

If Smalltalk wants to rewrite them so features can be
mixed and match as needed, instead of based on an limiting
taxonomy, that would be great. (I am working on 
my version of such a spec, BTW.)

OO thinking is taxonomy-happy. I see it in collections,
streams, and GUI's. One can have many more variations if
you treat features as indepedent. Sure, not all combinations
are valid, but the pattern of invalidity is not a
tree pattern for the most part.


> > And, this crap that only OO can do implimentation-hiding interfaces is
> > pure propaganda.
> 
> It is possible to hide the implementation behind any procedure. But
> it requires some kind of OO mechanism to be polymorphic. That
> is what OO is, by definition.
> 

Whatever. It is not much difference to the user of such API's, mostly a 
matter of verb position.

x.movenext
movenext(x)

tamato 
tamato


> --
> Patrick Logan
> mailto:patrickdlogan at home.com
> 
> 

-tmind-


From jarober at mail.com  Mon Feb 26 14:24:16 2001
From: jarober at mail.com (James A. Robertson)
Date: Mon, 26 Feb 2001 13:24:16 GMT
Subject: Collection interfaces (Was: New to OO concepts - re-usability)
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net>
Message-ID: <3A9A59BD.6E0AF59@mail.com>

Topmind wrote:
> 

> > > You mean something like SQL or ODBC?
> >
> > No, I mean collections, as with, say, the Smalltalk collection classes.
> >
> 
> Smalltalk uses collection taxonomies, which I frown on.
> Collection needs change, morph, and grow. If you tie your
> application to a specific "type", then the chances of
> getting screwed are high IME.
> 

You can mix and match Smalltalk collections (and have been able to for
decades).  You might try downloading one of the free or cheap Smalltalks
and looking for yourself.


> If Smalltalk wants to rewrite them so features can be
> mixed and match as needed, instead of based on an limiting
> taxonomy, that would be great. (I am working on
> my version of such a spec, BTW.)
> 
> OO thinking is taxonomy-happy. I see it in collections,
> streams, and GUI's. One can have many more variations if
> you treat features as indepedent. Sure, not all combinations
> are valid, but the pattern of invalidity is not a
> tree pattern for the most part.
> 
> > > And, this crap that only OO can do implimentation-hiding interfaces is
> > > pure propaganda.
> >
> > It is possible to hide the implementation behind any procedure. But
> > it requires some kind of OO mechanism to be polymorphic. That
> > is what OO is, by definition.
> >
> 
> Whatever. It is not much difference to the user of such API's, mostly a
> matter of verb position.
> 
> x.movenext
> movenext(x)
> 
> tamato
> tamato
> 
> > --
> > Patrick Logan
> > mailto:patrickdlogan at home.com
> >
> >
> 
> -tmind-

-- 
James A. Robertson
Product Manager (Smalltalk), Cincom
jarober at mail.com
<Talk Small and Carry a Big Class Library>


From topmind at technologist.com  Tue Feb 27 06:09:37 2001
From: topmind at technologist.com (Topmind)
Date: Tue, 27 Feb 2001 05:09:37 GMT
Subject: Collection interfaces (Was: New to OO concepts - re-usability)
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A59BD.6E0AF59@mail.com>
Message-ID: <MPG.1504d8d5ed9b8cf59896b1@news.earthlink.net>

> Topmind wrote:
> > 
> 
> > > > You mean something like SQL or ODBC?
> > >
> > > No, I mean collections, as with, say, the Smalltalk collection classes.
> > >
> > 
> > Smalltalk uses collection taxonomies, which I frown on.
> > Collection needs change, morph, and grow. If you tie your
> > application to a specific "type", then the chances of
> > getting screwed are high IME.
> > 
> 
> You can mix and match Smalltalk collections (and have been able to for
> decades).  You might try downloading one of the free or cheap Smalltalks
> and looking for yourself.
> 

"Mix" implies they start out differently. They may signify
something.

It *is* on my (long) list of things to look into.

BTW, any web links to ST collection interface
catalogs?

[snip]

-tmind-


From tgagne at ix.netcom.com  Mon Feb 26 17:39:13 2001
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Mon, 26 Feb 2001 11:39:13 -0500
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net>
Message-ID: <3A9A86B1.9A8C69A2@ix.netcom.com>

Topmind wrote:

> <snip>
>
> Smalltalk uses collection taxonomies, which I frown on.
> Collection needs change, morph, and grow. If you tie your
> application to a specific "type", then the chances of
> getting screwed are high IME.
>
> If Smalltalk wants to rewrite them so features can be
> mixed and match as needed, instead of based on an limiting
> taxonomy, that would be great. (I am working on
> my version of such a spec, BTW.)

My imagination may be limited, so you can suggest what some of your suggestions
would be to change, morph, and grow collections and streams?


--
.tom





From topmind at technologist.com  Tue Feb 27 05:37:24 2001
From: topmind at technologist.com (Topmind)
Date: Tue, 27 Feb 2001 04:37:24 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com>
Message-ID: <MPG.1504d14563dce6499896af@news.earthlink.net>

> Topmind wrote:
> 
> > <snip>
> >
> > Smalltalk uses collection taxonomies, which I frown on.
> > Collection needs change, morph, and grow. If you tie your
> > application to a specific "type", then the chances of
> > getting screwed are high IME.
> >
> > If Smalltalk wants to rewrite them so features can be
> > mixed and match as needed, instead of based on an limiting
> > taxonomy, that would be great. (I am working on
> > my version of such a spec, BTW.)
> 
> My imagination may be limited, so you can suggest what some of your suggestions
> would be to change, morph, and grow collections and streams?
> 
> 
> --
> .tom
> 

Regarding collections, I am working on that one like I said. 
The main point is that you *do* stack operations on a collection.
You don't make the collection ONLY a stack, for example.
Thus, there is no "morphing" per se. It is a view, not
a type.

As far as streams, make a list of all possible stream
interfaces and attributes. When you do your part, I
will do mine. (I am still working on collection API's
right now, streams will have to wait unless you wish
to help.)

-tmind-


From aleaxit at yahoo.com  Tue Feb 27 08:06:57 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 08:06:57 +0100
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net>
Message-ID: <97fjla128ir@news2.newsguy.com>

"Topmind" <topmind at technologist.com> wrote in message
news:MPG.1504d14563dce6499896af at news.earthlink.net...
    [snip]
> The main point is that you *do* stack operations on a collection.

I may be missing your point.  In Standard C++, you may
_adapt_ *some* collections for use as a stack -- that is
what std::stack is for, it's an ADAPTER.  But *NOT* all
collections are usable as 'stacks' -- only those on which
order of insertion is significant and recorded; what would
it make to 'pop' the 'top' (what top?) of a multiset, say?


Alex





From topmind at technologist.com  Tue Feb 27 20:51:10 2001
From: topmind at technologist.com (Topmind)
Date: Tue, 27 Feb 2001 19:51:10 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com>
Message-ID: <MPG.1505a76831c34b749896b3@news.earthlink.net>

> "Topmind" <topmind at technologist.com> wrote in message
> news:MPG.1504d14563dce6499896af at news.earthlink.net...
>     [snip]
> > The main point is that you *do* stack operations on a collection.
> 
> I may be missing your point.  In Standard C++, you may
> _adapt_ *some* collections for use as a stack -- that is
> what std::stack is for, it's an ADAPTER.  But *NOT* all
> collections are usable as 'stacks' -- only those on which
> order of insertion is significant and recorded; what would
> it make to 'pop' the 'top' (what top?) of a multiset, say?
> 
> 
> Alex
> 
> 

Why can't the nodes of a multi-set ALSO have time-stamps
or a sequantial record/node number in them? 
If so, then in T.O.P. thinking it should be
a sinch to use it as a stack also. (Even if by chance the the API
does not provide stack operations, writing a 
Push and Pop function/method should be a snap.)

IOW, a "stack" or "multiset" is *how you use a
collection*, not how it is implemented or what
the original API used on it was. The HAS-A approach
to collections is just plain more flexible because
you are less locked into this-or-that thinking
and fences.

Requirements change and morph and merge.

-tmind-


From aleaxit at yahoo.com  Tue Feb 27 22:23:22 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 22:23:22 +0100
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net>
Message-ID: <97h5s40vhc@news2.newsguy.com>

"Topmind" <topmind at technologist.com> wrote in message
news:MPG.1505a76831c34b749896b3 at news.earthlink.net...
    [snip]
> Why can't the nodes of a multi-set ALSO have time-stamps
> or a sequantial record/node number in them?

Because that would be unwarranted overhead on 99.947%
of the uses of a multiset...?!

Why doesn't insertion into the multiset compute "priority"
value according to _all_ possible formulas, as long as it's
at it, so I can later decide that my multiset is actually a
priority queue, too...?


> Requirements change and morph and merge.

And when they do, one refactors appropriately (and reruns
one's tests -- that goes without saying).  If I need my items
timestamped, prioritized, turned into lowercase, or maybe
translated into Elbonian, as I insert them into a collection,
I'll be much happier providing or choosing the appropriate
collection, rather than have all collections try to be all things
to all items, thank you very much.


Alex





From topmind at technologist.com  Wed Feb 28 03:11:10 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 02:11:10 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com>
Message-ID: <MPG.1506007591cf818b9896bc@news.earthlink.net>

> "Topmind" <topmind at technologist.com> wrote in message
> news:MPG.1505a76831c34b749896b3 at news.earthlink.net...
>     [snip]
> > Why can't the nodes of a multi-set ALSO have time-stamps
> > or a sequantial record/node number in them?
> 
> Because that would be unwarranted overhead on 99.947%
> of the uses of a multiset...?!
> 
> Why doesn't insertion into the multiset compute "priority"
> value according to _all_ possible formulas, as long as it's
> at it, so I can later decide that my multiset is actually a
> priority queue, too...?
> 


I am only saying that IF there are timestamps. I did NOT say
it was a prerequisite for a multi-set. You misunderstood
me.


> 
> > Requirements change and morph and merge.
> 
> And when they do, one refactors appropriately 

Refactor is a great PHB-directed euphemism for
code rework. Call it whatever you want, but it is
still unnecessary work.

The divisions of collections into types *is*
arbitrary for the most part.


> (and reruns
> one's tests -- that goes without saying).  If I need my items
> timestamped, prioritized, turned into lowercase, or maybe
> translated into Elbonian, as I insert them into a collection,
> I'll be much happier providing or choosing the appropriate
> collection, rather than have all collections try to be all things
> to all items, thank you very much.
> 

I never said that the *engine* had to provide all those
services. 

I am only selling the idea of not changing the existing
calls when collection needs are added to or change.

Born a stack != Die a stack

> 
> Alex
> 
> 
> 
> 

-tmind-


From amk at mira.erols.com  Wed Feb 28 03:34:23 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 28 Feb 2001 02:34:23 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net>
Message-ID: <slrn99opop.gr.amk@mira.erols.com>

On Wed, 28 Feb 2001 02:11:10 GMT, Topmind <topmind at technologist.com> wrote:
>> And when they do, one refactors appropriately 
>
>Refactor is a great PHB-directed euphemism for
>code rework. Call it whatever you want, but it is
>still unnecessary work.

But given that requirements change and lacking a reliable Delphic
Oracle, it's impossible to divine the correct solution at the
beginning, so the code rework is unavoidable and necessary.

It would certainly be better if a writer like Leacock knew always what was
best to do and what would look best in the eyes of posterity, but such
unnatural foresight cannot be required of any man.
  -- Robertson Davies, _A Voice from the Attic_

You write a great program, regardless of language, by redoing it over & over &
over & over, until your fingers bleed and your soul is drained. But if you
tell newbies *that*, they might decide to go off and do something sensible,
like bomb defusing<wink>.
  -- Tim Peters, 5 Jun 1998

--amk







From topmind at technologist.com  Wed Feb 28 05:52:47 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 04:52:47 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <slrn99opop.gr.amk@mira.erols.com>
Message-ID: <MPG.1506265ebf6d934d9896be@news.earthlink.net>

> On Wed, 28 Feb 2001 02:11:10 GMT, Topmind <topmind at technologist.com> 
wrote:
> >> And when they do, one refactors appropriately 
> >
> >Refactor is a great PHB-directed euphemism for
> >code rework. Call it whatever you want, but it is
> >still unnecessary work.
> 
> But given that requirements change and lacking a reliable Delphic
> Oracle, it's impossible to divine the correct solution at the
> beginning, so the code rework is unavoidable and necessary.
> 
> [snip]

The point is to *minimize* the changes needed. If the
changes don't match a method's boundaries, you have MORE
work than what you usually have in task-oriented
procedural/relational programming.

For more details, please see:

http://www.geocities.com/tablizer/boundary.htm

-tmind-



From aleaxit at yahoo.com  Wed Feb 28 09:08:44 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 28 Feb 2001 09:08:44 +0100
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <slrn99opop.gr.amk@mira.erols.com> <MPG.1506265ebf6d934d9896be@news.earthlink.net>
Message-ID: <97ibmc0k47@news1.newsguy.com>

"Topmind" <topmind at technologist.com> wrote in message
news:MPG.1506265ebf6d934d9896be at news.earthlink.net...
> > On Wed, 28 Feb 2001 02:11:10 GMT, Topmind <topmind at technologist.com>
> wrote:
> > >> And when they do, one refactors appropriately
> > >
> > >Refactor is a great PHB-directed euphemism for
> > >code rework. Call it whatever you want, but it is
> > >still unnecessary work.
> >
> > But given that requirements change and lacking a reliable Delphic
> > Oracle, it's impossible to divine the correct solution at the
> > beginning, so the code rework is unavoidable and necessary.
> >
> > [snip]
>
> The point is to *minimize* the changes needed. If the
> changes don't match a method's boundaries, you have MORE
> work than what you usually have in task-oriented
> procedural/relational programming.

Not with a good (minimal) process, helpful tools, and, most
particularly, the right mindset (which is both cause and
consequence of the good/minimal process).  Before you knock
it, TRY it -- pick a pilot project of suitable size, a team
ditto, and have a go at it with eXtreme Programming -- code
test-first, *refactor mercilessly*, and *do the simplest thing
that can possibly work*, being the three keys here... and do
note that they work together, *sinergically*.  (No doubt you
will consider this "a euphemism", too, but, as usual, I'm just
striving to use words precisely; I don't particularly care if
others use should-be-precise words, such as 'sinergy' or
'refactor', as euphemistic cover-up for lack of substance).


> For more details, please see:
>
> http://www.geocities.com/tablizer/boundary.htm

This seems to show what blind alleys classical approaches can
lead one to -- ```Reorganization can be risky. Many managers do
not want programs altered except to add or fix features. If an
significant error occurs due to reorganization, then heads roll.
I have been told many times, "If it ain't broke, please don't
fix it."''' being the key quote from that paper.

>From that unproductive mindset, among many other ills, comes
the "boundary fixation" with which the rest of the article
appears to be obsessed.  (But the ills are many indeed; the
"lava flow" anti-pattern, for example, is likely to be quite
pervasive in a development shops to which this article could
apply).

But the right unit of encapsulation is NOT the method, nor
the object, nor the class; it's a *component*, a small to
middling set of classes/objects/methods (and, in languages
which allow other organizing structures, functions, modules,
etc), which are developed and tested and released _and, thus,
reused_ *together*.  Component boundaries are indeed of very
high relevance.  Boundaries _within_ a components are not.

[Restructuring in a way that alters component relationships
is one thing -- it is, indeed, *re-architecting* a whole
software *system*; only *system-level* (aka 'integration')
tests will catch re-architecting problems, a good part of
test-suites that unit-test specific components or small
groups thereof will have to be scrapped, etc, etc.]

Restructuring code, design, and even architecture *behind*
the abstract interfaces which relate components is (should
be!) a low-risk, high-frequence activity.  Test suites MUST
be good enough to provide very high ensurance that a group
of components is still working to specs -- there is just no
working around THAT; if your test suites are so deficient
at the component and unit level, "not fixing what isn't
broken" is not an option, as you CAN'T KNOW what IS and what
isn't broken!


*Analogy warning, proceed at your own risk*...!-)

```Crossing a street can be risky. Many parents do not want
streets crossed except to reach a school or doctor. If an
significant accident occurs due to crossing, then heads roll.
I have been told many times, "If it ain't truly indispensable
to be on the other side, please don't cross the street"'''.

Adopting the 'software that happens to appear to work when
not thoroughly tested is sacred and must not be touched' idea
is parallel to keeping a kid always on his/her side of the
street (except perhaps for emergencies such as schools and
doctors:-).  Not exactly conducive to optimal development
(be it of social skill and a healthy life, OR good software).

Providing safer approaches to street-crossing (underpasses
and overpasses on high-traffic thoroughfares, lots more
dead-end streets with speed bumps in residential areas,
etc) is a far preferable attitude to trying to build, on
the one block a kid is going to be confined to, all he or
she could possibly ever need -- ice-cream stand, burger
joint, public library, and so on -- which seems to me to
be what you're advocating in the software (specifically,
collection-interfaces) realm.


Alex





From fredrik at pythonware.com  Wed Feb 28 10:12:13 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 09:12:13 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <slrn99opop.gr.amk@mira.erols.com> <MPG.1506265ebf6d934d9896be@news.earthlink.net> <97ibmc0k47@news1.newsguy.com>
Message-ID: <Nh3n6.15937$Qb7.2602111@newsb.telia.net>

> Before you knock it, TRY it

the martelli battlebot vs. the "galileo of
programming" crackpot?

now this could be really entertaining ;-)

Cheers /F




From topmind at technologist.com  Wed Feb 28 17:53:09 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 16:53:09 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <slrn99opop.gr.amk@mira.erols.com> <MPG.1506265ebf6d934d9896be@news.earthlink.net> <97ibmc0k47@news1.newsguy.com>
Message-ID: <MPG.1506cf3293b094b49896c1@news.earthlink.net>

>
> Before you knock
> it, TRY it -- pick a pilot project of suitable size, a team
> ditto, and have a go at it with eXtreme Programming

Sure, hack-it-til-it-works indeed does work after enough
time. But whether it produces a better system in the
long run is another issue. Without better metrics, it is
just Yet Another Methodology to add to the heap.

How about you try heavy table-oriented programming for a 
few years.

> code test-first, *refactor mercilessly*,

I would rather use task-orientation and avoid
excessive refactoring to begin with.

"Refactoring" is a euphemism for "OO is not
change-friendly after all".

>
> This seems to show what blind alleys classical approaches can
> lead one to -- ```Reorganization can be risky. Many managers do
> not want programs altered except to add or fix features. If an
> significant error occurs due to reorganization, then heads roll.
> I have been told many times, "If it ain't broke, please don't
> fix it."''' being the key quote from that paper.

Well, they are the boss. Sell rework to them, not me.

> (But the ills are many indeed; the
> "lava flow" anti-pattern, for example, is likely to be quite
> pervasive in a development shops to which this article could
> apply).

What is this "lava flow" thing? I keep hearing it mentioned. 
More OO lore that nobody can keep up with. Scriptures 
build fast when the agnostics are silenced.

I have shown in great detail why procedural seems better able 
to deal with non-boundaried differences, and I have yet to
see counter-arguments other than "you get used to code
rework if you just give it a fancy name." 
I won't sell out to wasteful effort.


> But the right unit of encapsulation is NOT the method, nor
> the object, nor the class; it's a *component*, a small to
> middling set of classes/objects/methods (and, in languages
> which allow other organizing structures, functions, modules,
> etc), which are developed and tested and released _and, thus,
> reused_ *together*.  Component boundaries are indeed of very
> high relevance.  Boundaries _within_ a components are not.

But the easy-to-componentize stuff is already for sale. Why 
would they bring in programmers if plug-and-play solutions
were already available? (Report writers, sorters, DB's,
GUI IDE's, communications, etc.)

What is left is stuff that has a jillion interconnections
to other things such that putting a clean box around it
is not practical. Components usually work for things where
the algorithm is complex but the interfaces are small.
In biz stuff the interfaces/links to many entities is the
largest part.

Besides, you can characterize my development philosophy
as "task componentization". AKA, a procedural module.


> Adopting the 'software that happens to appear to work when
> not thoroughly tested is sacred and must not be touched' idea
> is parallel to keeping a kid always on his/her side of the
> street (except perhaps for emergencies such as schools and
> doctors:-).  Not exactly conducive to optimal development
> (be it of social skill and a healthy life, OR good software).

I am not against re-arrangement. What I am complaining about
is paradigms that *increase* the amount of re-arrangement that
is required. OO appears to do such, partly because of the 
boundary/granularity issues.

The whole refactoring thing is side-effect of OO. The paradigm
is starting to barf, and now the same people who sold you
the barfers are now also selling barf bags.

Whadda Racket!

If I was not so merit-minded, I would probably happily
cash in on it too. 

-tmind-



From joconnor at cybermesa.com  Wed Feb 28 18:03:57 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Wed, 28 Feb 2001 10:03:57 -0700
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <slrn99opop.gr.amk@mira.erols.com> <MPG.1506265ebf6d934d9896be@news.earthlink.net> <97ibmc0k47@news1.newsguy.com> <MPG.1506cf3293b094b49896c1@news.earthlink.net>
Message-ID: <3A9D2F7D.A5FB0A30@cybermesa.com>

Topmind wrote:

>
> The whole refactoring thing is side-effect of OO.

The whole concept of refactoring comes fom SA/SD. It's also known as
"Functional Decomposition". With statements like this you clearly
illustrate that a) you don't know OO and b) you don't even understand
your own chosen tools either

--
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython




From aleaxit at yahoo.com  Wed Feb 28 10:12:04 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 28 Feb 2001 10:12:04 +0100
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net>
Message-ID: <97ifd60sen@news1.newsguy.com>

"Topmind" <topmind at technologist.com> wrote in message
news:MPG.1506007591cf818b9896bc at news.earthlink.net...
> > "Topmind" <topmind at technologist.com> wrote in message
> > news:MPG.1505a76831c34b749896b3 at news.earthlink.net...
> >     [snip]
> > > Why can't the nodes of a multi-set ALSO have time-stamps
> > > or a sequantial record/node number in them?
> >
> > Because that would be unwarranted overhead on 99.947%
> > of the uses of a multiset...?!
> >
> > Why doesn't insertion into the multiset compute "priority"
> > value according to _all_ possible formulas, as long as it's
> > at it, so I can later decide that my multiset is actually a
> > priority queue, too...?
>
> I am only saying that IF there are timestamps. I did NOT say
> it was a prerequisite for a multi-set. You misunderstood me.

IF there are to be timestamps on insertion of items into a
collection, a rare but far from unthinkable requirement, I do
not want the collection to do it behind my back -- just as I
do not want "prioritizing" to be happening without me knowing
that (and thus being able, if need be, to customize the way
items are timestamped, prioritized, etc).


> > > Requirements change and morph and merge.
> >
> > And when they do, one refactors appropriately
>
> Refactor is a great PHB-directed euphemism for
> code rework. Call it whatever you want, but it is
> still unnecessary work.
>
> The divisions of collections into types *is*
> arbitrary for the most part.

There is definitely a lot of arbitrariness if the division
is made into a _taxonomy_ -- as single-inheritance thinking
may suggest.  I'm definitely not advocating or defending
the exclusive use of taxonomy, which IMHO tends to breed a
somewhat warped worldview -- the natural view of abstract
algebra does not lead to that, for example; in chains of
specialization such as groupoid->semigroup->monoid->group,
one may add a condition such as 'commutative' at different
steps, so there is no taxonomic 'single-inheritance' chain
leading to 'Abelian group'.  However, reacting to the
modeling over-restrictiveness of taxonomy with a bid to
drop all _classification_ is, IMHO, an over-reaction; there
IS a lot of structure in classification-_DAG_s, after all
(and I'm not entering, here, into the issue of whether
taxonomy is 'close enough' to effectively model *real-world*
domains for programming purposes -- the indisputable success
of Smalltalk does show that single-inheritance is *practically
workable* for many tasks -- a general discussion of collection
classes, IMHO, gets us closer to domains such as algebra, than
to real-world domain modeling, anyway).


> I never said that the *engine* had to provide all those
> services.
>
> I am only selling the idea of not changing the existing
> calls when collection needs are added to or change.
>
> Born a stack != Die a stack

It's unclear whether, here, you're talking about an object
(keeping its _identity_) dynamically changing its _class_ (I
didn't think Smalltalk supported that, but I may easily be
wrong on that score; Python certainly does).  In my humble
opinion, though, such extreme dynamism, while sometimes it
proves VERY pragmatically useful (e.g., change a "proxy"
placeholder class to a full-blown implementation class in
a demand-driven, just-in-time manner), is not normally very
useful in the _modeling_ (analysis) phase -- reasoning in
terms of 'unchanging' class-membership tends to simplify
the analysis, even if practical needs drive one to make use
of the extremely-dynamic approaches in design & coding.

Assuming you're not talking of runtime class-changes, but
of modeling-time (analysis-time) 'behavior', then I'd still
want to refactor when the needs have changed so much as to
make me need (e.g.) a multiset where I used to use a stack.
But then, I addressed the refactoring issue (and what it
implies in terms of process and mindset) in another post to
this thread a very short time ago -- and I haven't read
anything from you that would change my mind on that (surely
not the lava-flow-prone approach in the URL you gave:-).


Alex





From topmind at technologist.com  Wed Feb 28 16:55:36 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 15:55:36 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <97ifd60sen@news1.newsguy.com>
Message-ID: <MPG.1506c16edd0cb6159896bf@news.earthlink.net>

> "Topmind" <topmind at technologist.com> wrote in message
> news:MPG.1506007591cf818b9896bc at news.earthlink.net...
> > > "Topmind" <topmind at technologist.com> wrote in message
> > > news:MPG.1505a76831c34b749896b3 at news.earthlink.net...
> > >     [snip]
> > > > Why can't the nodes of a multi-set ALSO have time-stamps
> > > > or a sequantial record/node number in them?
> > >
> > > Because that would be unwarranted overhead on 99.947%
> > > of the uses of a multiset...?!
> > >
> > > Why doesn't insertion into the multiset compute "priority"
> > > value according to _all_ possible formulas, as long as it's
> > > at it, so I can later decide that my multiset is actually a
> > > priority queue, too...?
> >
> > I am only saying that IF there are timestamps. I did NOT say
> > it was a prerequisite for a multi-set. You misunderstood me.
> 
> IF there are to be timestamps on insertion of items into a
> collection, a rare but far from unthinkable requirement, I do
> not want the collection to do it behind my back -- just as I
> do not want "prioritizing" to be happening without me knowing
> that (and thus being able, if need be, to customize the way
> items are timestamped, prioritized, etc).
> 


No no no! I am not advocating forced time-stamps. (Although,
a sequential unique record is quite often needed. But nothing
has to be forced.)

However, timestamping *is* a common need in my domain.


> 
> > > > Requirements change and morph and merge.
> > >
> > > And when they do, one refactors appropriately
> >
> > Refactor is a great PHB-directed euphemism for
> > code rework. Call it whatever you want, but it is
> > still unnecessary work.
> >
> > The divisions of collections into types *is*
> > arbitrary for the most part.
> 
> There is definitely a lot of arbitrariness if the division
> is made into a _taxonomy_ -- as single-inheritance thinking
> may suggest.  I'm definitely not advocating or defending
> the exclusive use of taxonomy, which IMHO tends to breed a
> somewhat warped worldview -- the natural view of abstract
> algebra does not lead to that, for example; in chains of
> specialization such as groupoid->semigroup->monoid->group,
> one may add a condition such as 'commutative' at different
> steps, so there is no taxonomic 'single-inheritance' chain
> leading to 'Abelian group'.  However, reacting to the
> modeling over-restrictiveness of taxonomy with a bid to
> drop all _classification_ is, IMHO, an over-reaction; 

When in doubt, leave it out.

Unless one can see and show some natural law of the Universe that
forces things into mutually-exclusive taxonomies, it is
bad, arbitrary modeling, and OO sould be ashamed of itself for
over promoting it. I hope the backlash is as strong as
the hype that propelled it onto the packaging buzzword
standards.

Some domains like chemistry, math, etc. might be able
to take advantage of natural divisions of "sub-things".
However, many domains don't, and OO seems like nothing
but added syntactical and modeling clutter without
sub-divisioning. Two out of the 3 OO Cononical Triads
depend on it. There is no polymorphism if there is
only one "type" with a bunch of independant attributes -
HAS-A reigns and sub-IS-A is rare.
(which is how it should be in biz modeling IME.)

> ....
> a general discussion of collection
> classes, IMHO, gets us closer to domains such as algebra, than
> to real-world domain modeling, anyway).

Algebra? Programming paradigms are mostly for human
communiction, not machine-to-machine. If many humans
don't model the world via algebra, then I don't see
what that gets one.


> It's unclear whether, here, you're talking about an object
> (keeping its _identity_) dynamically changing its _class_ 

No! There is no class to change to. There would be *one*
and only one collection protocol. Not all engines would
implement every feature of the protocol, but the idea
is that you don't pick a protocol that STICKS you with
only say stack-centric protocols or commands.

I don't know about your domain, but I observe collections
changing and growing in my domain. Starting out with
just a dictionary (hash) can be a real pain if you later
need more columns or more indexes, for example.

Look at even a call stack in an interpreter. It would
be nice to "X-ray" it and view it with different
sorts and filters for complex debugging. This is what
I like about tables. Studying the "state" of all kinds
of angles is a sinch: you don't have to re-invent 
micro querie languages, write save-to-file utilities,
etc., because they are usually built in. I like
plug-and-play power. Perhaps some bit-diddling personalities
like to re-invent little wheels, but I like to push simple
buttons and get powerful results.


> and I haven't read
> anything from you that would change my mind on that (surely
> not the lava-flow-prone approach in the URL you gave:-).

It seems because you are USED TO refactoring when your
collection needs change. I am spoiled. I know better.

Or, your domain may be different. 

> the indisputable success
> of Smalltalk does show that ....

I would hardly call it "indisputed". Smalltalk is sort of
the Frank Zappa of programming languages. It gets rave
reviews and it is intellectually stimulating. However,
it never quite catches on with general programmers.

I think because its power does not live up to it's haughtiness.

Using procedural and table techniques, I have matched
Smalltalk "can-you-do-this" challenges in the past,
and usually with more K.I.S.S.

Thanks for your feedback
-tmind-




From mickeyz at qwest.net  Wed Feb 28 17:37:11 2001
From: mickeyz at qwest.net (Randy A. Ynchausti)
Date: Wed, 28 Feb 2001 09:37:11 -0700
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <97ifd60sen@news1.newsguy.com> <MPG.1506c16edd0cb6159896bf@news.earthlink.net>
Message-ID: <SJ9n6.2187$Yw2.141138@news.uswest.net>

Topmind

> Using procedural and table techniques, I have matched
> Smalltalk "can-you-do-this" challenges in the past,
> and usually with more K.I.S.S.

Wow -- what are you takin?  It sure must be potent!

Maybe I am wrong, maybe it's not chemical -- maybe you have just linked up
with the guru who mentored the Beatles?  Anyway it sure seems to work for
you.

Randy




From topmind at technologist.com  Wed Feb 28 18:06:20 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 17:06:20 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <97ifd60sen@news1.newsguy.com> <MPG.1506c16edd0cb6159896bf@news.earthlink.net> <SJ9n6.2187$Yw2.141138@news.uswest.net>
Message-ID: <MPG.1506d244a87ac34b9896c3@news.earthlink.net>

> Topmind
> 
> > Using procedural and table techniques, I have matched
> > Smalltalk "can-you-do-this" challenges in the past,
> > and usually with more K.I.S.S.
> 
> Wow -- what are you takin?  It sure must be potent!
> 
> Maybe I am wrong, maybe it's not chemical -- maybe you have just linked up
> with the guru who mentored the Beatles?  Anyway it sure seems to work for
> you.
> 
> Randy
> 


Lookin in the mirror?

OO is far more Guru-oriented than anything before it. It is
generating tons of lore, scriptures, and meditation materials
without any objective proof or metrics that it is better.

One has to have years of "procedural DEprogramming" to
"get it"? The Patty Herst paradigm?

We have not seen such self-feeding intellectual fraud since
ecomonic communism, Elron Hubbard, and Utah Cold Fusion.

-tmind-


From Brian.Gridley at LibertyMutual.com  Wed Feb 28 19:21:40 2001
From: Brian.Gridley at LibertyMutual.com (Brian Gridley)
Date: Wed, 28 Feb 2001 12:21:40 -0600
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <97ifd60sen@news1.newsguy.com> <MPG.1506c16edd0cb6159896bf@news.earthlink.net> <SJ9n6.2187$Yw2.141138@news.uswest.net> <MPG.1506d244a87ac34b9896c3@news.earthlink.net>
Message-ID: <97jfj3$9e2@gazette.lmig.com>

I can imagine some limited places where this would be so, or some
constraints that would make this so, but it is hard to imagine on a project
with any depth.  I did procedural and 4GL for 15 years, and Smalltalk for 5.
I helped write a major system in Smalltalk at the same time our competitor
wrote something from nearly identical specs.  Ours cost 1/3 of theirs,
supplied more features, and is far more flexible.  Perhaps you can provide
specifics to enlighten us as to what you have found procedural to do better.

"Topmind" <topmind at technologist.com> wrote in message
news:MPG.1506d244a87ac34b9896c3 at news.earthlink.net...
> > > Using procedural and table techniques, I have matched
> > > Smalltalk "can-you-do-this" challenges in the past,
> > > and usually with more K.I.S.S.
> >
...> OO is far more Guru-oriented than anything before it. It is
> generating tons of lore, scriptures, and meditation materials
> without any objective proof or metrics that it is better...





From tgagne at ix.netcom.com  Tue Feb 27 16:12:57 2001
From: tgagne at ix.netcom.com (Thomas Gagne)
Date: Tue, 27 Feb 2001 10:12:57 -0500
Subject: Stream interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net>
Message-ID: <3A9BC3F9.1929EB9A@ix.netcom.com>

I don't think it's any of the programmer's business what the attributes are,
unless of course, you asking for Stream's heirarchy:
    Steam
        PeekableStream
            PositionableStream
                ExternalStream
                    BufferedExternalStream
                        ExternalReadStream
                            ExternalReadWriteStream
                        ExternalWriteStream
                InternalStream
                    WriteStream
                        ReadWriteStream

There may be other subclasses floating around I'm not aware of.  It should be
noted all of these classes, except the External* variety can be created on any
collection (other Smalltalker's can chirp in here if I'm going astray).

To list all the interfaces is a bit of a chore.  It may be easier to check to
see if there's something missing you believe should be provided.

Just checking Stream's method dictionary I find:

Stream MethodDictionary (#nextPut: #contents #atEnd #next: #do: #nextPutAll:
#nextMatchFor: #next:put: #next #space #nextChunkPut: #timeStamp #close #cr
#print: #store: #contentsSpecies #upToAll1: #next:into:startingAt: #emphasis
#emphasis: #crtab: #tab #crtab #isReadable #through: #policy #upTo:
#next:putAll:startingAt: #commit #throughAll: #upToEnd #flush
#nextAvailable:into:startingAt: #needsFileLineEndConversion
#computePrefixFunctionFor: #isWritable #isExternalStream #skipThrough:
#throughAll1: #nextAvailable: #skipToAll1: #tab: #ioConnection
#upToAll1:returnMatch:includePattern: )

There are others implemented in Stream's subclasses, but I expect this is the
majority of them.  I've been able to find everything I've been looking for
somewhere in Stream's methods, or at least one of its descendents.

--
.tom





From topmind at technologist.com  Tue Feb 27 21:07:13 2001
From: topmind at technologist.com (Topmind)
Date: Tue, 27 Feb 2001 20:07:13 GMT
Subject: Stream interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <3A9BC3F9.1929EB9A@ix.netcom.com>
Message-ID: <MPG.1505ab2efd31b1639896b4@news.earthlink.net>

[quote]
Steam
        PeekableStream
            PositionableStream
                ExternalStream
                    BufferedExternalStream
                        ExternalReadStream
                            ExternalReadWriteStream
                        ExternalWriteStream
                InternalStream
                    WriteStream
                        ReadWriteStream
[end quote]

I admit that I am not an expert on streams, but this 
hierarchy seams tree-happy to me.

Why not have something like this instead:

s = newStream(....)
s.peekable = true
s.positionable = false
s.external = blah
s.buffered = blah
s.readAble = true
s.writeAble = true
[use the stream now.....]

(Note that the dot syntax could be a 
dictionary instead of an object, BTW.
Also note that these would have *defaults*
so that not every setting needs to be
explicitly set.)

This setup is much easier to manage and grok IMO.  It also allows *more*
potential combinations, like buffered internal streams. (I can't think of
why anybody would want such, but why limit the future to one's
current lack of imagination. Perhaps they want to test buffering 
locally.)

I want to flatten GUI's and Collections in a similar way. 

Flatten flatten flatten!

-flatmind-

(But I am not flat-footed. I have a very prominent arch.)




From jarober at cc36623-b.hwrd1.md.home.com  Tue Feb 27 23:03:04 2001
From: jarober at cc36623-b.hwrd1.md.home.com (jarober at cc36623-b.hwrd1.md.home.com)
Date: Tue, 27 Feb 2001 22:03:04 GMT
Subject: Stream interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.ea <MPG.1505ab2efd31b1639896b4@news.earthlink.net>
Message-ID: <97h8d3$l03$1@cc36623-b.hwrd1.md.home.com>

In article <MPG.1505ab2efd31b1639896b4 at news.earthlink.net>,
Topmind <topmind at technologist.com> wrote:
>
>

<snip>

>I admit that I am not an expert on streams, but this 
>hierarchy seams tree-happy to me.
>

stream := (String new: 1000) readStream.

stream := 'myFile' asFileName writeStream.

Same idea, but simpler...


>Why not have something like this instead:
>
>s = newStream(....)
>s.peekable = true
>s.positionable = false
>s.external = blah
>s.buffered = blah
>s.readAble = true
>s.writeAble = true
>[use the stream now.....]
>
>(Note that the dot syntax could be a 
>dictionary instead of an object, BTW.
>Also note that these would have *defaults*
>so that not every setting needs to be
>explicitly set.)
>
>This setup is much easier to manage and grok IMO.  It also allows *more*
>potential combinations, like buffered internal streams. (I can't think of
>why anybody would want such, but why limit the future to one's
>current lack of imagination. Perhaps they want to test buffering 
>locally.)
>
>I want to flatten GUI's and Collections in a similar way. 
>
>Flatten flatten flatten!
>
>-flatmind-
>
>(But I am not flat-footed. I have a very prominent arch.)
>
>


-- 
James A. Robertson
Product Manager (Smalltalk), Cincom
jarober at mail.com
<Talk Small and Carry a Big Class Library>


From topmind at technologist.com  Wed Feb 28 03:03:49 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 02:03:49 GMT
Subject: Stream interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.ea <MPG.1505ab2efd31b1639896b4@news.earthlink.net> <97h8d3$l03$1@cc36623-b.hwrd1.md.home.com>
Message-ID: <MPG.1505febb4e41f7f69896bb@news.earthlink.net>

> In article <MPG.1505ab2efd31b1639896b4 at news.earthlink.net>,
> Topmind <topmind at technologist.com> wrote:
> >
> >
> 
> <snip>
> 
> >I admit that I am not an expert on streams, but this 
> >hierarchy seams tree-happy to me.
> >
> 
> stream := (String new: 1000) readStream.
> 
> stream := 'myFile' asFileName writeStream.
> 
> Same idea, but simpler...
> 

Simpler than what?

What I showed was illustrative of the concept, not the
utmost fewest syntax. Names can be given to common
combinations, for example. (This could be in the 
form of functions, or a built-in configuration names.)

You are not selling hierarchies, but shortcut names.

My approach has WAY more possible combos than a tree.

Make a little grid and see for yourself.

> 
> >Why not have something like this instead:
> >
> >s = newStream(....)
> >s.peekable = true
> >s.positionable = false
> >s.external = blah
> >s.buffered = blah
> >s.readAble = true
> >s.writeAble = true
> >[use the stream now.....]
> >
> >(Note that the dot syntax could be a 
> >dictionary instead of an object, BTW.
> >Also note that these would have *defaults*
> >so that not every setting needs to be
> >explicitly set.)
> >
> >This setup is much easier to manage and grok IMO.  It also allows *more*
> >potential combinations, like buffered internal streams. (I can't think of
> >why anybody would want such, but why limit the future to one's
> >current lack of imagination. Perhaps they want to test buffering 
> >locally.)
> >
> >I want to flatten GUI's and Collections in a similar way. 
> >
> >Flatten flatten flatten!
> >
> >-flatmind-
> >
> >(But I am not flat-footed. I have a very prominent arch.)
> >
> >

-tmind-


From sholden at holdenweb.com  Wed Feb 28 00:11:49 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 27 Feb 2001 18:11:49 -0500
Subject: Stream interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <3A9BC3F9.1929EB9A@ix.netcom.com> <MPG.1505ab2efd31b1639896b4@news.earthlink.net>
Message-ID: <JyWm6.834$zL.28252@e420r-atl1.usenetserver.com>

"Topmind" <topmind at technologist.com> wrote in message
news:MPG.1505ab2efd31b1639896b4 at news.earthlink.net...
>
[ ... ]
> I want to flatten GUI's and Collections in a similar way.
>
> Flatten flatten flatten!
>
> -flatmind-
>
> (But I am not flat-footed. I have a very prominent arch.)
>
Troll, troll, troll (but I am not a troll).

regards
 Steve





From topmind at technologist.com  Wed Feb 28 02:55:33 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 01:55:33 GMT
Subject: Stream interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <3A9BC3F9.1929EB9A@ix.netcom.com> <MPG.1505ab2efd31b1639896b4@news.earthlink.net> <JyWm6.834$zL.28252@e420r-atl1.usenetserver.com>
Message-ID: <MPG.1505fcd56fa604e49896ba@news.earthlink.net>

> > I want to flatten GUI's and Collections in a similar way.
> >
> > Flatten flatten flatten!
> >
> > -flatmind-
> >
> > (But I am not flat-footed. I have a very prominent arch.)
> >
> Troll, troll, troll (but I am not a troll).
> 
> regards
>  Steve
> 
> 


I will admit I am a troll if you admit that there is no proof
that OO is better for most domains.

-tmind-


From dwig at advancedmp.net  Wed Feb 28 23:40:12 2001
From: dwig at advancedmp.net (Don Dwiggins)
Date: 28 Feb 2001 14:40:12 -0800
Subject: Stream interfaces
In-Reply-To: topmind@technologist.com's message of "Tue, 27 Feb 2001 20:07:13 GMT"
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com>
	<CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net>
	<220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com>
	<MPG.150312ceb68f1ada98969f@news.earthlink.net>
	<8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com>
	<MPG.1503548620e632919896a4@news.earthlink.net>
	<MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com>
	<MPG.150394195f7f04a19896a7@news.earthlink.net>
	<3A9A86B1.9A8C69A2@ix.netcom.com>
	<MPG.1504d14563dce6499896af@news.earthlink.net>
	<3A9BC3F9.1929EB9A@ix.netcom.com>
	<MPG.1505ab2efd31b1639896b4@news.earthlink.net>
Message-ID: <ud7c2a0tv.fsf@advancedmp.net>

Topmind writes:
> Why not have something like this instead:

> s = newStream(....)
> s.peekable = true
> s.positionable = false
> s.external = blah
> s.buffered = blah
> s.readAble = true
> s.writeAble = true
> [use the stream now.....]

> (Note that the dot syntax could be a 
> dictionary instead of an object, BTW.
> Also note that these would have *defaults*
> so that not every setting needs to be
> explicitly set.)

A couple of reactions:

- Could this kind of thing be achieved using multiple inheritance with
  properly constructed mixins?  (For example, do the Common Lisp or Eiffel
  folks use mixins to this effect?)

- I've looked a very little at Aspect-Oriented Programming, which was pretty
  much at the research stage when I peeked.  Based on that little bit of
  knowledge, this looks like it might be along the same lines (maybe an
  application of AOP).  Sorry, I don't have a good URL handy, but it should
  be a good search term for your favorite engine.

-- 
Don Dwiggins                    "It fills my head with ideas,
Advanced MP Technology           but I don't quite know what they are"
dwig at advancedmp.net               -- Alice, after hearing 'Jabberwocky'




From ben.hutchings at roundpoint.com  Fri Feb 23 03:02:03 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 22 Feb 2001 18:02:03 -0800
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com>
Message-ID: <uae7ekvhg.fsf@roundpoint.com>

Chris Hanson <cmh at bDistributed.com> writes:

> In article <CgJk6.515$v77.66568 at newsread2.prod.itd.earthlink.net>,
> Chris Tavares <ctavares at develop.com> wrote:
> 
> > While the world is totally different (static vs. dynamic typing), the
> > concepts are in many ways the same. Studying the ideas behind the STL are
> > well worth doing for everyone I would think.
> 
> It's a great idea for learning what *not* to do.
> 
> The STL is horribly designed.  There's no hierarchy,

I think you mean it's not an OO design, and that's true.

There *are* common interfaces (random lookup, list-ops, front-ops,
stack-ops, iteration), but those interfaces aren't explicit so you
can't put off your choice of implementation to run-time - unless you
use wrappers.

> no real interchangeability in the container classes.

Some algorithms require different interfaces from others.  You can
interchange containers as long as you keep a clear idea of which
interfaces your algorithm(s) need and pick the right containers.  The
most common one is iteration, and almost all containers support that.

> And they force you to do all sorts of extra work (some of which just
> isn't feasible) to be able to store your objects in containers.

Storing polymorphic objects certainly seems to be too difficult, since
they have to be stored by reference and there is no ref-counting
smart-pointer class in the standard library (but there is
boost::shared_ptr).

> STL was designed by number crunchers with the broken assumption that
> programs are designed from the algorithm out and that efficiency is the
> absolute most important goal.  Software in the real world is designed
> from the business requirements out; this usually starts with a human
> interface and a set of user requirements, and efficiency is only
> measured compared to human use rather than the processor's maximum
> throughput.
<snip>

STL is not meant to do everything *for* the C++ programmer, but it is
meant to be used *by* every C++ programmer.  That's why it has to be
efficient and flexible, and not too abstract.  You can always write
wrapper classes on top of it that are more abstract but less flexible
or less efficient, but you could never write wrapper classes to make
an abstract container library more efficient!

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From lopemanc at swbell.net  Wed Feb 21 15:13:36 2001
From: lopemanc at swbell.net (Chris Lopeman)
Date: Wed, 21 Feb 2001 08:13:36 -0600
Subject: New to OO concepts - re-usability
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com>
Message-ID: <3A93CD10.C2123CE4@swbell.net>

Actually much of what you have talked about here is simply covered under different
topics.  Mostly it is referred to as polymorphism.  This is one of the three basic
concepts behind OO.  Another subtopic that touches on this is "inheritance based on
behavior".  While much inheritance is based on data (and this is how people are
usually taught) the more advanced you become you will learn how to recognize where
inheritance should take place based on behavior.

Chris Lopeman
Object Link Inc.

Thomas Gagne wrote:

> I just had an interesting conversation today with my brother who is learning
> programming from scratch.  He's learning Smalltalk, and was watching me debug a
> program I'm working on and we visited the subject of reuse.
>
> One of the things I learned recently (over in comp.lang.python, I think) was
> that it is as important to reuse idioms as it is to reuse code.  By designing
> your classes to work within the idioms of the language you're using your
> program ultimately reuses more, and becomes more reusable (and more easily
> factored) itself.
>
> For instance, I could create all kinds of classes with their own protocol for
> doing something, or I could endow them with behaviors that make them great
> collection citizens.  In Smalltalk, that would be thinking of how using do:,
> select:, collect:, detect:, and inject:into: on collection containing my
> objects would be either easier or more useful.
>
> Long-time Smalltalkers, and apparently Python programmers, may have already had
> this drilled into their heads.  I don't remember reading anything about
> designing software to work well within the idiom of any language in any of the
> books I've read.  Maybe others have?
>
> x wrote:
>
> > Hi,
> >
> > I have been reading up on the OO environment, focusing on the use of UML to
> > speed up creation of objects, or even to create superior designs of objects.
> >
> > Its great. Everyone is creating their reusable objects, no efforts are being
> > wasted, these huge libraries are being built....are they?
> >
> > Does everyone create their own libraries, with a little sharing when
> > requested, or is there some huge repository of objects available to all and
> > sundry where the benefits of mankind's OO efforts are being stored nice and
> > neat... you know, like the Human Genome project?
> >
> > I'm curious to know.
> >
> > Thanks
> >
> > Michael
> > michael.wijtenburg at iafrica.com
>
> --
> .tom



From brangdon at cix.co.uk  Sat Feb 24 07:09:00 2001
From: brangdon at cix.co.uk (Dave Harris)
Date: Sat, 24 Feb 2001 06:09 +0000 (GMT Standard Time)
Subject: New to OO concepts - re-usability
References: <3A93CD10.C2123CE4@swbell.net>
Message-ID: <memo.20010224060903.1005A@brangdon.madasafish.com>

lopemanc at swbell.net (Chris Lopeman) wrote (abridged):
> Actually much of what you have talked about here is simply covered 
> under different topics.  Mostly it is referred to as polymorphism.

The principle would be something like:
    Maximise polymorphism by avoiding gratuitous differences of
    syntax/protocol.

This is a bit more than polymorphism itself. I think it is a fairly 
subtle and difficult guideline to follow. People get gratuitous 
differences without realising it.

One variation is Meyer's "Uniform Reference Principle". For example, in 
C++ and Java (and Smalltalk), direct access to an instance variable has 
different syntax to a no-argument message send. Thus:

    a = b.c; // Variable.
    a = b.c(); // Function call.

The function call has spurious brackets. If we can drop them we get more 
polymorphism (if only at compile/link time).

Meyer takes this uniformity to fairly extreme lengths in the Eiffel 
libraries. For example, the STACK protocol doesn't use the usual push/pop 
names, instead it copies the names used by other collection types. It can 
be surprising at first but the benefits are apparently worth it.

There are benefits even if there is not likely to be an opportunity for 
polymorphism. If, say, both windows and files use open/close terminology, 
it is easier for programmers who know one to learn the other.

  Dave Harris, Nottingham, UK | "Weave a circle round him thrice,
      brangdon at cix.co.uk      |   And close your eyes with holy dread,
                              |  For he on honey dew hath fed
 http://www.bhresearch.co.uk/ |   And drunk the milk of Paradise."


From n8gray at caltech.edu.is.my.e-mail.address  Wed Feb 21 04:56:59 2001
From: n8gray at caltech.edu.is.my.e-mail.address (Nathan Gray)
Date: Tue, 20 Feb 2001 19:56:59 -0800
Subject: execvpe weirdness
Message-ID: <3A933C8B.8040103@caltech.edu.is.my.e-mail.address>

When I run the following script I get this error message:

Error: Cannot find images in /usr/local/bin
Apparently, Webots is not properly installed
You must set the environment variable _HOME to the
location of the  directory and re-run .
For example: export _HOME=/usr/local/
or:          setenv _HOME /usr/local/ 

This is the error you should get if WEBOTS_HOME is not set correctly, 
but the error should complain about WEBOTS_HOME, not _HOME!

Can somebody tell me where I've gone wrong?

Thanks,
-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
               Nathaniel Gray
     California Institute of Technology
      Computation and Neural Systems
       n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._

------------------- Text of runwebots ----------------------
#!/usr/bin/env python

import os, sys

# Parse command line
if len(sys.argv) < 2:
     print "Usage: runwebots (203|201)"
     sys.exit(1)

elif sys.argv[1] == "201":
     webotsDir = "/usr/local/webots"
     rcFile = ".webotsrc201"

elif sys.argv[1] == "203":
     webotsDir = "/home1/webots"
     rcFile = ".webotsrc203"

else:
     print "Usage: runwebots (203|201)"
     sys.exit(-1)

# Set up filenames and env. variables
execName = webotsDir + "/webots"
os.environ["WEBOTS_HOME"]=webotsDir
os.environ["LD_LIBRARY_PATH"]=""

# Make a link to the appropriate config file
home = os.environ["HOME"]
try:
     os.remove(home + "/.webotsrc")
except OSError:
     pass #link doesn't exist
os.symlink(rcFile, home + "/.webotsrc")

# And away we go
print "Running " + execName
os.execvpe(execName, [''], os.environ)



From n8gray at caltech.edu.is.my.e-mail.address  Wed Feb 21 05:43:40 2001
From: n8gray at caltech.edu.is.my.e-mail.address (Nathan Gray)
Date: Tue, 20 Feb 2001 20:43:40 -0800
Subject: execvpe weirdness
References: <3A933C8B.8040103@caltech.edu.is.my.e-mail.address>
Message-ID: <3A93477C.9050701@caltech.edu.is.my.e-mail.address>

Nathan Gray wrote:

> When I run the following script I get this error message:
> 
> Error: Cannot find images in /usr/local/bin
> Apparently, Webots is not properly installed
> You must set the environment variable _HOME to the
> location of the  directory and re-run .
> For example: export _HOME=/usr/local/
> or:          setenv _HOME /usr/local/
> This is the error you should get if WEBOTS_HOME is not set correctly, 
> but the error should complain about WEBOTS_HOME, not _HOME!
> 
> Can somebody tell me where I've gone wrong?
> 
> Thanks,
> -n8

OK, thanks to a dejanews search I figured out that I need to say:
	execvpe(execName, [execName], os.environ)
or
	execle(execName, execName, os.environ)

and I think I even understand why -- it's so you can call an executable 
with something other than the name of the executable as argv[0].  I can 
sorta see why this is useful, but it would be nice if the documentation 
for exec* was a little more explicit about the nature of "arg0".  A line 
like:  "Note:  arg0 is traditionally the name of the executable, and 
some executables will depend on this." would have saved me a lot of 
hair-pulling.

Or have I misunderstood something?

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
               Nathaniel Gray
     California Institute of Technology
      Computation and Neural Systems
       n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From n8gray at caltech.edu.is.my.e-mail.address  Wed Feb 21 06:09:46 2001
From: n8gray at caltech.edu.is.my.e-mail.address (Nathan Gray)
Date: Tue, 20 Feb 2001 21:09:46 -0800
Subject: execvpe weirdness
References: <3A933C8B.8040103@caltech.edu.is.my.e-mail.address> <3A93477C.9050701@caltech.edu.is.my.e-mail.address>
Message-ID: <3A934D9A.1040401@caltech.edu.is.my.e-mail.address>

Nathan Gray wrote


> OK, thanks to a dejanews search I figured out that I need to say:
>     execvpe(execName, [execName], os.environ)
> or
>     execle(execName, execName, os.environ)
> 
> and I think I even understand why -- it's so you can call an executable 
> with something other than the name of the executable as argv[0].  I can 
> sorta see why this is useful, but it would be nice if the documentation 
> for exec* was a little more explicit about the nature of "arg0".  A line 
> like:  "Note:  arg0 is traditionally the name of the executable, and 
> some executables will depend on this." would have saved me a lot of 
> hair-pulling.
> 
> Or have I misunderstood something?
> 
> -n8

Since we've already decided that flaming onesself is inappropriate on 
this newsgroup <wink>, I'll simply point out to poor Nathan that the 
documentation in 2.0 is much better on the exec* functions.

"""
The various exec*() functions take a list of arguments for the new 
program loaded into the process. In each case, the first of these 
arguments is passed to the new program as its own name rather than as an 
argument a user may have typed on a command line. For the C programmer, 
this is the argv[0] passed to a program's main(). For example, 
"os.execv('/bin/echo', ['foo', 'bar'])" will only print "bar" on 
standard output; "foo" will seem to be ignored.
"""

Maybe I'll make a witty remark about the time machine, too.

In response, Nathan will say that although the documentation is indeed 
better for 2.0, it's still a bit confusing for those who don't program 
in C, and he thinks that his suggestion still might help.

To which I'll respond that if he wants to see this happen he should 
really send his suggestion to python-docs at python.org.  Depending on my 
mood, I may also chastise him for not looking for the correct venue 
before posting.

With a mild air of indignity he'll reply that he already *has* sent his 
suggestion to python-docs at python.org.

I'll then apologize, and the two of us will co-write, co-direct, and 
co-star in a movie about long-separated twin cops who reunite to track 
down a wacky gang of incompetent jewel thieves.

Thanks for tuning in,
-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
               Nathaniel Gray
     California Institute of Technology
      Computation and Neural Systems
       n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From jacobkm at cats.ucsc.edu  Wed Feb 21 07:28:58 2001
From: jacobkm at cats.ucsc.edu (Jacob Kaplan-Moss)
Date: Tue, 20 Feb 2001 22:28:58 -0800
Subject: Rounding Question
Message-ID: <jacobkm-07145C.22285820022001@news.ucsc.edu>

Hello All --

So I've got a number between 40 and 130 that I want to round up to the 
nearest 10.  That is:
   
   40 --> 40
   41 --> 50
   42 --> 50
   ...
   49 --> 50
   50 --> 50
   51 --> 60
   etc.

The way I'm doing it now is:

>>> import math
>>> depth = <<whatever>>
>>> tableDepth = int(math.ceil( depth/10.0 ) * 10)

(NAUI divers will recognize me translating from actual depth to a table 
depth entry)

This works just fine for my purposes, but its somewhat counterintuitive; 
I wonder if there is a more self-explanatory way...

Anyone have any ideas?

TIA,

Jacob


From aleaxit at yahoo.com  Wed Feb 21 09:31:50 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 09:31:50 +0100
Subject: Rounding Question
References: <jacobkm-07145C.22285820022001@news.ucsc.edu>
Message-ID: <96vudj02hnf@news1.newsguy.com>

"Jacob Kaplan-Moss" <jacobkm at cats.ucsc.edu> wrote in message
news:jacobkm-07145C.22285820022001 at news.ucsc.edu...
> Hello All --
>
> So I've got a number between 40 and 130 that I want to round up to the
> nearest 10.  That is:
>
>    40 --> 40
>    41 --> 50

So, not really to the *nearest* 10, but to the *least multiple of 10
that is >= the original number*.  OK.  An *excellent* not-quite-toy
example on which to hinge some general-interest discussion, IMHO...


> >>> import math
> >>> depth = <<whatever>>
> >>> tableDepth = int(math.ceil( depth/10.0 ) * 10)
>
> (NAUI divers will recognize me translating from actual depth to a table
> depth entry)
>
> This works just fine for my purposes, but its somewhat counterintuitive;
> I wonder if there is a more self-explanatory way...

It is sometimes hard to say what people will find "intuitive" or
"self-explanatory", but, what about:

    tableDepth = (depth/10) * 10
    if tableDepth < depth: tableDepth += 10

The first step insures that tableDepth is a multiple of 10, <= depth;
the second step changes the < part of it into a >, as required.

People who are allergic to one-line if-statements might perhaps
want to change the second step to:

    tableDepth += 10*(tableDepth<depth)

but that seems slightly less intuitive to me, as it depends on
the fact that '<' returns 0 or 1, rather than just on its true/
false behavior which is what is normally required; similarly "not
really all that clear" (IMHO) are alternative expressions such as

    tableDepth += (0,10)[tableDepth<depth]

or

    tableDepth += tableDepth<depth and 10

although the latter does go back to the simpler true/false idea,
the subtlety it depends on is the fact that Python's "and" does
NOT just return true/false (like C's "&&", which it resembles
in other ways), but, rather, its RHS operand if its LHS one is
true, else its (false) LHS operand (without, in this latter
case, evaluating its RHS, but that's immaterial in this case).

I would not consider breaking the if-statement into two lines
a plus, either -- it's bad enough that all of these ideas
split a single concept into two steps, without making it
appear like THREE steps are somehow involved:-).


Getting back to ONE concept for any of these forms is not
hard at all, but, again, intuitiveness is in the eye of
the beholder.  Take, for example, the latter form:
    tableDepth = (depth/10) * 10
    tableDepth += tableDepth<depth and 10
how do we map this back into one expression? I think the
clearest conceptual transformation relies on recognizing
that an intermediate variable is involved, although we
chose to name it "tableDepth" (the same as the final target
value...):
    temp = (depth/10)*10
    tableDepth = temp + (temp<depth and 10)
and we can expand 'temp' to its defining expression in
the second step:
    tableDepth = (depth/10)*10 + ((depth/10)*10<depth and 10)

This may not look like much of a progress, but it does
enable us to recognize that the key test:
    (depth/10)*10<depth
means "is depth a multiple of 10"...?  And we can
express THAT one more directly:
    tableDepth = (depth/10)*10 + (depth%10 and 10)
and of course we could factor out the 10 multiplier:
    tableDepth = 10 * ((depth/10) + (depth%10 and 1))
or
    tableDepth = 10 * ((depth/10) + depth%10 != 0)


Whenever we have a division-with truncation, AND a
modulo-operation, with the same operands, the builtin
divmod function always comes to mind, leading to such
further alternative expressions as:
    quotient, remainder = divmod(depth, 10)
    if remainder: quotient += 1
    tableDepth = quotient * 10
or, trying to gain back some compactness for the
last two statements,
    tableDepth = (quotient + (remainder!=0)) * 10
which is really similar to the last expression in
the previous paragraph.


But the most classic way to express this would be:
    tableDepth = 10 * ((depth+9)/10)
which also happens to be just about the most compact
and concise one; its "intuitiveness" is probably
dependent on having seen it a zillion times before:-).

Deep down, it does depend on some moderately subtle
theorems of modulo-arithmetic, but familiarity does
have a way to compensate for that.  So, the "eye
of the beholder" effect may dominate (it's hard for
me to judge the actual "clarity" of something I have
written or read so many thousands of times in the
past).

Depending on your audience, a simple comment or
docstring:
    # set tableDepth to the least multiple of 10 >= depth
may suffice to make any or all of these expressions
acceptable (this is a controversial statement, as a
school exists that says too-complex code must be made
simpler, rather than explained in a comment, since
comments have a nasty way to get out of sync with code,
etc; however, there are arguments on both sides when
something has a certain "minimum irreducible intrinsic
complexity" and particularly when that "complexity" is
actually more of a *subtlety*...).

So, all in all, I would suggest avoiding the floating
point maths that your original approach used -- your
intuition, IMHO, was *QUITE* on target suggesting that
such an approach is sub-optimal... the complexities of
FP, after all, make the subtleties of modulo arithmetic
pale in comparison!  In the end, something like:

def depthToTableDepth(depth):
    "compute the least multiple of 10 >= depth"
    return int((depth+9)/10) * 10

may be the best compromise.  The 'int' call will be
redundant when depth is an integer, but it does not
hurt, and may in fact clarify the code by reminding
the reader that we are using TRUNCATING division --
and it will let this work unaltered if and when
depth should become a FP number at some later time
in the program's lifecycle, just like your original
FP-based expression would keep working then.  One
should not introduce unwarranted generality when
it (as is often the case) complicates things (the
important "you ain't gonna need it" principle), but,
at times, higher generality actually CLARIFIES and
SIMPLIFIES things wrt a more targeted/ad-hoc approach;
and I opine we're in such a case here, with the
clarity coming from the _explicitness_ of 'int()'.


Note that putting this into a function of its own,
rather than inline code, may often be best (although,
here, that's not wholly clear).  A further _delicate_
judgment decision is whether generalizing from '10'
is a gain or loss in clarity...:

def nextHigherOrEqualMultiple(depth, N):
    "compute the least multiple of N >= depth"
    return int((depth+N-1)/N) * N

with

tableDepth = nextHigherOrEqualMultiple(depth, 10)

I *think* it would be a substantial gain *IF* a name
that is both clear AND concise could be found for the
function, but not if we DO have to use a 25-characters
name for clarity:-).


Alex





From jacobkm at cats.ucsc.edu  Wed Feb 21 19:27:58 2001
From: jacobkm at cats.ucsc.edu (Jacob Kaplan-Moss)
Date: Wed, 21 Feb 2001 10:27:58 -0800
Subject: Rounding Question
References: <jacobkm-07145C.22285820022001@news.ucsc.edu> <96vudj02hnf@news1.newsguy.com>
Message-ID: <jacobkm-4E8467.10275821022001@news.ucsc.edu>

In article <96vudj02hnf at news1.newsguy.com>, "Alex Martelli" 
<aleaxit at yahoo.com> wrote:

[snip a whole bunch of great ideas...]

Thanks so much for the great discussion; although this is really a 
trivial part of what I'm doing, it's nice to be able to do it right...

In the end, I decided that:

> def nextHigherOrEqualMultiple(depth, N):
>     "compute the least multiple of N >= depth"
>     return int((depth+N-1)/N) * N
> 
> with
> 
> tableDepth = nextHigherOrEqualMultiple(depth, 10)

is what I'm going to use; I really only need to make this calculation 
once, so a long function name really isn't a problem, and although I'm 
writing for NAUI tables now, I know that certain other tables use 
increments of 5 feet, so it makes sense to have a generic function 
available.

Thanks once again,

Jacob


From scarblac at pino.selwerd.nl  Wed Feb 21 15:43:39 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 21 Feb 2001 14:43:39 GMT
Subject: Rounding Question
References: <jacobkm-07145C.22285820022001@news.ucsc.edu>
Message-ID: <slrn997ktk.h0.scarblac@pino.selwerd.nl>

Jacob Kaplan-Moss <jacobkm at cats.ucsc.edu> wrote in comp.lang.python:
> Hello All --
> 
> So I've got a number between 40 and 130 that I want to round up to the 
> nearest 10.  That is:
>    
>    40 --> 40
>    41 --> 50
>    42 --> 50
>    ...
>    49 --> 50
>    50 --> 50
>    51 --> 60

Rounding like this is the same as adding 5 to the number and then rounding
down. Rounding down is substracting the remainder if you were to divide by
10, for which we use the % operator in Python.

rounded = (number+5)-(number+5)%10

-- 
Remco Gerlich


From mikael at isy.liu.se  Wed Feb 21 16:49:26 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 21 Feb 2001 16:49:26 +0100 (MET)
Subject: Rounding Question
In-Reply-To: <slrn997ktk.h0.scarblac@pino.selwerd.nl>
Message-ID: <XFMail.010221164926.mikael@isy.liu.se>

On 21-Feb-01 Remco Gerlich wrote:
 >  Jacob Kaplan-Moss <jacobkm at cats.ucsc.edu> wrote in comp.lang.python:
 > > Hello All --
 > > 
 > > So I've got a number between 40 and 130 that I want to round up to the 
 > > nearest 10.  That is:
 > >    
 > >    40 --> 40
 > >    41 --> 50
 > >    42 --> 50
 > >    ...
 > >    49 --> 50
 > >    50 --> 50
 > >    51 --> 60
 >  
 >  Rounding like this is the same as adding 5 to the number and then rounding
 >  down. Rounding down is substracting the remainder if you were to divide by
 >  10, for which we use the % operator in Python.
 >  
 >  rounded = (number+5)-(number+5)%10

Actually, that should be

  rounded = (number+9)-(number+9)%10

to get the behaviour Jacob asked for, provided that number is an integer.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    21-Feb-01
Time:    16:46:56

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From aleaxit at yahoo.com  Wed Feb 21 17:05:36 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 17:05:36 +0100
Subject: Rounding Question
References: <jacobkm-07145C.22285820022001@news.ucsc.edu> <slrn997ktk.h0.scarblac@pino.selwerd.nl>
Message-ID: <970p0j0ri3@news1.newsguy.com>

"Remco Gerlich" <scarblac at pino.selwerd.nl> wrote in message
news:slrn997ktk.h0.scarblac at pino.selwerd.nl...
> Jacob Kaplan-Moss <jacobkm at cats.ucsc.edu> wrote in comp.lang.python:
> > Hello All --
> >
> > So I've got a number between 40 and 130 that I want to round up to the
> > nearest 10.  That is:
> >
> >    40 --> 40
> >    41 --> 50
> >    42 --> 50
> >    ...
> >    49 --> 50
> >    50 --> 50
> >    51 --> 60
>
> Rounding like this is the same as adding 5 to the number and then rounding
> down. Rounding down is substracting the remainder if you were to divide by
> 10, for which we use the % operator in Python.
>
> rounded = (number+5)-(number+5)%10

This will work if you use +9 in each case rather than +5 (note that he
doesn't really want rounding -- he wants 41 to 'round' to 50, for ex),
and it's Yet Another Way To Do It similar but not identical to the dozen
or so I explained in my previous post on this thread...:-)


Alex





From scarblac at pino.selwerd.nl  Wed Feb 21 18:04:23 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 21 Feb 2001 17:04:23 GMT
Subject: Rounding Question
References: <jacobkm-07145C.22285820022001@news.ucsc.edu> <slrn997ktk.h0.scarblac@pino.selwerd.nl>
Message-ID: <slrn997t5i.rb.scarblac@pino.selwerd.nl>

Remco Gerlich <scarblac at pino.selwerd.nl> wrote in comp.lang.python:
> Jacob Kaplan-Moss <jacobkm at cats.ucsc.edu> wrote in comp.lang.python:
> > Hello All --
> > 
> > So I've got a number between 40 and 130 that I want to round up to the 
> > nearest 10.  That is:

So I read 'round to the nearest 10' and Alex read it correctly...

> >    40 --> 40
> >    41 --> 50
> >    42 --> 50
> >    ...
> >    49 --> 50
> >    50 --> 50
> >    51 --> 60
> 
> Rounding like this is the same as adding 5 to the number and then rounding
> down. Rounding down is substracting the remainder if you were to divide by
> 10, for which we use the % operator in Python.
> 
> rounded = (number+5)-(number+5)%10
 
I'd go for 
rounded = number-(number%10)+10
now.

-- 
Remco Gerlich


From mikael at isy.liu.se  Thu Feb 22 08:23:19 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 22 Feb 2001 08:23:19 +0100 (MET)
Subject: Rounding Question
In-Reply-To: <slrn997t5i.rb.scarblac@pino.selwerd.nl>
Message-ID: <XFMail.010222082319.mikael@isy.liu.se>

On 21-Feb-01 Remco Gerlich wrote:
 >  I'd go for 
 >  rounded = number-(number%10)+10
 >  now.

I'm sorry, but this also behaves badly. It rounds 10 up to 20. To get 
Jacob's desired functionality, you could modify your idea to

  rounded = number-((number-1)%10)+9

still assuming only integer input.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    22-Feb-01
Time:    08:18:49

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From aleaxit at yahoo.com  Thu Feb 22 09:43:14 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 09:43:14 +0100
Subject: Rounding Question
References: <XFMail.010222082319.mikael@isy.liu.se>
Message-ID: <972jev0huk@news2.newsguy.com>

"Mikael Olofsson" <mikael at isy.liu.se> wrote in message
news:XFMail.010222082319.mikael at isy.liu.se...
On 21-Feb-01 Remco Gerlich wrote:
 >  I'd go for
 >  rounded = number-(number%10)+10
 >  now.

I'm sorry, but this also behaves badly. It rounds 10 up to 20. To get
Jacob's desired functionality, you could modify your idea to

  rounded = number-((number-1)%10)+9

still assuming only integer input.


...which of course is totally equivalent to the earlier
proposed
    rounded = (number+9) - ((number+9)%10)
since addition is commutative and associative (you can
move the +9 from the right of the expression) AND so
is addition in modulo-arithmetic, so that:

(X-1)%N==(X%N)+((-1)%N)==(X%N)+((N-1)%N)==(X+N-1)%N


Alex





From mikael at isy.liu.se  Thu Feb 22 14:37:56 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 22 Feb 2001 14:37:56 +0100 (MET)
Subject: Rounding Question
In-Reply-To: <972jev0huk@news2.newsguy.com>
Message-ID: <XFMail.010222143756.mikael@isy.liu.se>

I corrected Remco:
 >  I'm sorry, but this also behaves badly. It rounds 10 up to 20. To get
 >  Jacob's desired functionality, you could modify your idea to
 >  
 >    rounded = number-((number-1)%10)+9
 >  
 >  still assuming only integer input.


Alex Martelli wrote:
 >  ...which of course is totally equivalent to the earlier
 >  proposed
 >      rounded = (number+9) - ((number+9)%10)
 >  since addition is commutative and associative (you can
 >  move the +9 from the right of the expression) AND so
 >  is addition in modulo-arithmetic, so that:
 >  
 >  (X-1)%N==(X%N)+((-1)%N)==(X%N)+((N-1)%N)==(X+N-1)%N

Of course! I just couldn't stay silent when Remco stumbled again.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    22-Feb-01
Time:    14:35:06

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From mcalla at home.com  Thu Feb 22 20:35:46 2001
From: mcalla at home.com (Mike Callahan)
Date: Thu, 22 Feb 2001 19:35:46 GMT
Subject: Rounding Question
References: <XFMail.010222143756.mikael@isy.liu.se>
Message-ID: <mSdl6.313504$iy3.66178474@news1.rdc1.tn.home.com>

How about this, (taking advantange of the round function)?

>>> def rounded(n):
 return int(round(n+4, -1))

>>> rounded(10)
10
>>> rounded(11)
20
>>> rounded(19)
20
>>> rounded(111)
120
>>>

Mike Callahan




From scarblac at pino.selwerd.nl  Fri Feb 23 01:52:06 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 23 Feb 2001 00:52:06 GMT
Subject: Rounding Question
References: <XFMail.010222143756.mikael@isy.liu.se>
Message-ID: <slrn99bcuc.3gs.scarblac@pino.selwerd.nl>

Mikael Olofsson <mikael at isy.liu.se> wrote in comp.lang.python:
> I corrected Remco:
>  >  I'm sorry, but this also behaves badly. It rounds 10 up to 20. To get
>  >  Jacob's desired functionality, you could modify your idea to
>  >  
>  >    rounded = number-((number-1)%10)+9
>  >  
>  >  still assuming only integer input.
> 
> 
> Alex Martelli wrote:
>  >  ...which of course is totally equivalent to the earlier
>  >  proposed
>  >      rounded = (number+9) - ((number+9)%10)
>  >  since addition is commutative and associative (you can
>  >  move the +9 from the right of the expression) AND so
>  >  is addition in modulo-arithmetic, so that:
>  >  
>  >  (X-1)%N==(X%N)+((-1)%N)==(X%N)+((N-1)%N)==(X+N-1)%N
> 
> Of course! I just couldn't stay silent when Remco stumbled again.

Yeah yeah, rub it in... :)

Thing is I wasn't assuming integers, I wanted to write something that works
on floats as well. He only gave integer examples, but his example code with
math.ceil and dividing by 10.0 works on floats as well, and I thought it
likely that he'd get non-integer depth measurements once.

Therefore, I didn't want to use the /10*10 trick and using +9 didn't enter
my mind...

So now I'm proposing, for float input,

if number % 10 == 0:
   return number
else return number+10-(number%10)

Otherwise you'll have to find out what the highest float smaller than 10 is.

Since it is, again, far too late to be posting, I know full well that I'll
get it wrong again...

-- 
Remco Gerlich

PS By now his original math.ceil solution starts to look better than mine...


From adam at deprince.net  Fri Feb 23 08:20:48 2001
From: adam at deprince.net (Adam DePrince)
Date: Fri, 23 Feb 2001 02:20:48 -0500
Subject: Rounding Question
References: <XFMail.010222143756.mikael@isy.liu.se> <slrn99bcuc.3gs.scarblac@pino.selwerd.nl>
Message-ID: <3A960F50.CCEA969E@deprince.net>

Remco Gerlich wrote:
> Mikael Olofsson <mikael at isy.liu.se> wrote in comp.lang.python:
> > I corrected Remco:
> >  >  I'm sorry, but this also behaves badly. It rounds 10 up to 20. To get
> >  >  Jacob's desired functionality, you could modify your idea to
> >  >    rounded = number-((number-1)%10)+9
> >  >  still assuming only integer input.
> > Alex Martelli wrote:
> >  >  ...which of course is totally equivalent to the earlier
> >  >  proposed
> >  >      rounded = (number+9) - ((number+9)%10)
> >  >  since addition is commutative and associative (you can
> >  >  move the +9 from the right of the expression) AND so
> >  >  is addition in modulo-arithmetic, so that:
> >  >  (X-1)%N==(X%N)+((-1)%N)==(X%N)+((N-1)%N)==(X+N-1)%N
> > Of course! I just couldn't stay silent when Remco stumbled again.
> Yeah yeah, rub it in... :)
> Thing is I wasn't assuming integers, I wanted to write something that works
> on floats as well. He only gave integer examples, but his example code with
> math.ceil and dividing by 10.0 works on floats as well, and I thought it
> likely that he'd get non-integer depth measurements once.
                        ^^^^^^^^^

Keep in mind the dangers associated with floats.  Imprecision. 
Rounding.  Truncation.  Underflow.  

Are you generating the numbers to be rounded up via some other
computation?  If so, then what happens when a rounding error makes your
answer 40.0000001 when on paper the answer should be 40.  Your ceil
function will return 50 instead of the correct value of 40.

In your first post you mentioned that this is for diving table
computation.  Very saftey critical application.  Floating point numbers
have limitations.  Equality doesn't always work when pencil and paper
computations indicate it should.  IEEE floats can't represent numbers
like 0.4 precisely (0.4 is a repeating fraction in base 2).  The answers
will not be consistent from one type of machine to the other, or even
from one compiler to the other.  Entire careers are made out of dealing
with the weirdness of floating point numbers near boundry conditions. 
And a function that chooses an answer by checking if a number of exactly
40 or the smallest number representable that is greater than 40,
presuming you expect computed values exactly equal to 40, qualifies as a
boundry condition.  Stick with integer math.  You will live longer.

-- 
Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From vze2f978 at mail.verizon.net  Sat Feb 24 01:05:56 2001
From: vze2f978 at mail.verizon.net (Lloyd Hugh Allen)
Date: Sat, 24 Feb 2001 00:05:56 GMT
Subject: Rounding Question
References: <XFMail.010222143756.mikael@isy.liu.se> <slrn99bcuc.3gs.scarblac@pino.selwerd.nl>
Message-ID: <3A96FB2C.60235B39@mail.verizon.net>

Remco Gerlich wrote:
> 
> Mikael Olofsson <mikael at isy.liu.se> wrote in comp.lang.python:
> > I corrected Remco:
> >  >  I'm sorry, but this also behaves badly. It rounds 10 up to 20. To get
> >  >  Jacob's desired functionality, you could modify your idea to
> >  >
> >  >    rounded = number-((number-1)%10)+9
> >  >
> >  >  still assuming only integer input.
> >
> >
> > Alex Martelli wrote:
> >  >  ...which of course is totally equivalent to the earlier
> >  >  proposed
> >  >      rounded = (number+9) - ((number+9)%10)
> >  >  since addition is commutative and associative (you can
> >  >  move the +9 from the right of the expression) AND so
> >  >  is addition in modulo-arithmetic, so that:
> >  >
> >  >  (X-1)%N==(X%N)+((-1)%N)==(X%N)+((N-1)%N)==(X+N-1)%N
> >
> > Of course! I just couldn't stay silent when Remco stumbled again.
> 
> Yeah yeah, rub it in... :)
> 
> Thing is I wasn't assuming integers, I wanted to write something that works
> on floats as well. He only gave integer examples, but his example code with
> math.ceil and dividing by 10.0 works on floats as well, and I thought it
> likely that he'd get non-integer depth measurements once.
> 
> Therefore, I didn't want to use the /10*10 trick and using +9 didn't enter
> my mind...
> 
> So now I'm proposing, for float input,
> 
> if number % 10 == 0:
>    return number
> else return number+10-(number%10)
...
> 
> PS By now his original math.ceil solution starts to look better than mine...

Assuming integers, this'll do the same thing in one line:

>>> def tableDepth(x):
	return 10 * ((x + 9) / 10)

>>> tableDepth(49)
50
>>> tableDepth(40)
40

without integers, though, it'll give a runtime error. Or something.


From maildsb at glbgrp.com  Wed Feb 21 08:14:45 2001
From: maildsb at glbgrp.com (=?ISO-2022-JP?B?GyRCOXFGYjpHQmc1aSEiOkdEY05BNmIkTiVhITwla0dbPy4lNSE8JVMlORsoQg==?=)
Date: Wed, 21 Feb 2001 16:14:45 +0900
Subject: =?ISO-2022-JP?B?GyRCOXFGYjpHQmc1aSEiOkdEY05BNmIkTiVhGyhC?=
 =?ISO-2022-JP?B?GyRCITwla0dbPy4lNSE8JVMlOSROJDQwRkZiGyhC?=
Message-ID: <0221101161445.1217@WEBS>

$B9qFb:GBg5i!":GDcNA6b$N%a!<%kG[?.%5!<%S%9$N$40FFb(B

$B%[!<%`%Z!<%8$N at kEA!"$*E9$N at kEA!"?7>&IJ$N9pCN!"5a?MJg=8!"D4::!"9TJ}ITL@<TA\:w(B
$BEy!#(B

$B!!=i$a$F!"%a!<%k$5$;$F$$$?$@$-$^$9!#(B

  $B%0%m!<%V%$%s%U%)!J3t!K$N>pJs%a%G%#%";v6HIt$H?=$7$^$9!#(B
$B#U#R#L!'(B http://www.glbgrp.com$BFb(B

$BEv<R$O!XE*3N$J>pJsDs6!%5!<%S%9!Y!X>pJs4IM}Ds6!%7%9%F%`$N9=C[!YEy!">pJs$K4X$9(B
$B$kA4HL$N%5!<%S%9$rDs6!$9$k2q<R$G$9!#(B

$B!!$3$NEY!"?75,;v6H$H$7$F!"#1%a!<%k%"%I%l%9C10L#21_$GG[?.%5!<%S%9$r9T$&$3$H$H(B
$B$J$j$^$7$?!#(B
$B!JC"$7!":GDc#5#0#0#07oC10L$H$7!"DI2C$b#5#0#0#07oC10L$G$*<u$1CW$7$^$9!K(B

  $B%5!<%S%9$NG[?.NA$O8=:_2f!9$,CN$jF@$kG[?.%5!<%S%9$NCf$G:G$bDcNA6b$G$9!#(B

$B!!%a!<%k9-9pG[?.$O?7J99-9p$J$I$H0c$$!"9XFIN($O!"?7J9!"@^$j9~$_9-9p$N?tG\$N3N(B
$BN($G8z2LE*$G$9!#(B

  $B@'Hs$4MxMQ$/$@$5$$!#(B

  $BMxMQMQES!'(B
A:$B%[!<%`%Z!<%8$N at kEA!JNI$$9-9p$G$"$l$P!"3N<B$K%"%/%;%9?t$OA}$($^$9!K(B
B:$B$*E9$N at kEA!JA49qE*$K9-$,$j$^$9!K(B
C:$B?7>&IJ$N9pCN(B
D:$B5a?MJg=8!J9-HO0O$KJg=8$,=PMh$^$9!K(B
E:$BD4::!J%"%s%1!<%HD4::!"Ey!K"(C"$7JLES2C9)NA6b$,$+$+$j$^$9!#(B
F:$B9TJ}ITL@<TA\:w4j$$!J7Y;!$X$NA\:w4j$$FO$,$5$l$F$$$k;v$,A0Ds$H$J$j$^$9!K(B
$B!!(B $B!J$"$kDxEY$N<L??Ey$rE:IU=PMh$^$9$N$G%S%i!"D%$j;f$h$j3NN($,9b$/$J$j$^$9!K(B


$B!!Nc$($P#5!$#0#0#07o$K%a!<%k$rG[?.$7$?>l9g$O(B
$B!!(B      $B#1#0!$#0#0#01_!\>CHq at G!\?69~<j?tNA!K$H$J$j$^$9!#(B


$B!!?d>)G[?.%G!<%?NL!'#1#6#6(BKbyt$B!JE:IU%U%!%$%k$=$NB>4^$`!K(B
$B"($3$l$O$"$/$^$G?d>)$G$9$,!"$"$^$jBg$-$$$H%a!<%k<u?.;~$K%-%c%s%;%k$5$l$F$7$^(B
$B$&2DG=@-$,9b$/$J$j$^$9!#(B

$B!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\(B

$B4pK\5,Ls!'(B
  $BK\%5!<%S%9$O!"Ev<R4IM}$N%a!<%k%"%I%l%9$KBP$7>pJs!J9-9p!"J8=q(B  $BEy!K$rDj$a$i(B
$B$l$?6b3[$GG[?.$9$k%5!<%S%9$G$9!#(B

$B$*<u$1$G$-$J$$FbMF(B
  $B>pJs$NFbMF$O!"HH:a$K4X78$9$k$b$N!"Hpkn!&Cf=}!&7y$,$i$;!"5u56!&(B $B%o%$%;%DJ8(B
$B=q!&%o%$%;%D9-9p!"K=F0!"Ey$rM6H/$9$k2DG=@-!#Kt$O0cK!@-$,$"$k$HH=CG$5$l$k$b$N(B
$B$O$*<u$1$G$-$^$;$s!#(B

$B%&%$%k%9$NG[?.$rKI;_$9$k0Y$K!'(B
$B!!E:IU%U%!%$%k$O!"A4$F%&%$%k%9!&%9%-%c%s$r9T$$3NG'CW$7$^$9!#(B

$B"(G[?.NA0J30$NNA6b!'!J%"%s%1!<%HD4::Ey!"JV?.$r4uK>$9$k>l9g!K(B

$BJV?.!&2sEz$5$l$?$b$N$O!"8D?M$rFCDj$G$-$k>pJs!J%"%I%l%9!&;aL>!&Ey!K$r=|$$$?(B
$B%G!<%?$r%F%-%9%H(BCSV$BEy$GA4$F$*EO$7$$$?$7$^$9!#(B
$BKt2sEz%G!<%?$O#17o#5#01_$N<j?tNA$,JLES$+$+$j$^$9!#!J=87W!"Ey$O9T$$$^$;$s!K(B
  $B$^$?%"%s%1!<%HD4::Ey$NJV?.$O!"2sEz<T$N<+M30U;V$K4p$E$/$N0Y(B,$B$=$N2sEz$NM-L5(B
$B!&7o?tFbMF$K$D$$$F$O!"Ev<R$O0l at ZJ]>Z=PMh$+$M$^$9$N$G$4N;>5$/$@$5$$!#(B

 $BA\:w4j$$Ey$N2sEz!&O"Mm%a!<%k$OL5NA$G%G!<%?$r$*EO$7$7$^$9!#(B
$B!!!JKtG[?.NA6b$b!"(B1$B%a!<%k%"%I%l%9C10L#11_$K$5$;$F$$$?$@$-$^$9!#!K(B


 $BLd$$9g$o$;!'(Bmaildsb at g-con.com

$B!!(B $B!|$3$N%a!<%k$,ITMW$N>l$O!"$*<j?t$G$9$,$3$N%a!<%k(Bmailnothnk at g-con.com$B%"%I%l%9(B
$B!!!!(B $B08$K$4JV?.$/$@$5$k$h$&$*4j$$?=$7$"$2$^$9!#(B
$B!!!!!!(Bmailnothnk at g-con.com
$B!!(B $B!!(B
$B!!(B
 $B!!!|$*FO$1$9$k%a!<%k%"%I%l%9$NJQ99$O!"2<5-$X$4O"Mm2<$5$$!#(B
      mailchng at g-con.com
$B!!!!"(%a!<%kITMW!&%"%I%l%9JQ99$N$4O"Mm$r$$$?$@$$$?$K$b4X$o$i$:!"(B
$B!!!!!!9T$-0c$$$K$h$j!"%a!<%kAw?.$5$l$k>l9g$,$4$6$$$^$9!#(B
$B!!!!!!$"$7$+$i$:!"$4N;>5$/$@$5$$!#(B
$B!!!!!!!!!!!!!!!!(B
                $B!!!!!!!!!!!!!!H/9T!'%0%m!<%V%$%s%U%)3t<02q<R(B


$B!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\!v!\(B
$B!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!]!v!](B

 $B!!%a!<%kG[?.%5!<%S%9$NMxMQ$NJ}K!!!(B

$B-!$^$:!"G[?.MQ%a!<%k$r:n at .$/$@$5$$!#(B

$B-":n at .$7$?%a!<%k!JE:IU%U%!%$%k$b4^$`!K$r(B  maildsb at glbgrp.com $B$KAw?.$/$@$5(B
$B$$!#(B
   $BLd$$9g$o$;$$$?$@$1$l$P!"?=9~=q$rAw?.CW$7$^$9!#(B

$B-#%a!<%k$NFbMF$rEv<R$G%A%'%C%/CW$7$^$9(B

$B-$#O#K$J$iAw?.$5$l$F$-$?%a!<%k%"%I%l%908$K!"G[?.NA6b(B,$B?69~$_ at hEy$r!"$4O"MmCW(B
$B$7$^$9!#(B
  $B"(G[?.;XDjF|$N#6F|A0$K$O?69~$r40N;$7$F$/$@$5$k$h$&$*4j$$CW$7$^$9!#G/KvG/(B
$B;O!"G/EY;O$a=*$o$j!":.$_9g$$$^$9$N$G#2=54VA0$K$*4j$$$7$^$9!#(B

$B-%?69~3NG'$,$G$-$^$7$?$iG[?.MQ%a!<%k$NG[?.=`Hw$rCW$7$^$9!#(B
$B!!!!(B
$B-&G[?.%a!<%k$N:G=*3NG'$N%a!<%k$r;XDj$5$l$?%a!<%k%"%I%l%908$KAw?.CW$7$^$9!#JQ(B
$B99Ey$NO"Mm%a!<%k$,L5$$>l9g!"G[?.%a!<%k$N3NG'$,=PMh$?$b$N$H$_$J$7$^$9!#(B

$B-'G[?.;XDjF|$K%a!<%k$NG[?.(B
$B!!!!$=$N;~$K0MMj<T$NJ}$K$b%a!<%kAw?.$5$l$^$9!#(B

$B!!(B $B!!!!!!(B
$BCm0U!'8=:_!"FCDj;XDj$K$D$$$F$O!"J,N`;XDj$,$G$-$^$;$s!#(B
      $BF10l%a!<%k$rJ,3d$7$?>l9g!"F10l?MJ*$KF1$8%a!<%k$,G[?.$5$l$k2DG=@-$,$"$k(B
$B$3$H$r$4N;>5$/$@$5$$!#(B


$B!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t!t(B






From karl.grammer at univie.ac.at  Wed Feb 21 09:45:43 2001
From: karl.grammer at univie.ac.at (grammer)
Date: Wed, 21 Feb 2001 09:45:43 +0100
Subject: Newbie question: Python on Mac
Message-ID: <B6B93EC6.153B%karl.grammer@univie.ac.at>

Hi out there, I hope somebody can help me with this:

Whenever I open a Python script using Tkinter on my Mac, it is not possible
to enter text or numbers into fields, although the scripts run on a Windows
NT machine, where entering numbers or text in the fields is possible .

I wandered through many faqs and website. I did not find this problem  -  am
I completely dumb or what is this ?


Karl (slowly getting mad)


-- 
---------------------------------------------------------------------------
Karl Grammer 
Institute for Urban Ethology at the Institute for Antropology
Althanstrasse 14
A-1090 Vienna/Austria
Tel: +43-1-4277-54766 Fax:  +43-1-4277-9547
http://evolution.anthro.univie.ac.at/lbi.html

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010221/0d4345df/attachment.htm>

From sandipan at vsnl.com  Wed Feb 21 11:13:05 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Wed, 21 Feb 2001 15:43:05 +0530
Subject: Netfilter
Message-ID: <005501c09bee$e2670d20$020000c3@node02>

I control ipchains on my network firewall over python with os.system. Now,
Linux 2.4 onwards will have a new and more powerful (and more featureful)
packet filtering capability with netfilter (See
http://www.linux-mag.com/1999-10/bestdefense_01.html for an intro.) Is there
any plans for netfilter support in python that we can look forward to use ?

Regards,

Sandipan




From precisa at uai.com.br  Wed Feb 21 11:20:46 2001
From: precisa at uai.com.br (precisa at uai.com.br)
Date: Wed, 21 Feb 2001 10:20:46 -0000
Subject: Scott Ambler's persistence layer in Python ?
Message-ID: <9704pu+d6nr@eGroups.com>

Hi,

Has someone implemented Scott Ambler's persistence layer in Python ?

http://www.ambysoft.com/persistenceLayer.html

Thanks in advance

Geraldo Lopes de Souza






From hamish_lawson at yahoo.co.uk  Wed Feb 21 18:40:28 2001
From: hamish_lawson at yahoo.co.uk (hamish_lawson at yahoo.co.uk)
Date: Wed, 21 Feb 2001 17:40:28 -0000
Subject: Scott Ambler's persistence layer in Python ?
In-Reply-To: <9704pu+d6nr@eGroups.com>
Message-ID: <970uic+rlq2@eGroups.com>

> Has someone implemented Scott Ambler's persistence layer in Python ?
> 
> http://www.ambysoft.com/persistenceLayer.html

The forthcoming Webware 0.5 has a persistence component called 
MiddleKit. The user's guide has a reference to Scott Ambler's papers. 
The system is currently only available via CVS, but below I've 
included cvsweb links to the introduction and the user's guide.

http://cvs.sourceforge.net/cgi-
bin/cvsweb.cgi/~checkout~/Webware/MiddleKit/Docs/Intro.html?
cvsroot=webware

http://cvs.sourceforge.net/cgi-
bin/cvsweb.cgi/~checkout~/Webware/MiddleKit/Docs/UsersGuide.html?
cvsroot=webware

Webware is an suite of application-server components, but some of 
them - including, I think, MiddleKit - are designed to be able to be 
used outside of that system.

Hamish Lawson





From timo.becker at inosoft.de  Wed Feb 21 11:23:22 2001
From: timo.becker at inosoft.de (timo.becker at inosoft.de)
Date: 21 Feb 2001 10:23:22 GMT
Subject: Python, MFC and the good old Hello World
Message-ID: <9704uq$4iu$1@news.netmar.com>

hi,
can anybody give me a brief introduction in using mfc from python? Is it
possible to port the following hello world to python and what would it look
like?

----- 8< ----- 8< ----- 8< ----- 8< ----- 8< -----

class CMainWindow : public CFrameWnd {
public:
  CMainWindow() { Create(NULL,"Hello World!"); }
};

class CHelloApp : public CWinApp {
public:
  virtual BOOL InitInstance() {
    m_pMainWnd = new CMainWindow();
    m_pMainWnd->ShowWindow(m_nCmdShow);
    m_pMainWnd->UpdateWindow();
    return TRUE;
  }
};

CHelloApp HelloApp;

----- 8< ----- 8< ----- 8< ----- 8< ----- 8< -----


thanks in advance
Timo



 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse at newsone.net


From MarkH at ActiveState.com  Wed Feb 21 23:44:23 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 21 Feb 2001 22:44:23 GMT
Subject: Python, MFC and the good old Hello World
References: <9704uq$4iu$1@news.netmar.com>
Message-ID: <3A9443EE.3070904@ActiveState.com>

timo.becker at inosoft.de wrote:

> hi,
> can anybody give me a brief introduction in using mfc from python? Is it
> possible to port the following hello world to python and what would it look
> like?

I'm afraid I don't have the time to do and test this - but check out "pywin\framework\intpyapp.py" - that is where you can find Pythonwin's main frame and app code.

Note that if you dont have significant MFC experience or existing code you must interface with, wxPython would be a better choice.

Mark.



From Timo.Becker at INOSOFT.de  Thu Feb 22 00:35:18 2001
From: Timo.Becker at INOSOFT.de (Timo Becker)
Date: Thu, 22 Feb 2001 00:35:18 +0100
Subject: Python, MFC and the good old Hello World
References: <9704uq$4iu$1@news.netmar.com> <3A9443EE.3070904@ActiveState.com>
Message-ID: <971jbf$mptke$1@ID-76757.news.dfncis.de>

Thanks for your help, Mark.

> I'm afraid I don't have the time to do and test this - but check out
"pywin\framework\intpyapp.py" - that is
> where you can find Pythonwin's main frame and app code.

Radim Pol?sek just told me that PythonWin comes with a helloapp.py in the
demos/app directory (I must have been blind :-)) which is almost a one to
one translation of the code I posted.


> Note that if you dont have significant MFC experience or existing code you
must interface with, wxPython
> would be a better choice.

I just started using MFC under C++ and only wanted to find out how much i
can transfer of that new knowledge to Python. Most time I use Python Scripts
as COM Servers and don't mind about a gui at all.

Timo




From aleaxit at yahoo.com  Thu Feb 22 10:05:02 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 10:05:02 +0100
Subject: Python, MFC and the good old Hello World
References: <9704uq$4iu$1@news.netmar.com> <3A9443EE.3070904@ActiveState.com> <971jbf$mptke$1@ID-76757.news.dfncis.de>
Message-ID: <972knt0j5n@news2.newsguy.com>

"Timo Becker" <Timo.Becker at INOSOFT.de> wrote in message
news:971jbf$mptke$1 at ID-76757.news.dfncis.de...
> Thanks for your help, Mark.
>
> > I'm afraid I don't have the time to do and test this - but check out
> "pywin\framework\intpyapp.py" - that is
> > where you can find Pythonwin's main frame and app code.
>
> Radim Pol?sek just told me that PythonWin comes with a helloapp.py in the
> demos/app directory (I must have been blind :-)) which is almost a one to
> one translation of the code I posted.

Emphasis on the "almost" -- e.g., it does NOT use a frame as
its "frame", which your code did (and so did my translation
thereof) -- it uses a generic window; it ignores the show
parameter with which the program is invoked, forcing instead
the use of SW_SHOWNORMAL (the OS compensates for this specific
application bug in some versions, e.g. on NT4 SP4, but not on
all versions), which your code and my translation thereof did
not; etc.


> > Note that if you dont have significant MFC experience or existing code
you
> must interface with, wxPython
> > would be a better choice.
>
> I just started using MFC under C++ and only wanted to find out how much i
> can transfer of that new knowledge to Python. Most time I use Python
Scripts
> as COM Servers and don't mind about a gui at all.

Unless you *already* have substantial legacy MFC code or knowledge,
I second the suggestion -- go for wxPython instead (or, Tkinter,
or, Dynamic HTML, ...).  The MFC architecture, besides not being
portable, is "legacy" even from Microsoft's viewpoint.


Alex





From NoSpam at NoSpam.com  Thu Feb 22 17:57:16 2001
From: NoSpam at NoSpam.com (Tom)
Date: Thu, 22 Feb 2001 16:57:16 GMT
Subject: Python, MFC and the good old Hello World
References: <9704uq$4iu$1@news.netmar.com> <3A9443EE.3070904@ActiveState.com> <971jbf$mptke$1@ID-76757.news.dfncis.de>
Message-ID: <Mxbl6.3830$XZ6.2210295@news4.rdc1.on.home.com>

I was an MFC/C++ developer.  I am now using the wxWindows GUI for both
Python and C++.  I find it to be similar to MFC - just better.  I would
recommend it to you.

Tom.

"Timo Becker" <Timo.Becker at INOSOFT.de> wrote in message
news:971jbf$mptke$1 at ID-76757.news.dfncis.de...
> Thanks for your help, Mark.
>
> > I'm afraid I don't have the time to do and test this - but check out
> "pywin\framework\intpyapp.py" - that is
> > where you can find Pythonwin's main frame and app code.
>
> Radim Pol?sek just told me that PythonWin comes with a helloapp.py in the
> demos/app directory (I must have been blind :-)) which is almost a one to
> one translation of the code I posted.
>
>
> > Note that if you dont have significant MFC experience or existing code
you
> must interface with, wxPython
> > would be a better choice.
>
> I just started using MFC under C++ and only wanted to find out how much i
> can transfer of that new knowledge to Python. Most time I use Python
Scripts
> as COM Servers and don't mind about a gui at all.
>
> Timo
>
>




From hamish_lawson at yahoo.co.uk  Wed Feb 21 11:34:18 2001
From: hamish_lawson at yahoo.co.uk (=?iso-8859-1?q?Hamish=20Lawson?=)
Date: Wed, 21 Feb 2001 10:34:18 +0000 (GMT)
Subject: Specifying links in Python-URL bulletin
Message-ID: <20010221103418.29467.qmail@web205.mail.yahoo.com>

The recent acquiring of the Deja service by Google will have given the
compiler of the Python-URL some headaches in trying to create links to
Usenet items. I applaud the inventive solution using appropriate search
terms for Google Groups, as in the example below.

http://groups.google.com/groups?q=Dustin+Boswell+how+to+compile+https+support+into+Python2.0&hl=en&lr=&safe=off&rnum=1&ic=1

However, aside from the funky coloured highlighting which decorates the
resulting message (and which some might regard as added value!), I
wonder if there's the possibility that such a search may return
multiple results or the wrong result.

Instead it's possible to specify a particular message in Google Groups:

http://groups.google.com/groups?q=group:comp.lang.python&seld=925202150&ic=1

For a message that has been located in Google using search terms, the
message ID can be obtained by following the "View thread" link.

Alternatively both Yahoo Groups and Python.org have archives of the
mailing list:

http://groups.yahoo.com/group/python-list/message/124006

http://mail.python.org/pipermail/python-list/2001-February/029488.html


Hamish Lawson


____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From phd at phd.pp.ru  Wed Feb 21 11:56:11 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 21 Feb 2001 13:56:11 +0300 (MSK)
Subject: Specifying links in Python-URL bulletin
In-Reply-To: <20010221103418.29467.qmail@web205.mail.yahoo.com>
Message-ID: <Pine.LNX.4.30.0102211351110.2991-100000@fep132.fep.ru>

On Wed, 21 Feb 2001, [iso-8859-1] Hamish Lawson wrote:
> The recent acquiring of the Deja service by Google will have given the
> compiler of the Python-URL some headaches in trying to create links to

   I was invited to be the editor 1 (one!) day before Google announced the
deal! The day I started the job Deja interface worked and I collected some
links. Next day I wanted to continue... Guess what I found on deja.com? :)))

> Usenet items. I applaud the inventive solution using appropriate search
> terms for Google Groups, as in the example below.
>
> http://groups.google.com/groups?q=Dustin+Boswell+how+to+compile+https+support+into+Python2.0&hl=en&lr=&safe=off&rnum=1&ic=1

   :)))

> However, aside from the funky coloured highlighting which decorates the
> resulting message (and which some might regard as added value!), I
> wonder if there's the possibility that such a search may return
> multiple results or the wrong result.
>
> Instead it's possible to specify a particular message in Google Groups:
>
> http://groups.google.com/groups?q=group:comp.lang.python&seld=925202150&ic=1

   My short experience showed that these "seld" values changed after some
time. If you repeat yor search next day, e.g. :(

> For a message that has been located in Google using search terms, the
> message ID can be obtained by following the "View thread" link.
>
> Alternatively both Yahoo Groups and Python.org have archives of the
> mailing list:
>
> http://groups.yahoo.com/group/python-list/message/124006
> http://mail.python.org/pipermail/python-list/2001-February/029488.html

   Yes, this, probably, is a good approach. Thank you.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From hamish_lawson at yahoo.co.uk  Wed Feb 21 12:35:01 2001
From: hamish_lawson at yahoo.co.uk (=?iso-8859-1?q?Hamish=20Lawson?=)
Date: Wed, 21 Feb 2001 11:35:01 +0000 (GMT)
Subject: Specifying links in Python-URL bulletin
In-Reply-To: <Pine.LNX.4.30.0102211351110.2991-100000@fep132.fep.ru>
Message-ID: <20010221113501.20088.qmail@web214.mail.yahoo.com>

Hello Oleg

> My short experience showed that these "seld" values changed after
> some time. If you repeat yor search next day, e.g. :(

Really? That's poor! They'll presumably have to settle on some kind of
permanent ID scheme when they roll out their forthcoming browsing
facilities.

Hamish Lawson


____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie



From phd at phd.pp.ru  Wed Feb 21 12:39:11 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 21 Feb 2001 14:39:11 +0300 (MSK)
Subject: Specifying links in Python-URL bulletin
In-Reply-To: <20010221113501.20088.qmail@web214.mail.yahoo.com>
Message-ID: <Pine.LNX.4.30.0102211438220.2991-100000@fep132.fep.ru>

On Wed, 21 Feb 2001, [iso-8859-1] Hamish Lawson wrote:
> > My short experience showed that these "seld" values changed after
> > some time. If you repeat yor search next day, e.g. :(
>
> Really? That's poor! They'll presumably have to settle on some kind of
> permanent ID scheme when they roll out their forthcoming browsing
> facilities.

   I think they are rather "session IDs", not permanent.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From meowbot at meowing.net  Thu Feb 22 10:13:37 2001
From: meowbot at meowing.net (Consolidated Meowbot)
Date: 22 Feb 2001 04:13:37 -0500
Subject: Specifying links in Python-URL bulletin
References: <mailman.982769210.31302.python-list@python.org>
Message-ID: <87elwrw05a.fsf@slip-129-37-99-9.tx.us.prserv.net>

HOW ARE YOU GENTLEMEN!!!
ALL YOUR NEWS ARE BELONG TO GOOGLE.
YOU ARE ON THE WAY TO CONFUSION.

>> My short experience showed that these "seld" values changed after
>> some time. If you repeat yor search next day, e.g. :(

> Really? That's poor! They'll presumably have to settle on some kind of
> permanent ID scheme when they roll out their forthcoming browsing
> facilities.

The Deja Fscking Googleborg does allow using the real message ID.  For
example, to see the article <news-answers/introduction_980967874 at rtfm.mit.edu>
you would use

<http://groups.google.com/groups?as_umsgid=%3Cnews-answers%2Fintroduction_980967874%40rtfm.mit.edu%3E>

Appending an &ic=1 to that may or may not take you directly to the
article instead of the summary, depending on the Googleborg's mood.

The syntax for these URLs seems to be changing as the DFGoogleborg
develops its news interface, so for the short term it may be more
useful to simply provide the message ID so that one can enter the
query `msgid:<blah>' at the main groups.google.com page, or paste it
into the msgid field on the advanced search form.


From hzhu at users.sourceforge.net  Wed Feb 21 12:32:03 2001
From: hzhu at users.sourceforge.net (Huaiyu Zhu)
Date: Wed, 21 Feb 2001 03:32:03 -0800 (PST)
Subject: Alternative iterator syntax
Message-ID: <Pine.LNX.4.10.10102210114040.1917-100000@mars.localdomain>

Following suggestion by Jeff Petkau <jpet at eskimo.com>,
(http://mail.python.org/pipermail/python-list/2001-February/029944.html),
here is a different proposal for iterators.

Like the colon-syntax proposal, it achieves the following

1. It provides an extensible iterator interface without pretending to
   provide random access to elements.
2. It resolves the endless "i indexing sequence" debate.
3. It allows performance enhancements to dictionary iteration.
4. It is completely backward compatible.

The gist of this proposal is that iterator syntax and semantics are not
special.  In particular,

1. Their semantics are obvious regardless of their origins and context.
2. They are generally applicable, not tied to sequences and mappings.
3. They reduce existing special syntax without introducing new ones.
4. They are backward compatible in a straightforward way.
5. The same syntax is applicable to "for..in"  and "in".
6. It provides a magic hook for the list() function.

The following text contains six sections: definitions, predefined usage,
builtins, issues, rationale, appendix.


-------------------------------------------------------

Definitions of Iterators:

Iterators MUST define three magic methods:
    __next__
    __contains__
    __list__

They MAY also define additional magic methods:
    __arity__
    __prev__
    __first__
    __last__
    __reset__

These magic methods SHOULD satisfy the following semantic restrictions
when they are defined
    
    __next__() returns items of the iterator one by one.  It raises
    IndexError at the end.
    
    __contains__(x) returns true iff __next__() would return x at some
    point.
    
    __list__() returns a list of all x that would be returned by
    __next__().

    __arity__() returns n iff x is an n-tuple for every x in __list__().

    __prev__() iterates in the opposite direction of __next__()

    __first__() and __last__() return the first and last elements
    __next__() would have returned.
    
    __reset__() resets the iterator pointer to the start.
    
They MAY but NEED NOT define a __len__ method, which may or may not
return semantically correct result.

A callable iterator x MUST define x.__call__() == x.__list__().  An easy
way to make an iterator x callable is to set

    x.__call__ = x.__list__
    
Iterator usage in the following contexts are pre-defined:
    1. for loops
    2. in operator
    3. list() function

Since the semantics of each iterator x is completely specified by
x.__list__(), we shall use such description when convenient.


-------------------------------------------------------

Predefined Usage of Iterators:

1. The looping structure:

  Consider the structure
  
    for a, b, c, ... in A:
        do something
    else:
        do otherthing

  If A defines __next__, the above is interpreted as the equivalence of
  (__got is an invisible variable):
    
    __got = 0
    while 1:
        try:
            a, b, c, ... = A.__next__()
            __got = 1
            do something
        except IndexError:
            break
    if not __got:
        do otherthing

  In addition, if A.__arity__ is defined, it will be checked before the
  loop to avoid ValueError: unpack tuple of wrong size.  This concept
  itself may also be generalized to allow safe tuple unpacking (see
  below).

  Else if A defines __list__, then "for x in A:" is interpreted as "for
  x in A.__list__():".  Note that every list is an iterator (see below).

  Else if A is a callable object, then "for x in A:" is interpreted as
  "for x in A():".

    
2. The in operator:

  Consider the expression
    (a, b, c, ...) in A:

  If A defines __contains__, the above is interpreted as the equivalence
  of:
    A.__contains__((a,b,c, ...))

  Else if A defines __list__, then "x in A" is interpreted as "x in
  A.__list__()".  Note that every list is an iterator (see below).

  Else if A is a callable object, then "x in A" is interpreted as "x in
  A()".


3. Get full list:

  Consider the expression
    list(A)

  If a.__list__ is defined, the above is interpreted as
    A.__list__().

  Else it is interpreted as in Python 2.0.
  
  This could be used to short circuit costly __getitem__ iterations
  whenever possible.

  If A is callable iterator, the above could also be written as
    A()


-------------------------------------------------------

Applications to Builtins and Other Common Situations:

1. Sequences.  The builtin sequence types (list, tuple, string) are
  changed to iterators by adding the three magic methods __next__,
  __contains__ and __list__.  The following is always true:

    x.__list__() == list(x)

  In addition, Each sequence object x defines two attributes (indexes,
  items) that are themselves iterators.  The arity of indexes is 1, that
  of items is 2.

    x.indexes.__list__() == range(len(x))
    x.items.__list__() == zip(range(len(x)), x)

  They can be used in the following way

    a = "abc"
    for i in a.indexes:   # i iterate over 0, 1, 2
    for i,x in a.items:   # i, x iterate over (0,'a'), (1,'b'), (2,'c')
    if i in a.indexes:    # equiv. to i in range(len(a))
    if i,x in a.items:    # equiv. to i in a.indexes and a[i]==x.
    b = a.items()         # equiv. to b = zip(range(len(a)), a)


2. Mappings.  Each mapping object x defines three attributes (keys,
  values, items) that are callable iterators.  The arity of keys and
  values is 1, that of items is 2.
  
    x.keys.__list__() == x.keys()
    x.values.__list__() == x.values()
    x.items.__list__() == x.items()
    
  They can be used in the following way

    d = {...}
    for k in d.keys:        # lazy iteration
    for v in d.values:      # lazy iteration
    for k,v in d.items:     # lazy iteration
    if k in d.keys:         # equiv. to d.has_key(k)
    if v in d.values:       # equiv. to v in d.values()
    if k, v in d.items:     # equiv. to d.has_key(k) and d[k]==v.

  Furthermore, all existing codes which treat them as methods continue
  to work, thanks to equivalences like

    d.keys() == d.keys.__call__() == d.keys.__list__()

  It is easy to convert existing code to the new syntax with global
  search and replacement in an editor:

      Replace all ".items():" with ".items:" 
      Replace all ".keys():" with ".keys:" 
      Replace all ".values():" with ".values:"

  Such replacements have net effect of reducing code clutter.  Except
  the last one, they may also improve performance.


3. Splitting of texts.  Several line-reading methods can be consolidated:

    for line in file.readlines:    # equivalent to file.xreadlines()
    for line in x.splitlines:      # lazy split
    lines = file.readlines():      # just as now
    line = file.readline()         # equiv. to file.readlines.__next__()
    for c in sys.stdin.readchars:  # if the os allows this
        
  It is possible to define additional attributes of iterators that are
  also iterators.  Here are some possibilities:
  
    for x,y,z in file.readlines.splitwords:
                          # iterate over lines split as words 

    for block in file.read.split("\n\n"):
                          # split a file into empty-line separated blocks
                          # without reading the whole file in.

4. The Iterator class.  An iterator class could be included with
  standard python, so that one could write:

    a = Iterator([1,2,3])       # Or a = Iterator(); a.__list__ = [1,2,3]
    for x in a: do something

  This is more useful compared with plain "for x in [1,2,3]" in
  situations where iterators are passed around, where iterations could
  be interrupted, or where iterators have attributes that are also
  iterators.  Users could subclass this to achieve the effects they
  want.

  One example is the above read.split.  Another example is when you want
  a server to start returning some results before it gets them all

    for record in database.query(question):

  This is by no means tied to existing sequence and mapping classes.  It
  also provides easy syntax to express ideas like

    for k in dict.keys.sorted:  # If sorting is done in a different thread
                                # it does not need to be completed before
                                # the first key is returned.

  Many operations on list could be well defined on iterators without
  generating a new list.  Subclasses of Iterator can be defined to do
  things like

    map(func, iterator)         # return an iterator without a new list
    [f(x) for x in iterator]    # ditto

    filter(func, iterator)      # return an iterator without a new list
    [x for x in iterator if func(x)] # ditto


-------------------------------------------------------

Compatibility and Conversion Issues:

This proposal introduces net gains for users without additional cost.
It is completely backward compatible in application code.  But to take
advantage of the new features, it is necessary to make certain changes
in user code.

First consider what it takes to use iterators.  For example, consider
UserDict.  To use the iterators, it is necessary to change

    for x in d.keys():
to
    for x in d.keys:

For old code applied to new objects (where d.keys is iterator), there is
no advantage, because d.keys() would generate a new list.  There is no
harm either.  This is why d.keys must be a callable iterator.

For new code applied to old objects (where d.keys is a method), the
looping structure is interpreted as over d.keys(), which is a list.
There is no harm either.  This is why we defines for loop specifically
for callable objects.  Otherwise this would be a major obstacle for
users converting old code to new ones, because they would have to
remember which subclass of UserDict had be converted so that keys is now
an iterator.

Now consider what it takes to produce iterators.  Still take UserDict as
example.  It is necessary to change

    def keys(self): return self.data.keys()
to
    keys = data.keys

Any subclass that redefines keys, values and items also need to be
converted if they want to use iterators.  It might be necessary to
introduce one more level of indirection using __getattr__, as
illustrated by the following idiom

    def __getattr__(self, name):
        if name == "keys": return self.data.keys
        elif name == "values": return self.data.values
        elif items == "items": return self.data.items

Analogous treatment applies to sequences.  For example, in UserList it
is necessary to define

    __next__ = data.__next__
    def __list__(self): return self.data
    __contains__ = data.__contains__

This covers cases like

    for x in []:
    for x in UserList():
    for x in {}.keys:
    for x in UserDict().keys:

In conclusion, users can use "x in X" without remembering whether X is
an iterator, a method, or a sequence.  Implementers can replace all list
returning methods with iterators whenever necessary, assured that user
codes expecting lists would still get what they expect.


-------------------------------------------------------

Rationale:

1. Following Python tradition, using magic methods is preferred to
   adding special syntaxes.  Special syntaxes are only added if existing
   syntax space is not large enough to call these magic methods, which
   is not the case here.

2. Iterators should not be limited to what they can emit, so the items
   should be allowed to be tuples of any desired arity.  For example, an
   iterator over spatial points may be used as "x,y,z in points".  This
   generality is not possible with the colon based special syntax that
   is only applicable to lists and dicts.

3. Iterators should be completely clear about what they iterate over
   (keys, values, items, indexes, lines, ...) without the assistance of
   context.  Not depending on context is one of the major advantages of
   python over Perl, and should be adhered to if at all possible.  The
   proposed syntax do not make any implicit assumptions about what is to
   be iterated over.

4. This is completely backward compatible.  Any conversion can be done
   one by one. It can also be performed by global search and
   replacement.  This is not so for the special colon syntax.

5. Any conversion to the new syntax has the immediate effect of removing
   an empty () while possibly improving performance.  Nothing else
   changes.  The conversion does not require implementers and users to
   keep in sync.

6. It plays well with list comprehension, which has a syntax lacking in
   delimiters.  The alternative colon syntax would have added delimiters
   at the wrong places.

7. The arity attribute also solves the problem of tuple unpacking where
   the correctness of syntax cannot be determined in advance of an
   iteration.

8. The similarity between for loops and in operators offer such a
   conceptual simplification that it should not be dropped unless
   absolutely necessary.


-------------------------------------------------------

Appendix: Application of the __arity__ syntax to general tuple unpacking.

This is independent of the iterator issue, but it is closely related to
the outputs of iterators.  The __arity__ syntax proposed above appears
to also solve tuple unpacking problems elsewhere.

Tuple unpacking is perhaps the only place where the validity of python
assignment syntax depends on the value.  It is therefore often desirable
to convey this information before it is too late.  The __arity__
attribute may be used for this task.

    def func():
        __arity__ = 2
        x = (1,2)
        return x

    a, b = func()        # Guaranteed success for unpacking
    a = func()           # a becomes a tuple
    a, b, c = func()     # UnpackError: attempt to change arity from 2 to 3.

    def func():
        __arity__ = 2
        x = 2
        return x
    a, b = func()        # ArityError: func arity is 2, but returns 1.
    a = func()           # ArityError: func arity is 2, but returns 1.
    a, b, c = func()     # UnpackError: attempt to change arity from 2 to 3.

Both UnpackError and ArityError are subclasses of ValueError.  This
could be implemented this way: If the left hand side of an assignment is
a tuple while the right hand side is a call to f, check that the arity
of the tuple matches f.__arity__ unless f.__arity__ is None.

All functions not defining __arity__ attribute default to
__arity__==None.  Therefore one can write things like

    if f.__arity__ == 1: x = f()
    elif f.__arity__ == 2: x,y = f()
    else: items = f()


Huaiyu Zhu   <hzhu at users.sourceforge.net>





From aahz at panix.com  Wed Feb 21 17:15:02 2001
From: aahz at panix.com (Aahz Maruch)
Date: 21 Feb 2001 08:15:02 -0800
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org>
Message-ID: <970pi6$i4q$1@panix3.panix.com>

In article <mailman.982769240.31345.python-list at python.org>,
Huaiyu Zhu  <hzhu at users.sourceforge.net> wrote:
>
>Following suggestion by Jeff Petkau <jpet at eskimo.com>,
>(http://mail.python.org/pipermail/python-list/2001-February/029944.html),
>here is a different proposal for iterators.

This is really great; with a tiny bit of work, what you're calling an
iterator is almost the same as an Icon generator, something many of us
have been wishing for a long time.  I hope you cc'd this to Ka-Ping; I'd
also suggest posting this to python-dev.

>Applications to Builtins and Other Common Situations:
>
>1. Sequences.  The builtin sequence types (list, tuple, string) are
>  changed to iterators by adding the three magic methods __next__,
>  __contains__ and __list__.  The following is always true:
>
>    x.__list__() == list(x)
>
>  In addition, Each sequence object x defines two attributes (indexes,
>  items) that are themselves iterators.  The arity of indexes is 1, that
>  of items is 2.
>
>    x.indexes.__list__() == range(len(x))
>    x.items.__list__() == zip(range(len(x)), x)

I think I would change the text to read:

    In addition, each sequence object x defines two methods (indexes,
    items) that each return an iterator object.  The arity of indexes is
    1, that of items is 2.

The reason is that we want to keep hammering on the generality of
iterators, for example in passing a single iterator object to a bunch of
threads to generate keys.

That brings up the point of how iterator.__list__() should work with an
unbounded iterator.

>Both UnpackError and ArityError are subclasses of ValueError.  This
>could be implemented this way: If the left hand side of an assignment is
>a tuple while the right hand side is a call to f, check that the arity
>of the tuple matches f.__arity__ unless f.__arity__ is None.
>
>All functions not defining __arity__ attribute default to
>__arity__==None.  Therefore one can write things like
>
>    if f.__arity__ == 1: x = f()
>    elif f.__arity__ == 2: x,y = f()
>    else: items = f()

I have no clue how to design/implement this, but it would be really neat
if functions could *read* the desired arity and change the return value
based on it.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From aahz at panix.com  Wed Feb 21 17:43:02 2001
From: aahz at panix.com (Aahz Maruch)
Date: 21 Feb 2001 08:43:02 -0800
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org>
Message-ID: <970r6m$a4$1@panix2.panix.com>

In article <mailman.982769240.31345.python-list at python.org>,
Huaiyu Zhu  <hzhu at users.sourceforge.net> wrote:
>
>  [...]

I'm not sure where to put this question, so I'm just deleting all of
Huaiyu's text.

How does one get UserDict.keys (note: no parens here) to return an
iterator?  Normally this will simply return a reference to the method.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From mwh21 at cam.ac.uk  Wed Feb 21 18:36:16 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 21 Feb 2001 17:36:16 +0000
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <970r6m$a4$1@panix2.panix.com>
Message-ID: <m37l2kszu7.fsf@atrus.jesus.cam.ac.uk>

aahz at panix.com (Aahz Maruch) writes:

> In article <mailman.982769240.31345.python-list at python.org>,
> Huaiyu Zhu  <hzhu at users.sourceforge.net> wrote:
> >
> >  [...]
> 
> I'm not sure where to put this question, so I'm just deleting all of
> Huaiyu's text.
> 
> How does one get UserDict.keys (note: no parens here) to return an
> iterator?  Normally this will simply return a reference to the method.

class _Keys:
   def __init__(self, dict):
       self.dict = dict
   def __contains__(self, item):
       return self.dict.has_key(item)
   def __call__(self):
       return self.dict.keys()
   ... etc ...

class UserDict:
    def __init__(self, dict=None):
        if data is None:
            data = {}
        self.data = dict
        self.keys = _Keys(dict)
        ... etc ...

Not *very* different from how you'd implement this in C, really.

Hmm, you might want to have the _Keys object hang on the instance
rather than the dict if assigning to the .data member of UserDict is
allowed.  Which would create a circular reference, but we don't have
to worry about that too much now...

Cheers,
M.

-- 
  > Look I don't know.  Thankyou everyone for arguing me round in
  > circles.  
  No need for thanks, ma'am; that's what we're here for.
                                    -- LNR & Michael M Mason, cam.misc


From tim.hochberg at ieee.org  Wed Feb 21 20:24:28 2001
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 21 Feb 2001 19:24:28 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <970r6m$a4$1@panix2.panix.com> <m37l2kszu7.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <MBUk6.126$a_2.14850@news1.rdc1.az.home.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m37l2kszu7.fsf at atrus.jesus.cam.ac.uk...
> aahz at panix.com (Aahz Maruch) writes:
>
> > In article <mailman.982769240.31345.python-list at python.org>,
> > Huaiyu Zhu  <hzhu at users.sourceforge.net> wrote:
> > >
> > >  [...]
> >
> > I'm not sure where to put this question, so I'm just deleting all of
> > Huaiyu's text.
> >
> > How does one get UserDict.keys (note: no parens here) to return an
> > iterator?  Normally this will simply return a reference to the method.
>
> class _Keys:
>    def __init__(self, dict):
>        self.dict = dict
>    def __contains__(self, item):
>        return self.dict.has_key(item)
>    def __call__(self):
>        return self.dict.keys()
>    ... etc ...
>
> class UserDict:
>     def __init__(self, dict=None):
>         if data is None:
>             data = {}
>         self.data = dict
>         self.keys = _Keys(dict)
>         ... etc ...
>
> Not *very* different from how you'd implement this in C, really.
>
> Hmm, you might want to have the _Keys object hang on the instance
> rather than the dict if assigning to the .data member of UserDict is
> allowed.  Which would create a circular reference, but we don't have
> to worry about that too much now...

I don't think this will work right though. Consider:

for k in aUserDict.keys:
   # Do some stuff
for k in aUserDict.keys:
    # Do some more stuff

The way this is implemented here the second loop would bail out immediately
because there is only a single iterator for a given instance and it is
already exhausted. One could imagine even odder instances where one broke
out halfway through the first loop; the second loop would then start
iterating halfway through the dictionaries keys. To get this to work "right"
one would need to generate a new iterator using getattr magic every time the
attribute was asked. Something like:

def UserDict:
   def __attr_keys_(self, op, value):
      if "op" == "get": return _Keys(dict)
      else: raise ValueError("keys not writable")

Here I've used the syntax from the draft "Attribute Access Handlers" PEP.
This makes the keys attribute sufficiently magical that it makes me a bit
nervous. On the other hand, one could do kinda cool stuff like:

myKeys = aUserDict.keys # Get my own copy of the keys iterator
for key in myKeys:
    # Do Stuff
    if key == someSpecialKey:
        break
# Do some processing out of the loop
for key in myKeys: # resume where we left off
    # Do some other stuff

Sure that example's a little contrived, but I'm sure that must be usefule
somewhere <0.8 wink>.

-tim






From mwh21 at cam.ac.uk  Wed Feb 21 20:34:59 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 21 Feb 2001 19:34:59 +0000
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <970r6m$a4$1@panix2.panix.com> <m37l2kszu7.fsf@atrus.jesus.cam.ac.uk> <MBUk6.126$a_2.14850@news1.rdc1.az.home.com>
Message-ID: <m31ysru8ws.fsf@atrus.jesus.cam.ac.uk>

"Tim Hochberg" <tim.hochberg at ieee.org> writes:

[schnipp]
> I don't think this will work right though. Consider:
> 
> for k in aUserDict.keys:
>    # Do some stuff
> for k in aUserDict.keys:
>     # Do some more stuff
> 
> The way this is implemented here the second loop would bail out
> immediately because there is only a single iterator for a given
> instance and it is already exhausted.

Nah, reset the allocator every time item 0 is asked for.  I'd have to
admit I haven't read Huaiyu Zhu's proposal carefully enough to be sure
this would work, but something like this ought to be possible, I'd
have thought.

I mean,

l = range(10)
for i in l:
   # do some stuff
for i in l:
   # do some other stuff

will have to do something along these lines, no?

Cheers,
M.

-- 
  In general, I'd recommend injecting LSD directly into your temples,
  Syd-Barret-style, before mucking with Motif's resource framework.
  The former has far lower odds of leading directly to terminal
  insanity.                                            -- Dan Martinez


From tim.hochberg at ieee.org  Wed Feb 21 21:29:34 2001
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 21 Feb 2001 20:29:34 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <970r6m$a4$1@panix2.panix.com> <m37l2kszu7.fsf@atrus.jesus.cam.ac.uk> <MBUk6.126$a_2.14850@news1.rdc1.az.home.com> <m31ysru8ws.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <OyVk6.349$a_2.40263@news1.rdc1.az.home.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m31ysru8ws.fsf at atrus.jesus.cam.ac.uk...
> "Tim Hochberg" <tim.hochberg at ieee.org> writes:
>
> [schnipp]
> > I don't think this will work right though. Consider:
> >
> > for k in aUserDict.keys:
> >    # Do some stuff
> > for k in aUserDict.keys:
> >     # Do some more stuff
> >
> > The way this is implemented here the second loop would bail out
> > immediately because there is only a single iterator for a given
> > instance and it is already exhausted.
>
> Nah, reset the allocator every time item 0 is asked for.  I'd have to
> admit I haven't read Huaiyu Zhu's proposal carefully enough to be sure
> this would work, but something like this ought to be possible, I'd
> have thought.

But the loop psuedocode that Huaiyu Zhu proposed has no concept of asking
for item zero:

   for a, b, c, ... in A:
        do something

becomes:

    while 1:
        try:
            a, b, c, ... = A.__next__()
            do something
        except IndexError:
            break

One could add an A.__reset__() call to the start of the loop, but I think
that that would hamstring some other applications one would like for
iterators. Resetting at the end doesn't appear to be any better. For the
moment, I'll stand my original claim that the The Right Thing To Do (tm) is
to get a new iterator each time.

On the other hand (it was a short moment), this isn't really the behaviour
you want for, for instance,  file.readlines. Let's look at two cases:

One would like aDict.keys to return distinct iterators:

for key in aDict.keys:
  doSomething
for key in aDict.keys:
  doSomething

This should iterate over aDict's keys twice.

However aFile can really have one iterator:

for line in aFile.readlines:
    doSomething
for line in aFile.readlines:
    doSomething

This should almost certainly only iterate over the file once (in the first
loop) and bail out immediately in the second loop. I suppose allowing
different types of objects to return new iterators or not as appropriate
would fix this.

OK, so my new position is that objects should return new iterators when they
feel it's appropriate <0.1 wink>

Not-really-concerned-that-I'll-be-mistaken-for-the-real-tim'ly yours,

-tim

> I mean,
>
> l = range(10)
> for i in l:
>    # do some stuff
> for i in l:
>    # do some other stuff
>
> will have to do something along these lines, no?
>
> Cheers,
> M.
>
> --
>   In general, I'd recommend injecting LSD directly into your temples,
>   Syd-Barret-style, before mucking with Motif's resource framework.
>   The former has far lower odds of leading directly to terminal
>   insanity.                                            -- Dan Martinez




From wolfson at midway.uchicago.edu  Wed Feb 21 22:51:15 2001
From: wolfson at midway.uchicago.edu (Ben Wolfson)
Date: Wed, 21 Feb 2001 21:51:15 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <MBUk6.126$a_2.14850@news1.rdc1.az.home.com> <m31ysru8ws.fsf@atrus.jesus.cam.ac.uk> <OyVk6.349$a_2.40263@news1.rdc1.az.home.com>
Message-ID: <nLWk6.171$l25.2240@newsfeed.uchicago.edu>

In article <OyVk6.349$a_2.40263 at news1.rdc1.az.home.com>,
Tim Hochberg <tim.hochberg at ieee.org> wrote:
>
>"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
>news:m31ysru8ws.fsf at atrus.jesus.cam.ac.uk...
>> "Tim Hochberg" <tim.hochberg at ieee.org> writes:
>>
>> [schnipp]
>> > I don't think this will work right though. Consider:
>> >
>> > for k in aUserDict.keys:
>> >    # Do some stuff
>> > for k in aUserDict.keys:
>> >     # Do some more stuff
>> >
>> > The way this is implemented here the second loop would bail out
>> > immediately because there is only a single iterator for a given
>> > instance and it is already exhausted.
>>
>> Nah, reset the allocator every time item 0 is asked for.  I'd have to
>> admit I haven't read Huaiyu Zhu's proposal carefully enough to be sure
>> this would work, but something like this ought to be possible, I'd
>> have thought.
>
>But the loop psuedocode that Huaiyu Zhu proposed has no concept of asking
>for item zero:
>
>   for a, b, c, ... in A:
>        do something
>
>becomes:
>
>    while 1:
>        try:
>            a, b, c, ... = A.__next__()
>            do something
>        except IndexError:
>            break
>
>One could add an A.__reset__() call to the start of the loop, but I think
>that that would hamstring some other applications one would like for
>iterators. Resetting at the end doesn't appear to be any better. For the
>moment, I'll stand my original claim that the The Right Thing To Do (tm) is
>to get a new iterator each time.

One could also specify that, when appropriate, an iterator calls
__reset__ before raising an IndexError.

-- 
Barnabas T. Rumjuggler
For the world, I count it not an inn, but an hospital; and a place not to
live, but to die in.
 -- Thomas Browne, _Religio Medici_


From qrczak at knm.org.pl  Thu Feb 22 00:21:10 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 21 Feb 2001 23:21:10 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <MBUk6.126$a_2.14850@news1.rdc1.az.home.com> <m31ysru8ws.fsf@atrus.jesus.cam.ac.uk> <OyVk6.349$a_2.40263@news1.rdc1.az.home.com> <nLWk6.171$l25.2240@newsfeed.uchicago.edu>
Message-ID: <slrn998jb6.28a.qrczak@qrnik.zagroda>

Wed, 21 Feb 2001 21:51:15 GMT, Ben Wolfson <wolfson at midway.uchicago.edu> pisze:

> One could also specify that, when appropriate, an iterator calls
> __reset__ before raising an IndexError.

It doesn't work. I can have several iterators walking over a single
sequence at once.

Obtaining a stateful iterator is a stateful operation. You can't make
the iterator once and keep it with the sequence, but you must produce
an iterator each time an iteration is requested - unless its usage
is functional, i.e. for the next element you don't call a method
with side effects keeping the same iterator handle, but rebind the
iterator handle after each iteration.

If such operation is spelled as an attribute access, it must still be a
function call internally. Disadvantages are twofold: The implementation
must use __getattr__ to let attribute access have function call
semantics. And it's confusing to programmers, which expect that such
syntax accesses the same object each time (at least logically the same)
- but they are stateful and have different identities.

<ftp://ftp.netcom.com/pub/hb/hbaker/Iterator.html>

For comparison, in the tradition of functional languages linked lists
are very common. They are their own iterators, i.e. it is not needed
to have a separate iterator concept.

Elements are not accessed by indices (unless someone really wants it),
because such element access is O(n). Instead recursive functions are
written to traverse lists. There are functions which wrap common
patterns of such iteration, e.g. map, filter, reduce.

Iteration relies on the fact that obtaining a tail of a list (the
same list but without its first element) is O(1). It is safe to not
perform any copying, because lists are generally immutable. In lazy
languages lists are of course lazy. Functional programming is really
about immutability; functions are only the second important thing.

I would strongly advise to not invent a complicated stateful iterator
framework, with all C++ features like stepping backwards, stepping
the specified number of elements, computing a distance between two
iterators over the same sequence, sorting the part of a sequence
between two iterators, and all weird rules about which iterators are
invalidated when you modify the sequence in interesting ways.

Instead, if you want to have lazy reading of files expressed as an
iteration over the sequence of lines, lazy iteration over dictionary
keys without building their physical list, etc. - either design a
simple iterator protocol to keep close to the current Python, or build
a rich framework for lazy lists. In lazy languages all data structures
are lazy, but for Python it should be enough to have lazy lists.

Please don't stress mutability of such sequences. You should not
mutate a sequence you are iterating over.

In any case, the thing a 'for' loop is looping over should be
kind-of-sequence. Not iterator (a new concept for Python) nor a pair
of them (ugh!). Various things can be expressed as kind-of-sequences,
e.g. reversing adaptor which does not make a physical copy, or a
subsequence, or things discussed before: range(), readlines(), keys().

By kind-of-sequence I mean that it's not sufficient to use current
sequence interface. It does not allow to distinguish between sequential
access and random index access. It can be extended to primarily have
sequential access in mind.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From aahz at panix.com  Wed Feb 21 21:29:35 2001
From: aahz at panix.com (Aahz Maruch)
Date: 21 Feb 2001 12:29:35 -0800
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <m37l2kszu7.fsf@atrus.jesus.cam.ac.uk> <MBUk6.126$a_2.14850@news1.rdc1.az.home.com> <m31ysru8ws.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <9718ff$13a$1@panix6.panix.com>

In article <m31ysru8ws.fsf at atrus.jesus.cam.ac.uk>,
Michael Hudson  <mwh21 at cam.ac.uk> wrote:
>
>Nah, reset the allocator every time item 0 is asked for.  I'd have to
>admit I haven't read Huaiyu Zhu's proposal carefully enough to be sure
>this would work, but something like this ought to be possible, I'd
>have thought.
>
>I mean,
>
>l = range(10)
>for i in l:
>   # do some stuff
>for i in l:
>   # do some other stuff
>
>will have to do something along these lines, no?

Nope.  In Huaiyu's proposal, l will be a real list and each of the two
for loops will create a new iterator.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From root at rainerdeyke.com  Wed Feb 21 21:29:21 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Wed, 21 Feb 2001 20:29:21 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <970r6m$a4$1@panix2.panix.com> <m37l2kszu7.fsf@atrus.jesus.cam.ac.uk> <MBUk6.126$a_2.14850@news1.rdc1.az.home.com>
Message-ID: <ByVk6.277990$ge4.95753828@news2.rdc2.tx.home.com>

"Tim Hochberg" <tim.hochberg at ieee.org> wrote in message
news:MBUk6.126$a_2.14850 at news1.rdc1.az.home.com...
> "Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
> news:m37l2kszu7.fsf at atrus.jesus.cam.ac.uk...
> > aahz at panix.com (Aahz Maruch) writes:
> > class _Keys:
> >    def __init__(self, dict):
> >        self.dict = dict
> >    def __contains__(self, item):
> >        return self.dict.has_key(item)
> >    def __call__(self):
> >        return self.dict.keys()
> >    ... etc ...
> >
> > class UserDict:
> >     def __init__(self, dict=None):
> >         if data is None:
> >             data = {}
> >         self.data = dict
> >         self.keys = _Keys(dict)
> >         ... etc ...

> I don't think this will work right though. Consider:
>
> for k in aUserDict.keys:
>    # Do some stuff
> for k in aUserDict.keys:
>     # Do some more stuff

UserDict.__getattr__ could construct a new 'keys' object on demand instead.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From KatMouse at gmx.de  Wed Feb 21 21:18:43 2001
From: KatMouse at gmx.de (Eduard Hiti)
Date: Wed, 21 Feb 2001 21:18:43 +0100
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org>
Message-ID: <9717su$db4$02$1@news.t-online.com>

| Predefined Usage of Iterators:
|
| 1. The looping structure:
|
|   Consider the structure
|
|     for a, b, c, ... in A:
|         do something
|     else:
|         do otherthing
|
|   If A defines __next__, the above is interpreted as the equivalence of
|   (__got is an invisible variable):
|
|     __got = 0
|     while 1:
|         try:
|             a, b, c, ... = A.__next__()
|             __got = 1
|             do something
|         except IndexError:
|             break
|     if not __got:
|         do otherthing
|

I think there is an error in the code above. Shouldn't it rather go like:

     __exhausted = 0
     while 1:
         try:
             a, b, c, ... = A.__next__()
             do something
         except IndexError:
             __exhausted = 1
             break
     if __exhausted:
         do otherthing


The 'do something' block can break to avoid executing 'do otherthing'.





From root at rainerdeyke.com  Wed Feb 21 21:29:21 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Wed, 21 Feb 2001 20:29:21 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org>
Message-ID: <ByVk6.277989$ge4.95754105@news2.rdc2.tx.home.com>

"Huaiyu Zhu" <hzhu at users.sourceforge.net> wrote in message
news:mailman.982769240.31345.python-list at python.org...
> Following suggestion by Jeff Petkau <jpet at eskimo.com>,
> (http://mail.python.org/pipermail/python-list/2001-February/029944.html),
> here is a different proposal for iterators.
>
> Like the colon-syntax proposal, it achieves the following
>
> 1. It provides an extensible iterator interface without pretending to
>    provide random access to elements.
> 2. It resolves the endless "i indexing sequence" debate.
> 3. It allows performance enhancements to dictionary iteration.
> 4. It is completely backward compatible.

5. It will allow iteration over "sequences" longer than 'sys.maxint'.

> They MAY but NEED NOT define a __len__ method, which may or may not
> return semantically correct result.

What is the point in allowing a semantically incorrect '__len__'?

>   If A defines __next__, the above is interpreted as the equivalence of
>   (__got is an invisible variable):

<snip>

>   Else if A defines __list__, then "for x in A:" is interpreted as "for
>   x in A.__list__():".  Note that every list is an iterator (see below).

(see below)

>   Else if A is a callable object, then "for x in A:" is interpreted as
>   "for x in A():".

Won't this interfere with existing user-defined sequences that are also
callable but don't support the iterator protocol?

> 2. The in operator:
>
>   Consider the expression
>     (a, b, c, ...) in A:
>
>   If A defines __contains__, the above is interpreted as the equivalence
>   of:
>     A.__contains__((a,b,c, ...))
>
>   Else if A defines __list__, then "x in A" is interpreted as "x in
>   A.__list__()".  Note that every list is an iterator (see below).

Why is '__next__' not used?

> Applications to Builtins and Other Common Situations:
>
> 1. Sequences.  The builtin sequence types (list, tuple, string) are
>   changed to iterators by adding the three magic methods __next__,
>   __contains__ and __list__.  The following is always true:
>
>     x.__list__() == list(x)

This is really really bad.  Iterators have state.  Tuples and strings do
not.  Consider the following:

a = 'Hell, O world!'
a.__next__()
b = 'Hell, O world!'
print '"%s", "%s", %d' % (a, b, a == b) # Are they the same?
print a.__next__(), b.__next__()

Don't do it.  Instead, either the original iteration protocol should be left
in place as backup, or there should be some automagic way of generating
iterators form sequences.


>   In addition, Each sequence object x defines two attributes (indexes,
>   items) that are themselves iterators.  The arity of indexes is 1, that
>   of items is 2.


NOOooooOoooOo!!!!!!!!

>
>     x.indexes.__list__() == range(len(x))

'range(len(x))' returns a list of integers, not a list of 1-tuples.  This
may not be significant for sequences, but consider mappings:

> 2. Mappings.  Each mapping object x defines three attributes (keys,
>   values, items) that are callable iterators.  The arity of keys and
>   values is 1, that of items is 2.

The following is currently valid Python code:

d = {('cheese', 'shop') : ('dead', 'parrot') }

for a, b in d.keys():
  print a, b # Prints 'cheese shop'

for a, b in d.values():
  print a, b # Prints 'dead parrot'

Setting the arity of 'keys' and 'values' to 1 breaks this.

> 4. The Iterator class.  An iterator class could be included with
>   standard python, so that one could write:
>
>     a = Iterator([1,2,3])       # Or a = Iterator(); a.__list__ = [1,2,3]

I thought the '__list__' attribute of iterators was a method that returns a
list, not a list.

> 1. Following Python tradition, using magic methods is preferred to
>    adding special syntaxes.  Special syntaxes are only added if existing
>    syntax space is not large enough to call these magic methods, which
>    is not the case here.

I thought all magic methods were merely ways of implementing the
functionality of built-in types in classes, i.e. e.g. '__add__' is never
called directly, but merely used by the built-in types 'class instance' to
implement the syntax 'a + b'.

> 2. Iterators should not be limited to what they can emit, so the items
>    should be allowed to be tuples of any desired arity.  For example, an
>    iterator over spatial points may be used as "x,y,z in points".  This
>    generality is not possible with the colon based special syntax that
>    is only applicable to lists and dicts.

Why must iterators emit tuples at all?

> Appendix: Application of the __arity__ syntax to general tuple unpacking.
>
> This is independent of the iterator issue, but it is closely related to
> the outputs of iterators.  The __arity__ syntax proposed above appears
> to also solve tuple unpacking problems elsewhere.

<snip>

I am totally against the '__arity__' syntax.   It only seems to replace one
thrown exception for another.  Besides, in my experience, code like the
following is not uncommon:

def return_tuple(n):
  return (0,) * n

a, b = return_tuple(2)
x, y, z = return_tuple(3)


Such code could not use '__arity__', reducing it to special case syntax
sugar.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From qrczak at knm.org.pl  Wed Feb 21 22:46:33 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 21 Feb 2001 21:46:33 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org>
Message-ID: <slrn998dpo.ch1.qrczak@qrnik.zagroda>

Wed, 21 Feb 2001 03:32:03 -0800 (PST), Huaiyu Zhu <hzhu at users.sourceforge.net> pisze:

> Following suggestion by Jeff Petkau <jpet at eskimo.com>,
> (http://mail.python.org/pipermail/python-list/2001-February/029944.html),
> here is a different proposal for iterators.

I dislike it for the following reasons:

* It requires "magic attributes", i.e. implementing __getattr__ in
  each user class that wants to be a sequence with the same interface
  as list. Python already has a way used for things like obtaining
  iterators: method calls. Use that. Using d.keys to mean a flavor
  of d.keys() is weird.

* It is too complex. There are many special methods, and syntactic
  sugar like for loops expands to one of many forms depending on
  what operations are provided. Python succeeded being quite simple.
  Please don't ruin that.

* The __arity__ thing does not fit dynamically typed Python at all.
  When you want some static typing, go for a more general type system
  (which would be optional in the context of Python) instead of
  adding a kludge like this.


What do I propose instead? Choose one:

1. Design a simple iterator protocol and add methods like
   xitems(), xkeys() and xvalues() which produce such iterators.
   I mean simple protocol, used only for 'for' loops! A single
   method for __next__ suffices (it can be __call__).

2. Design a lazy list framework. Let items(), keys(), values(),
   range(), readlines() etc. return lazy lists. Laziness is not as easy
   in imperative languages as in functional ones. The framework should
   take care to keep the current behavior of
       k = d.keys()
       d[spam] = eggs
       use(k)
   It can be done by putting weak references (avoiding reference loops)
   to iterators obtained from a mutable object inside the object, to
   notify iterators when the object is about to mutate, so they suck
   all elements immediately before the mutation. Perhaps design a more
   general mutability notification framework before that.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From Denys.Duchier at ps.uni-sb.de  Wed Feb 21 23:39:05 2001
From: Denys.Duchier at ps.uni-sb.de (Denys Duchier)
Date: 21 Feb 2001 23:39:05 +0100
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org> <slrn998dpo.ch1.qrczak@qrnik.zagroda>
Message-ID: <is1ysrk6eu.fsf@elk.ps.uni-sb.de>

qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk) writes:

> 1. Design a simple iterator protocol and add methods like
>    xitems(), xkeys() and xvalues() which produce such iterators.

I completely agree with this suggestion.  I don't think that the hocus
pocus of magic attributes is a good semantic foundation for language
design.  This suggestion is simpler: it's business as usual with added
value.

In all other respects, the iterators idea is much more sensible than
the proposed `:' syntax extension.

Cheers,

-- 
Dr. Denys Duchier			Denys.Duchier at ps.uni-sb.de
Forschungsbereich Programmiersysteme	(Programming Systems Lab)
Universitaet des Saarlandes, Geb. 45	http://www.ps.uni-sb.de/~duchier
Postfach 15 11 50			Phone: +49 681 302 5618
66041 Saarbruecken, Germany		Fax:   +49 681 302 5615


From jpet at eskimo.com  Thu Feb 22 06:41:19 2001
From: jpet at eskimo.com (Jeff Petkau)
Date: Thu, 22 Feb 2001 05:41:19 GMT
Subject: Alternative iterator syntax
References: <mailman.982769240.31345.python-list@python.org>
Message-ID: <3E1l6.2278$3d5.446687@paloalto-snr1.gtei.net>

Huaiyu Zhu <hzhu at users.sourceforge.net> wrote in message
news:mailman.982769240.31345.python-list at python.org...
> Following suggestion by Jeff Petkau <jpet at eskimo.com>,
> (http://mail.python.org/pipermail/python-list/2001-February/029944.html),
> here is a different proposal for iterators.
...

As proposed, this would break existing classes that define
only spam.__getitem__() and expect 'for x in spam' to work.

It would be a much smaller change to leave 'for' alone,
using __getitem__(0,1,2...).  Or maybe adding the fallback
to __call__ if __getitem__ is not defined.

But I'm not sure this whole thing is a good idea.  How about
just some builtin functions for items(), keys(), values(),
and indexed() that return lazy iterators?

for k,v in items(mydict)
for k in keys(mydict)
for v in values(mydict)
for i,v in indexed(mysequence)

indexed() is easy:

    class indexed:
        def __init__(self,seq):
            self.seq = seq
        def __getitem__(self,i):
            return i,self.seq[i]

The others would require a special hook in dict. I'd prefer
an integer-indexed access function, but having it return an
iterator works too if you're into that sort of thing.

    class items:
        def __init__(self,d):
            self.dict = d
        def __getitem__(self,i):
            return i,self.dict.__getitem_n__(i)

--Jeff Petkau (jpet at eskimo.com)





From alet at unice.fr  Wed Feb 21 13:16:45 2001
From: alet at unice.fr (Jerome Alet)
Date: Wed, 21 Feb 2001 13:16:45 +0100
Subject: [ANNOUNCE] JAXML version 2.20
Message-ID: <3A93B1AD.56713213@unice.fr>

I'm very pleased to announce you the new 2.20 version of the JAXML
module.

With the JAXML module, generating XML documents has never been easier,
this module:

        * Takes care of the indentation for you: your XML documents are
readable by humans.
        * Takes care of closing tags for you.
        * Can generate every tag you want in an incredibly easy way.
        * Offers two templating facilities (see below)
        * Can generate XHTML documents and be used in CGI scripts
because it
          takes care of the HTTP headers too.

Here's a sample:

--- CUT ---
import jaxml
x = jaxml.XML_document()
x.something(value=5).dummy("some dummy text", isntit="YES")
x['dummy'] = 'interesting' # new templating facility
print x

will produce:

<?xml version="1.0" encoding="iso-8859-1"?>
<something value="5">
    <interesting isntit="YES">some interesting text</interesting>
</something>

and

x._updatemapping({'dummy': 'boring'})
print x

will produce:

<?xml version="1.0" encoding="iso-8859-1"?>
<something value="5">
    <boring isntit="YES">some boring text</boring>
</something>
--- CUT ---

The JAXML module is GPLed, you can download it freely from:

        http://cortex.unice.fr/~jerome/jaxml/

Changes:

        * A new _updatemapping() method was added to initialise
          or update the internal mapping for the new templating
facility.
	  This mapping is only used when rendering the document.
        * An infinite loop bug which was triggered when you called dir()
          on an XML_document instance was corrected.
        * The basic '+' and '*' operators can now be used to concatenate
          or repeat documents. Nota Bene: this is not a real
concatenation,
	  the second document is inserted in the first one at the current
	  position.

The well commented test program and module's sources are your best
friends.

Comments are very welcome, and NO the licence won't be changed to other
than pure GPL, BUT we can discuss special agreements if you are really 
interested (this don't necessary mean money).

-- 
Jerome Alet - alet at unice.fr - http://cortex.unice.fr/~jerome


From syver at NOSPAMcyberwatcher.com  Wed Feb 21 13:38:44 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 21 Feb 2001 13:38:44 +0100
Subject: Python program stopping execution wating for keypress
Message-ID: <dFOk6.857$t21.9975@news3.oke.nextra.no>

I've experienced some strange behaviour from some Python console programs
that are running a long time. The incidents I can remember are all
multithreaded and uses win32com.

Some times the programs just stop in the middle of the execution,
essentially sleeping as they use no processor time. The times I've observed
this is on applications running in a console, so I tried pressing a key and
then the application starts executing again. Sometimes needing more than one
key press per run to keep it going.

Have anyone witnessed this behaviour, I would be glad to know.




From fredrik at pythonware.com  Wed Feb 21 19:33:41 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 18:33:41 GMT
Subject: Python program stopping execution wating for keypress
References: <dFOk6.857$t21.9975@news3.oke.nextra.no>
Message-ID: <9STk6.14843$Qb7.2370627@newsb.telia.net>

Syver Enstad wrote:
> Some times the programs just stop in the middle of the execution,
> essentially sleeping as they use no processor time. The times I've observed
> this is on applications running in a console, so I tried pressing a key and
> then the application starts executing again. Sometimes needing more than one
> key press per run to keep it going.
>
> Have anyone witnessed this behaviour, I would be glad to know.

does the program print stuff to the console, and stop at
a print statement?

do you have "quick edit" (or whatever it's called in your
version of windows) enabled?

did you click in the window?

(this switches the console to "selection mode", and blocks
output until you press a key or click the right button.  the
titlebar also changes in this mode)

Cheers /F




From erno-news at erno.iki.fi  Wed Feb 21 20:21:00 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 21 Feb 2001 21:21:00 +0200
Subject: Python program stopping execution wating for keypress
References: <dFOk6.857$t21.9975@news3.oke.nextra.no>
Message-ID: <kud7cbsuzn.fsf@lasipalatsi.fi>

In article <dFOk6.857$t21.9975 at news3.oke.nextra.no>, "Syver Enstad"
<syver at NOSPAMcyberwatcher.com> writes:

| Some times the programs just stop in the middle of the execution,
| essentially sleeping as they use no processor time. The times I've
| observed this is on applications running in a console, so I tried
| pressing a key and then the application starts executing again.

some piece of code you are using probably is buggy and tries
to read from stdin. try booby-trapping it by assigning a non-file
object to sys.stdin (or doing sys.stdin.close()).

  -- erno


From thomas at cintra.no  Wed Feb 21 13:55:19 2001
From: thomas at cintra.no (Thomas Weholt)
Date: Wed, 21 Feb 2001 13:55:19 +0100
Subject: Anybody doing irc-stuff with Python / status of irc-modules for Python
Message-ID: <ZUOk6.867$t21.9660@news3.oke.nextra.no>

Hi,

Question #1 : Is anybody doing anything IRC-related in Python?
Question #2 : Does anybody know what module to use based on features and
ease of use?

I need a library to connect to a server and act as a information provider in
a channel ( as a BOT ), also serving files if asked, like an fserv. The
thing must answer queries, commands and send info to users as they log in.

Any clues? Any information would be appreciated.

PS! Also interested in libraries related to ICQ with similar goals in mind.

Thomas




From claird at starbase.neosoft.com  Wed Feb 21 17:32:34 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Feb 2001 10:32:34 -0600
Subject: Anybody doing irc-stuff with Python / status of irc-modules for Python
References: <ZUOk6.867$t21.9660@news3.oke.nextra.no>
Message-ID: <3D3B6B6D10077B24.500FF5674E90754F.3A301FA60722AC3A@lp.airnews.net>

In article <ZUOk6.867$t21.9660 at news3.oke.nextra.no>,
Thomas Weholt <thomas at cintra.no> wrote:
>Hi,
>
>Question #1 : Is anybody doing anything IRC-related in Python?
			.
			.
			.
Sean Reifschneider does <URL: ftp://ftp.tummy.com:/pub/tummy/ >
Maybe he'll check in and explain a bit more of his work.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From u603515625 at spawnkill.ip-mobilphone.net  Wed Feb 21 15:48:15 2001
From: u603515625 at spawnkill.ip-mobilphone.net (u603515625 at spawnkill.ip-mobilphone.net)
Date: Wed, 21 Feb 2001 14:48:15 GMT
Subject: A working setup.py for DCOracle 1.3.2 /Oracle 8.1.6/HP-UX 11
Message-ID: <l.982766896.1613098144@[193.103.148.194]>

To whom it may concern !

every now and then questions arise about the compilation of DCOracle in several contexts.

Here is a working setup.py to install DCoracle.py in a system consisting of the following components:
- HP-UX 11
- Oracle 8.1.6
- Python 2.0
- DCOracle 1.3.2

I took the the setup.py created by Andrew M.Kuchling in his ZODB-Distribution and changed the ORACLE_LIBS accordingly.
The result follows below the ------- dashes.

The python 2.0 had to be compiled according the hints given in
http://sourceforge.net/bugs/?func=detailbug&bug_id=110665&group
_id=5470 .


Hope that helps somebody

Norbert Klamann

Norbert.Klamann at klamann-software.de


--------------------------------------------------------------------
#!/usr/bin/env python
 
# Setup file for DCOracle
# Python modules written by Digital Creations <http://www.digicool.com>
# setup.py written by A.M. Kuchling <amk1 at bigfoot.com>

import os
from distutils.core import setup
from distutils.extension import Extension

# Set Oracle include and library names

ORACLE_HOME = os.environ['ORACLE_HOME']


# XXX add definitions for other versions here
if 1:
    # Definitions for Oracle 8.0.5, shared library build on Linux
    ORACLE_INCLUDES = ['rdbms/demo', 'plsql/public', 'network/public']
    ORACLE_LIBS = ['clntsh']
    ORACLE_LIBDIRS = [ 'lib' ]

# Prefix all the directories with the Oracle home directory
ORACLE_INCLUDES = map(lambda x: os.path.join(ORACLE_HOME, x), ORACLE_INCLUDES)
ORACLE_LIBDIRS = map(lambda x: os.path.join(ORACLE_HOME, x), ORACLE_LIBDIRS)

# Definitions for extension modules

buffer = Extension(name = 'Buffer',
                   sources= ['src/Buffer.c']
                   )
oci_ = Extension(name = 'oci_',
                         include_dirs = ORACLE_INCLUDES,
                         libraries    = ORACLE_LIBS,
                         library_dirs = ORACLE_LIBDIRS,
                         sources= ['src/oci_.c']
                         )

setup(name = "DCOracle", 
      version = "1.3.1b1",
      description = "Python interface to Oracle",
      author = "Digital Creations",
      author_email = "zope-dev at zope.org",
      url = "http://www.zope.org",

      packages=['DCOracle'],
      
      ext_modules = [buffer, oci_],
      )


 



-- 
Sent by  Norbert.Klamann  from klamann-software subdomain of de
This is a spam protected message. Please answer with reference header.
Posted via http://www.usenet-replayer.com/cgi/content/new


From alexlong at earthlink.net  Wed Feb 21 16:52:46 2001
From: alexlong at earthlink.net (Alex)
Date: Wed, 21 Feb 2001 15:52:46 GMT
Subject: Graphical python development tool for Linux?
Message-ID: <3a93e509.1040040@news.earthlink.net>

Would anyone be kind enough to recommend a graphical Python
development tool, perhaps similar to Pythonwin for Windows, only for
Linux?


From fredrik at pythonware.com  Wed Feb 21 19:06:31 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 18:06:31 GMT
Subject: Graphical python development tool for Linux?
References: <3a93e509.1040040@news.earthlink.net>
Message-ID: <HsTk6.14832$Qb7.2369801@newsb.telia.net>

"Alex" wrote:
> Would anyone be kind enough to recommend a graphical Python
> development tool, perhaps similar to Pythonwin for Windows, only for
> Linux?

http://www.faqts.com/knowledge_base/view.phtml/aid/6433/fid/199

Cheers /F




From chrisw at nipltd.com  Wed Feb 21 17:17:59 2001
From: chrisw at nipltd.com (Chris Withers)
Date: Wed, 21 Feb 2001 16:17:59 +0000
Subject: Compiling Python 2.0 on Win32
Message-ID: <3A93EA37.5E78BA5A@nipltd.com>

Hi,

I want to run a CVS version of Zope on my WinNT 4.0 workstation.

Apparently, that means I need to compile Python from source on my box too.

Where should I look to find out how to do this?

cheers,

Chris



From fredrik at pythonware.com  Wed Feb 21 18:22:52 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 17:22:52 GMT
Subject: Compiling Python 2.0 on Win32
References: <mailman.982772306.7650.python-list@python.org>
Message-ID: <MPSk6.14824$Qb7.2368195@newsb.telia.net>

Chris Withers wrote:
> Apparently, that means I need to compile Python from
> source on my box too.
>
> Where should I look to find out how to do this?

how about reading the README file?

Cheers /F




From chrisw at nipltd.com  Wed Feb 21 18:50:51 2001
From: chrisw at nipltd.com (Chris Withers)
Date: Wed, 21 Feb 2001 17:50:51 +0000
Subject: Compiling Python 2.0 on Win32
References: <mailman.982772306.7650.python-list@python.org> <MPSk6.14824$Qb7.2368195@newsb.telia.net>
Message-ID: <3A93FFFB.E583F32D@nipltd.com>

Fredrik Lundh wrote:
> 
> Chris Withers wrote:
> > Apparently, that means I need to compile Python from
> > source on my box too.
> >
> > Where should I look to find out how to do this?
> 
> how about reading the README file?

I did look through that, but looked again.

I now have an opticians appointment :-)

*embarrassed (and probably badly spelled) grinz*

Chris



From gnana at mips.biochem.mpg.de  Wed Feb 21 18:05:23 2001
From: gnana at mips.biochem.mpg.de (Gnanasekaran Thoppae)
Date: Wed, 21 Feb 2001 17:05:23 -0000
Subject: newbie - concatanating 2 lists
Message-ID: <970sgj+rrf7@eGroups.com>

hi,

i am just beginning to use python.

i have:

li1 = ['a', 'b', 'c']
li2 = ['x', 'y', 'z']

i want:

li3 = ['ax', 'by', 'cz']

how do i do it?

thanks

-gnana




From shaleh at valinux.com  Wed Feb 21 18:29:48 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Wed, 21 Feb 2001 09:29:48 -0800 (PST)
Subject: newbie - concatanating 2 lists
In-Reply-To: <970sgj+rrf7@eGroups.com>
Message-ID: <XFMail.20010221092948.shaleh@valinux.com>

On 21-Feb-2001 Gnanasekaran Thoppae wrote:
> hi,
> 
> i am just beginning to use python.
> 
> i have:
> 
> li1 = ['a', 'b', 'c']
> li2 = ['x', 'y', 'z']
> 
> i want:
> 
> li3 = ['ax', 'by', 'cz']
> 
> how do i do it?
> 

python 2 has a lovely function called zip(), maybe it could help.  Otherwise, a
quick map would probably do this.



From mwh21 at cam.ac.uk  Wed Feb 21 18:40:41 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 21 Feb 2001 17:40:41 +0000
Subject: newbie - concatanating 2 lists
References: <mailman.982775239.14135.python-list@python.org>
Message-ID: <m34rxoszmu.fsf@atrus.jesus.cam.ac.uk>

"Gnanasekaran Thoppae" <gnana at mips.biochem.mpg.de> writes:

> hi,
> 
> i am just beginning to use python.
> 
> i have:
> 
> li1 = ['a', 'b', 'c']
> li2 = ['x', 'y', 'z']
> 
> i want:
> 
> li3 = ['ax', 'by', 'cz']
> 
> how do i do it?

map(lambda x,y: x+y, li1, li2) will do.

More generally,

map(''.join, zip(li1, li2))

will extend to >2 lists.

Cheers,
M.

-- 
  But since I'm not trying to impress  anybody in The Software Big 
  Top, I'd rather walk the wire using a big pole, a safety harness, 
  a net, and with the wire not more than 3 feet off the ground.
                                   -- Grant Griffin, comp.lang.python


From glenfant at equod.com.nospam  Wed Feb 21 18:47:36 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Wed, 21 Feb 2001 18:47:36 +0100
Subject: newbie - concatanating 2 lists
References: <mailman.982775239.14135.python-list@python.org>
Message-ID: <970ukb$t8j$1@reader1.imaginet.fr>

This works with Python 1.5.2 and should with any other

>>> li1 = ['a', 'b', 'c']
>>> li2 = ['x', 'y', 'z']
>>> l3 = map(lambda s1,S2:s1+s2, li1, li2)
>>> l3
['ax', 'by', 'cz']

You're welcome

"Gnanasekaran Thoppae" <gnana at mips.biochem.mpg.de> a ?crit dans le message
news: mailman.982775239.14135.python-list at python.org...
> hi,
>
> i am just beginning to use python.
>
> i have:
>
> li1 = ['a', 'b', 'c']
> li2 = ['x', 'y', 'z']
>
> i want:
>
> li3 = ['ax', 'by', 'cz']
>
> how do i do it?
>
> thanks
>
> -gnana
>
>



From brettlea at earthlink.net  Thu Feb 22 05:51:20 2001
From: brettlea at earthlink.net (Brett Leach)
Date: Thu, 22 Feb 2001 04:51:20 GMT
Subject: newbie - concatanating 2 lists
References: <mailman.982775239.14135.python-list@python.org>
Message-ID: <3A949B0C.31507A4@earthlink.net>

Gnanasekaran Thoppae wrote:
> 
> hi,
> 
> i am just beginning to use python.
> 
> i have:
> 
> li1 = ['a', 'b', 'c']
> li2 = ['x', 'y', 'z']
> 
> i want:
> 
> li3 = ['ax', 'by', 'cz']
> 
> how do i do it?
> 
> thanks
> 
> -gnana

Try this:

li1 = ['a','b','c']
li2 = ['x','y','z']
li3 = []

for i in range( len( li1 ) ):
    li3.append( li1[i] + li2[i] )


From stephen_purcell at yahoo.com  Wed Feb 21 18:41:27 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 21 Feb 2001 18:41:27 +0100
Subject: newbie - concatanating 2 lists
References: <970sgj+rrf7@eGroups.com>
Message-ID: <20010221184127.A25981@freedom.puma-ag.com>

Gnanasekaran Thoppae wrote:
> i am just beginning to use python.

Good choice!

> li1 = ['a', 'b', 'c']
> li2 = ['x', 'y', 'z']
> 
> i want:
> 
> li3 = ['ax', 'by', 'cz']

There may be more readable ways, but the shortest is probably this:

>>> li1 = ['a', 'b', 'c']
>>> li2 = ['x', 'y', 'z']
>>> li3 = map(lambda a,b: a+b, ['a','b','c'], ['x','y','z'])
>>> li3
['ax', 'by', 'cz']


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From gnana at mips.biochem.mpg.de  Wed Feb 21 18:48:43 2001
From: gnana at mips.biochem.mpg.de (Gnanasekaran Thoppae)
Date: Wed, 21 Feb 2001 17:48:43 -0000
Subject: newbie - concatanating 2 lists
In-Reply-To: <20010221184127.A25981@freedom.puma-ag.com>
Message-ID: <970v1r+70d1@eGroups.com>

Hi,

Thanks Steve and Perry.

I just realised that sometimes simple things can make one think a lot!
Python rocks!

-gnana

--- In python-list at y..., Steve Purcell <stephen_purcell at y...> wrote:
> Gnanasekaran Thoppae wrote:
> > i am just beginning to use python.
> 
> Good choice!
> 
> > li1 = ['a', 'b', 'c']
> > li2 = ['x', 'y', 'z']
> > 
> > i want:
> > 
> > li3 = ['ax', 'by', 'cz']
> 
> There may be more readable ways, but the shortest is probably this:
> 
> >>> li1 = ['a', 'b', 'c']
> >>> li2 = ['x', 'y', 'z']
> >>> li3 = map(lambda a,b: a+b, ['a','b','c'], ['x','y','z'])
> >>> li3
> ['ax', 'by', 'cz']
> 
> 
> -Steve
> 
> -- 
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Get servlets at http://pyserv.sourceforge.net/
> "Even snakes are afraid of snakes." -- Steven Wright
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From aleaxit at yahoo.com  Thu Feb 22 09:47:00 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 09:47:00 +0100
Subject: newbie - concatanating 2 lists
References: <970sgj+rrf7@eGroups.com> <mailman.982777311.18345.python-list@python.org>
Message-ID: <972jm10i4p@news2.newsguy.com>

"Steve Purcell" <stephen_purcell at yahoo.com> wrote in message
news:mailman.982777311.18345.python-list at python.org...
> Gnanasekaran Thoppae wrote:
> > i am just beginning to use python.
>
> Good choice!
>
> > li1 = ['a', 'b', 'c']
> > li2 = ['x', 'y', 'z']
> >
> > i want:
> >
> > li3 = ['ax', 'by', 'cz']
>
> There may be more readable ways, but the shortest is probably this:
>
> >>> li1 = ['a', 'b', 'c']
> >>> li2 = ['x', 'y', 'z']
> >>> li3 = map(lambda a,b: a+b, ['a','b','c'], ['x','y','z'])
> >>> li3
> ['ax', 'by', 'cz']

It's not the shortest way in Python2 -- rather, I'd code it:

      li3 = [ a+b for a,b in zip(li1, li2) ]


Alex





From stephen_purcell at yahoo.com  Thu Feb 22 11:22:01 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Thu, 22 Feb 2001 11:22:01 +0100
Subject: newbie - concatanating 2 lists
In-Reply-To: <972jm10i4p@news2.newsguy.com>; from aleaxit@yahoo.com on Thu, Feb 22, 2001 at 09:47:00AM +0100
References: <970sgj+rrf7@eGroups.com> <mailman.982777311.18345.python-list@python.org> <972jm10i4p@news2.newsguy.com>
Message-ID: <20010222112201.A353@freedom.puma-ag.com>

Alex Martelli wrote:
> Steve Purclel wrote:
> > >>> li3 = map(lambda a,b: a+b, ['a','b','c'], ['x','y','z'])
> 
> It's not the shortest way in Python2 -- rather, I'd code it:
> 
>       li3 = [ a+b for a,b in zip(li1, li2) ]

Hang on, it *is* still the shortest way in Python 2:

        li3 = [a+b for a,b in zip(li1,li2)]
        li3 = map(lambda a,b:a+b,li1,li2)

Any excuse to put off using list comprehensions for a while... :-)

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From scarblac at pino.selwerd.nl  Thu Feb 22 21:32:04 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 22 Feb 2001 20:32:04 GMT
Subject: newbie - concatanating 2 lists
References: <970sgj+rrf7@eGroups.com> <mailman.982777311.18345.python-list@python.org> <972jm10i4p@news2.newsguy.com> <mailman.982837323.11982.python-list@python.org>
Message-ID: <slrn99atmt.2rf.scarblac@pino.selwerd.nl>

Steve Purcell <stephen_purcell at yahoo.com> wrote in comp.lang.python:
> Hang on, it *is* still the shortest way in Python 2:
> 
>         li3 = [a+b for a,b in zip(li1,li2)]
>         li3 = map(lambda a,b:a+b,li1,li2)

          li3 = map(''.join,zip(li1,li2))
(as mentioned before)

Is two strokes smaller.

That's just my luck. First time I ever win at Perl Golf, it's in Python...

-- 
Remco Gerlich


From scarblac at pino.selwerd.nl  Thu Feb 22 11:01:26 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 22 Feb 2001 10:01:26 GMT
Subject: newbie - concatanating 2 lists
References: <mailman.982776626.16974.python-list@python.org>
Message-ID: <slrn999ood.219.scarblac@pino.selwerd.nl>

Sean 'Shaleh' Perry <shaleh at valinux.com> wrote in comp.lang.python:
> 
> On 21-Feb-2001 Gnanasekaran Thoppae wrote:
> > i am just beginning to use python.
> > 
> > i have:
> > 
> > li1 = ['a', 'b', 'c']
> > li2 = ['x', 'y', 'z']
> > 
> > i want:
> > 
> > li3 = ['ax', 'by', 'cz']
> > 
> > how do i do it? 
> 
> python 2 has a lovely function called zip(), maybe it could help.  Otherwise, a
> quick map would probably do this.
 
For Python 2, with zip and list comprehensions, it would look like

li3 = [a+b for a,b in zip(li1,li2)]

Hmm, or even
li3 = map(''.join, zip(l1,l2))

I like
li3 = map(operator.add, li1, li2)
best (why do people write their own lambda for operator.add? :)).

But all of these are trying to get it to fit in the least number of
characters (and they're pretty fast, although I have doubts about the zip
things).

A newbie should also understand the basic way:

A = ['a','b','c']
B = ['d','e','f']
C = []
for i in range(len(A)):
   C.append(A[i]+B[i])
   
This will give an exception if l1 is longer than l2, but all the above
methods have problems with different length lists as well.


zip(), list comprehensions, string methods etc are cool, but we end up 
with a lot of ways to do it...
-- 
Remco Gerlich


From adam at deprince.net  Wed Feb 21 21:56:57 2001
From: adam at deprince.net (Adam DePrince)
Date: Wed, 21 Feb 2001 15:56:57 -0500
Subject: newbie - concatanating 2 lists
References: <mailman.982776626.16974.python-list@python.org> <slrn999ood.219.scarblac@pino.selwerd.nl>
Message-ID: <3A942B99.490C286C@deprince.net>

Remco Gerlich wrote:
> Sean 'Shaleh' Perry <shaleh at valinux.com> wrote in comp.lang.python:
> > On 21-Feb-2001 Gnanasekaran Thoppae wrote:
> > > i am just beginning to use python.
> > > i have:
> > > li1 = ['a', 'b', 'c']
> > > li2 = ['x', 'y', 'z']
> > > i want:
> > > li3 = ['ax', 'by', 'cz']
> > > how do i do it?
> > python 2 has a lovely function called zip(), maybe it could help.  Otherwise, a
> > quick map would probably do this.
> For Python 2, with zip and list comprehensions, it would look like
> li3 = [a+b for a,b in zip(li1,li2)]
> Hmm, or even
> li3 = map(''.join, zip(l1,l2))
> I like
> li3 = map(operator.add, li1, li2)
> best (why do people write their own lambda for operator.add? :)).
> But all of these are trying to get it to fit in the least number of
> characters (and they're pretty fast, although I have doubts about the zip
> things).
> A newbie should also understand the basic way:
> 
> A = ['a','b','c']
> B = ['d','e','f']
> C = []
> for i in range(len(A)):

And if you want to handle mismatched lists gracefully ... 

  for i in range(min((len(A),len(B)))):


>    C.append(A[i]+B[i])
> 
> This will give an exception if l1 is longer than l2, but all the above
> methods have problems with different length lists as well.
> 
> zip(), list comprehensions, string methods etc are cool, but we end up
> with a lot of ways to do it...
> --
> Remco Gerlich


From fimafeng at yahoo.com  Wed Feb 21 18:43:25 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Wed, 21 Feb 2001 17:43:25 GMT
Subject: What's the best way???--installing 2.0 and IDLE on RH 7
Message-ID: <3A93FE0E.9070606@yahoo.com>

Well, my first attempt has landed me one buggered up system...

I need to rebuild my linux box anyways so this serves as good reason.

What I want to do is have BeOpen python 2.0 work with IDLE..

My last attempt always had IDLE running with 1.5.2

Before I go installing I need to know...

Should I install the 8.3 versions of tcl and tk that come with RedHat
as well as the 1.5.2 version of TKinter? (The one listed with the Python 
2.0 is also newer)

when I did this last time and then installed BeOpen 2.0 the RPM 
complained about needing the libraries from 8.0 tcl and tk.....

What is the best way to proceed? RPM's makefiles please help!!!

Many thanks,
Chris



From Max.Haas at unibas.ch  Wed Feb 21 19:25:30 2001
From: Max.Haas at unibas.ch (Max Haas)
Date: Wed, 21 Feb 2001 19:25:30 +0100
Subject: FW: newbie - concatanating 2 lists
In-Reply-To: <B6B9C63A.4114%Max.Haas@unibas.ch>
Message-ID: <B6B9C6AA.412B%Max.Haas@unibas.ch>

----------
Von: Max Haas <Max.Haas at unibas.ch>
Datum: Wed, 21 Feb 2001 19:23:40 +0100
An: Gnanasekaran Thoppae <gnana at mips.biochem.mpg.de>
Betreff: Re: newbie - concatanating 2 lists

Hi gnana,

perhaps the first step could be:

li1 = ['a', 'b', 'c']
li2 = ['x', 'y', 'z']
p = 0
m = []

for item in li1:
    m.append(item + li2[p])
    p = p + 1
print m

Max

am 21.2.2001 18:05 Uhr schrieb Gnanasekaran Thoppae unter
gnana at mips.biochem.mpg.de:

> hi,
> 
> i am just beginning to use python.
> 
> i have:
> 
> li1 = ['a', 'b', 'c']
> li2 = ['x', 'y', 'z']
> 
> i want:
> 
> li3 = ['ax', 'by', 'cz']
> 
> how do i do it?
> 
> thanks
> 
> -gnana
> 




From zzizz_ at notmail.com  Sat Feb 24 08:28:01 2001
From: zzizz_ at notmail.com (zzzzz)
Date: Sat, 24 Feb 2001 18:28:01 +1100
Subject: FW: newbie - concatanating 2 lists
References: <B6B9C63A.4114%Max.Haas@unibas.ch> <mailman.982780104.23132.python-list@python.org>
Message-ID: <v6oe9tc97d425qrbrvamnvri0u6suf1b2f@4ax.com>

On Wed, 21 Feb 2001 19:25:30 +0100, Max Haas <Max.Haas at unibas.ch>
wrote:

>> i have:
>> 
>> li1 = ['a', 'b', 'c']
>> li2 = ['x', 'y', 'z']
>> 
>> i want:
>> 
>> li3 = ['ax', 'by', 'cz']
try:
map(lambda a,b: a+b,li1,li2)

>
>Hi gnana,
>
>perhaps the first step could be:
>
>li1 = ['a', 'b', 'c']
>li2 = ['x', 'y', 'z']
>p = 0
>m = []
>
>for item in li1:
>    m.append(item + li2[p])
>    p = p + 1
>print m
>
>Max

rgds z.



From jeremy at alum.mit.edu  Wed Feb 21 20:09:43 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Wed, 21 Feb 2001 14:09:43 -0500 (EST)
Subject: Update to PEP 227 (static scoping)
Message-ID: <14996.4727.604581.858363@w221.z064000254.bwi-md.dsl.cnc.net>

There has been renewed discussion of backwards compatibility issues
introduced by nested scopes.  Following some discussion on python-dev,
I have updated the discussion of these issues in the PEP.  Of course,
more comments are welcome.  I am particularly interested in reports of
actual compatibility issues with existing code, as opposed to
hypotheticals. 

The particular concerns raised lately have to do with previously legal
code that will fail with a SyntaxError with nested scopes.  Early in
the design process, there was discussion of code that will behave
differently with nested scopes.  At the time, the subtle behavior
change was considered acceptable because it was believed to occur
rarely in practice and was probably hard to understand to begin with.

A related issue, already discussed on both lists, was the restrictions
added in Python 2.1a2 on the use of import * in functions and exec
with nested scope.  The former restriction was always documented in
the reference manual, but never enforced.  Subsequently, we decided to
allow import * and exec except in cases where the meaning was
ambiguous with respect to nested scopes.

This probably sounds a bit abstract; I hope the PEP (included below)
spells out the issues more clearly.  If you have code that currently
depends on any of the three following behaviors, I'd like to hear
about it:

    - A function is contained within another function.  The outer
      function contains a local name that shadows a global name.  The
      inner function uses the global.  The one case of this I have
      seen in the wild was caused by a local variable named str in the
      outer function and a use of builtin str in the inner function.

    - A function that contains a nested function with free variables
      and also uses exec that does not specify a namespace, e.g.
          def f():
              exec foo
              def g():
                  ...
      "exec foo in ns" should be legal, although the current CVS code
      base does not yet allow it.

    - A function like the one above, except that is uses import *
      instead of exec.

Jeremy

PEP: 227
Title: Statically Nested Scopes
Version: $Revision: 1.6 $
Author: jeremy at digicool.com (Jeremy Hylton)
Status: Draft
Type: Standards Track
Python-Version: 2.1
Created: 01-Nov-2000
Post-History: XXX what goes here?

Abstract

    This PEP proposes the addition of statically nested scoping
    (lexical scoping) for Python 2.1.  The current language definition
    defines exactly three namespaces that are used to resolve names --
    the local, global, and built-in namespaces.  The addition of
    nested scopes would allow resolution of unbound local names in
    enclosing functions' namespaces.

    One consequence of this change that will be most visible to Python
    programs is that lambda statements could reference variables in
    the namespaces where the lambda is defined.  Currently, a lambda
    statement uses default arguments to explicitly creating bindings
    in the lambda's namespace.

Introduction

    This proposal changes the rules for resolving free variables in
    Python functions.  The Python 2.0 definition specifies exactly
    three namespaces to check for each name -- the local namespace,
    the global namespace, and the builtin namespace.  According to
    this defintion, if a function A is defined within a function B,
    the names bound in B are not visible in A.  The proposal changes
    the rules so that names bound in B are visible in A (unless A
    contains a name binding that hides the binding in B).

    The specification introduces rules for lexical scoping that are
    common in Algol-like languages.  The combination of lexical
    scoping and existing support for first-class functions is
    reminiscent of Scheme.

    The changed scoping rules address two problems -- the limited
    utility of lambda statements and the frequent confusion of new
    users familiar with other languages that support lexical scoping,
    e.g. the inability to define recursive functions except at the
    module level.

    The lambda statement introduces an unnamed function that contains
    a single statement.  It is often used for callback functions.  In
    the example below (written using the Python 2.0 rules), any name
    used in the body of the lambda must be explicitly passed as a
    default argument to the lambda.

    from Tkinter import *
    root = Tk()
    Button(root, text="Click here",
           command=lambda root=root: root.test.configure(text="..."))

    This approach is cumbersome, particularly when there are several
    names used in the body of the lambda.  The long list of default
    arguments obscure the purpose of the code.  The proposed solution,
    in crude terms, implements the default argument approach
    automatically.  The "root=root" argument can be omitted.

Specification

    Python is a statically scoped language with block structure, in
    the traditional of Algol.  A code block or region, such as a
    module, class defintion, or function body, is the basic unit of a
    program.

    Names refer to objects.  Names are introduced by name binding
    operations.  Each occurrence of a name in the program text refers
    to the binding of that name established in the innermost function
    block containing the use.

    The name binding operations are assignment, class and function
    definition, and import statements.  Each assignment or import
    statement occurs within a block defined by a class or function
    definition or at the module level (the top-level code block).

    If a name binding operation occurs anywhere within a code block,
    all uses of the name within the block are treated as references to
    the current block.  (Note: This can lead to errors when a name is
    used within a block before it is bound.)

    If the global statement occurs within a block, all uses of the
    name specified in the statement refer to the binding of that name
    in the top-level namespace.  Names are resolved in the top-level
    namespace by searching the global namespace, the namespace of the
    module containing the code block, and the builtin namespace, the
    namespace of the module __builtin__.  The global namespace is
    searched first.  If the name is not found there, the builtin
    namespace is searched.

    If a name is used within a code block, but it is not bound there
    and is not declared global, the use is treated as a reference to
    the nearest enclosing function region.  (Note: If a region is
    contained within a class definition, the name bindings that occur
    in the class block are not visible to enclosed functions.)

    A class definition is an executable statement that may uses and
    definitions of names.  These references follow the normal rules
    for name resolution.  The namespace of the class definition
    becomes the attribute dictionary of the class.

    The following operations are name binding operations.  If they
    occur within a block, they introduce new local names in the
    current block unless there is also a global declaration.

    Function defintion: def name ...
    Class definition: class name ...
    Assignment statement: name = ...    
    Import statement: import name, import module as name,
        from module import name
    Implicit assignment: names are bound by for statements and except
        clauses

    The arguments of a function are also local.

    There are several cases where Python statements are illegal when
    used in conjunction with nested scopes that contain free
    variables.

    If a variable is referenced in an enclosing scope, it is an error
    to delete the name.  The compiler will raise a SyntaxError for
    'del name'.

    If the wildcard form of import (import *) is used in a function
    and the function contains a nested block with free variables, the
    compiler will raise a SyntaxError.

    If exec is used in a function and the function contains a nested
    block with free variables, the compiler will raise a SyntaxError
    unless the exec explicit specifies the local namespace for the
    exec.  (In other words, "exec obj" would be illegal, but 
    "exec obj in ns" would be legal.)

Discussion

    The specified rules allow names defined in a function to be
    referenced in any nested function defined with that function.  The
    name resolution rules are typical for statically scoped languages,
    with three primary exceptions:

        - Names in class scope are not accessible.
        - The global statement short-circuits the normal rules.
        - Variables are not declared.

    Names in class scope are not accessible.  Names are resolved in
    the innermost enclosing function scope.  If a class defintion
    occurs in a chain of nested scopes, the resolution process skips
    class definitions.  This rule prevents odd interactions between
    class attributes and local variable access.  If a name binding
    operation occurs in a class defintion, it creates an attribute on
    the resulting class object.  To access this variable in a method,
    or in a function nested within a method, an attribute reference
    must be used, either via self or via the class name.

    An alternative would have been to allow name binding in class
    scope to behave exactly like name binding in function scope.  This
    rule would allow class attributes to be referenced either via
    attribute reference or simple name.  This option was ruled out
    because it would have been inconsistent with all other forms of
    class and instance attribute access, which always use attribute
    references.  Code that used simple names would have been obscure.

    The global statement short-circuits the normal rules.  Under the
    proposal, the global statement has exactly the same effect that it
    does for Python 2.0.  It's behavior is preserved for backwards
    compatibility.  It is also noteworthy because it allows name
    binding operations performed in one block to change bindings in
    another block (the module).

    Variables are not declared.  If a name binding operation occurs
    anywhere in a function, then that name is treated as local to the
    function and all references refer to the local binding.  If a
    reference occurs before the name is bound, a NameError is raised.
    The only kind of declaration is the global statement, which allows
    programs to be written using mutable global variables.  As a
    consequence, it is not possible to rebind a name defined in an
    enclosing scope.  An assignment operation can only bind a name in
    the current scope or in the global scope.  The lack of
    declarations and the inability to rebind names in enclosing scopes
    are unusual for lexically scoped languages; there is typically a
    mechanism to create name bindings (e.g. lambda and let in Scheme)
    and a mechanism to change the bindings (set! in Scheme).

    XXX Alex Martelli suggests comparison with Java, which does not
    allow name bindings to hide earlier bindings.  

Examples

    A few examples are included to illustrate the way the rules work.

    XXX Explain the examples

    >>> def make_adder(base):
    ...     def adder(x):
    ...         return base + x
    ...     return adder
    >>> add5 = make_adder(5)
    >>> add5(6)
    11

    >>> def make_fact():
    ...     def fact(n):
    ...         if n == 1:
    ...             return 1L
    ...         else:
    ...             return n * fact(n - 1)
    ...     return fact
    >>> fact = make_fact()
    >>> fact(7)    
    5040L

    >>> def make_wrapper(obj):
    ...     class Wrapper:
    ...         def __getattr__(self, attr):
    ...             if attr[0] != '_':
    ...                 return getattr(obj, attr)
    ...             else:
    ...                 raise AttributeError, attr
    ...     return Wrapper()
    >>> class Test:
    ...     public = 2
    ...     _private = 3
    >>> w = make_wrapper(Test())
    >>> w.public
    2
    >>> w._private
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    AttributeError: _private

    An example from Tim Peters of the potential pitfalls of nested scopes
    in the absence of declarations:

    i = 6
    def f(x):
        def g():
            print i
        # ...
        # skip to the next page
        # ...
        for i in x:  # ah, i *is* local to f, so this is what g sees
            pass
        g()

    The call to g() will refer to the variable i bound in f() by the for
    loop.  If g() is called before the loop is executed, a NameError will
    be raised.

    XXX need some counterexamples

Backwards compatibility

    There are two kinds of compatibility problems caused by nested
    scopes.  In one case, code that behaved one way in earlier
    versions, behaves differently because of nested scopes.  In the
    other cases, certain constructs interact badly with nested scopes
    and will trigger SyntaxErrors at compile time.

    The following example from Skip Montanaro illustrates the first
    kind of problem:

    x = 1
    def f1():
        x = 2
        def inner():
            print x
        inner()

    Under the Python 2.0 rules, the print statement inside inner()
    refers to the global variable x and will print 1 if f1() is
    called.  Under the new rules, it refers to the f1()'s namespace,
    the nearest enclosing scope with a binding.

    The problem occurs only when a global variable and a local
    variable share the same name and a nested function uses that name
    to refer to the global variable.  This is poor programming
    practice, because readers will easily confuse the two different
    variables.  One example of this problem was found in the Python
    standard library during the implementation of nested scopes.

    To address this problem, which is unlikely to occur often, a
    static analysis tool that detects affected code will be written.
    The detection problem is straightfoward.

    The other compatibility problem is casued by the use of 'import *'
    and 'exec' in a function body, when that function contains a
    nested scope and the contained scope has free variables.  For
    example:

    y = 1
    def f():
        exec "y = 'gotcha'" # or from module import *
        def g():
            return y
        ...

    At compile-time, the compiler cannot tell whether an exec that
    operators on the local namespace or an import * will introduce
    name bindings that shadow the global y.  Thus, it is not possible
    to tell whether the reference to y in g() should refer to the
    global or to a local name in f().

    In discussion of the python-list, people argued for both possible
    interpretations.  On the one hand, some thought that the reference
    in g() should be bound to a local y if one exists.  One problem
    with this interpretation is that it is impossible for a human
    reader of the code to determine the binding of y by local
    inspection.  It seems likely to introduce subtle bugs.  The other
    interpretation is to treat exec and import * as dynamic features
    that do not effect static scoping.  Under this interpretation, the
    exec and import * would introduce local names, but those names
    would never be visible to nested scopes.  In the specific example
    above, the code would behave exactly as it did in earlier versions
    of Python.

    Since each interpretation is problemtatic and the exact meaning
    ambiguous, the compiler raises an exception.

    A brief review of three Python projects (the standard library,
    Zope, and a beta version of PyXPCOM) found four backwards
    compatibility issues in approximately 200,000 lines of code.
    There was one example of case #1 (subtle behavior change) and two
    examples of import * problems in the standard library.

    (The interpretation of the import * and exec restriction that was
    implemented in Python 2.1a2 was much more restrictive, based on
    language that in the reference manual that had never been
    enforced.  These restrictions were relaxed following the release.)

locals() / vars()

    These functions return a dictionary containing the current scope's
    local variables.  Modifications to the dictionary do not affect
    the values of variables.  Under the current rules, the use of
    locals() and globals() allows the program to gain access to all
    the namespaces in which names are resolved.

    An analogous function will not be provided for nested scopes.
    Under this proposal, it will not be possible to gain
    dictionary-style access to all visible scopes.

Rebinding names in enclosing scopes

    There are technical issues that make it difficult to support
    rebinding of names in enclosing scopes, but the primary reason
    that it is not allowed in the current proposal is that Guido is
    opposed to it.  It is difficult to support, because it would
    require a new mechanism that would allow the programmer to specify
    that an assignment in a block is supposed to rebind the name in an
    enclosing block; presumably a keyword or special syntax (x := 3)
    would make this possible.

    The proposed rules allow programmers to achieve the effect of
    rebinding, albeit awkwardly.  The name that will be effectively
    rebound by enclosed functions is bound to a container object.  In
    place of assignment, the program uses modification of the
    container to achieve the desired effect:

    def bank_account(initial_balance):
        balance = [initial_balance]
        def deposit(amount):
            balance[0] = balance[0] + amount
            return balance
        def withdraw(amount):
            balance[0] = balance[0] - amount
            return balance
        return deposit, withdraw

    Support for rebinding in nested scopes would make this code
    clearer.  A class that defines deposit() and withdraw() methods
    and the balance as an instance variable would be clearer still.
    Since classes seem to achieve the same effect in a more
    straightforward manner, they are preferred.

Implementation

    The implementation for C Python uses flat closures [1].  Each def
    or lambda statement that is executed will create a closure if the
    body of the function or any contained function has free
    variables.  Using flat closures, the creation of closures is
    somewhat expensive but lookup is cheap.

    The implementation adds several new opcodes and two new kinds of
    names in code objects.  A variable can be either a cell variable
    or a free variable for a particular code object.  A cell variable
    is referenced by containing scopes; as a result, the function
    where it is defined must allocate separate storage for it on each
    invocation.  A free variable is reference via a function's closure.

    XXX Much more to say here

References

    [1] Luca Cardelli.  Compiling a functional language.  In Proc. of
    the 1984 ACM Conference on Lisp and Functional Programming,
    pp. 208-217, Aug. 1984
        http://citeseer.nj.nec.com/cardelli84compiling.html




From aahz at panix.com  Wed Feb 21 21:24:09 2001
From: aahz at panix.com (Aahz Maruch)
Date: 21 Feb 2001 12:24:09 -0800
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org>
Message-ID: <971859$dq$1@panix6.panix.com>

[p&e]

[I don't really have any comments on this proposal, but to prove that
I've read it ;-), I'm including a list proofreading fixes.]

In article <mailman.982783412.29104.python-list at python.org>,
Jeremy Hylton  <jeremy at alum.mit.edu> wrote:
>
>Introduction
>
>    This proposal changes the rules for resolving free variables in
>    Python functions.  The Python 2.0 definition specifies exactly
>    three namespaces to check for each name -- the local namespace,
>    the global namespace, and the builtin namespace.  According to
>    this defintion, if a function A is defined within a function B,
>    the names bound in B are not visible in A.  The proposal changes
>    the rules so that names bound in B are visible in A (unless A
>    contains a name binding that hides the binding in B).

s/defintion/definition/

>Specification
>
>    Python is a statically scoped language with block structure, in
>    the traditional of Algol.  A code block or region, such as a
>    module, class defintion, or function body, is the basic unit of a
>    program.

s/defintion/definition/

>    If the global statement occurs within a block, all uses of the
>    name specified in the statement refer to the binding of that name
>    in the top-level namespace.  Names are resolved in the top-level
>    namespace by searching the global namespace, the namespace of the
>    module containing the code block, and the builtin namespace, the
>    namespace of the module __builtin__.  The global namespace is
>    searched first.  If the name is not found there, the builtin
>    namespace is searched.

Clarification: if the global statement is not the first executable
statement in a block, the results are undefined.

>    If exec is used in a function and the function contains a nested
>    block with free variables, the compiler will raise a SyntaxError
>    unless the exec explicit specifies the local namespace for the
>    exec.  (In other words, "exec obj" would be illegal, but 
>    "exec obj in ns" would be legal.)

s/explicit/explicitly/

>    Names in class scope are not accessible.  Names are resolved in
>    the innermost enclosing function scope.  If a class defintion
>    occurs in a chain of nested scopes, the resolution process skips
>    class definitions.  This rule prevents odd interactions between
>    class attributes and local variable access.  If a name binding
>    operation occurs in a class defintion, it creates an attribute on
>    the resulting class object.  To access this variable in a method,
>    or in a function nested within a method, an attribute reference
>    must be used, either via self or via the class name.

s/defintion/definition/

>    The global statement short-circuits the normal rules.  Under the
>    proposal, the global statement has exactly the same effect that it
>    does for Python 2.0.  It's behavior is preserved for backwards
>    compatibility.  It is also noteworthy because it allows name
>    binding operations performed in one block to change bindings in
>    another block (the module).

s/It's/Its/

>    An example from Tim Peters of the potential pitfalls of nested scopes
>    in the absence of declarations:

Clarification: s/Peters of/Peters demonstrates/

>Backwards compatibility
>
>    There are two kinds of compatibility problems caused by nested
>    scopes.  In one case, code that behaved one way in earlier
>    versions, behaves differently because of nested scopes.  In the
>    other cases, certain constructs interact badly with nested scopes
>    and will trigger SyntaxErrors at compile time.

s/versions,/versions/

>    At compile-time, the compiler cannot tell whether an exec that
>    operators on the local namespace or an import * will introduce
>    name bindings that shadow the global y.  Thus, it is not possible
>    to tell whether the reference to y in g() should refer to the
>    global or to a local name in f().

s/operators/operates/

>    The implementation adds several new opcodes and two new kinds of
>    names in code objects.  A variable can be either a cell variable
>    or a free variable for a particular code object.  A cell variable
>    is referenced by containing scopes; as a result, the function
>    where it is defined must allocate separate storage for it on each
>    invocation.  A free variable is reference via a function's closure.

s/reference/referenced/
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From fredrik at pythonware.com  Wed Feb 21 21:45:00 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 20:45:00 GMT
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org>
Message-ID: <gNVk6.14871$Qb7.2374663@newsb.telia.net>

Jeremy wrote:

>     - A function that contains a nested function with free variables
>       and also uses exec that does not specify a namespace, e.g.
>           def f():
>               exec foo
>               def g():
>                   ...
>       "exec foo in ns" should be legal, although the current CVS code
>       base does not yet allow it.
>
>     - A function like the one above, except that is uses import *
>       instead of exec.

Note that in real life, "free" means "global or builtin".  For example,
this common Tkinter pattern will no longer compile:

import sys
def makeui():
    from Tkinter import *
    def callback():
        print "bye"
        sys.exit(1)
    Button(command=callback).pack()

# gives an error on line two (the first def statement)

If you have 2.1a2 installed, the following script will check for exec
and import problems.  Run it from the top of your development tree,
and let us know what you find...  (running it on the 2.0 source tree
is also somewhat entertaining ;-)

#
# look for exec/import scope problems

import os
import fnmatch
import string, traceback

class GlobDirectoryWalker:
    # a forward iterator that traverses a directory tree

    def __init__(self, directory, pattern="*"):
        self.stack = [directory]
        self.pattern = pattern
        self.files = []
        self.index = 0

    def __getitem__(self, index):
        while 1:
            try:
                file = self.files[self.index]
                self.index = self.index + 1
            except IndexError:
                # pop next directory from stack
                self.directory = self.stack.pop()
                self.files = os.listdir(self.directory)
                self.index = 0
            else:
                # got a filename
                fullname = os.path.join(self.directory, file)
                if os.path.isdir(fullname) and not os.path.islink(fullname):
                    self.stack.append(fullname)
                if fnmatch.fnmatch(file, self.pattern):
                    return fullname

count = 0
for file in GlobDirectoryWalker(".", "*.py"):
    try:
        # print file, "..."
        count = count + 1
        print count, "\r",
        compile(open(file).read() + "\n", file, "exec")
    except SyntaxError, v:
        v = str(v)
        if string.find(v, "nested scope") > 0:
            print "-" * 68
            print file
            traceback.print_exc()
print count, "files checked"

# Cheers /F




From jeremy at alum.mit.edu  Wed Feb 21 22:10:46 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Wed, 21 Feb 2001 16:10:46 -0500 (EST)
Subject: Update to PEP 227 (static scoping)
In-Reply-To: <gNVk6.14871$Qb7.2374663@newsb.telia.net>
References: <mailman.982783412.29104.python-list@python.org>
	<gNVk6.14871$Qb7.2374663@newsb.telia.net>
Message-ID: <14996.11990.285306.871815@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "FL" == Fredrik Lundh <fredrik at pythonware.com> writes:

  FL> If you have 2.1a2 installed, the following script will check for
  FL> exec and import problems.  Run it from the top of your
  FL> development tree, and let us know what you find...  (running it
  FL> on the 2.0 source tree is also somewhat entertaining ;-)

Unless your idea of entertainment is vastly different than mine, I
don't understand what you mean.  The output I see from your handy
script is:

1025 files checked

If you see problems on your machine, it would be helpful if you could
submit a bug report.

Jeremy



From fredrik at pythonware.com  Thu Feb 22 08:51:18 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 07:51:18 GMT
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org> <gNVk6.14871$Qb7.2374663@newsb.telia.net>
Message-ID: <Wx3l6.14915$Qb7.2385994@newsb.telia.net>

I wrote:
> Note that in real life, "free" means "global or builtin".  For example,
> this common Tkinter pattern will no longer compile:
>
> import sys
> def makeui():
>     from Tkinter import *
>     def callback():
>         print "bye"
>         sys.exit(1)
>     Button(command=callback).pack()
>
> # gives an error on line two (the first def statement)

also note that one easy way to fix this is to use the
default argument hack (!):

    import sys
    def makeui():
        from Tkinter import *
        def callback(sys=sys):
            print "bye"
            sys.exit(1) # this works!?
        Button(command=callback).pack()

I'm glad I didn't sign up to do that Python book ;-)

Cheers /F




From Denys.Duchier at ps.uni-sb.de  Wed Feb 21 22:40:00 2001
From: Denys.Duchier at ps.uni-sb.de (Denys Duchier)
Date: 21 Feb 2001 22:40:00 +0100
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org>
Message-ID: <is8zmzk95b.fsf@elk.ps.uni-sb.de>

Jeremy Hylton <jeremy at alum.mit.edu> writes:

> Rebinding names in enclosing scopes
> 
>     There are technical issues that make it difficult to support
>     rebinding of names in enclosing scopes, but the primary reason
>     that it is not allowed in the current proposal is that Guido is
>     opposed to it.  It is difficult to support, because it would
>     require a new mechanism that would allow the programmer to specify
>     that an assignment in a block is supposed to rebind the name in an
>     enclosing block; presumably a keyword or special syntax (x := 3)
>     would make this possible.

if "global x" was instead allowed to mean "the x in the innermost
enclosing scope that defines one (or the global scope otherwise)",
wouldn't that solve the problem.  It wouldn't be semantically backward
compatible, but, in practice, would that break a lot of existing
programs and require too high a cost for fixing them?  Just wondering.

Cheers,

-- 
Dr. Denys Duchier			Denys.Duchier at ps.uni-sb.de
Forschungsbereich Programmiersysteme	(Programming Systems Lab)
Universitaet des Saarlandes, Geb. 45	http://www.ps.uni-sb.de/~duchier
Postfach 15 11 50			Phone: +49 681 302 5618
66041 Saarbruecken, Germany		Fax:   +49 681 302 5615


From jeremy at alum.mit.edu  Wed Feb 21 23:43:32 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Wed, 21 Feb 2001 17:43:32 -0500 (EST)
Subject: Update to PEP 227 (static scoping)
In-Reply-To: <is8zmzk95b.fsf@elk.ps.uni-sb.de>
References: <mailman.982783412.29104.python-list@python.org>
	<is8zmzk95b.fsf@elk.ps.uni-sb.de>
Message-ID: <14996.17556.654376.765535@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "DD" == Denys Duchier <Denys.Duchier at ps.uni-sb.de> writes:

  DD> Jeremy Hylton <jeremy at alum.mit.edu> writes:
  >> Rebinding names in enclosing scopes
  [...]

  DD> if "global x" was instead allowed to mean "the x in the
  DD> innermost enclosing scope that defines one (or the global scope
  DD> otherwise)", wouldn't that solve the problem.  It wouldn't be
  DD> semantically backward compatible, but, in practice, would that
  DD> break a lot of existing programs and require too high a cost for
  DD> fixing them?  Just wondering.

This was an option considered early on, but abandoned because
... Guido didn't like it?  I know he didn't, but I don't recall
exactly why.  A few issues come to mind:

 - It is odd to use the name "global" for something that does not mean
   global.  This alone is enough to reject it.  If we're going to add
   a special declaration statement, it ought to have a clear name.

 - It will break code that depends on global meaning "global" (or
   should I have the scare quotes the other way <wink>).  If we're
   going to break code, there's no point in breaking code that was
   being explicit about what it was doing.

As for using some other word, I'm neutral.  I'm not sure how much the
rebinding will be used, and I'm reluctant to get into a debate about
how to spell things and what to call them.  

the-dict-iteration-blues-are-getting-to-me-ly y'rs,
Jeremy
 



From tim.one at home.com  Thu Feb 22 02:32:48 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 21 Feb 2001 20:32:48 -0500
Subject: Update to PEP 227 (static scoping)
In-Reply-To: <is8zmzk95b.fsf@elk.ps.uni-sb.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEBOJBAA.tim.one@home.com>

[Denys Duchier]
> if "global x" was instead allowed to mean "the x in the innermost
> enclosing scope that defines one (or the global scope otherwise)",
> wouldn't that solve the problem.

You answered that next:

> It wouldn't be semantically backward compatible,

Bingo.

> but, in practice, would that break a lot of existing programs and
> require too high a cost for fixing them?  Just wondering.

It depends on who you ask, of course.

>From a cost/benefit view, the crushingly overwhelming source of scope
complaints comes from people using Python's little lambdas.  Indeed, there
were exceedingly few scope complaints before little lambdas were introduced,
and non-stop howling ever since.

I say "little" because they're restricted to a single expression, and
expressions in Python can never rebind a name (all binding constructs are
statements in Python, not expressions).  So the people who want lexical
scoping in Python the most (in practice, not the ones who just talk about it
or are idly pining for Scheme <wink>) have no possibility of rebinding a
name anyway in the context they most want it.

Doesn't mean Python will never allow intermediate-level bindings, but
there's very little call for it now, so it just raises more issues that
needn't be solved now.

the-ironic-thing-is-that-if-guido-were-ignoring-users-he-wouldn't-
    have-gotten-within-a-10-foot-snake-of-lexical-nesting-ly y'rs  - tim




From owen at astrono.junkwashington.emu  Thu Feb 22 20:30:44 2001
From: owen at astrono.junkwashington.emu (Russell E. Owen)
Date: Thu, 22 Feb 2001 11:30:44 -0800
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org>
Message-ID: <973pd4$jhq$1@nntp6.u.washington.edu>

In article <mailman.982783412.29104.python-list at python.org>, 
jeremy at alum.mit.edu wrote:

(much of generally excellent description of PEP 227 omitted)

>    A class definition is an executable statement that may uses and
>    definitions of names.

Might you rephrase this? I hope there is a typo here.

>Discussion
>...
>    Variables are not declared.  If a name binding operation occurs
>    anywhere in a function, then that name is treated as local to the
>    function and all references refer to the local binding.  If a
>    reference occurs before the name is bound, a NameError is raised.
>    The only kind of declaration is the global statement, which allows
>    programs to be written using mutable global variables.  As a
>    consequence, it is not possible to rebind a name defined in an
>    enclosing scope.  An assignment operation can only bind a name in
>    the current scope or in the global scope.  The lack of
>    declarations and the inability to rebind names in enclosing scopes
>    are unusual for lexically scoped languages; there is typically a
>    mechanism to create name bindings (e.g. lambda and let in Scheme)
>    and a mechanism to change the bindings (set! in Scheme).

I confess to being nervous about this proposed change. Static scoping 
does add some very nice functionality, but it seems to come at a 
significant cost in
- complexity (the new rules have some tricky exceptions)
- safety (Tim's example is a good one; also typos can do more damage)

I realize this has probably been discussed at length by people 
intimately connected with the language, so I'm on thin ice here, but 
I'll speak up anyway.

Firstly, I am wondering if we might not be able to obtain a most of the 
gains of this proposal with less of the negative aspects by scaling it 
back?

It seems that the most important driver for these new rules is lambda 
functions. It also helps in other areas, but somebody pointed out that 
folks where generally content with the old scoping rules until lambda 
functions were introduced. So...perhaps one could improve the syntax for 
lambda functions, instead of messing with scoping? Here is one 
possibility:

lambda list-of-args : list-of-non-local-variables : code

This basically does the same as the default argument trick. It would be 
nice if it was an error to try to refer to non-local variables as 
arguments when calling lambda, but this is by no means necessary.



If the more general solution of static scoping is desired (and it does 
have some advantages, of course), then here are some further thoughts:

How about requiring an explicit statement that "this variable is 
non-local"? This is reminiscent of the existing "global" keyword. This 
has several advantages, to my mind:
- It works equally well in functions and in class methods. Hence classes 
don't need to be special-cased in the new rules.
- The common and most-desirable cases (all variables are local or were 
in the argument list) is the default; if you want to access non-local 
variables (which is less common and potentially more dangerous) you have 
to be explicit about it.

I think a keyword is best, as it matches the existing "global" keyword, 
providing nice regularity.

The one problem I see with a keyword is lambda functions (which I don't 
think could use it). For that, perhaps again one could adopt special 
notation, such as listed above, for specifying the non-local variables 
it will use.



If it is not deemed desirable to require declaring non-local variables, 
then I do have a final plea: some *optional* means of explicitly 
declaring a variable as local.

Such notation was not needed with the old scoping rules, but I think 
it'll be very useful for writing safe code with the new rules. Good 
coding practice (especially in large projects) would suggest using it 
for all programs -- to avoid typos causing mysterious bugs.

Optional explicit declaration also enables the ability to warn of the 
use of non-declared variables. Something that can be very nice for large 
projects -- since it catches typos.


-- Russell


From qrczak at knm.org.pl  Fri Feb 23 00:34:32 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Feb 2001 23:34:32 GMT
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org> <973pd4$jhq$1@nntp6.u.washington.edu>
Message-ID: <slrn99b8g7.67d.qrczak@qrnik.zagroda>

Thu, 22 Feb 2001 11:30:44 -0800, Russell E. Owen <owen at astrono.junkwashington.emu> pisze:

> I confess to being nervous about this proposed change.

I am very happy about static scoping. It's very natural and found in
most languages.

> - complexity (the new rules have some tricky exceptions)

For me these are those exceptions which should be changed.
But I can live with them.

> - safety (Tim's example is a good one; also typos can do more damage)

It breaks only poorly written code, which relies on the fact that
a scope has holes for nested functions. Shadowing a global name and
then accessing it from inside of the function which shadowed it is
a questionable practice.

Typos don't do *more* damage. Only different typos do different damage.
With old rules wrapping a piece of code in a function silently changes
the meaning of its free variables which are not global.

> So...perhaps one could improve the syntax for lambda functions,
> instead of messing with scoping? Here is one possibility:
> 
> lambda list-of-args : list-of-non-local-variables : code

I would not call this an improvement. It's easy to forget a variable,
in which case I silently get a legal but incorrect code. Variables
are already present in the expression, no need to repeat them.

> It would be nice if it was an error to try to refer to non-local
> variables as arguments when calling lambda, but this is by no
> means necessary.

In this case a translation between the new scheme and your scheme is
fully automatic, and the new scheme has a simpler syntax.

> How about requiring an explicit statement that "this variable
> is non-local"?

The same arguments as for lambda above apply here.

It would make sense if it allowed rebinding that variable from inner
functions. It's unnecessary when we only refer to the variable,
because there is no other choice than being non-local if it's not
assigned to here.

> If it is not deemed desirable to require declaring non-local variables, 
> then I do have a final plea: some *optional* means of explicitly 
> declaring a variable as local.

It's already there:
    variable = None

> Optional explicit declaration also enables the ability to warn of
> the use of non-declared variables. Something that can be very nice
> for large projects -- since it catches typos.

Python already could warn about use of variables which are not defined,
except that it does not work with dynamic binding: from module import *,
exec, and modifying globals().

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From cfelling at iae.nl  Fri Feb 23 00:40:21 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 23 Feb 2001 00:40:21 +0100
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org> <973pd4$jhq$1@nntp6.u.washington.edu>
Message-ID: <974815$b1o$1@animus.fel.iae.nl>

Russell E. Owen <owen at astrono.junkwashington.emu> wrote:
...
> I confess to being nervous about this proposed change. Static scoping 
> does add some very nice functionality, but it seems to come at a 
> significant cost in
> - complexity (the new rules have some tricky exceptions)
> - safety (Tim's example is a good one; also typos can do more damage)

I'm worried too by the increasing complexity of this scoping thing!

...
> lambda list-of-args : list-of-non-local-variables : code

maybe this could be spelled (as is suggested before me thinks) like:

  lambda list-of-args ; list-of-non-local-variables : code

and then functions could use the same spelling for closures, like

  def f(list-of-args ;
        list-of-closures):
      pass

This would prevent the nasty Pascal habbit of deeply nested
references, or actual it would spell it out, as each and every
function needs to pass on all vars that are going to be referenced
somewhere inside that function being it in the function body itself or
in some locally declared one.  (Probably my upbringing in Eindhoven
(and subsequent admiration for Dijkstra) shines through here:)

Ofcourse nested functions do not recurse easily in such a scheme,
but neither do recursive methods in the PEP that worries us so much.
(I know self.method will do, but why do I have to spell recursion
differently for methods? yak)

Or maybe the semantics of that semi-colon inside the argument list
could be such that it allowed to pass the function name into the
visible scope of that function?

<rambling on>
We could use that semicolon again to create a place to define function
attributes like in

   def f(arg-list ;
         closures ) ;
         function-attributes :
       pass

Spelled like this, I would expect to be able to address function
attributes inside the function body.  I think The BDFL is against
this, but why?
<\rambling off>
-- 
groetjes, carel


From fredrik at pythonware.com  Wed Feb 21 22:36:14 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 21:36:14 GMT
Subject: Update to PEP 227 (static scoping)
References: <mailman.982783412.29104.python-list@python.org><gNVk6.14871$Qb7.2374663@newsb.telia.net> <mailman.982790554.12466.python-list@python.org>
Message-ID: <ixWk6.14889$Qb7.2376269@newsb.telia.net>

Jeremy Hylton wrote:
> Unless your idea of entertainment is vastly different than mine

dunno.  have you read "Popul?rmusik fr?n Vittula"?  excellent book ;-)

> I don't understand what you mean.  The output I see from your handy
> script is:
>
> 1025 files checked
>
> If you see problems on your machine, it would be helpful if you could
> submit a bug report.

the Python-2.0 source tree, untarred from BeOpen-Python-2.0.tar.gz:

--------------------------------------------------------------------
.\Tools\idle\IOBinding.py
Traceback (most recent call last):
  File "\pythonware\scan.py", line 38, in ?
    compile(open(file).read() + "\n", file, "exec")
SyntaxError: test: exec or 'import *' makes names ambiguous in nested
scope (IOBinding.py, line 202)
--------------------------------------------------------------------
.\Tools\idle\Percolator.py
Traceback (most recent call last):
  File "\pythonware\scan.py", line 38, in ?
    compile(open(file).read() + "\n", file, "exec")
SyntaxError: main: exec or 'import *' makes names ambiguous in nested
scope (Percolator.py, line 55)

both are fixed in the 2.1a2 source tree, of course.

Cheers /F




From tuttledon at hotmail.com  Thu Feb 22 19:24:24 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Thu, 22 Feb 2001 18:24:24 GMT
Subject: Update to PEP 227 (static scoping)
References: <mailman.982833927.2343.python-list@python.org>
Message-ID: <sPcl6.5905$h52.1437390@typhoon.southeast.rr.com>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.982833927.2343.python-list at python.org...
>
> From a cost/benefit view, the crushingly overwhelming source of scope
> complaints comes from people using Python's little lambdas.  Indeed, there
> were exceedingly few scope complaints before little lambdas were
introduced,
> and non-stop howling ever since.
>
> I say "little" because they're restricted to a single expression, and
> expressions in Python can never rebind a name (all binding constructs are
> statements in Python, not expressions).  So the people who want lexical
> scoping in Python the most (in practice, not the ones who just talk about
it
> or are idly pining for Scheme <wink>) have no possibility of rebinding a
> name anyway in the context they most want it.

Not knowing what I'm talking about, but...could lexical scoping be
implemented as a command line switch?  '-L'

Don




From jeremy at alum.mit.edu  Fri Feb 23 02:23:15 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Thu, 22 Feb 2001 20:23:15 -0500 (EST)
Subject: Update to PEP 227 (static scoping)
In-Reply-To: <sPcl6.5905$h52.1437390@typhoon.southeast.rr.com>
References: <mailman.982833927.2343.python-list@python.org>
	<sPcl6.5905$h52.1437390@typhoon.southeast.rr.com>
Message-ID: <14997.48003.217156.62550@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "DT" == Don Tuttle <tuttledon at hotmail.com> writes:

  DT> could lexical scoping be implemented as a command line switch?  '-L'

It would have to be -N, because Python is already lexically scoped.
The question of the moment is whether the scopes should nest.

But, no, it can't be done via the command-line, because code written
with nested scopes will not necessarily work with the old scoping
rules and vice versa.  If I hand you a module, it is possible it will
work with one and only one set of scoping rules.  If you run it with
the wrong command line argument, it will blow up.  Worse, that module
can't be combined with another module that depends on the opposite
rules.  

An alternative would be to add a notation to the module telling the
compiler which rules it expects.

Jeremy






From tuttledon at hotmail.com  Fri Feb 23 03:56:58 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Fri, 23 Feb 2001 02:56:58 GMT
Subject: Update to PEP 227 (static scoping)
References: <mailman.982833927.2343.python-list@python.org><sPcl6.5905$h52.1437390@typhoon.southeast.rr.com> <mailman.982893604.1444.python-list@python.org>
Message-ID: <_jkl6.6220$h52.2081013@typhoon.southeast.rr.com>

>   DT> could lexical scoping be implemented as a command line switch?  '-L'
>
"Jeremy Hylton"
> But, no... An alternative would be to add a notation to the module telling
the
> compiler which rules it expects.

Is this approch still an option for 2.1 or is always nested a done deal?

Don





From tuttledon at hotmail.com  Fri Feb 23 05:19:30 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Fri, 23 Feb 2001 04:19:30 GMT
Subject: Update to PEP 227 (static scoping)
References: <mailman.982833927.2343.python-list@python.org><sPcl6.5905$h52.1437390@typhoon.southeast.rr.com> <mailman.982893604.1444.python-list@python.org> <_jkl6.6220$h52.2081013@typhoon.southeast.rr.com>
Message-ID: <mxll6.6264$h52.2209038@typhoon.southeast.rr.com>

> >   DT> could lexical scoping be implemented as a command line switch?
'-L'
> >
> "Jeremy Hylton"
> > But, no... An alternative would be to add a notation to the module
telling
> the
> > compiler which rules it expects.
>
> Is this approch still an option for 2.1 or is always nested a done deal?
>
Never mind. ;-)




From hinsen at cnrs-orleans.fr  Fri Feb 23 11:05:07 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 23 Feb 2001 11:05:07 +0100
Subject: Update to PEP 227 (static scoping)
References: <mailman.982833927.2343.python-list@python.org> <sPcl6.5905$h52.1437390@typhoon.southeast.rr.com> <mailman.982893604.1444.python-list@python.org>
Message-ID: <m3r90pbtpo.fsf@chinon.cnrs-orleans.fr>

Jeremy Hylton <jeremy at alum.mit.edu> writes:

> It would have to be -N, because Python is already lexically scoped.
> The question of the moment is whether the scopes should nest.

There is one thing that I have missed in PEP 227: a motivation for the
proposed change. All I can find is the statement that other languages
have nested scopes. That sounds like a valid reason when designing a
new language, but is it sufficient to modify an existing language in a
way that risks breaking working code? I can't remember seeing
excessive complaints about Python's current scoping rules.

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From aahz at panix.com  Fri Feb 23 17:21:51 2001
From: aahz at panix.com (Aahz Maruch)
Date: 23 Feb 2001 08:21:51 -0800
Subject: Update to PEP 227 (static scoping)
References: <mailman.982833927.2343.python-list@python.org> <sPcl6.5905$h52.1437390@typhoon.southeast.rr.com> <mailman.982893604.1444.python-list@python.org> <m3r90pbtpo.fsf@chinon.cnrs-orleans.fr>
Message-ID: <9762mv$iu9$1@panix2.panix.com>

In article <m3r90pbtpo.fsf at chinon.cnrs-orleans.fr>,
Konrad Hinsen  <hinsen at cnrs-orleans.fr> wrote:
>Jeremy Hylton <jeremy at alum.mit.edu> writes:
>>
>> It would have to be -N, because Python is already lexically scoped.
>> The question of the moment is whether the scopes should nest.
>
>There is one thing that I have missed in PEP 227: a motivation for the
>proposed change. All I can find is the statement that other languages
>have nested scopes. That sounds like a valid reason when designing a
>new language, but is it sufficient to modify an existing language in a
>way that risks breaking working code? I can't remember seeing
>excessive complaints about Python's current scoping rules.

You've missed all the lambda-whiners.  PEP 227 doesn't quite come out
and say so explicitly, but you'll note that the abstract specifically
mentions lambda as an issue.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From jeremy at alum.mit.edu  Fri Feb 23 18:14:29 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Fri, 23 Feb 2001 12:14:29 -0500 (EST)
Subject: Update to PEP 227 (static scoping)
In-Reply-To: <m3r90pbtpo.fsf@chinon.cnrs-orleans.fr>
References: <mailman.982833927.2343.python-list@python.org>
	<sPcl6.5905$h52.1437390@typhoon.southeast.rr.com>
	<mailman.982893604.1444.python-list@python.org>
	<m3r90pbtpo.fsf@chinon.cnrs-orleans.fr>
Message-ID: <14998.39541.803383.217822@w221.z064000254.bwi-md.dsl.cnc.net>

You're right that the PEP doesn't spend much time explaining the
motivation.  Just this paragraph in the introduction:

    The changed scoping rules address two problems -- the limited
    utility of lambda statements and the frequent confusion of new
    users familiar with other languages that support lexical scoping,
    e.g. the inability to define recursive functions except at the
    module level.

I'll expand this in the next draft.

Jeremy



From neo at thezion.net  Wed Feb 21 20:10:04 2001
From: neo at thezion.net (Neo)
Date: Wed, 21 Feb 2001 20:10:04 +0100
Subject: Unicode error with ado in win 32
Message-ID: <9713oe$mse8b$1@ID-35431.news.dfncis.de>

I've created a simple script which read and print on the interactive window
the data from a database field (sql server 7).
It work properly but when into the rs.fields('azienda') theres the string
"abb?" the script returns the following message.

Traceback (most recent call last):
  File "g:\programmi\Python20\Pythonwin\pywin\framework\scriptutils.py",
line 298, in RunScript
    debugger.run(codeObject, __main__.__dict__, start_stepping=0)
  File "g:\programmi\Python20\Pythonwin\pywin\debugger\__init__.py", line
60, in run
    _GetCurrentDebugger().run(cmd, globals,locals, start_stepping)
  File "g:\programmi\Python20\Pythonwin\pywin\debugger\debugger.py", line
582, in run
    _doexec(cmd, globals, locals)
  File "g:\programmi\Python20\Pythonwin\pywin\debugger\debugger.py", line
921, in _doexec
    exec cmd in globals, locals
  File "G:\Programmi\Python20\fede.py", line 12, in ?
    print unicode(a)
  File "g:\programmi\Python20\win32com\client\dynamic.py", line 165, in
__str__
    return str(self.__call__())
UnicodeError: ASCII encoding error: ordinal not in range(128)

I've try to translate the string into my regional unicode character set, but
it not work.

Help me please.

The strange kind is that using an ODBC connection this error not appear.
Thanks.



--
Neo
****************************
Follow the White Rabbit...
Knock Knock Neo...
www.thezion.net
****************************

Below there's the block code that i've used.

Bye


dataconn.Open (strConn)
sql="select * from GpAzienda ;"
rs.Open(sql, dataconn)
while not rs.eof:
    a = gettext.gettext(rs.Fields('azienda'))
    print a
    rs.MoveNext()

del rs
del dataconn





From fredrik at pythonware.com  Wed Feb 21 20:17:35 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 19:17:35 GMT
Subject: Unicode error with ado in win 32
References: <9713oe$mse8b$1@ID-35431.news.dfncis.de>
Message-ID: <jvUk6.19182$AH6.2594262@newsc.telia.net>

Keanu wrote:
> It work properly but when into the rs.fields('azienda') theres the string
> "abb?" the script returns the following message.

>   File "G:\Programmi\Python20\fede.py", line 12, in ?
>     print unicode(a)

try one of:

    # always works, but may not look right
    print a.encode("utf-8")

    # replace non-ascii with "?"
    print a.encode("ascii", "replace")

    # skip non-ascii characters
    print a.encode("ascii", "ignore")

    # latin-1 (western europe)
    print a.encode("latin-1", "replace")

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From neo at thezion.net  Wed Feb 21 23:47:20 2001
From: neo at thezion.net (Neo)
Date: Wed, 21 Feb 2001 23:47:20 +0100
Subject: Unicode error with ado in win 32
References: <9713oe$mse8b$1@ID-35431.news.dfncis.de> <jvUk6.19182$AH6.2594262@newsc.telia.net>
Message-ID: <971gfp$attcb$1@ID-35431.news.dfncis.de>

Thanks!
This method work properly but only if you have a like a string.

Unfortunally i've mistake the pointer aritmetics.

Writing
a=rs.fields("azienda")
return a pointer to rs.fields("azienda") that is not useful to work like
string.

Also writing
a=rs.fields("azienda").value
return a string that is simply encodable.

Tnx 4 all ^_^


--
Neo
****************************
Follow the White Rabbit...
Knock Knock Neo...
www.thezion.net
****************************



>     # always works, but may not look right
>     print a.encode("utf-8")
>
>     # replace non-ascii with "?"
>     print a.encode("ascii", "replace")
>
>     # skip non-ascii characters
>     print a.encode("ascii", "ignore")
>
>     # latin-1 (western europe)
>     print a.encode("latin-1", "replace")
>
> Cheers /F
>





From spamers at must.die  Wed Feb 21 20:32:11 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Wed, 21 Feb 2001 20:32:11 +0100
Subject: bug in string.join()?
Message-ID: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com>

Hi!

>>> string.join(("abcde"))
'a b c d e'

Is this the expected behavior, or is it a bug? O:-)




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From thorsten at mats.gmd.de  Wed Feb 21 21:17:22 2001
From: thorsten at mats.gmd.de (Thorsten Horstmann)
Date: Wed, 21 Feb 2001 21:17:22 +0100
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com>
Message-ID: <3A942252.E5D1B221@mats.gmd.de>

"Fernando Rodr?guez" wrote:
> >>> string.join(("abcde"))
> 'a b c d e'
> 
> Is this the expected behavior, or is it a bug? O:-)

it's the expected behavior.
string.join concatenate all sequence types. So it iterates
over all characters in the string and joins them with the
default separator ' '. 
It's  not a bug - it's a feature!  ;-)

  -Thorsten


-- 
Anybody else on the list got an opinion? 
Should I change the language or not?  (Guido van Rossum, 28 Dec 1991)
from http://www.amk.ca/quotations/python-quotes/page-1.html


From jmarshal at mathworks.com  Wed Feb 21 21:37:09 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 21 Feb 2001 20:37:09 GMT
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com> <3A942252.E5D1B221@mats.gmd.de>
Message-ID: <9718tl$dtv$1@news.mathworks.com>

Thorsten Horstmann <thorsten at mats.gmd.de> wrote:
> "Fernando Rodrguez" wrote:
>> >>> string.join(("abcde"))
>> 'a b c d e'
>> 
>> Is this the expected behavior, or is it a bug? O:-)

> it's the expected behavior.
> string.join concatenate all sequence types. So it iterates
> over all characters in the string and joins them with the
> default separator ' '. 
> It's  not a bug - it's a feature!  ;-)

Also of note is that you have an unnecessary set of parens.  If you
meant to pass a tuple of one element into string.join, you needed to
add a trailing comma.  And you'd get a different behavior:

>>> string.join(("abcde"))
'a b c d e'
>>> string.join(("abcde",))
'abcde'


From danielk at aracnet.com  Thu Feb 22 04:06:13 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Wed, 21 Feb 2001 19:06:13 -0800
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com>
Message-ID: <qe099tcnql5nnlbl4r0enalcq09np11mup@4ax.com>

On Wed, 21 Feb 2001 20:32:11 +0100, Fernando Rodr?guez <spamers at must.die>
wrote:

>>>> string.join(("abcde"))
>'a b c d e'
>
>Is this the expected behavior, or is it a bug? O:-)

What would you expect it to do? It's already 'joined' without spaces to start
with.

Dan




From adom at i.am  Thu Feb 22 16:39:20 2001
From: adom at i.am (luc lam)
Date: Thu, 22 Feb 2001 15:39:20 GMT
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com> <qe099tcnql5nnlbl4r0enalcq09np11mup@4ax.com>
Message-ID: <973bqv$iub$1@m1.cs.man.ac.uk>

Coffee went everywhere when I read that
 Daniel Klein <danielk at aracnet.com> had written:

>On Wed, 21 Feb 2001 20:32:11 +0100, Fernando Rodr?guez <spamers at must.die>
>wrote:
>
>>>>> string.join(("abcde"))
>>'a b c d e'
>>
>>Is this the expected behavior, or is it a bug? O:-)
>
>What would you expect it to do? It's already 'joined' without spaces to start
>with.
>
>Dan

It threw me somewhat at first:

>>> import string
>>> list = ["i've","come","for","an","argument"]
>>> print string.join(list)
i've come for an argument
>>> print string.join(list," ")
i've come for an argument
>>> print string.join(list,"")
i'vecomeforanargument
>>>                                      

so the " " is implicit. i've sometimes thought it would be more intuitive to 
have the "" behaviour as the implicit behaviour, but it's simple enough either 
way i guess.

luc


From adam at deprince.net  Wed Feb 21 20:33:59 2001
From: adam at deprince.net (Adam DePrince)
Date: Wed, 21 Feb 2001 14:33:59 -0500
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com> <qe099tcnql5nnlbl4r0enalcq09np11mup@4ax.com> <973bqv$iub$1@m1.cs.man.ac.uk>
Message-ID: <3A941827.20C9F0F9@deprince.net>

luc lam wrote:
> 
> Coffee went everywhere when I read that
>  Daniel Klein <danielk at aracnet.com> had written:
> >On Wed, 21 Feb 2001 20:32:11 +0100, Fernando Rodr?guez <spamers at must.die>
> >wrote:
> >>>>> string.join(("abcde"))
> >>'a b c d e'
> >>Is this the expected behavior, or is it a bug? O:-)
> >What would you expect it to do? It's already 'joined' without spaces to start
> >with.
> >Dan
> It threw me somewhat at first:
> >>> import string
> >>> list = ["i've","come","for","an","argument"]
> >>> print string.join(list)
> i've come for an argument
> >>> print string.join(list," ")
> i've come for an argument
> >>> print string.join(list,"")
> i'vecomeforanargument
> so the " " is implicit. i've sometimes thought it would be more intuitive to
> have the "" behaviour as the implicit behaviour, but it's simple enough either


When constructing strings from lists the default value of ' ' as a
delimiter is a pain.  Consider that string.join is often used to replace
string interpolation.  Consider this straw man:

queries = []
for ....
	queries.append("insert into table xyz values '%s','%s','%s' .... 

One might prefer to say 

queries = []
for .....
	queries.append( map( str, ["insert into table xyz values '", d1, "','"
... 

queries = map( string.join, queries )

Of course, for this to work, you would have to say:

queries = map( string.join, queries, ("",)*len(queries))


So, should we propose a join0 that has a delimiter of '', or a version
of map that provides default parameters :-)


Now, to play devils advocate:

' ' makes a convenient delimiter for seperating words in text; it also
matches the defalt delimiter in split.  Notice that:

>>> from string import * 
>>> split( join( ["i've","come","for","an","arguement"]))
["i've", 'come', 'for', 'an', 'arguement']

If join used '' then word lists could not be broken appart and rejoined
so easily.

> way i guess.
> luc

Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From tim.hochberg at ieee.org  Thu Feb 22 22:13:23 2001
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Thu, 22 Feb 2001 21:13:23 GMT
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com> <qe099tcnql5nnlbl4r0enalcq09np11mup@4ax.com> <973bqv$iub$1@m1.cs.man.ac.uk> <3A941827.20C9F0F9@deprince.net>
Message-ID: <Thfl6.2686$Qd4.324192@news1.rdc1.az.home.com>

"Adam DePrince" <adam at deprince.net> wrote in message
[SNIP]
> When constructing strings from lists the default value of ' ' as a
> delimiter is a pain.  Consider that string.join is often used to replace
> string interpolation.  Consider this straw man:
>
> queries = []
> for ....
> queries.append("insert into table xyz values '%s','%s','%s' ....
>
> One might prefer to say
>
> queries = []
> for .....
> queries.append( map( str, ["insert into table xyz values '", d1, "','"
> ...
> 
> queries = map( string.join, queries )
>
> Of course, for this to work, you would have to say:
>
> queries = map( string.join, queries, ("",)*len(queries))

Death to map! Use list comprehensions instead. And string methods! New is
good, old is bad <0.9 wink>.

   queries = [''.join(q) for q in queeries]

[SNAP]

-tim




From jwbnews at scandaroon.com  Fri Feb 23 06:16:19 2001
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Thu, 22 Feb 2001 21:16:19 -0800
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com> <qe099tcnql5nnlbl4r0enalcq09np11mup@4ax.com> <973bqv$iub$1@m1.cs.man.ac.uk> <3A941827.20C9F0F9@deprince.net> <Thfl6.2686$Qd4.324192@news1.rdc1.az.home.com>
Message-ID: <jwbnews-64D9C0.21161922022001@corp.supernews.com>

In article <Thfl6.2686$Qd4.324192 at news1.rdc1.az.home.com>,
 "Tim Hochberg" <tim.hochberg at ieee.org> wrote:

> Death to map! Use list comprehensions instead.

As soon as my 1940s brain comprehends them.  ;-)  --john

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From spamers at must.die  Sat Feb 24 00:47:20 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Sat, 24 Feb 2001 00:47:20 +0100
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com> <qe099tcnql5nnlbl4r0enalcq09np11mup@4ax.com>
Message-ID: <6jtd9t0b25925bs865a5t796q1iof1ao65@4ax.com>

On Wed, 21 Feb 2001 19:06:13 -0800, Daniel Klein <danielk at aracnet.com> wrote:

>On Wed, 21 Feb 2001 20:32:11 +0100, Fernando Rodr?guez <spamers at must.die>
>wrote:
>
>>>>> string.join(("abcde"))
>>'a b c d e'
>>
>>Is this the expected behavior, or is it a bug? O:-)
>
>What would you expect it to do? It's already 'joined' without spaces to start
>with.

That's why I expected it to do nothing. Instead, it did something weird.




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From BrianQ at ActiveState.com  Sat Feb 24 02:20:11 2001
From: BrianQ at ActiveState.com (Brian Quinlan)
Date: Fri, 23 Feb 2001 17:20:11 -0800
Subject: bug in string.join()?
In-Reply-To: <6jtd9t0b25925bs865a5t796q1iof1ao65@4ax.com>
Message-ID: <001c01c09dff$ef83afa0$b503a8c0@activestate.ca>

OK, first of all, according to the docs, string.join operates on tuples and
lists, not strings. If you change your code to string.join(("abcde",)), you
get what you expected.

But string.join will actually except anything sequence-type like and, by
default, insert a space between it's items. Which is exactly what it did.

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Fernando Rodr?guez
Sent: Friday, February 23, 2001 3:47 PM
To: python-list at python.org
Subject: Re: bug in string.join()?


On Wed, 21 Feb 2001 19:06:13 -0800, Daniel Klein <danielk at aracnet.com>
wrote:

>On Wed, 21 Feb 2001 20:32:11 +0100, Fernando Rodr?guez <spamers at must.die>
>wrote:
>
>>>>> string.join(("abcde"))
>>'a b c d e'
>>
>>Is this the expected behavior, or is it a bug? O:-)
>
>What would you expect it to do? It's already 'joined' without spaces to
start
>with.

That's why I expected it to do nothing. Instead, it did something weird.




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------
--
http://mail.python.org/mailman/listinfo/python-list




From spamers at must.die  Wed Feb 21 20:32:12 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Wed, 21 Feb 2001 20:32:12 +0100
Subject: Non single-line comments?
Message-ID: <pp589t4k4s5a2uh3qkp5t9i3atoblur2jm@4ax.com>

Hi!

	Are there any plans to support non single-line comments (a la /* */ in
C)? O:-)  I sometimes miss them while debugging...




//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From wolfgang.grafen at marconi.com  Wed Feb 21 20:45:47 2001
From: wolfgang.grafen at marconi.com (Wolfgang Grafen)
Date: Wed, 21 Feb 2001 20:45:47 +0100
Subject: Non single-line comments?
References: <pp589t4k4s5a2uh3qkp5t9i3atoblur2jm@4ax.com>
Message-ID: <3A941AEB.5ECDD353@marconi.com>

Use """ to comment out blocks:

>>> def ignored_python_code():
...     print "comment out from next line"
...     """
...     { some python code to comment out }
... """
...     print "continue now from here"
... 
>>> ignored_python_code()
comment out from next line
continue now from here


Regards

Wolfgang



"Fernando Rodr?guez" wrote:
> 
> Hi!
> 
>         Are there any plans to support non single-line comments (a la /* */ in
> C)? O:-)  I sometimes miss them while debugging...
> 
> //-----------------------------------------------
> //      Fernando Rodriguez Romero
> //
> //      frr at mindless dot com
> //------------------------------------------------


From bobhicks at adelphia.net  Wed Feb 21 20:47:37 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Wed, 21 Feb 2001 19:47:37 GMT
Subject: Non single-line comments?
References: <pp589t4k4s5a2uh3qkp5t9i3atoblur2jm@4ax.com>
Message-ID: <B6B98585.BF8%bobhicks@adelphia.net>

What is wrong with the # sign? I think it clearly distinguishes any
comments, even multi-line.

My .02

- Bob

> From: Fernando Rodr?guez <spamers at must.die>
> Organization: T.I.A.
> Newsgroups: comp.lang.python
> Date: Wed, 21 Feb 2001 20:32:12 +0100
> Subject: Non single-line comments?
> 
> Hi!
> 
> Are there any plans to support non single-line comments (a la /* */ in
> C)? O:-)  I sometimes miss them while debugging...
> 
> 
> 
> 
> //-----------------------------------------------
> // Fernando Rodriguez Romero
> //
> // frr at mindless dot com
> //------------------------------------------------



From phlip_cpp at my-deja.com  Wed Feb 21 20:50:21 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: Wed, 21 Feb 2001 11:50:21 -0800
Subject: IDLE - change the font size & background color
Message-ID: <p0Vk6.18$0Q5.43640@news.pacbell.net>

Pytho-logicians:

Before I crack open the source, I'm wondering if I overlooked some simple 
way to make the font bigger and set the background color to a light grey. 
Does anyone have, like, a patch?

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  DARE to resist drug-war profiteering  --


From butcuz at yahoo.com  Wed Feb 21 21:56:52 2001
From: butcuz at yahoo.com (sa)
Date: Wed, 21 Feb 2001 15:56:52 -0500
Subject: IDLE - change the font size & background color
References: <p0Vk6.18$0Q5.43640@news.pacbell.net>
Message-ID: <3a942c61_4@corp.newsfeeds.com>

changing font size is pretty simple:

- goto the directory where IDLE is usually stored (It's
C:\Python20\Tools\idle on my windoze system)

- look to one of two files that directory:
    - config-win.txt [for windoze]
    - config-unix.txt [for unix]

and change the font size accordingly. [I like Verdana, 8pt]

[for the background color, I'm affraid you're gong to have to dig into the
source]

sa

"Phlip" <phlip_cpp at my-deja.com> wrote in message
news:p0Vk6.18$0Q5.43640 at news.pacbell.net...
> Pytho-logicians:
>
> Before I crack open the source, I'm wondering if I overlooked some simple
> way to make the font bigger and set the background color to a light grey.
> Does anyone have, like, a patch?
>
> --
>   Phlip                          phlip_cpp at my-deja.com
> ============ http://c2.com/cgi/wiki?PhlIp ============
>   --  DARE to resist drug-war profiteering  --




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From phlip_cpp at my-deja.com  Wed Feb 21 23:57:35 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: Wed, 21 Feb 2001 14:57:35 -0800
Subject: IDLE - change the font size & background color
References: <p0Vk6.18$0Q5.43640@news.pacbell.net> <3a942c61_4@corp.newsfeeds.com>
Message-ID: <VLXk6.2442$i04.109027@news.pacbell.net>

sa wrote:

> changing font size is pretty simple:
> 
> - goto the directory where IDLE is usually stored (It's
> C:\Python20\Tools\idle on my windoze system)
> 
> - look to one of two files that directory:
>     - config-win.txt [for windoze]
>     - config-unix.txt [for unix]
> 
> and change the font size accordingly. [I like Verdana, 8pt]
> 
> [for the background color, I'm affraid you're gong to have to dig into the
> source]

we got it.

Tip 1: background color's in 'config.txt'.

Tip 2: Get the latest IDLE ;-)

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  My opinions are those of your employer  --


From joseamartin at venmex.com  Wed Feb 21 20:58:17 2001
From: joseamartin at venmex.com (Jose Martin)
Date: Wed, 21 Feb 2001 15:58:17 -0400
Subject: Simplest Way to call Stored Procs In DCOracle == cur.execute(' call myproc() '  )  
Message-ID: <t9886fmfesb6c6@corp.supernews.com>

Thats it's








From fwittenburg at myokay.net  Wed Feb 21 21:00:11 2001
From: fwittenburg at myokay.net (Florian W.)
Date: Wed, 21 Feb 2001 21:00:11 +0100
Subject: Comparison with Ruby ?
Message-ID: <9716n5$2lln$1@news.okay.net>

Hi Python-Fans !

A friend told me about the programming language "Ruby".
It's seems that it's a fine one.

Will the high number of features shrink the Python-community to
a very low amount ?

Mfg Florian





From claird at starbase.neosoft.com  Wed Feb 21 21:31:21 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 21 Feb 2001 14:31:21 -0600
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net>
Message-ID: <9457E37E39EE85FF.533763A1B69129A1.704A78AD6E685E9C@lp.airnews.net>

In article <9716n5$2lln$1 at news.okay.net>,
Florian W. <fwittenburg at myokay.net> wrote:
>Hi Python-Fans !
>
>A friend told me about the programming language "Ruby".
>It's seems that it's a fine one.
>
>Will the high number of features shrink the Python-community to
>a very low amount ?
			.
			.
			.
For the most part, the people in c.l.p are more practiced
in making the best of Python's features (and making Pyfea-
tures the best), than speculating about how they might
evaporate.

If you need comparisons of Ruby and Python beyond what
<URL: http://starbase.neosoft.com/~claird/comp.lang.misc/language_comparisons.html>
provides, let me know; I have a few more I haven't annotated
yet.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From dfan at harmonixmusic.com  Wed Feb 21 23:21:00 2001
From: dfan at harmonixmusic.com (Dan Schmidt)
Date: 21 Feb 2001 17:21:00 -0500
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net> <9457E37E39EE85FF.533763A1B69129A1.704A78AD6E685E9C@lp.airnews.net>
Message-ID: <wk3dd7ad9v.fsf@turangalila.harmonixmusic.com>

I hesitate to mention this lest it be construed as advocacy, but the
PROGRAMMING RUBY book (by David Thomas and Andrew Hunt, the PRAGMATIC
PROGRAMMER guys) is now online at <http://www.rubycentral.com/book/>.
Whether or not you like Ruby, it's a very well-written book and makes
learning the language orders of magnitude easier for English speakers.

-- 
http://www.dfan.org


From claird at starbase.neosoft.com  Thu Feb 22 16:38:10 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 22 Feb 2001 09:38:10 -0600
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net> <9457E37E39EE85FF.533763A1B69129A1.704A78AD6E685E9C@lp.airnews.net> <wk3dd7ad9v.fsf@turangalila.harmonixmusic.com>
Message-ID: <12C9A77C1B9B010B.158D24368FADE26B.D347411B0EEA8942@lp.airnews.net>

In article <wk3dd7ad9v.fsf at turangalila.harmonixmusic.com>,
Dan Schmidt  <dfan at harmonixmusic.com> wrote:
>I hesitate to mention this lest it be construed as advocacy, but the
>PROGRAMMING RUBY book (by David Thomas and Andrew Hunt, the PRAGMATIC
>PROGRAMMER guys) is now online at <http://www.rubycentral.com/book/>.
>Whether or not you like Ruby, it's a very well-written book and makes
>learning the language orders of magnitude easier for English speakers.
			.
			.
			.
I think the book is even better than that:  it manages
also to convey The Ruby Mentality.  Dave and Andrew are
good at what they do.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From bobhicks at adelphia.net  Wed Feb 21 21:45:13 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Wed, 21 Feb 2001 20:45:13 GMT
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net>
Message-ID: <B6B99307.E23%bobhicks@adelphia.net>

troll troll troll

> From: "Florian W." <fwittenburg at myokay.net>
> Organization: okay.net Newsserver @ Jippii!
> Newsgroups: comp.lang.python
> Date: Wed, 21 Feb 2001 21:00:11 +0100
> Subject: Comparison with Ruby ?
> 
> Hi Python-Fans !
> 
> A friend told me about the programming language "Ruby".
> It's seems that it's a fine one.
> 
> Will the high number of features shrink the Python-community to
> a very low amount ?
> 
> Mfg Florian
> 
> 
> 



From Randy.L.Kemp at motorola.com  Wed Feb 21 22:14:06 2001
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Wed, 21 Feb 2001 15:14:06 -0600
Subject: Comparison with Ruby ?
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84E6E@il93exp01.css.mot.com>

If you go to www.rubycentral.com, you will find all you need to know about Ruby.  If you go to www.python.org or www.activestate.com, you will find all you need to know about Python.  Ruby is very popular in Japan, but has not caught on as big in other parts of the world.  Perl is only popular because it was at the right place at the right time with web usage.  It's hard to predict what will catch on.  Are you implying we might be singing a revamped Rolling Stones song Rube Tuesday?  How do the words go?
  "Goodbye to programming with Ruby on Tuesday.
  Python's the name for all that's new,
  Since Perl changes with every new release day,
  I'm not going to miss you."

-----Original Message-----
From: Florian W. [mailto:fwittenburg at myokay.net]
Sent: Wednesday, February 21, 2001 2:00 PM
To: python-list at python.org
Subject: Comparison with Ruby ?


Hi Python-Fans !

A friend told me about the programming language "Ruby".
It's seems that it's a fine one.

Will the high number of features shrink the Python-community to
a very low amount ?

Mfg Florian






From juergen.bocklage at gmx.net  Thu Feb 22 00:00:46 2001
From: juergen.bocklage at gmx.net (Juergen Bocklage)
Date: Thu, 22 Feb 2001 00:00:46 +0100
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net>
Message-ID: <3A94489D.5B65ED19@gmx.net>

Aehmmm, I'm just a beginner, but some remarks on the ruby-python
comparision.


Python and Ruby deal with similar problem areas. Now, let's compare
Python with Ruby:

     Ruby uses conservative statement structure, using `end'.
JB> Python syntax is the best I ever saw, indenting is just smart.

     You don't have to say `self.' to access attribute of the object.
Ruby's instance variables are prefixed by "@".
JB> Then I prefere self. It's more descriptive.

     Python separates types and classes, while Ruby treats them the
same. Python types are more limited (no
     inheritance/subclassing; cannot add methods to existing types).
JB> Don't know

     Ruby has a better (or "real") closure feature.
JB> what does this mean

     Object attributes are not accessible by default in Ruby.
JB> That's bad. Wrinting __ in front of a instance attribute, is okay
for me.

     Ruby converts small integers and long integers automatically.
JB> Okay, nice for you

     Ruby does not have tuples.
JB> Bad for Ruby. They are fast.

     Statements in Ruby have values; you can say

          def max(a,b)
            if a>b then a else b end
          end
JB> Where is the return? How can someone see when sth. is returned?

     Ruby has more natural operator overloading; e.g. you can define

          def +(x)
           self.to_i + x
          end

JB> I know this from C++, (Syntactical sugar or a good source for
errors)

     Ruby uses mark-and-sweep GC, unlike Python's ref-counting GC. So
that:
          No INCREF, DECREF are required for extensions.
     Extensions for Ruby written in C/C++ can easily define Ruby
classes.
     Ruby has a loop abstraction using block; e.g.

          10.times do
            ...
          end

JB> What's this?

     You can define your own arbitrary iterator.
JB> Why?

     Ruby's block for method can be used more than iterator; e.g.

          mutex.synchronize do
            .. critical process ..
          end

JB> Don't know

     Ruby provides method combination using `super'.
JB> this is a good feature, in python I think you have to use the name
of parent class.

     Ruby is often faster than Python.
JB> Often??

After a short look at the language I have to say, python is much more
self explanatory then ruby. Also this parenteses less calling of
methods, does not look good. Looks like calling an attribute.

Python is for me the language for beginners and designers.
I teached my flat mates in one eveing how to prototype there C++ app.
with python. They where just ;-)) and concentrated on the design.

- See ya JB

"Florian W." wrote:

> Hi Python-Fans !
>
> A friend told me about the programming language "Ruby".
> It's seems that it's a fine one.
>
> Will the high number of features shrink the Python-community to
> a very low amount ?
>
> Mfg Florian

--
Juergen Bocklage
================================
mailto:juergen.bocklage at gmx.net




From aleaxit at yahoo.com  Thu Feb 22 10:19:34 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 10:19:34 +0100
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net>
Message-ID: <972lk10k0n@news2.newsguy.com>

"Juergen Bocklage" <juergen.bocklage at gmx.net> wrote in message
news:3A94489D.5B65ED19 at gmx.net...
> Aehmmm, I'm just a beginner, but some remarks on the ruby-python
> comparision.
    [snip]
>      Python separates types and classes, while Ruby treats them the
> same. Python types are more limited (no
>      inheritance/subclassing; cannot add methods to existing types).
> JB> Don't know

A Ruby win here.


>      Ruby has a better (or "real") closure feature.
> JB> what does this mean

Nothing, actually.  Some people are SO keen on nested lexical
scopes, that they'll (mistakenly) argue that a closure is not
one unless it supports _nested_ lexical scopes.  Until 2.0,
Python's scopes don't nest, so, of course, they keep not
nesting in closures; at 2.1, they will (alas) nest, so, of
course, they'll keep nesting in closures.

>      Ruby converts small integers and long integers automatically.
> JB> Okay, nice for you

No way.  Explicit is better than implicit.  I'd rather NOT
have such 'automatic' (automagick) conversions anywhere!

>      Ruby does not have tuples.
> JB> Bad for Ruby. They are fast.

Disagree.  'Fast' should be handled by the compiler and runtime.
Tuples should be lists with an "immutable" setting (and there
should be similar "immutable" settings for dictionaries) but
otherwise indistinguishable.

>           def +(x)
>            self.to_i + x
>           end
>
> JB> I know this from C++, (Syntactical sugar or a good source for
> errors)

It's not more error-prone than using __add__ as the magic name
for overloading addition, which is why Python does.  Looks like
a total wash to me.

>      Ruby has a loop abstraction using block; e.g.
>
>           10.times do
>             ...
>           end
>
> JB> What's this?

A method 'times' on integer objects, that takes a code-block.


>      You can define your own arbitrary iterator.
> JB> Why?

To iterate arbitrarily.  Why else?

>      Ruby provides method combination using `super'.
> JB> this is a good feature, in python I think you have to use the name
> of parent class.

Disagree -- I _like_ it that the SPECIFIC parent-class you
mean has to be explicitly mentioned (multiple-inheritance,
remember).  I don't like the complex rules of Dylan to allow
magick implicitness in this, nor constraining inheritance to
single just to allow 'super' to work:-).


> Python is for me the language for beginners and designers.

Not just for them.  Lots of Hackers and eXtreme Programmers
are also pretty happy with it.


Alex





From emile at fenx.com  Thu Feb 22 14:11:17 2001
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 22 Feb 2001 05:11:17 -0800
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net> <972lk10k0n@news2.newsguy.com>
Message-ID: <97337o$nnk22$1@ID-11957.news.dfncis.de>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:972lk10k0n at news2.newsguy.com...
> "Juergen Bocklage" <juergen.bocklage at gmx.net> wrote in
message
> news:3A94489D.5B65ED19 at gmx.net...
<snip>
> >      Ruby converts small integers and long integers
automatically.
> > JB> Okay, nice for you
>
> No way.  Explicit is better than implicit.  I'd rather NOT
> have such 'automatic' (automagick) conversions anywhere!
>

What *magic* requires my attention between 2147483647 and
2147483648?  I'd rather have a single type for a python
number, and let C do what it needs to do to allow it.

I learned a long time ago to store monetary values as whole
numbers in order to avoid rounding and precision related
problems.  Here, that means that the maximum value a sum of
numbers can have is $21,474,836.47.  On a trial balance
where you add up the debits for the period, even small
companies hit this limit quickly.  This simply means that
all monetary values must be longs and is certainly not a
problem, but I'd much prefer not having to think of it.

--

Emile van Sebille
emile at fenx.com
-------------------






From othello at javanet.com  Fri Feb 23 04:03:23 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Thu, 22 Feb 2001 22:03:23 -0500
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net> <972lk10k0n@news2.newsguy.com>
Message-ID: <3A95D2FB.5D7937A9@javanet.com>

Alex Martelli wrote:

> "Juergen Bocklage" <juergen.bocklage at gmx.net> wrote in message
> news:3A94489D.5B65ED19 at gmx.net...
> > Aehmmm, I'm just a beginner, but some remarks on the ruby-python
> > comparision.
>     [snip]
> >      Python separates types and classes, while Ruby treats them the
> > same. Python types are more limited (no
> >      inheritance/subclassing; cannot add methods to existing types).
> > JB> Don't know
>
> A Ruby win here.

Yes!

> >      Ruby does not have tuples.
> > JB> Bad for Ruby. They are fast.
>
> Disagree.  'Fast' should be handled by the compiler and runtime.
> Tuples should be lists with an "immutable" setting (and there
> should be similar "immutable" settings for dictionaries) but
> otherwise indistinguishable.

Double Yes!

>
> >      Ruby provides method combination using `super'.
> > JB> this is a good feature, in python I think you have to use the name
> > of parent class.
>
> Disagree -- I _like_ it that the SPECIFIC parent-class you
> mean has to be explicitly mentioned (multiple-inheritance,
> remember).  I don't like the complex rules of Dylan to allow
> magick implicitness in this, nor constraining inheritance to
> single just to allow 'super' to work:-).

I would like to see a version of 'super' that searches inherited classes
in the same order that methods are searched for.  Super.method, should act
like it would if the method were not found in the current class definition.
'nuff said.

>
> > Python is for me the language for beginners and designers.
>
> Not just for them.  Lots of Hackers and eXtreme Programmers
> are also pretty happy with it.
>

I'm not a beginner or designer and I'm  happy.  See my big smile:

@ @
\_/


Raymond



From jeremy at alum.mit.edu  Fri Feb 23 04:10:48 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Thu, 22 Feb 2001 22:10:48 -0500 (EST)
Subject: Comparison with Ruby ?
In-Reply-To: <3A95D2FB.5D7937A9@javanet.com>
References: <9716n5$2lln$1@news.okay.net>
	<3A94489D.5B65ED19@gmx.net>
	<972lk10k0n@news2.newsguy.com>
	<3A95D2FB.5D7937A9@javanet.com>
Message-ID: <14997.54456.899428.840814@w221.z064000254.bwi-md.dsl.cnc.net>

On the issue of super and overriding: I'd like to recommend again the
use of explicit renaming to make this a little clearer.  It's an idea
I first saw in MzScheme --
    http://www.cs.rice.edu/CS/PLT/packages/mzscheme/index.html
-- and it has a very natural mapping to Python.

class Foo(Bar, Baz):
    __super_meth1 = Bar.meth1
    __super_meth2 = Baz.meth2

    def meth1(self, arg):
        self.__super_meth1(arg)

    def meth2(self, arg):
        self.__super_meth2(arg)

The idea is to create new bindings for the method objects in the
current class.  Then invocation of the method in the parent class
looks like a normal method call.  It's important to add the __ first.
If you don't and more than one class in the inheritance hierarchy
overrides the same method, you'll be in trouble.  It's merely
convention to make it __super, but it helps to indicate what is going
on.

A nice side-effect of this style is that it localizes the information
about what dependencies there are on parent classes.  If you change
the base classes of Foo, you've got all the information about specific
methods of base classes that are being called -- as opposed to have
Baz.meth2(self, arg) scattered throughout the body of Foo.

Jeremy




From aleaxit at yahoo.com  Fri Feb 23 11:33:04 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 23 Feb 2001 11:33:04 +0100
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net> <972lk10k0n@news2.newsguy.com> <3A95D2FB.5D7937A9@javanet.com>
Message-ID: <975ea40314d@news1.newsguy.com>

"Raymond Hettinger" <othello at javanet.com> wrote in message
news:3A95D2FB.5D7937A9 at javanet.com...
    [snip]
> I would like to see a version of 'super' that searches inherited classes
> in the same order that methods are searched for.  Super.method, should act
> like it would if the method were not found in the current class
definition.
> 'nuff said.

Generalizing this a bit (to find all attributes, rather
than just methods), you could just use the following
free-standing function:

def findSuperclassForAttributeName(self, attributeName):
    for base in self.__class__.__bases__:
        if hasattr(base, attributeName): return base
    else: return None

or, you could use a convenient mixin class, by having
your classes inherit from it:

class SuperFinder:
    [snip: method same as above]

but that's just for the syntax-sugar satisfaction of
being able to use this in a 'cooler' way, such as:

class Whatever(Abase,Anotherbase,SuperFinder):
    def floop(self, plik, plok):
        preFlooper(plok)
        super = self.findSuperclassForAttributeName('floop')
        if super is None: raise AttributeError
        super.floop(self, plok, plik)
        postFlooper(plik+23.7)

and if you're into 'cool' solutions (rather than simple
ones), then you can do even "better" -- e.g., here's a
less-general, but potentially interesting, approach:

import new
def bindInSuper(boundMethod):
    obj = boundMethod.im_self
    name = boundMethod.__name__
    for base in obj.__class__.__bases__:
        if hasattr(base, name): break
    else: raise AttributeError
    return new.instancemethod(getattr(base, name), obj, base)

leading to the usage pattern:

class Whatever(Abase,Anotherbase,SuperFinder):
    def floop(self, plik, plok):
        preFlooper(plok)
        bindInSuper(self.floop)(plok, plik)
        postFlooper(plik+23.7)


Note that this only works ONCE, to find ONE 'super', or
build one bound-method; e.g., the returned bound-method
of bindInSuper is a fixedpoint for it.  It would not
be TOO hard to work around this, by having bindInSuper
return an instance of a suitable special-purpose class,
that is both callable AND specialcased for further
searches on the inheritance DAG (you'd have to decide
exactly what sequence of [bound-methods wrapped into
suitable callable instances] you want to get from a
sequence of calls to bindInSuper in various cases, of
course:-).

But I'm going to leave this as an exercise to the
interested reader.  The real value of this fancy stuff
is not in truly making use of it in production code,
anyway -- it's in showing off on comp.lang.python,
ahem, I *mean*, in better understanding what IS going
on "behind the scenes" (but not "secretly" at all --
Python is wonderfully open about what its mechanisms
ARE doing on your behalf!!!).  In *practice*, one
should use *the simplest thing that can possibly work* --
and, unless one is building the next great development
and debugging environment, it is seldom indeed that
the "simplest thing" will require introspection and
metaprogramming facilities.


Alex





From moshez at zadka.site.co.il  Fri Feb 23 11:49:44 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri, 23 Feb 2001 12:49:44 +0200 (IST)
Subject: Comparison with Ruby ?
In-Reply-To: <972lk10k0n@news2.newsguy.com>
References: <972lk10k0n@news2.newsguy.com>, <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net>
Message-ID: <20010223104944.BEB4AA840@darjeeling.zadka.site.co.il>

On Thu, 22 Feb 2001 10:19:34 +0100, "Alex Martelli" <aleaxit at yahoo.com> wrote:

> >      Ruby converts small integers and long integers automatically.
> > JB> Okay, nice for you
> 
> No way.  Explicit is better than implicit.  I'd rather NOT
> have such 'automatic' (automagick) conversions anywhere!

Err....I don't agree. See PEP 0228
(http://python.sourceforge.net/peps/pep-0228.html).
I think it could be done in a very Pythonic way (of course, for full disclosure,
I am the author of said PEP...)
Numbers are numbers: I shouldn't care about how to store numbers in 
machine-level integers more then I should care how to reallocate lists.
This PEP is a bit backwards incompatible, but in my not so humble opinion,
I think it does have a good chance of being in Py3K.
There is a related session in IPC9 given by Tim Peters, by the way.

-- 
"I'll be ex-DPL soon anyway so I'm        |LUKE: Is Perl better than Python?
looking for someplace else to grab power."|YODA: No...no... no. Quicker,
   -- Wichert Akkerman (on debian-private)|      easier, more seductive.
For public key, finger moshez at debian.org  |http://www.{python,debian,gnu}.org



From aleaxit at yahoo.com  Fri Feb 23 12:13:07 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 23 Feb 2001 12:13:07 +0100
Subject: Comparison with Ruby ?
References: <972lk10k0n@news2.newsguy.com>, <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net> <20010223104944.BEB4AA840@darjeeling.zadka.site.co.il>
Message-ID: <021101c09d89$9a418740$102b2bc1@cadlab.it>

"Moshe Zadka" <moshez at zadka.site.co.il> writes:

> > >      Ruby converts small integers and long integers automatically.
> > > JB> Okay, nice for you
> >
> > No way.  Explicit is better than implicit.  I'd rather NOT
> > have such 'automatic' (automagick) conversions anywhere!
>
> Err....I don't agree. See PEP 0228
> (http://python.sourceforge.net/peps/pep-0228.html).
> I think it could be done in a very Pythonic way (of course, for full
disclosure,
> I am the author of said PEP...)

Guess I should have mentioned I _do_ like PEP 228 -- reworking the
entire conceptual model of Python numbers in such a simplifying
way seems a very worthy goal.  Automatic conversions behind my
back are a different thing...

> Numbers are numbers: I shouldn't care about how to store numbers in
> machine-level integers more then I should care how to reallocate lists.

But I _do_ care about how list are allocated -- somebody mentioned
recently that a simple 'optimization' I and others had suggested,
where a (huge) list of known length N is "pre-allocated" as N*[None],
then the various items set, gave them a factor-of-2 boost in time
to run their program over the N-times-append alternative.  Speed is
not the crucial thing when one uses Python, but factors of 2 or more
are still of some interest, in many cases.

> This PEP is a bit backwards incompatible, but in my not so humble opinion,
> I think it does have a good chance of being in Py3K.
> There is a related session in IPC9 given by Tim Peters, by the way.

Alas, I won't be there, but I hope I'll be able to read some report
about it on the net later...


Alex





From johann at physics.berkeley.edu  Fri Feb 23 21:28:17 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 23 Feb 2001 12:28:17 -0800
Subject: Comparison with Ruby ?
References: <972lk10k0n@news2.newsguy.com> <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net> <mailman.982925467.28685.python-list@python.org>
Message-ID: <mt7l2h87q6.fsf@astron.berkeley.edu>

Moshe Zadka writes:

> Numbers are numbers: I shouldn't care about how to store numbers in 
> machine-level integers more then I should care how to reallocate lists.
> This PEP is a bit backwards incompatible, but in my not so humble opinion,
> I think it does have a good chance of being in Py3K.

I just read that; looks nice.  Inspired by Scheme?  Then again, it's
not suprising that I like it; at least someone's Python humor file
quotes me twice going off on the same issue.  I'll have to be more
circumspect this time.  :-)

In any case, as someone who does a lot of numerics, I've got to say
that folding this under "numbers are numbers" is unwholesome.  Inexact
numbers are not exact numbers.  That lets the number-crunchers use
their types happily.

Integers (fixnums) and rationals and bignums, on the other hand, are
"numbers" and there should be no walls between them.

--J

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From fwittenburg at myokay.net  Thu Feb 22 21:48:35 2001
From: fwittenburg at myokay.net (Florian W.)
Date: Thu, 22 Feb 2001 21:48:35 +0100
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net>
Message-ID: <973tte$1056$1@news.okay.net>

First of all... thanks for the comments

Sometimes the finger is faster than the brain.
So I didn't searched with google the usenet about
ruby<->Python comments.

Sorry, Florian

P.S. I prefer to mark blocks with tabs :-)

Florian W. <fwittenburg at myokay.net> schrieb in im Newsbeitrag:
9716n5$2lln$1 at news.okay.net...
> Hi Python-Fans !
>
> A friend told me about the programming language "Ruby".
> It's seems that it's a fine one.
>
> Will the high number of features shrink the Python-community to
> a very low amount ?
>
> Mfg Florian
>
>
>




From faraway at mhtc.net  Wed Feb 21 21:28:20 2001
From: faraway at mhtc.net (Craig Dunigan)
Date: Wed, 21 Feb 2001 14:28:20 -0600
Subject: base64.decodestring or xmlrpc (in zope) is truncating strings
Message-ID: <GzVk6.77$S94.629@client>

If anyone knows of any problems with base64 or xmlrpc, or has experienced
any similar behavior, I'd really like to hear from you!

I am extending a zope/xmlrpc client on Win32 that was written by a much
brighter programmer than me.  Basically, it uses xmlrpc to call external
methods on the zope server, and synchronizes a local web tree with a users
remote tree.  It makes its rpc call to an external method on the server
(RedHat 7, Zope 2.3) containing this bit of code (it returns file contents
from the zodb):

  meta_type = fileObject.meta_type
  if meta_type == 'DTML Document':
    return ("OK", base64.encodestring(fileObject.document_src()))
  elif meta_type == 'File':
    return ("OK", base64.encodestring(fileObject.data))
  elif meta_type == 'Image':
    return ("OK", base64.encodestring(fileObject.data))
  else:
    return ("FAIL", 'Unsupported meta_type')


This is the client's method to receive the file contents:

  def getFile(self, pathSpec, fileName):
    parms = {
      'relativePathSpec':pathSpec,
      'fileName': fileName
    }
    retval = None
    try:
      retval = eval("self.serverObject.getConnection()." +
self.remoteBasePathSpec + ".EWSyncGetFile(parms)")
      # EWSyncGetFile is the remote method that contains the snippet above
    except:
      raise "ZopeFileSystemException", "type: %s\nvalue: %s" %
(sys.exc_type, sys.exc_value)
    if retval[0] != "OK":
      raise "ZopeFileSystemException", "Failure from EWSyncGetFile()\n  %s"
% retval[1]
    return base64.decodestring(retval[1])

I've put in some debug code to spit print statements to the console, then
run the client window and server console side by side, and visually verified
that string lengths of the server's return after encode and the client's
retval[1] prior to decode are identical, meaning we're receiving the encode
string unchanged.  I've also compared the pre-encode string length on the
server to the decode string length immediately after the base64.decodestring
in the client, and they aren't the same in about a third of my test
documents.  Of course, the received file contents (the calling object of
"getFile" does a write to the local filesystem) are corrupt and unusable.
I've done a compare of the original files and the corrupted ones in a hex
editor, and in every case the corrupted file simply stops somewhere in the
middle.  There doesn't appear to be a set length that it fails at, since the
received corrupted files vary in byte size from ~80KB to ~120KB.  Larger
files have also come through unchanged.  The files that get truncated are MS
PowerPoint and Adobe PDF, which are the bulk of what this site will contain
(of course!).  Some PowerPoint and PDF files come through just fine, and
I've been unable to determine what the difference is between the files that
work and the ones that don't.

If anyone can make any suggestions, I'd be extremely grateful!




From fredrik at pythonware.com  Wed Feb 21 21:45:01 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 20:45:01 GMT
Subject: base64.decodestring or xmlrpc (in zope) is truncating strings
References: <GzVk6.77$S94.629@client>
Message-ID: <hNVk6.14872$Qb7.2374586@newsb.telia.net>

Craig Dunigan wrote:
>   meta_type = fileObject.meta_type
>   if meta_type == 'DTML Document':
>     return ("OK", base64.encodestring(fileObject.document_src()))
>   elif meta_type == 'File':
>     return ("OK", base64.encodestring(fileObject.data))
>   elif meta_type == 'Image':
>     return ("OK", base64.encodestring(fileObject.data))
>   else:
>     return ("FAIL", 'Unsupported meta_type')

xmlrpc usage note: instead of encoding and decoding yourself,
you can use the xmlrpclib.Binary wrapper:

    if meta_type == 'DTML Document':
        return ("OK", xmlrpclib.Binary(fileObject.document_src())
    elif meta_type in ('File', 'Image'):
        return ("OK", xmlrpclib.Binary(fileObject.data))
    else:
        return ("FAIL", 'Unsupported meta_type')

the receiver gets a Binary object as well, and can pick out
the original string from the data attribute:

    if retval[0] == "OK":
        return retval[1].data

> the calling object of "getFile" does a write to the local filesystem

...which usually means that you forgot to open the file in
binary mode:

    file = open(filename, "w") # open in text mode
    file = open(filename, "wb") # open in binary mode

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From faraway at mhtc.net  Wed Feb 21 22:42:12 2001
From: faraway at mhtc.net (Craig Dunigan)
Date: Wed, 21 Feb 2001 15:42:12 -0600
Subject: base64.decodestring or xmlrpc (in zope) is truncating strings
References: <GzVk6.77$S94.629@client> <hNVk6.14872$Qb7.2374586@newsb.telia.net>
Message-ID: <YEWk6.89$S94.611@client>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:hNVk6.14872$Qb7.2374586 at newsb.telia.net...
> Craig Dunigan wrote:
> >   meta_type = fileObject.meta_type
> >   if meta_type == 'DTML Document':
> >     return ("OK", base64.encodestring(fileObject.document_src()))
> >   elif meta_type == 'File':
> >     return ("OK", base64.encodestring(fileObject.data))
> >   elif meta_type == 'Image':
> >     return ("OK", base64.encodestring(fileObject.data))
> >   else:
> >     return ("FAIL", 'Unsupported meta_type')
>
> xmlrpc usage note: instead of encoding and decoding yourself,
> you can use the xmlrpclib.Binary wrapper:
>     if meta_type == 'DTML Document':
>         return ("OK", xmlrpclib.Binary(fileObject.document_src())
>     elif meta_type in ('File', 'Image'):
>         return ("OK", xmlrpclib.Binary(fileObject.data))
>     else:
>         return ("FAIL", 'Unsupported meta_type')
>
> the receiver gets a Binary object as well, and can pick out
> the original string from the data attribute:
>
>     if retval[0] == "OK":
>         return retval[1].data
>

I wasn't familiar with this, I'll look into it.  It may resolve the problem.

> > the calling object of "getFile" does a write to the local filesystem
>
> ...which usually means that you forgot to open the file in
> binary mode:

Often true, but not in this case.  And, as I noted in my original post, I
can verify that the decoded string isn't the right length before it even
gets passed to the calling object.  Anyway, the writing object code is:

  def addFile(self, pathSpec, fileName, contents, timeStamp, title,
description):
    fullFileSpec = os.path.join(pathSpec, fileName)
    file = open(fullFileSpec, "wb")
    file.write(contents)
    file.close()
    os.utime(fullFileSpec, (int(timeStamp), int(timeStamp)))

"contents" ultimately gets passed from the base64 decode, which is wrong
before its object ever passes it anywhere.  But, thanks for the xmlrpclib
tip, I'll test that out.





From joonas at olen.to  Wed Feb 21 22:38:31 2001
From: joonas at olen.to (Joonas Paalasmaa)
Date: Wed, 21 Feb 2001 23:38:31 +0200
Subject: Confused about extending.
Message-ID: <3A943557.B9893C03@olen.to>

I have made the following C-function and now I would like to
use it via Python in that way, that it would
manipulate Python lists smoothedValues and originalValues.
But I am just a beginner with C and I got confused with extending HowTo.
I would be glad if someone would explain how to manipulate the code to
make it extendingable.

Joonas.



void SmoothedList(double *originalValues, double *smoothedValues,
        int listSize, int filterSize)
{
        double tmpResult;
        int i, j;
        int iterations = listSize - filterSize + 1;

        for(i = 0; i < iterations; i++)
        {
                for(j = 0, tmpResult = 0.0; j < filterSize; j++)
                        tmpResult += originalValues[i + j];
                smoothedValues[i] = tmpResult / (double)filterSize;
        }
}


From chris.gonnerman at usa.net  Thu Feb 22 06:02:45 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Wed, 21 Feb 2001 23:02:45 -0600
Subject: Confused about extending.
References: <3A943557.B9893C03@olen.to>
Message-ID: <018b01c09c8c$ba95ae20$a100000a@local>

----- Original Message -----
From: "Joonas Paalasmaa" <joonas at olen.to>
Subject: Confused about extending.


> I have made the following C-function and now I would like to
> use it via Python in that way, that it would
> manipulate Python lists smoothedValues and originalValues.
> But I am just a beginner with C and I got confused with extending HowTo.
> I would be glad if someone would explain how to manipulate the code to
> make it extendingable.
>
> Joonas.

I'm puzzled as to why you have implemented this function in C instead of
Python.  Is it for
speed?  Just how big ARE your lists?

> void SmoothedList(double *originalValues, double *smoothedValues,
>         int listSize, int filterSize)
> {
>         double tmpResult;
>         int i, j;
>         int iterations = listSize - filterSize + 1;
>
>         for(i = 0; i < iterations; i++)
>         {
>                 for(j = 0, tmpResult = 0.0; j < filterSize; j++)
>                         tmpResult += originalValues[i + j];
>                 smoothedValues[i] = tmpResult / (double)filterSize;
>         }
> }

def SmoothedList(originalValues, smoothedValues, listSize, filterSize)
    iterations  =  listSize - filterSize + 1;
    for i in range(0,iterations):
        tmpResult = 0.0
        for j in range(0, filterSize):
            tmpResult  =  tmpResult + originalValues[i + j]
            smoothedValues[i] = tmpResult / filterSize

------------------------------------
Have I missed something here?







From djc at object-craft.com.au  Thu Feb 22 11:58:33 2001
From: djc at object-craft.com.au (Dave Cole)
Date: 22 Feb 2001 21:58:33 +1100
Subject: Confused about extending.
References: <3A943557.B9893C03@olen.to>
Message-ID: <m3u25n56hy.fsf@vole.object-craft.com.au>

>>>>> "Joonas" == Joonas Paalasmaa <joonas at olen.to> writes:

Joonas> I have made the following C-function and now I would like to
Joonas> use it via Python in that way, that it would manipulate Python
Joonas> lists smoothedValues and originalValues.  But I am just a
Joonas> beginner with C and I got confused with extending HowTo.  I
Joonas> would be glad if someone would explain how to manipulate the
Joonas> code to make it extendingable.

I am not sure if you are asking the question that I am answering, but
I recently posted a small extension module which has processing which
is quite similar to what you would need to do in a C extension.

Look back a couple of days for the thread:

        Re: Optimisation problem - in C?

- Dave

-- 
http://www.object-craft.com.au


From joonas at olen.to  Thu Feb 22 14:44:59 2001
From: joonas at olen.to (Joonas Paalasmaa)
Date: Thu, 22 Feb 2001 15:44:59 +0200
Subject: Confused about extending.
References: <3A943557.B9893C03@olen.to> <mailman.982834567.4502.python-list@python.org>
Message-ID: <3A9517DB.8000501@olen.to>

>> I have made the following C-function and now I would like to
>> use it via Python in that way, that it would
>> manipulate Python lists smoothedValues and originalValues.
>> But I am just a beginner with C and I got confused with extending HowTo.
>> I would be glad if someone would explain how to manipulate the code to
>> make it extendingable.
>> 
>> Joonas.
> 
> 
> I'm puzzled as to why you have implemented this function in C instead of
> Python.  Is it for
> speed?  Just how big ARE your lists?

My lists are big so that is the reason why I have to make
the processing with C. But the frontend must be Python and C-module
would be used from Python. For example:

##############
import cListprocessor

list = [1,2,3,4,5,6,7,8,9,15,9,8,7,6,5,4,3,2,1]

processedList = cListprocessor.SmoothedList(list)
##############


- Joonas Paalasmaa

>> void SmoothedList(double *originalValues, double *smoothedValues,
>>         int listSize, int filterSize)
>> {
>>         double tmpResult;
>>         int i, j;
>>         int iterations = listSize - filterSize + 1;
>> 
>>         for(i = 0; i < iterations; i++)
>>         {
>>                 for(j = 0, tmpResult = 0.0; j < filterSize; j++)
>>                         tmpResult += originalValues[i + j];
>>                 smoothedValues[i] = tmpResult / (double)filterSize;
>>         }
>> }
> 
> 
> def SmoothedList(originalValues, smoothedValues, listSize, filterSize)
>     iterations  =  listSize - filterSize + 1;
>     for i in range(0,iterations):
>         tmpResult = 0.0
>         for j in range(0, filterSize):
>             tmpResult  =  tmpResult + originalValues[i + j]
>             smoothedValues[i] = tmpResult / filterSize
> 
> ------------------------------------
> Have I missed something here?




From chris.gonnerman at usa.net  Sat Feb 24 15:30:13 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 24 Feb 2001 08:30:13 -0600
Subject: Confused about extending. (SOLVED)
References: <3A943557.B9893C03@olen.to> <mailman.982834567.4502.python-list@python.org> <3A9517DB.8000501@olen.to> <016801c09d53$5f023340$a100000a@local> <3A96CDD1.7AF41654@olen.to>
Message-ID: <004701c09e6e$5216f000$a100000a@local>

Joonas:

Attached are source files smoothing.c (the Python C module source),
Smoothing.py (the testrig), and joonas.cpp (your source code, marked up
by me).

I am posting this to the list also as the code might be instructive, and
also
in the hopes that if I have made any mistakes someone else might notice
them.

Read the markup in joonas.cpp; I think you have a logic error in your
function, as the last two values in the result list are never assigned by
the
smoothing function but you print them out as if you expect the values to be
meaningful.  (The number of values left unassigned appears to be filterSize
- 1).

The marked-up joonas.cpp (which initializes smoothedList to all zeros)
produces the same values as Smoothing.py with the smoothing.so module,
so I assume my code is correct.

I have been getting a headache reading the Python C API document.
IMHO the question of whether or not the *_SetItem functions "steal" the
reference I give them or not should be explained at least briefly in the
descriptive text of the function itself.  I don't THINK I am leaking
references (and/or memory) in the smoothing.c code but an expert might
be kind enough to double check for me.

Anyway, here they are.  You will need to find the Makefile.pre.in file for
your version of Python, put it in the folder with these files, and do:

    make -f Makefile.pre.in boot
    make

Makefile.pre.in is in /usr/lib/python1.5/config, but your mileage may vary.

If everything is cool you will wind up with smoothing.so (and a lot of other
junk you don't really need).  Look in Smoothing.py for an example of use.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: joonas.cpp
Type: application/octet-stream
Size: 1213 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010224/5e4aa977/attachment.obj>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: smoothing.c
URL: <http://mail.python.org/pipermail/python-list/attachments/20010224/5e4aa977/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Smoothing.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20010224/5e4aa977/attachment.asc>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Setup.in
Type: application/octet-stream
Size: 31 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010224/5e4aa977/attachment-0001.obj>

From s713221 at student.gu.edu.au  Wed Feb 21 22:59:33 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Thu, 22 Feb 2001 07:59:33 +1000
Subject: How does Subclassing Work?
Message-ID: <3A943A45.2E433767@student.gu.edu.au>

Recently I tried to write a piece of code that subclassed Canvas.py from
the Tkinter distribution, where I overwrote CanvasItem. Now I was hoping
that all calls to Arc, Line etc. (which weren't overwritten) would call
my modified CanvasItem, but it seems they were calling the unmodified
CanvasItem - which isn't what I wanted.

What I wanted was:

My module	  ______> calls		CanvasItem
		  |
Canvas.py	Arc			CanvasItem

What I got was:

My module				CanvasItem

Canvas.py	Arc _______> calls	CanvasItem

Other than copy all of Canvas.py into my module and redefine everything,
is there a way I can get this to work as expected without rewriting
Canvas.py? What I was trying to do was alter Canvas so that I could
create classes that would make complex items.

Joal Heagney/AncientHart


From aleaxit at yahoo.com  Thu Feb 22 11:36:12 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 11:36:12 +0100
Subject: How does Subclassing Work?
References: <3A943A45.2E433767@student.gu.edu.au>
Message-ID: <972q2u0ool@news2.newsguy.com>

<s713221 at student.gu.edu.au> wrote in message
news:3A943A45.2E433767 at student.gu.edu.au...
> Recently I tried to write a piece of code that subclassed Canvas.py from
> the Tkinter distribution, where I overwrote CanvasItem. Now I was hoping
> that all calls to Arc, Line etc. (which weren't overwritten) would call
> my modified CanvasItem, but it seems they were calling the unmodified
> CanvasItem - which isn't what I wanted.

It's what I would expect, though; the 'class' statement for (e.g.)
Arc was already executed, right when you imported Canvas, and,
at that time, it got the original CanvasItem class to use as the
base -- it didn't get your subclassed version.

> Other than copy all of Canvas.py into my module and redefine everything,
> is there a way I can get this to work as expected without rewriting
> Canvas.py? What I was trying to do was alter Canvas so that I could
> create classes that would make complex items.

I hesitate to suggest it, but... if you went through all the classes
of interest, currently set to inherit from the original CanvasItem,
and changed that portion of their __bases__ to your modified version;
and also set the CanvasItem attribute of the Canvas module to refer
to your modified version; things would probably work as you want.

Here's a toy example to indicate what I mean...:

== Original.py
class OriginalItem:
    def __init__(self, *whatever):
        print "OriginalItem",whatever

class Derived1(OriginalItem):
    def __init__(self, *whatever):
        print "Derived1",whatever
        apply(OriginalItem.__init__, (self,)+whatever)

class Derived2(OriginalItem):
    def __init__(self, *whatever):
        print "Derived2",whatever
        apply(OriginalItem.__init__, (self,)+whatever)
== end

== Tweak.py
import Original

baseClass = Original.OriginalItem
class TweakedItem(baseClass):
    def __init__(self, *whatever):
        print "TweakedItem",whatever

for key,val in Original.__dict__.items():
    if type(val)==type(baseClass):
        if val is baseClass:
            setattr(Original, key, TweakedItem)
        else:
            if val.__bases__ == (baseClass,):
                val.__bases__ = (TweakedItem,)

x = Original.Derived1("foo","bar")
y = Original.Derived2("bar","baz")
== end

This is not very general (though it would suffice for
your specific need), as it only handles single-base
classes etc, but it's not hard to generalize it along
this axis.

I do wonder about the appropriateness of such "black
magick" to actual production code, though.


Alex





From s713221 at student.gu.edu.au  Fri Feb 23 02:45:31 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Fri, 23 Feb 2001 11:45:31 +1000
Subject: How does Subclassing Work?
References: <3A943A45.2E433767@student.gu.edu.au> <972q2u0ool@news2.newsguy.com>
Message-ID: <3A95C0BB.E3DE90E2@student.gu.edu.au>

Alex Martelli wrote:
> 
> <s713221 at student.gu.edu.au> wrote in message
> news:3A943A45.2E433767 at student.gu.edu.au...
> > Recently I tried to write a piece of code that subclassed Canvas.py from
> > the Tkinter distribution, where I overwrote CanvasItem. Now I was hoping
> > that all calls to Arc, Line etc. (which weren't overwritten) would call
> > my modified CanvasItem, but it seems they were calling the unmodified
> > CanvasItem - which isn't what I wanted.
> 
> It's what I would expect, though; the 'class' statement for (e.g.)
> Arc was already executed, right when you imported Canvas, and,
> at that time, it got the original CanvasItem class to use as the
> base -- it didn't get your subclassed version.
> 
> > Other than copy all of Canvas.py into my module and redefine everything,
> > is there a way I can get this to work as expected without rewriting
> > Canvas.py? What I was trying to do was alter Canvas so that I could
> > create classes that would make complex items.
> 
> I hesitate to suggest it, but... if you went through all the classes
> of interest, currently set to inherit from the original CanvasItem,
> and changed that portion of their __bases__ to your modified version;
> and also set the CanvasItem attribute of the Canvas module to refer
> to your modified version; things would probably work as you want.
> 
> Here's a toy example to indicate what I mean...:
> 
> == Original.py
> class OriginalItem:
>     def __init__(self, *whatever):
>         print "OriginalItem",whatever
> 
> class Derived1(OriginalItem):
>     def __init__(self, *whatever):
>         print "Derived1",whatever
>         apply(OriginalItem.__init__, (self,)+whatever)
> 
> class Derived2(OriginalItem):
>     def __init__(self, *whatever):
>         print "Derived2",whatever
>         apply(OriginalItem.__init__, (self,)+whatever)
> == end
> 
> == Tweak.py
> import Original
> 
> baseClass = Original.OriginalItem
> class TweakedItem(baseClass):
>     def __init__(self, *whatever):
>         print "TweakedItem",whatever
> 
> for key,val in Original.__dict__.items():
>     if type(val)==type(baseClass):
>         if val is baseClass:
>             setattr(Original, key, TweakedItem)
>         else:
>             if val.__bases__ == (baseClass,):
>                 val.__bases__ = (TweakedItem,)
> 
> x = Original.Derived1("foo","bar")
> y = Original.Derived2("bar","baz")
> == end
> 
> This is not very general (though it would suffice for
> your specific need), as it only handles single-base
> classes etc, but it's not hard to generalize it along
> this axis.
> 
> I do wonder about the appropriateness of such "black
> magick" to actual production code, though.
> 
> Alex

To true. Maybe the cut-paste-alter method is better after all.

Joal Heagney/AncientHart


From joonas at olen.to  Wed Feb 21 23:03:31 2001
From: joonas at olen.to (Joonas Paalasmaa)
Date: Thu, 22 Feb 2001 00:03:31 +0200
Subject: Problems using Grail.
Message-ID: <3A943B33.CD618A9B@olen.to>

When I tried to start Grail by typing "python grail-0.6\grail.py"
Python gave the following error message.
Python version is 1.6 and platform is Windows95.
What could be the problem?

- Joonas

Traceback (innermost last):
  File "C:\Python16\grail-0.6\grail.py", line 499, in ?
    main()
  File "C:\Python16\grail-0.6\grail.py", line 108, in main
    app = Application(prefs=prefs, display=display)
  File "C:\Python16\grail-0.6\grail.py", line 248, in __init__
    self.stylesheet = Stylesheet.Stylesheet(self.prefs)
  File "C:\Python16\grail-0.6\Stylesheet.py", line 21, in __init__
    self.load()
  File "C:\Python16\grail-0.6\Stylesheet.py", line 45, in load
    massaged.append((g, c), v % fparms_dict)
TypeError: append requires exactly 1 argument; 2 given


From grante at visi.com  Thu Feb 22 00:51:54 2001
From: grante at visi.com (Grant Edwards)
Date: Wed, 21 Feb 2001 23:51:54 GMT
Subject: Problems using Grail.
References: <3A943B33.CD618A9B@olen.to>
Message-ID: <slrn998l4p.cbp.grante@isis.visi.com>

In article <3A943B33.CD618A9B at olen.to>, Joonas Paalasmaa wrote:

>When I tried to start Grail by typing "python grail-0.6\grail.py"
>Python gave the following error message.
>Python version is 1.6 and platform is Windows95.
>What could be the problem?
>
>- Joonas
>
>Traceback (innermost last):
>  File "C:\Python16\grail-0.6\grail.py", line 499, in ?
>    main()
>  File "C:\Python16\grail-0.6\grail.py", line 108, in main
>    app = Application(prefs=prefs, display=display)
>  File "C:\Python16\grail-0.6\grail.py", line 248, in __init__
>    self.stylesheet = Stylesheet.Stylesheet(self.prefs)
>  File "C:\Python16\grail-0.6\Stylesheet.py", line 21, in __init__
>    self.load()
>  File "C:\Python16\grail-0.6\Stylesheet.py", line 45, in load
>    massaged.append((g, c), v % fparms_dict)
>TypeError: append requires exactly 1 argument; 2 given

Grail is pretty old, and there have been language changes in 1.6 and later
that are causing problems.  The append() method for lists has changed:

1.5.2:

>>> print x
[1, 2, 3, 4]
>>> x.append(5,6,7)
>>> print x
[1, 2, 3, 4, (5, 6, 7)]

Newsgroups: comp.lang.python  
Subject: Re: Problems using Grail.  
References: <3A943B33.CD618A9B at olen.to>  
Organization: Vector Internet Services, Inc.  
Followup-To:  
  
In article <3A943B33.CD618A9B at olen.to>, Joonas Paalasmaa wrote:  
  
>When I tried to start Grail by typing "python grail-0.6\grail.py"  
>Python gave the following error message.  
>Python version is 1.6 and platform is Windows95.  
>What could be the problem?  
>  
>- Joonas  
>  
>Traceback (innermost last):  
>  File "C:\Python16\grail-0.6\grail.py", line 499, in ?  
>    main()  
>  File "C:\Python16\grail-0.6\grail.py", line 108, in main  
>    app = Application(prefs=prefs, display=display)  
>  File "C:\Python16\grail-0.6\grail.py", line 248, in __init__  
>    self.stylesheet = Stylesheet.Stylesheet(self.prefs)  
>  File "C:\Python16\grail-0.6\Stylesheet.py", line 21, in __init__  
>    self.load()  
>  File "C:\Python16\grail-0.6\Stylesheet.py", line 45, in load  
>    massaged.append((g, c), v % fparms_dict)  
>TypeError: append requires exactly 1 argument; 2 given  
  
Grail is pretty old, and there have been language changes in 1.6 and later
that are causing problems. (I never got Grail to run under 1.5.2 either.)

In your example, it's the append() method for lists has changed:
  
Python 1.5.2:

>>> print x
[1, 2, 3, 4]
>>> x.append(5,6,7)
>>> print x
[1, 2, 3, 4, (5, 6, 7)]

python 2.0

>>> print x
[1, 2, 3, 4]
>>> x.append(5,6,7)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: append requires exactly 1 argument; 3 given

>>> x.append((5,6,7))
>>> print x
[1, 2, 3, 4, (5, 6, 7)]

-- 
Grant Edwards                   grante             Yow!  ... I want a COLOR
                                  at               T.V. and a VIBRATING BED!!!
                               visi.com            


From hamish_lawson at yahoo.co.uk  Thu Feb 22 10:22:46 2001
From: hamish_lawson at yahoo.co.uk (hamish_lawson at yahoo.co.uk)
Date: Thu, 22 Feb 2001 09:22:46 -0000
Subject: Problems using Grail.
In-Reply-To: <3A943B33.CD618A9B@olen.to>
Message-ID: <972lp6+iiqq@eGroups.com>

Joonas Paalasmaa <joonas at o...> wrote:

> When I tried to start Grail by typing "python grail-0.6\grail.py"
> Python gave the following error message.

>   File "C:\Python16\grail-0.6\Stylesheet.py", line 45, in load
>     massaged.append((g, c), v % fparms_dict)
> TypeError: append requires exactly 1 argument; 2 given

The append() call is officially meant to take one argument, but in 
versions of Python before 1.6, multiple arguments would be understood 
as comprising a tuple. However this behaviour was never officially 
documented or approved, and was dropped in Python 1.6 to bring the 
actual behaviour in line with the official documentation. This line 
of code will therefore need to be modified to use an explicit tuple 
in the append argument:

    massaged.append(((g, c), v % fparms_dict))

Hamish Lawson





From costas at springmail.com  Wed Feb 21 23:08:38 2001
From: costas at springmail.com (costas at springmail.com)
Date: Wed, 21 Feb 2001 22:08:38 GMT
Subject: import statement is case sensitive
Message-ID: <3a943bea.525315589@News.CIS.DFN.DE>

Is there anyway to make the import statement NOT case sensitive?
sometimes directory names are in mixed case or different case from the
import statment.

Also what is the purpose of even having it case sensitive. I've yet to
see a file system that has case sensitive filenames....

costas


From fredrik at pythonware.com  Wed Feb 21 23:20:13 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 22:20:13 GMT
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE>
Message-ID: <xaXk6.14892$Qb7.2377523@newsb.telia.net>

Costas wrote:
> Is there anyway to make the import statement NOT case sensitive?
> sometimes directory names are in mixed case or different case from the
> import statment.

try:

    > python -h

and read the help text.

or read this article:

    http://python.sourceforge.net/peps/pep-0235.html

or rename the directories; even if your operating system is
case-insensitive, it's probably not case ignorant.

> Also what is the purpose of even having it case sensitive. I've yet to
> see a file system that has case sensitive filenames....

time to get a real operating systems? ;-)

Cheers /F




From costas at meezon.com  Thu Feb 22 04:31:39 2001
From: costas at meezon.com (Costas Menico)
Date: Thu, 22 Feb 2001 03:31:39 GMT
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <xaXk6.14892$Qb7.2377523@newsb.telia.net>
Message-ID: <3a94846a.3272811@News.CIS.DFN.DE>

Interesting article. However the PYTHONCASEOK as suggested in the
article does not seem to work. I set it to PYTHONCASEOK = 1 and import
still insists on being case-sensitive.

Also python -h has nothing helpful in this respect. Any other ideas
barring changing OS and renaming directories would be appreciated.

Thanks

Costas

>Costas wrote:
>> Is there anyway to make the import statement NOT case sensitive?
>> sometimes directory names are in mixed case or different case from the
>> import statment.
>
>try:
>
>    > python -h
>
>and read the help text.
>
>or read this article:
>
>    http://python.sourceforge.net/peps/pep-0235.html
>
>or rename the directories; even if your operating system is
>case-insensitive, it's probably not case ignorant.
>
>> Also what is the purpose of even having it case sensitive. I've yet to
>> see a file system that has case sensitive filenames....
>
>time to get a real operating systems? ;-)
>
>Cheers /F
>
>



From ljohnson at resgen.com  Thu Feb 22 00:05:31 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Wed, 21 Feb 2001 17:05:31 -0600
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE>
Message-ID: <t98i9hsgjkov26@corp.supernews.com>

> Also what is the purpose of even having it case sensitive. I've yet to
> see a file system that has case sensitive filenames....

There is an operating system called "Unix" that has case-sensitive file
names. There is a slight chance that others on this newsgroup have also
heard of this obscure operating system.




From costas at meezon.com  Thu Feb 22 03:58:29 2001
From: costas at meezon.com (Costas Menico)
Date: Thu, 22 Feb 2001 02:58:29 GMT
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <t98i9hsgjkov26@corp.supernews.com>
Message-ID: <3a947ac4.802440@News.CIS.DFN.DE>

"Lyle Johnson" <ljohnson at resgen.com> wrote:

>> Also what is the purpose of even having it case sensitive. I've yet to
>> see a file system that has case sensitive filenames....
>
>There is an operating system called "Unix" that has case-sensitive file
>names. There is a slight chance that others on this newsgroup have also
>heard of this obscure operating system.
>

Well given the fact that Linux and Windows is beating it to death I
can see why Unix is becoming obscure :)

Actually I would consider this a shortcoming of Unix. Shipping a
product that depends on the directory names and files being in the
right case must cause all sorts of headaches.

I am trying to run an an application under Windows and now I have to
worry if my directory name is in the correct case. I use PythonWin (a
fantastic product) to run and test it. There is no option for telling
it to ignore the case of the import filenames that I could see.

And renaming directories and files is not an easy option.. But anyway,
this is should automatically be handled in Python depending on the OS.


Why would asnyone make the same filenames with different cases? Is
there a good use?


From aleaxit at yahoo.com  Thu Feb 22 09:56:06 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 09:56:06 +0100
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <t98i9hsgjkov26@corp.supernews.com> <3a947ac4.802440@News.CIS.DFN.DE>
Message-ID: <972k770ik7@news2.newsguy.com>

"Costas Menico" <costas at meezon.com> wrote in message
news:3a947ac4.802440 at News.CIS.DFN.DE...
> "Lyle Johnson" <ljohnson at resgen.com> wrote:
>
> >> Also what is the purpose of even having it case sensitive. I've yet to
> >> see a file system that has case sensitive filenames....
> >
> >There is an operating system called "Unix" that has case-sensitive file
> >names. There is a slight chance that others on this newsgroup have also
> >heard of this obscure operating system.
>
> Well given the fact that Linux and Windows is beating it to death I
> can see why Unix is becoming obscure :)

Linux is identical to Unix in this respect (and in most others; it
IS a version of Unix in all respects except trademarks &c).


> Actually I would consider this a shortcoming of Unix. Shipping a
> product that depends on the directory names and files being in the
> right case must cause all sorts of headaches.

Debatable, although I would tend to agree.  Internationalization
issues make "case-insensitive" a big problem, though -- the uppercase
equivalent of, say, "e with an acute accent", depends so much
on the character-encoding in use.

I consider it a similar bug (and Windows and Mac share it with
Unix) to allow spaces, tabs, and other whitespace characters in
filenames, by the way.  The driving idea behind all such things
is, I guess, to make it possible for the user to name his or
her file as he/she wants, with entire sentences, etc; I do NOT
like this, I'd rather see filenames as "identifiers" for the
files.  But I guess this marks me as a dinosaur, once again.

> And renaming directories and files is not an easy option.. But anyway,
> this is should automatically be handled in Python depending on the OS.

In Windows, you cannot _directly_ rename a file or directory
to change its case (a Windows bug, I'd say); you have to
    rename Foo tempname
    rename tempname foo
because trying to directly do
    rename Foo foo
will fail.

It _would_ definitely be easier for the user if Python 'knew'
a given filesystem is case-insensitive and compensated for that.


> Why would asnyone make the same filenames with different cases? Is
> there a good use?

There are a few uses (e.g., some C++ compilers take xx.c as a file
in the C languace, xx.C as one in the C++ language) but I would not,
personally, call them "good".


Alex





From qrczak at knm.org.pl  Thu Feb 22 17:39:21 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Feb 2001 16:39:21 GMT
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <t98i9hsgjkov26@corp.supernews.com> <3a947ac4.802440@News.CIS.DFN.DE> <972k770ik7@news2.newsguy.com>
Message-ID: <slrn99ag5p.gsa.qrczak@qrnik.zagroda>

Thu, 22 Feb 2001 09:56:06 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:

> I consider it a similar bug (and Windows and Mac share it with Unix)
> to allow spaces, tabs, and other whitespace characters in filenames,
> by the way.  The driving idea behind all such things is, I guess,
> to make it possible for the user to name his or her file as he/she
> wants, with entire sentences, etc; I do NOT like this, I'd rather
> see filenames as "identifiers" for the files.

Actually the Unix tradition is to use mostly identifier-like filenames,
despite the fact that all characters but '\0' and '/' are legal,
and that filenames can be passed to programs intact (command line
arguments are physically passed separately, not split by applications
themselves as in DOS/Windows, and quoting filenames with spaces is
only the shell's business).

Since Windows "invented" long filenames in 1995 with a new set of
rules which characters are legal and lots of backwards compatibility
headaches wrt. charsets and case in short and long filenames, in
practice it and its programs really make use of spaces and other
characters in filenames.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From tim.one at home.com  Thu Feb 22 23:33:29 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 22 Feb 2001 17:33:29 -0500
Subject: import statement is case sensitive
In-Reply-To: <972k770ik7@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEFGJBAA.tim.one@home.com>

[Alex Martelli]
> ...
> In Windows, you cannot _directly_ rename a file or directory
> to change its case (a Windows bug, I'd say); you have to
>     rename Foo tempname
>     rename tempname foo
> because trying to directly do
>     rename Foo foo
> will fail.

Here under Win98SE:

>dir/b fo*

>md Foo
>dir/b fo*
Foo

>ren Foo foo
>dir/b fo*
foo

>rd foo
>echo blah > FOO
>dir/b fo*
FOO

>ren FOO fOo
>dir/b fo*
fOo

>
So, works fine for me, whether it's a directory or a file.  Certainly
possible that this varies across Windows flavors.




From db3l at fitlinxx.com  Sat Feb 24 01:26:12 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 23 Feb 2001 19:26:12 -0500
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <t98i9hsgjkov26@corp.supernews.com> <3a947ac4.802440@News.CIS.DFN.DE> <972k770ik7@news2.newsguy.com>
Message-ID: <un1bd7wpn.fsf@ctwd0143.fitlinxx.com>

"Alex Martelli" <aleaxit at yahoo.com> writes:

> In Windows, you cannot _directly_ rename a file or directory
> to change its case (a Windows bug, I'd say); you have to
>     rename Foo tempname
>     rename tempname foo
> because trying to directly do
>     rename Foo foo
> will fail.

Actually, I think that depends on filesystem (and perhaps OS).  On my
NT4 system with NTFS I can rename just to change case without a
problem.

But I'm pretty sure it fails on a FAT filesystem since at some level in
the directory structure the name is uppercase in either case.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From chris.gonnerman at usa.net  Thu Feb 22 14:49:58 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 22 Feb 2001 07:49:58 -0600
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <t98i9hsgjkov26@corp.supernews.com> <3a947ac4.802440@News.CIS.DFN.DE>
Message-ID: <002b01c09cd6$59c522c0$a100000a@local>

----- Original Message -----
From: "Costas Menico" <costas at meezon.com>
Subject: Re: import statement is case sensitive
> Actually I would consider this a shortcoming of Unix. Shipping a
> product that depends on the directory names and files being in the
> right case must cause all sorts of headaches.

Not really.  You are forgetting that the difficulty in getting file and
directory names in the right case in Windows simply doesn't trouble
Unixoid OS's.

> I am trying to run an an application under Windows and now I have to
> worry if my directory name is in the correct case. I use PythonWin (a
> fantastic product) to run and test it. There is no option for telling
> it to ignore the case of the import filenames that I could see.

When you save files in PythonWin it understands it must get the case
right, so I'd be surprised if you type them right if it doesn't save them
right.

> And renaming directories and files is not an easy option.. But anyway,
> this is should automatically be handled in Python depending on the OS.

I use MinGW32 tools on my computer; others like CygWin.  Both provide
Unix-like commands for the Windows environment, and either will make
your problems easier to handle.

> Why would asnyone make the same filenames with different cases? Is
> there a good use?

I don't know if it is a GOOD use, but the Python GTK bindings do that with
a module named gtk.py and one named GTK.py (the first is the binding module,
the second is it's constants) and this causes problems on Windows.  The most
common port of this code to Windows renames the second file GTKconst.py,
to avoid the name collision.

Anyway, many of the standard modules are named in mixed case.  Windows may
be the most popular desktop OS in the world, but in the realm of open-source
development it's often considered a (broken) niche environment.






From mikael at isy.liu.se  Thu Feb 22 14:50:18 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 22 Feb 2001 14:50:18 +0100 (MET)
Subject: import statement is case sensitive
In-Reply-To: <3a947ac4.802440@News.CIS.DFN.DE>
Message-ID: <XFMail.010222145018.mikael@isy.liu.se>

On 22-Feb-01 Costas Menico wrote:
 >  Actually I would consider this a shortcoming of Unix. Shipping a
 >  product that depends on the directory names and files being in the
 >  right case must cause all sorts of headaches.

Perhaps so, but...

I'm still upset over the fact that my Windows box at home doesn't allow
me to call a directory "FTP", but changes that to "Ftp" as soon as I 
hit return. Perhaps there is some setting that I can change, but as a 
default setting it sucks, IMO.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    22-Feb-01
Time:    14:45:38

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From aleaxit at yahoo.com  Thu Feb 22 18:00:16 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 18:00:16 +0100
Subject: import statement is case sensitive
References: <XFMail.010222145018.mikael@isy.liu.se>
Message-ID: <973gis01f0u@news2.newsguy.com>

"Mikael Olofsson" <mikael at isy.liu.se> wrote in message
news:XFMail.010222145018.mikael at isy.liu.se...

On 22-Feb-01 Costas Menico wrote:
 >  Actually I would consider this a shortcoming of Unix. Shipping a
 >  product that depends on the directory names and files being in the
 >  right case must cause all sorts of headaches.

Perhaps so, but...

I'm still upset over the fact that my Windows box at home doesn't allow
me to call a directory "FTP", but changes that to "Ftp" as soon as I
hit return. Perhaps there is some setting that I can change, but as a
default setting it sucks, IMO.


I suspect your Windows box DOES allow what you say it doesn't --
what may currently disallow creation (or display of) all-uppercase
names is more likely to be the shell GUI front-end (and, yes, it
IS just a checkbox you can turn off somewhere).

To check this, use any method whatsoever _except_ the shell GUI
front-ends to create and display your all-uppercase names (e.g.,
a "Command Prompt" aka "DOS Box" should do).  If, as I believe,
this will work just fine, then maybe you'll see it's not "the
Windows box" that you're railing against here, just a specific
cosmetic choice of its shell GUI front-end.


Alex





From ben.hutchings at roundpoint.com  Fri Feb 23 01:46:34 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 22 Feb 2001 16:46:34 -0800
Subject: import statement is case sensitive
References: <XFMail.010222145018.mikael@isy.liu.se> <973gis01f0u@news2.newsguy.com>
Message-ID: <ulmqykyz9.fsf@roundpoint.com>

"Alex Martelli" <aleaxit at yahoo.com> writes:

> "Mikael Olofsson" <mikael at isy.liu.se> wrote in message
> news:XFMail.010222145018.mikael at isy.liu.se...
> >
> > On 22-Feb-01 Costas Menico wrote:
> > >  Actually I would consider this a shortcoming of Unix. Shipping a
> > >  product that depends on the directory names and files being in the
> > >  right case must cause all sorts of headaches.
> >
> > Perhaps so, but...
> >
> > I'm still upset over the fact that my Windows box at home doesn't allow
> > me to call a directory "FTP", but changes that to "Ftp" as soon as I
> > hit return. Perhaps there is some setting that I can change, but as a
> > default setting it sucks, IMO.
> 
> 
> I suspect your Windows box DOES allow what you say it doesn't --
> what may currently disallow creation (or display of) all-uppercase
> names is more likely to be the shell GUI front-end (and, yes, it
> IS just a checkbox you can turn off somewhere).
<snip>

Correct.  Windows Explorer (or the underlying folder viewing widget)
does this to any file that doesn't have a "long filename" distinct
from its DOS-compatible file name, but at least some implementations
of the VFAT file-system do not store a separate "long filename" if
the chosen name is already DOS-compatible.

A file named, for example "CAPITALS WITH SPACES" will have that stored
as its "long filename" and something like "CAPITA~1" for DOS
compatibility.  Its name will be displayed correctly.  A file named
"CAPITALS" will have just that DOS-compatible name stored, and will
be displayed as "Capitals" by default.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From tim.one at home.com  Thu Feb 22 23:18:55 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 22 Feb 2001 17:18:55 -0500
Subject: import statement is case sensitive
In-Reply-To: <XFMail.010222145018.mikael@isy.liu.se>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEFEJBAA.tim.one@home.com>

[Mikael Olofsson]
> I'm still upset over the fact that my Windows box at home doesn't allow
> me to call a directory "FTP", but changes that to "Ftp" as soon as I
> hit return. Perhaps there is some setting that I can change, but as a
> default setting it sucks, IMO.

Start -> Settings -> Folder Options ... -> View

then check "Allow all uppercase names".  You'll probably be sorry you did
this, though (it makes stuff considerably harder to read, IMO).  Note that
this has no effect on the filename actually stored, it only affects how
ALLCAPS names get *displayed*, and even so just how Explorer displays them.

As long as you're there, check "Display the full path in title bar", uncheck
"Hide file extensions for known file types" and select "Show all files"
under "Hidden files" (some of these are named a little differently under
different flavors of Windows).  The default settings are optimized for end
users, not developers.  And like it or not, they probably picked the best
default values for end users in all respects.

pure-end-users-don't-care-about-the-same-things-you-do-ly y'rs  - tim




From tanzer at swing.co.at  Fri Feb 23 08:11:58 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Fri, 23 Feb 2001 08:11:58 +0100
Subject: import statement is case sensitive 
In-Reply-To: Your message of "Thu, 22 Feb 2001 17:18:55 EST."
             <LNBBLJKPBEHFEDALKOLCEEFEJBAA.tim.one@home.com> 
Message-ID: <m14WCOU-000wcEC@swing.co.at>

"Tim Peters" <tim.one at home.com> wrote:

> The default settings are optimized for end users, not developers. And
> like it or not, they probably picked the best default values for end
> users in all respects.
> 
> pure-end-users-don't-care-about-the-same-things-you-do-ly y'rs  - tim

In all respects but "safety" against social engineering by viruses
traveling under `.TXT.vbs' extensions <wink>.

PS: I'll readily admit that it might not matter much whether the
    `.vbs' of ILOVEYOU is shown or not.

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From mikael at isy.liu.se  Fri Feb 23 10:28:43 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 23 Feb 2001 10:28:43 +0100 (MET)
Subject: import statement is case sensitive
In-Reply-To: <LNBBLJKPBEHFEDALKOLCEEFEJBAA.tim.one@home.com>
Message-ID: <XFMail.010223102843.mikael@isy.liu.se>

On 22-Feb-01 Tim Peters wrote:
 >  [Mikael Olofsson]
 > > I'm still upset over the fact that my Windows box at home doesn't allow
 > > me to call a directory "FTP", but changes that to "Ftp" as soon as I
 > > hit return. Perhaps there is some setting that I can change, but as a
 > > default setting it sucks, IMO.
 >  
 >  Start -> Settings -> Folder Options ... -> View
 >  
 >  [snip details]

Thanks, I knew there was a way.

 >  pure-end-users-don't-care-about-the-same-things-you-do-ly y'rs  - tim

<OT level=10>
In this case, I consider myself an end user. I'm just brought up with
Unices and old Macs. None of those change things for me (as an end user)
without asking (at least I haven't notised any such cases). Today's M$ 
software (probably on Mac aswell as on Windows), however, always have 
default settings that automates everything that I do not want to have 
automated. You have probably tried M$ Word with all its default 
settings. It's like a freakin' disco. Hell, I want my computer to do 
what I tell it to do. If something is to be automated I want it to be 
automated only when and if I say so. 

Sigh, I probably should change to Linux... or dump my computer... I still
have my typewriter and my slide-rule somewhere

When will there be a computer that allows me to sit down and just do my
work? Not start applications, change file formats to comply with other
applications or other versions of the same application, change OS 
settings just because I need a certain application (read: help my kids 
start games that only works with 256 colors), change stupid default
settings, and all that jazz... 

I-guess-I'm-just-too-old-for-this-ly y'rs
</OT>

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    23-Feb-01
Time:    09:55:11

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From qrczak at knm.org.pl  Thu Feb 22 17:32:01 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 22 Feb 2001 16:32:01 GMT
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <t98i9hsgjkov26@corp.supernews.com> <3a947ac4.802440@News.CIS.DFN.DE>
Message-ID: <slrn99afo0.gsa.qrczak@qrnik.zagroda>

Thu, 22 Feb 2001 02:58:29 GMT, Costas Menico <costas at meezon.com> pisze:

> Actually I would consider this a shortcoming of Unix. Shipping a
> product that depends on the directory names and files being in the
> right case must cause all sorts of headaches.

Not at all. I never caused me any trouble. Case of filenames doesn't
change by accident, as filenames themselves don't change. Why would
a directory be in a wrong case?!

The reality is that this is Windows which has problems with filenames.
Unix has very simple filename rules; Windows has very complicated
and poorly implemented rules. Sorry for repeating myself from 16 Feb,
but you asked for that:

For example when you try to create a directory called Spam, Windows
Explorer will create SPAM and display it as Spam. And when asked to
create SPAM, it will create it but display Spam. It behaves correctly
when the name has more than 8 characters or when a character outside
ASCII is used.

This is Win95 - I don't know if they fixed this in later versions.

Worse: using the Explorer you can create a filename with some non-ASCII
characters (e.g. <<French quotes>> or bullet), which will be silently
converted to other characters, perhaps different in long and short
names (in these cases: '<<' and '>>' character pairs, and '\7' -
a control character which displays as bullet in DOS). Details depend
on the default codepages (the above is for CP-1250 and CP-852). This
can make the file inaccessible under either Windows or DOS or both
(because the converted character is sometimes not legal in a filename,
but the filename was checked for correctness before conversion). You
can't delete or rename such file from either Windows or DOS or both,
until you repair the filesystem. Scandisk is not always able to do it.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From tim.one at home.com  Thu Feb 22 01:35:13 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 21 Feb 2001 19:35:13 -0500
Subject: import statement is case sensitive
In-Reply-To: <3a943bea.525315589@News.CIS.DFN.DE>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEBMJBAA.tim.one@home.com>

> Is there anyway to make the import statement NOT case sensitive?

Assuming you're running some flavor of Windows, do

set PYTHONCASEOK=1

at a DOS prompt, autoexec.bat, or under NT/2000 use the Control Panel ->
System thingie.




From costas at meezon.com  Thu Feb 22 20:51:17 2001
From: costas at meezon.com (Costas Menico)
Date: Thu, 22 Feb 2001 19:51:17 GMT
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE>
Message-ID: <3a956af6.17279998@news.cis.dfn.de>

Just wanted to let you know what I have found so far using Python
under Windows/DOS.

1) I can get PythonWin import to be case-insensitive but only if I do
an os.environ["PYTHONCASEOK"]="1" before the import (but after I
import os in the correct case :)

2) I tried setting it at the DOS prompt (SET PYTHONCASEOK=1) but
python.exe seems to ignore it.

3) You can set View/Folder Options/View in Windows Explorer to allow
all upper case filenames. But it really looks screwed up since many
directories are in upper case. Very difficult to read.

4) python -h   scrolls off the screen and there is no way to either
redirect the output nor pipe it to MORE.EXE.I had to find the help in
the book.

The original problem i was faced had to do with mxODBC. He imports
from ODBC (upper case) directories in his sample code. I know I can
change his code but I this is the stuff that comes back to bite.. 

Anyway I found a solution around the case issue and for now I have to
deal with Tkinter and not working with PythonWin. Works with
python.exe. 

Does anyone have ideas why it doesnt work? 

Thanks for the help

Costas Menico



From tim.one at home.com  Thu Feb 22 23:27:11 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 22 Feb 2001 17:27:11 -0500
Subject: import statement is case sensitive
In-Reply-To: <3a956af6.17279998@news.cis.dfn.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEFFJBAA.tim.one@home.com>

[Costas Menico]
> Just wanted to let you know what I have found so far using Python
> under Windows/DOS.
>
> 1) I can get PythonWin import to be case-insensitive but only if I do
> an os.environ["PYTHONCASEOK"]="1" before the import (but after I
> import os in the correct case :)
>
> 2) I tried setting it at the DOS prompt (SET PYTHONCASEOK=1) but
> python.exe seems to ignore it.

Then you're not doing it correctly.  Explain every step in detail, else
nobody can guess which part you're missing.  Setting PYTHONCASEOK in some
random DOS box won't have any effect on PythonWin, if that's your complaint.

> 3) You can set View/Folder Options/View in Windows Explorer to allow
> all upper case filenames. But it really looks screwed up since many
> directories are in upper case. Very difficult to read.

Note that this option only affects how Explorer *displays* names.  It has no
effect on how they're stored.  BTW, I agree that this option makes things
harder to read, and I don't use it myself either.

> 4) python -h   scrolls off the screen and there is no way to either
> redirect the output nor pipe it to MORE.EXE.I had to find the help in
> the book.

Your DOS box is too small, then.  Right-click on the icon you use to launch
the DOS box, select Properties, then select the Screen tab, and select "50
lines" from the "Initial size" dropdown list (assuming you're running Win9X;
NT/2000 have more flexible options).

> The original problem i was faced had to do with mxODBC. He imports
> from ODBC (upper case) directories in his sample code. I know I can
> change his code but I this is the stuff that comes back to bite..
>
> Anyway I found a solution around the case issue and for now I have to
> deal with Tkinter and not working with PythonWin. Works with
> python.exe.
>
> Does anyone have ideas why it doesnt work?

Sorry, don't know to what "it" refers at this point.




From db3l at fitlinxx.com  Sat Feb 24 01:34:25 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 23 Feb 2001 19:34:25 -0500
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE> <3a956af6.17279998@news.cis.dfn.de>
Message-ID: <uk86g9awe.fsf@ctwd0143.fitlinxx.com>

costas at meezon.com (Costas Menico) writes:

> 1) I can get PythonWin import to be case-insensitive but only if I do
> an os.environ["PYTHONCASEOK"]="1" before the import (but after I
> import os in the correct case :)

Or just set PYTHONCASEOK in the environment from which you start
Pythonwin or your scripts.

> 2) I tried setting it at the DOS prompt (SET PYTHONCASEOK=1) but
> python.exe seems to ignore it.

That should work - are you positive you're running python under that
same environment (e.g., you can't set that variable in a DOS prompt,
then exit that shell, and restart another to run Python in).

> 3) You can set View/Folder Options/View in Windows Explorer to allow
> all upper case filenames. But it really looks screwed up since many
> directories are in upper case. Very difficult to read.

And that's actually what your filesystem contents contains - ugly
isn't it?  The whole mess with the "pretty" settings in stuff like
Explorer is just to hide how badly some of the older filesystem types
deal (or don't deal) with case.

> 4) python -h   scrolls off the screen and there is no way to either
> redirect the output nor pipe it to MORE.EXE.I had to find the help in
> the book.

This is a little tricky - the output goes to stderr rather than stdout
(the error output stream) which isn't normally included in your pipe.
You could fix this by routing stderr to stdout, as in:

    python -h 2>&1 | more

Of course, the output isn't all that big, so you only miss a few
lines.  Using something like "mode 80,30" to make a bigger window
would work as well.

Then again, the help doesn't mention PYTHONCASEOK so it's not going to
be too much help in that regard :-)

> The original problem i was faced had to do with mxODBC. He imports
> from ODBC (upper case) directories in his sample code. I know I can
> change his code but I this is the stuff that comes back to bite.. 

But uppercase is correct - so why would you want to change the imports
to lowercase (I'm guessing that's what you are trying?)

Note that one source of confusion is the existence of an "odbc.pyd"
module in the win32all distributions.  You should delete or rename
that if you're going to be using mxODBC or you might not get the
module you want (or might get a complaint about "import ODBC" if
Python finds the win32all "odbc" module first).

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From jurgen.defurne at philips.com  Fri Feb 23 07:57:50 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Fri, 23 Feb 2001 07:57:50 +0100
Subject: import statement is case sensitive
Message-ID: <0056900016131369000002L092*@MHS>

Welcome to the world in which Bill thinks he is the boss.

Jurgen




mikael at isy.liu.se@SMTP at python.org on 22/02/2001 15:21:39
Sent by:	python-list-admin at python.org
To:	python-list at python.org@SMTP
cc:	 
Subject:	Re: import statement is case sensitive
Classification:	


On 22-Feb-01 Costas Menico wrote:
 >  Actually I would consider this a shortcoming of Unix. Shipping a
 >  product that depends on the directory names and files being in the
 >  right case must cause all sorts of headaches.

Perhaps so, but...

I'm still upset over the fact that my Windows box at home doesn't allow
me to call a directory "FTP", but changes that to "Ftp" as soon as I=20
hit return. Perhaps there is some setting that I can change, but as a=20
default setting it sucks, IMO.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael              =20
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    22-Feb-01
Time:    14:45:38

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------
--
http://mail.python.org/mailman/listinfo/python-list





From sholden at holdenweb.com  Wed Feb 28 18:13:51 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 28 Feb 2001 12:13:51 -0500
Subject: import statement is case sensitive
References: <3a943bea.525315589@News.CIS.DFN.DE>
Message-ID: <5pan6.4921$zL.241848@e420r-atl1.usenetserver.com>

<costas at springmail.com> wrote in message
news:3a943bea.525315589 at News.CIS.DFN.DE...
> Is there anyway to make the import statement NOT case sensitive?
> sometimes directory names are in mixed case or different case from the
> import statment.
>
> Also what is the purpose of even having it case sensitive. I've yet to
> see a file system that has case sensitive filenames....
>
You might as well ask what is the purpose of having a filesystem retain the
case of filenames, but then ignore the case when looking them up, which is
what NT appears to do. Ever had two files in an NT filesystem with names
differing only in case? There are circumstances under which this can happen.
That's one good reason to assume that the user's case differences are
intended to be significant!

regards
 Steve





From rob at netzilient.com  Wed Feb 21 23:21:04 2001
From: rob at netzilient.com (Rob Eden)
Date: Wed, 21 Feb 2001 16:21:04 -0600
Subject: Interrupting blocking reads
Message-ID: <3a943dcb@news.advancenet.net>

I have an application that is reading from a number of different files
(well, external process' output streams to be precise). I need to be able to
interrupt the threads, but I don't see a way to do this other than with a
signal.  The problem is, I want to be able to interrupt an individual thread
while possibly leaving the others running.

So can you answer any of these questions:

  1) Is there a way to interrupt a blocking file read other than with a
signal?
  2) Is there a non-blocking file read?  read(0) seems to return no data.
     But it returns right away! (This is useful why?)
  3) Is there a way to get signals to work with multiple threads?
  4) Any other solutions to the problem?

Thanks!


Rob Eden




From spahievi at vega.bg  Wed Feb 21 23:35:54 2001
From: spahievi at vega.bg (Niki Spahiev)
Date: Thu, 22 Feb 2001 00:35:54 +0200
Subject: fileobj.write(buffer(...))
Message-ID: <1044353420.20010222003554@vega.bg>

Is this working as expected?

>>> open( 'test', 'w' ).write( buffer( 'abcd', 2 ) )
>>> from cStringIO import StringIO
>>> StringIO().write( buffer( 'abcd', 2 ) )
Traceback (innermost last):
  File "<pyshell#4>", line 1, in ?
    StringIO().write( buffer( 'abcd', 2 ) )
SystemError: bad argument to internal function
>>>   

-- 
Best regards,
 Niki Spahiev





From lisowski.tomasz at sssa.NOSPAM.com.pl  Thu Feb 22 09:33:53 2001
From: lisowski.tomasz at sssa.NOSPAM.com.pl (Tomasz Lisowski)
Date: Thu, 22 Feb 2001 09:33:53 +0100
Subject: fileobj.write(buffer(...))
References: <mailman.982795168.21257.python-list@python.org>
Message-ID: <972ip3$juj$1@news.tpi.pl>

Uzytkownik "Niki Spahiev" <spahievi at vega.bg> napisal w wiadomosci
news:mailman.982795168.21257.python-list at python.org...
> Is this working as expected?
>
> >>> open( 'test', 'w' ).write( buffer( 'abcd', 2 ) )
> >>> from cStringIO import StringIO
> >>> StringIO().write( buffer( 'abcd', 2 ) )
> Traceback (innermost last):
>   File "<pyshell#4>", line 1, in ?
>     StringIO().write( buffer( 'abcd', 2 ) )
> SystemError: bad argument to internal function

What about closing the file? Since you do not have a reference to the file
object, you are not able to close it. Perhaps it would be better to write:

>>> f=open( 'test', 'w' )
>>> f.write( buffer( 'abcd', 2 ) )
>>> f.close()

Don't know about StringIO, sorry :-(

Tomasz




From sav at ulmen.mv.ru  Thu Feb 22 19:42:02 2001
From: sav at ulmen.mv.ru (Alexander Semenov)
Date: Thu, 22 Feb 2001 21:42:02 +0300
Subject: fileobj.write(buffer(...))
References: <mailman.982795168.21257.python-list@python.org> <972ip3$juj$1@news.tpi.pl>
Message-ID: <973mhu$1r2e$1@news1.simtel.ru>

"Tomasz Lisowski" <lisowski.tomasz at sssa.NOSPAM.com.pl> wrote in message
news:972ip3$juj$1 at news.tpi.pl...
> Uzytkownik "Niki Spahiev" <spahievi at vega.bg> napisal w wiadomosci
> news:mailman.982795168.21257.python-list at python.org...
> > Is this working as expected?
> > >>> open( 'test', 'w' ).write( buffer( 'abcd', 2 ) )
> > >>> from cStringIO import StringIO
> > >>> StringIO().write( buffer( 'abcd', 2 ) )
> > Traceback (innermost last):
> >   File "<pyshell#4>", line 1, in ?
> >     StringIO().write( buffer( 'abcd', 2 ) )
> > SystemError: bad argument to internal function

> What about closing the file? Since you do not have a reference to the file
> object, you are not able to close it. Perhaps it would be better to write:

No, python python implements reference counting. Then file object loose all
references it will del-ed. I think file object handles del correctly and
closes
corresponding handle.

> Don't know about StringIO, sorry :-(

As for StringIO, my "Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit
(Intel)] on win32"
is more informatical: it says:

>>> from cStringIO import StringIO
>>> StringIO().write( buffer( 'abcd', 2 ) )
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: expected string or Unicode object, buffer found

It looks like StringIO have some limitation in supporting buffers.

WBR,
Alexander Semenov





From db3l at fitlinxx.com  Sat Feb 24 01:39:20 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 23 Feb 2001 19:39:20 -0500
Subject: fileobj.write(buffer(...))
References: <mailman.982795168.21257.python-list@python.org> <972ip3$juj$1@news.tpi.pl> <973mhu$1r2e$1@news1.simtel.ru>
Message-ID: <uhf1k9ao7.fsf@ctwd0143.fitlinxx.com>

"Alexander Semenov" <sav at ulmen.mv.ru> writes:

> No, python python implements reference counting. Then file object loose all
> references it will del-ed. I think file object handles del correctly and
> closes
> corresponding handle.

Just as a general point, you can't guarantee when _del_ will be
called.  While the current CPython implementations (even in 2.x I
think) trigger that when the RC hits zero, there's no guarantees, and
it might not be called until later (or even until Python's process
terminates).  I'm not sure if JPython/Jython does it immediately even
in their current releases.

So depending on the criticality of controlling the close, you may want
to handle it yourself explicitly, particularly if you are writing data
that might otherwise not get flushed until (possibly) much later.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From tdickenson at devmail.geminidataloggers.co.uk  Mon Feb 26 13:12:10 2001
From: tdickenson at devmail.geminidataloggers.co.uk (Toby Dickenson)
Date: Mon, 26 Feb 2001 12:12:10 +0000
Subject: fileobj.write(buffer(...))
References: <mailman.982795168.21257.python-list@python.org> <972ip3$juj$1@news.tpi.pl> <973mhu$1r2e$1@news1.simtel.ru> <uhf1k9ao7.fsf@ctwd0143.fitlinxx.com>
Message-ID: <f3hk9tsek2smno687h0kmrjf1jcsceu32n@4ax.com>

David Bolen <db3l at fitlinxx.com> wrote:

>"Alexander Semenov" <sav at ulmen.mv.ru> writes:
>
>> No, python python implements reference counting. Then file object loose all
>> references it will del-ed. I think file object handles del correctly and
>> closes
>> corresponding handle.
>
>Just as a general point, you can't guarantee when _del_ will be
>called.  While the current CPython implementations (even in 2.x I
>think) trigger that when the RC hits zero, there's no guarantees, and
>it might not be called until later (or even until Python's process
>terminates).  I'm not sure if JPython/Jython does it immediately even
>in their current releases.
>
>So depending on the criticality of controlling the close, you may want
>to handle it yourself explicitly, particularly if you are writing data
>that might otherwise not get flushed until (possibly) much later.

That is less true than it was before 2.0...

2.0 takes care not to be recursive when deallocating recursive data
structures. This is achieved by moving dead objects onto a separate
list, and deleting things iteratively.

If you have more than one thread manipulating deeply recursive data
structures then it is possible for deallocations to be delayed, and
then actually handled by a thread other than the one that released the
last reference.

Im still suprised noone else finds this as much of a problem as I do.


Toby Dickenson
tdickenson at geminidataloggers.com


From db3l at fitlinxx.com  Mon Feb 26 23:15:19 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 26 Feb 2001 17:15:19 -0500
Subject: fileobj.write(buffer(...))
References: <mailman.982795168.21257.python-list@python.org> <972ip3$juj$1@news.tpi.pl> <973mhu$1r2e$1@news1.simtel.ru> <uhf1k9ao7.fsf@ctwd0143.fitlinxx.com> <f3hk9tsek2smno687h0kmrjf1jcsceu32n@4ax.com>
Message-ID: <ug0h1147c.fsf@ctwd0143.fitlinxx.com>

Toby Dickenson <tdickenson at devmail.geminidataloggers.co.uk> writes:

> That is less true than it was before 2.0...

Implementation wise or specification wise?  I don't think anything
changed in 2.0 to guarantee destruction behavior with respect to the
language, which is really what I was point out.

CPython 1.5.2 is actually very regular in this respect - destroying
objects pretty much as soon as their reference count hits zero.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From flognat at flognat.myip.org  Thu Feb 22 20:45:58 2001
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Thu, 22 Feb 2001 19:45:58 GMT
Subject: fileobj.write(buffer(...))
References: <mailman.982795168.21257.python-list@python.org>
Message-ID: <m3d7caa4e4.fsf@master.athome>

You forgot one line, the middle line.. I think.. You have to
do the operations on an instantiated object, not the class.

>>> from cStringIO import StringIO
>>> fileobj=StringIO() 
>>> fileobj.write( buffer( 'abcd', 2 ) )

        /Andy

/ Niki Spahiev <spahievi at vega.bg> wrote:
| Is this working as expected?
| 
| >>> open( 'test', 'w' ).write( buffer( 'abcd', 2 ) )
| >>> from cStringIO import StringIO
| >>> StringIO().write( buffer( 'abcd', 2 ) )
| Traceback (innermost last):
|   File "<pyshell#4>", line 1, in ?
|     StringIO().write( buffer( 'abcd', 2 ) )
| SystemError: bad argument to internal function
| >>>   
| 
| -- 
| Best regards,
|  Niki Spahiev


From locka at iol.ie  Thu Feb 22 00:04:09 2001
From: locka at iol.ie (Adam Lock)
Date: Wed, 21 Feb 2001 23:04:09 GMT
Subject: Curses for Win32?
Message-ID: <3A9449BB.284DC79B@iol.ie>

I have a Python 2.0 program that uses Tkinter and curses for display. It
works fine on Unix because there is a curses package but not on Win32
because there isn't. Perversely even MacPython 2.0 has a curses package!

Does anyone know if a curses package exists for Win32? I would even be
happy if it compiled but didn't function since I want will be using the
Tkinter UI anyway.

Many thanks

-- 
Adam Lock - locka at iol.ie


From fredrik at pythonware.com  Thu Feb 22 00:12:24 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 21 Feb 2001 23:12:24 GMT
Subject: Curses for Win32?
References: <3A9449BB.284DC79B@iol.ie>
Message-ID: <sXXk6.14902$Qb7.2379140@newsb.telia.net>

Adam Lock wrote:
> I have a Python 2.0 program that uses Tkinter and curses for display. It
> works fine on Unix because there is a curses package but not on Win32
> because there isn't. Perversely even MacPython 2.0 has a curses package!
>
> Does anyone know if a curses package exists for Win32? I would even be
> happy if it compiled but didn't function since I want will be using the
> Tkinter UI anyway.

you only need the imports to work?

alternative 1: wrap the imports in try/except:

    try:
        import curses
    except ImportError:
        curses = None # let's hope nobody uses this

alternative 2: add python stub modules:

    > more curses.py
    # empty

if you decide that you want console output after all, you
could consider alternative 3: write a curses-compatible
wrapper for

    http://effbot.org/efflib/console

Cheers /F




From porter at et.byu.edu  Thu Feb 22 02:13:00 2001
From: porter at et.byu.edu (C. Porter Bassett)
Date: Wed, 21 Feb 2001 18:13:00 -0700 (MST)
Subject: Curses for Win32?
In-Reply-To: <3A9449BB.284DC79B@iol.ie>
References: <3A9449BB.284DC79B@iol.ie>
Message-ID: <Pine.GHP.4.21.0102211811220.28286-100000@don.et.byu.edu>

I myself would love to have a python2.0 windows curses module, but nobody
with the know-how seems to think that it is important.  

What you could do, though, is create a dummy module called curses.  You
say that you don't mind if it doesn't function.  Just create dummy
functions corresponding to the functions that your program calls.  

--------------------------------------------------------------------------
C. Porter Bassett     porter at et.byu.edu   http://www.cporterbassett.com
--------------------------------------------------------------------------
"Pretend like this is a really witty saying." - Anonymous
--------------------------------------------------------------------------

On Wed, 21 Feb 2001, Adam Lock wrote:

> I have a Python 2.0 program that uses Tkinter and curses for display. It
> works fine on Unix because there is a curses package but not on Win32
> because there isn't. Perversely even MacPython 2.0 has a curses package!
> 
> Does anyone know if a curses package exists for Win32? I would even be
> happy if it compiled but didn't function since I want will be using the
> Tkinter UI anyway.
> 
> Many thanks
> 
> -- 
> Adam Lock - locka at iol.ie
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From dsh8290 at rit.edu  Thu Feb 22 04:32:46 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 21 Feb 2001 22:32:46 -0500
Subject: Curses for Win32?
In-Reply-To: <3A9449BB.284DC79B@iol.ie>; from locka@iol.ie on Wed, Feb 21, 2001 at 11:04:09PM +0000
References: <3A9449BB.284DC79B@iol.ie>
Message-ID: <20010221223246.A26874@harmony.cs.rit.edu>

On Wed, Feb 21, 2001 at 11:04:09PM +0000, Adam Lock wrote:
| I have a Python 2.0 program that uses Tkinter and curses for display. It
| works fine on Unix because there is a curses package but not on Win32
| because there isn't. Perversely even MacPython 2.0 has a curses package!
| 
| Does anyone know if a curses package exists for Win32? I would even be

Check out cygwin -- a POSIX emulation layer on top of windows.  It has
ncurses.  If you can compile python using the cygwin tools you will
have a functioning curses library.

| happy if it compiled but didn't function since I want will be using the
| Tkinter UI anyway.

If it doesn't function, and you are using TKinter why do you need
curses to exist?  If it causes the import to fail, put the import in a
if statement, or a try-except block.

HTH,
-D



From chris.gonnerman at usa.net  Thu Feb 22 06:10:15 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Wed, 21 Feb 2001 23:10:15 -0600
Subject: Curses for Win32?
References: <3A9449BB.284DC79B@iol.ie>
Message-ID: <018f01c09c8d$bf57d4a0$a100000a@local>

A "cursory" (hey, stop throwing things!) search of the 'net shows no Python
curses modules for
Win32.  It would be some work, but I have a Win32 port of Borland's conio.h
interface to
Python at:

    http://newcenturycomputers.net/projects/python/wconio

and Pablo J. Vidal has a Unix port of the same thing at:

    http://crazylovetrain.hypermart.net/projects.htm

I am going to examine his work with an eye toward making them
source-compatible.

----- Original Message -----
From: "Adam Lock" <locka at iol.ie>
Subject: Curses for Win32?


> I have a Python 2.0 program that uses Tkinter and curses for display. It
> works fine on Unix because there is a curses package but not on Win32
> because there isn't. Perversely even MacPython 2.0 has a curses package!
>
> Does anyone know if a curses package exists for Win32? I would even be
> happy if it compiled but didn't function since I want will be using the
> Tkinter UI anyway.
>
> Many thanks
>
> --
> Adam Lock - locka at iol.ie





From mak at mikroplan.com.pl  Thu Feb 22 11:13:43 2001
From: mak at mikroplan.com.pl (Grzegorz Makarewicz)
Date: Thu, 22 Feb 2001 11:13:43 +0100
Subject: Curses for Win32?
In-Reply-To: <3A9449BB.284DC79B@iol.ie>
Message-ID: <NDBBIKNLJKPLOLAJJJAPMEHIGCAA.mak@mikroplan.com.pl>


Adam Lock wrote:
> 
> I have a Python 2.0 program that uses Tkinter and curses for display. It
> works fine on Unix because there is a curses package but not on Win32
> because there isn't. Perversely even MacPython 2.0 has a curses package!
> 
> Does anyone know if a curses package exists for Win32? I would even be
> happy if it compiled but didn't function since I want will be using the
> Tkinter UI anyway.
> 

Not so complicated as ncurses but always better than nothing,
try standard module curses (from python 1.4.x) connected to pdcurses:

http://www.mikroplan.com.pl/~mak/python

mak




From garabik-news at spam.melkor.dnp.fmph.uniba.sk  Thu Feb 22 16:33:42 2001
From: garabik-news at spam.melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: 22 Feb 2001 15:33:42 GMT
Subject: Curses for Win32?
References: <3A9449BB.284DC79B@iol.ie>
Message-ID: <982855925.989488@cdwork.cvt.stuba.sk>

Adam Lock <locka at iol.ie> wrote:
 : I have a Python 2.0 program that uses Tkinter and curses for display. It
 : works fine on Unix because there is a curses package but not on Win32
 : because there isn't. Perversely even MacPython 2.0 has a curses package!

 : Does anyone know if a curses package exists for Win32? I would even be
 : happy if it compiled but didn't function since I want will be using the
 : Tkinter UI anyway.

well, I am not going to answer this for the third time :-)
(not that I would not want to, but I do not have the URL
readily accessible just now)
serach the archive for "pdcurses"

-- 
 -----------------------------------------------------------
| Radovan Garabik http://melkor.dnp.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__    garabik @ melkor.dnp.fmph.uniba.sk     |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From aleaxit at yahoo.com  Thu Feb 22 18:18:35 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 18:18:35 +0100
Subject: Curses for Win32?
References: <3A9449BB.284DC79B@iol.ie> <982855925.989488@cdwork.cvt.stuba.sk>
Message-ID: <973hlf11g60@news2.newsguy.com>

"Radovan Garabik" <garabik-news at spam.melkor.dnp.fmph.uniba.sk> wrote in
message news:982855925.989488 at cdwork.cvt.stuba.sk...
> Adam Lock <locka at iol.ie> wrote:
>  : I have a Python 2.0 program that uses Tkinter and curses for display.
It
>  : works fine on Unix because there is a curses package but not on Win32
>  : because there isn't. Perversely even MacPython 2.0 has a curses
package!
>
>  : Does anyone know if a curses package exists for Win32? I would even be
>  : happy if it compiled but didn't function since I want will be using the
>  : Tkinter UI anyway.
>
> well, I am not going to answer this for the third time :-)
> (not that I would not want to, but I do not have the URL
> readily accessible just now) serach the archive for "pdcurses"

Haven't tried it, but,
http://www.mikroplan.com.pl/~mak/python/files/curses_src.zip
appears to be a zipfile containing a C source (to be compiled
to yield a for-Python module) and a curses.py module (that is
presumably meant to go with that extension module).  There
is a binary archive too, but I don't know what Python version
it is for.


Alex





From ylee12 at uiuc.edu  Thu Feb 22 05:05:48 2001
From: ylee12 at uiuc.edu (Young-Jin Lee)
Date: Wed, 21 Feb 2001 22:05:48 -0600
Subject: [Q] Best way to start learning Python
Message-ID: <U90l6.4131$dL4.51231@vixen.cso.uiuc.edu>

Hi, I'd like to know what is the best way to learn Python.
I downloaded Python tutorial from www.python.org, but it's very short and
simple.
Could you recommend the next step?
Thanks in advance.

--
Young-Jin Lee




From fuess at att.net  Thu Feb 22 05:42:53 2001
From: fuess at att.net (David Fuess)
Date: Thu, 22 Feb 2001 04:42:53 GMT
Subject: [Q] Best way to start learning Python
References: <U90l6.4131$dL4.51231@vixen.cso.uiuc.edu>
Message-ID: <4t599tgn1tn6jc6arhcig924ffgh32go89@4ax.com>

There are any number of excellent references on Python. For the
general language you might try:

	Python Developer's Handbook, Andre Lessa, Sams

	Programming Python, Mark Lutz, O'Reilly (cannonical language
reference)

	Learning Python, Mark Lutz, O'Reilly, 1999

	Internet Programming with Python, Aaron Watters, M&T Books

for a start ...

Dave

On Wed, 21 Feb 2001 22:05:48 -0600, "Young-Jin Lee" <ylee12 at uiuc.edu>
wrote:

>Hi, I'd like to know what is the best way to learn Python.
>I downloaded Python tutorial from www.python.org, but it's very short and
>simple.
>Could you recommend the next step?
>Thanks in advance.



From rapto at arrakis.es  Fri Feb 23 14:21:11 2001
From: rapto at arrakis.es (Marcos Sánchez Provencio)
Date: Fri, 23 Feb 2001 14:21:11 +0100
Subject: [Q] Best way to start learning Python
References: <U90l6.4131$dL4.51231@vixen.cso.uiuc.edu>
Message-ID: <newscache$u2p79g$l21$1@proxy.bitmailer.com>

> I downloaded Python tutorial from www.python.org, but it's very short and
> simple.

This is a feature, not a bug. Python _is_ short and simple.





From f8dy at yahoo.com  Mon Feb 26 15:43:55 2001
From: f8dy at yahoo.com (Mark Pilgrim)
Date: Mon, 26 Feb 2001 09:43:55 -0500
Subject: [Q] Best way to start learning Python
References: <U90l6.4131$dL4.51231@vixen.cso.uiuc.edu> <newscache$u2p79g$l21$1@proxy.bitmailer.com>
Message-ID: <97dq0b$oe881$1@ID-77331.news.dfncis.de>

O'Reilly recently reviewed several free online Python books for both
beginners and more experienced programmers.

http://www.oreillynet.com/pub/a/python/2001/02/07/pythonnews.html

-M
You're smart; why haven't you learned Python yet?
http://diveintopython.org/

"Marcos S?nchez Provencio" <rapto at arrakis.es> wrote in message
news:newscache$u2p79g$l21$1 at proxy.bitmailer.com...
> > I downloaded Python tutorial from www.python.org, but it's very short
and
> > simple.
>
> This is a feature, not a bug. Python _is_ short and simple.
>
>
>




From michael+gnus at trollope.org  Sun Feb 25 07:53:01 2001
From: michael+gnus at trollope.org (Michael Powe)
Date: 24 Feb 2001 22:53:01 -0800
Subject: [Q] Best way to start learning Python
References: <U90l6.4131$dL4.51231@vixen.cso.uiuc.edu>
Message-ID: <87pug7s182.fsf@cecilia.trollope.org>

>>>>> "Young-Jin" == Young-Jin Lee <ylee12 at uiuc.edu> writes:

    Young-Jin> Hi, I'd like to know what is the best way to learn
    Young-Jin> Python.  I downloaded Python tutorial from
    Young-Jin> www.python.org, but it's very short and simple.  Could
    Young-Jin> you recommend the next step?  Thanks in advance.

<Learning Python> by Mark Lutz & David Ascher

mp

-- 
    -----------Michael Powe         Portland, Oregon USA----------
"There are 100,000 total marijuana smokers in the U.S., and most are
Negroes, Hispanics, Filipinos and entertainers.  Their Satanic music,
jazz and swing, result from marijuana use.  This marijuana can cause
white women to seek sexual relations with Negroes, entertainers and
any others."  -- Harry J. Anslinger, Dir. Fed. Bureau of Narcotics,
Congressional Testimony, 1937



From mertz at gnosis.cx  Thu Feb 22 07:08:41 2001
From: mertz at gnosis.cx (Dr. David Mertz)
Date: Thu, 22 Feb 2001 01:08:41 -0500
Subject: The price of fame...
Message-ID: <pzKl6kKkXYRJ092yn@bellatlantic.net>                         

As shockingly many readers of this group will probably know (so server
logs indirectly say), I have written some articles on Python, mostly in
the series _Charming Python_, which lives, finally, at
ibm.com/developerworks/linux/.  But before and after they live there,
they live roundabout gnosis.cx/publish/.

Here's the story, some unexpected favorable mentions, mostly at
PythonWare's daily-url has caused my web-host providers to act
erratically and capriciously (let me put in a passing bad word against
CSoft.net, whom I will not have any more as soon as my DNS changes go
through).

But there's more...  I apologize for the Proustian quality...  I
contributed to the caprice, in a small way, by using a plain,
old-fashion CGI program (no fastCGI, mod_python, or the like) to
prettify documents that are named '*.txt' into splashy HTML thingies.
Is this a good time to mention how much I miss gopherspace?

So everyone who went to look at my links ran a moderately complicated
Python script, which made for some server load.  Excessive dynamism has
come to haunt me, after my self-steeping in the ways Pythonic.
Admittedly, the script was merely cute, but quite needless (I should
just convert once, and leave it... or at most on a cron).

Well, at the end of this narration is just a request.  Anyone who might
try to find any article that lives at the end of
"http://gnosis.cx/cgi/txt2html.cgi?source=" would be better served, for
the nonce at least, to leave off that first part, and just look at the
URLish stuff that comes later in the enormous CGI URL.  Not quite as
pretty... but at least you won't get a 404.

All the best, David...


--
 mertz@   _/_/_/_/_/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY:_/_/_/_/ v i
gnosis  _/_/                    Postmodern Enterprises         _/_/  s r
.cx    _/_/  MAKERS OF CHAOS....                              _/_/   i u
      _/_/_/_/_/ LOOK FOR IT IN A NEIGHBORHOOD NEAR YOU_/_/_/_/_/    g s




From tchur at optushome.com.au  Thu Feb 22 20:34:13 2001
From: tchur at optushome.com.au (Tim Churches)
Date: Fri, 23 Feb 2001 06:34:13 +1100
Subject: The price of fame...
References: <pzKl6kKkXYRJ092yn@bellatlantic.net>
Message-ID: <3A9569B5.D2540F58@optushome.com.au>

"Dr. David Mertz" wrote:
> the series _Charming Python_, which lives, finally, at
> ibm.com/developerworks/linux/.  But before and after they live there,
> they live roundabout gnosis.cx/publish/.

Ah, they live on Christmas Island! Maybe some Gecarcoidea natalis (red
crab) cut the cables to your Web server... see
http://www.christmas.net.au for more information on the home of Dr.
Mertz' Web site.

Tim Churches
Sydney, Australia



From Noah at noah.org  Thu Feb 22 09:27:02 2001
From: Noah at noah.org (Noah)
Date: Thu, 22 Feb 2001 00:27:02 -0800
Subject: Non-blocking pipe read under Windows NT
Message-ID: <q34l6.297$0Q5.207467@news.pacbell.net>

Hmmm... Windows Python2.0 does not have fcntl.
How do I do a non-blocking select on a Pipe file descriptor?

Under UNIX I can do something like this:

# hard way to read a file

import os, fcntl, FCNTL

# Open a pipe to read a file.
(fin, fout, ferr) = os.popen3 ('cat testfile')

# Turn off blocking on file. read() will then return -1 if no data
flags = fcntl.fcntl (fout.fileno(), FCNTL.F_GETFL, 0)
flags = flags | FCNTL.O_NONBLOCK
fcntl.fcntl (fout.fileno(), FCNTL.F_SETFL, flags)

# Print out the file without blocking.
done = 0
while !done:
    (r,w,e) = select.select ([fout], [], [], None)
    if len(r) > 0:
        data = r[0].read()
        if data == -1:
            done = 1
        else:
            print data




From NoSpam at NoSpam.com  Thu Feb 22 17:52:09 2001
From: NoSpam at NoSpam.com (Tom)
Date: Thu, 22 Feb 2001 16:52:09 GMT
Subject: Non-blocking pipe read under Windows NT
References: <q34l6.297$0Q5.207467@news.pacbell.net>
Message-ID: <Zsbl6.3820$XZ6.2201217@news4.rdc1.on.home.com>

That's a bit tricky.

Windows (at least NT/2K, which fully support named pipes) does support
non-blocking pipes, but I don't think that Python does.  But I don't think
that matters.  I looked into it a while ago (for C++) and decided that it
would be easier to implement an extra thread than to implement non-blocking
pipes for cross-platform purposes.  For one thing, cross-platform threading
libraries are common, whereas non-blocking pipes seem suprisingly uncommon.

So, I don't know what you are trying to do, but if threads are an
alternative soln to your problem (ie. thread + block == non-block), I would
recommend them.

If you still want to persue non-block windows pipes, you could look at the
article "
Named Pipe Type, Read, and Wait Modes" (
http://msdn.microsoft.com/library/psdk/winbase/pipes_60mr.htm).  The win32
implementation of Python's thread creation (for the popen fn's) is based on
MS knowledgebase article Q190351 (available from
http://msdn.microsoft.com/library/).  These articles are, unfortunately, in
'C'.

Good-luck,

Tom.

"Noah" <Noah at noah.org> wrote in message
news:q34l6.297$0Q5.207467 at news.pacbell.net...
> Hmmm... Windows Python2.0 does not have fcntl.
> How do I do a non-blocking select on a Pipe file descriptor?
>
> Under UNIX I can do something like this:
>
> # hard way to read a file
>
> import os, fcntl, FCNTL
>
> # Open a pipe to read a file.
> (fin, fout, ferr) = os.popen3 ('cat testfile')
>
> # Turn off blocking on file. read() will then return -1 if no data
> flags = fcntl.fcntl (fout.fileno(), FCNTL.F_GETFL, 0)
> flags = flags | FCNTL.O_NONBLOCK
> fcntl.fcntl (fout.fileno(), FCNTL.F_SETFL, flags)
>
> # Print out the file without blocking.
> done = 0
> while !done:
>     (r,w,e) = select.select ([fout], [], [], None)
>     if len(r) > 0:
>         data = r[0].read()
>         if data == -1:
>             done = 1
>         else:
>             print data
>
>




From corour01 at motorola.com  Thu Feb 22 10:02:27 2001
From: corour01 at motorola.com (O'Rourke Clodagh-corour01)
Date: Thu, 22 Feb 2001 09:02:27 -0000
Subject: Embedding setenv
Message-ID: <13EE655665F4D311AB4D0008C789498A01EE2C44@zei02exm02.cork.cig.mot.com>

Hi all,

I was wondering does anyone know how to embed the setenv command in python.
I've tried os.system("setenv  VARIABLE value")
But I get:  sh: setenv: not found

Any ideas?

Thanks,
Clodagh O Rourke



From boltong at europem01.nt.com  Thu Feb 22 11:23:13 2001
From: boltong at europem01.nt.com (Bolton, Gawain [ADC:4808:EXCH])
Date: Thu, 22 Feb 2001 11:23:13 +0100
Subject: Embedding setenv
References: <mailman.982834914.6002.python-list@python.org>
Message-ID: <3A94E891.93DF51E3@europem01.nt.com>

O'Rourke Clodagh-corour01 wrote:

> Hi all,
>
> I was wondering does anyone know how to embed the setenv command in python.
> I've tried os.system("setenv  VARIABLE value")
> But I get:  sh: setenv: not found
>
> Any ideas?
>
> Thanks,
> Clodagh O Rourke

Yeah try:

import os
os.environ[name] = value

Note that this will set the environment variable for the current process.

Your os.system() call did not work firstly because "setenv" is a csh command
and os.system() spawned a plain old shell session.  Secondly, this is the
wrong approach because the environment variable will be set only for the
sub-process created by the system() command.  Thirdly, this approach is very
inefficient.


Gawain

--
-------------------------------------------------------------------------------
Gawain Bolton                     | E-mail: boltong at nortelnetworks.com
Section 4808                      |
UMTS Development                  |
Nortel Networks                   | Voice:  ESN 579-3763   +33 1.39.44.37.63
Guyancourt, France                | FAX:    ESN 579-3009   +33 1.39.44.30.09
-------------------------------------------------------------------------------





From fredrik at pythonware.com  Thu Feb 22 12:28:05 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 11:28:05 GMT
Subject: Embedding setenv
References: <mailman.982834914.6002.python-list@python.org>
Message-ID: <9J6l6.19294$AH6.2613345@newsc.telia.net>

Clodagh O'Rourke wrote:
> I was wondering does anyone know how to embed the setenv command in python.
> I've tried os.system("setenv  VARIABLE value")
> But I get:  sh: setenv: not found

Even if the default shell had a setenv command (setenv is
a csh feature, not a unix command), that approach wouldn't
work.

try this instead:

    os.environ["VARIABLE"] = "value"

(for more info on why setting environment in one subprocess
won't affect environment variables in another subprocess, see
your nearest Unix FAQ).

Cheers /F




From chris.gonnerman at usa.net  Thu Feb 22 14:42:08 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Thu, 22 Feb 2001 07:42:08 -0600
Subject: Embedding setenv
References: <13EE655665F4D311AB4D0008C789498A01EE2C44@zei02exm02.cork.cig.mot.com>
Message-ID: <002301c09cd5$44157d40$a100000a@local>

----- Original Message -----
From: "O'Rourke Clodagh-corour01" <corour01 at motorola.com>
Subject: Embedding setenv


> Hi all,
>
> I was wondering does anyone know how to embed the setenv command in
python.
> I've tried os.system("setenv  VARIABLE value")
> But I get:  sh: setenv: not found
>
> Any ideas?
>
> Thanks,
> Clodagh O Rourke

In general, you can't do this.  setenv (in csh) is a builtin, and doesn't
exist in /bin/sh or /bin/bash
at all.  Are you trying to set variables in your parent processes'
environment?  Can't do that in
Unix, period.  If you are trying to set your own environment, including
controlling the environment
of future child processes, use os.environ, which is a mapping
(dictionary-like object):

    import os
    os.environ['VARIABLE'] = 'value'






From corour01 at motorola.com  Thu Feb 22 15:11:01 2001
From: corour01 at motorola.com (O'Rourke Clodagh-corour01)
Date: Thu, 22 Feb 2001 14:11:01 -0000
Subject: Embedding setenv
Message-ID: <13EE655665F4D311AB4D0008C789498A01EE2C51@zei02exm02.cork.cig.mot.com>

>>I've tried os.system("setenv  VARIABLE value")
>>But I get:  sh: setenv: not found


>n general, you can't do this.  setenv (in csh) is a builtin, and doesn't
>xist in /bin/sh or /bin/bash
>t all.  Are you trying to set variables in your parent processes'
>nvironment?  Can't do that in
>nix, period.  If you are trying to set your own environment, including
>ontrolling the environment
>o future child processes, use os.environ, which is a mapping
>dictionary-like object):

    import os
    os.environ['VARIABLE'] = 'value'


I'm trying to set environment variables in a GSM network environment I tried 

os.environ['VARIABLE'] = 'value'

And I got.....

SyntaxError: can't assign to function call 

I'm working on putting a front end on the Network-Controller Environment variables reading data to/from Informix and setting the environment.

If I can't do this directly from the python back end of the application, I might have to write the variables/values to another script and keep the setenv stuff separate....

Thanks,
Clodagh.



From emile at fenx.com  Thu Feb 22 15:45:46 2001
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 22 Feb 2001 06:45:46 -0800
Subject: Embedding setenv
References: <mailman.982851124.2766.python-list@python.org>
Message-ID: <9738p2$ncbrn$1@ID-11957.news.dfncis.de>

"O'Rourke Clodagh-corour01" <corour01 at motorola.com> wrote in
message
news:mailman.982851124.2766.python-list at python.org...
>
> >>I've tried os.system("setenv  VARIABLE value")
> >>But I get:  sh: setenv: not found
>
>
> >n general, you can't do this.  setenv (in csh) is a
builtin, and doesn't
> >xist in /bin/sh or /bin/bash
> >t all.  Are you trying to set variables in your parent
processes'
> >nvironment?  Can't do that in
> >nix, period.  If you are trying to set your own
environment, including
> >ontrolling the environment
> >o future child processes, use os.environ, which is a
mapping
> >dictionary-like object):
>
>     import os
>     os.environ['VARIABLE'] = 'value'
>

this works for me...


>
> I'm trying to set environment variables in a GSM network
environment I tried
>
> os.environ['VARIABLE'] = 'value'
>
> And I got.....
>
> SyntaxError: can't assign to function call
>

gotta be something else going on...


> I'm working on putting a front end on the
Network-Controller Environment variables reading data
to/from Informix and setting the environment.
>
> If I can't do this directly from the python back end of
the application, I might have to write the variables/values
to another script and keep the setenv stuff separate....
>
> Thanks,
> Clodagh.
>


I'm not sure it'll do what you want it to do, but it
certainly does what you say you're doing...


--

Emile van Sebille
emile at fenx.com
-------------------






From fredrik at pythonware.com  Thu Feb 22 16:10:53 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 15:10:53 GMT
Subject: Embedding setenv
References: <mailman.982851124.2766.python-list@python.org>
Message-ID: <1_9l6.19330$AH6.2620037@newsc.telia.net>

Clodagh O'Rourke wrote:
> os.environ['VARIABLE'] = 'value'
>
> And I got.....
>
> SyntaxError: can't assign to function call

Are you sure you used square brackets?

>>> os.environ('VARIABLE') = 'value'
SyntaxError: can't assign to function call

>>> os.environ['VARIABLE'] = 'value'
>>> os.environ['VARIABLE']
'value'

Cheers /F




From costas at meezon.com  Thu Feb 22 20:58:20 2001
From: costas at meezon.com (Costas Menico)
Date: Thu, 22 Feb 2001 19:58:20 GMT
Subject: Embedding setenv
References: <mailman.982851124.2766.python-list@python.org>
Message-ID: <3a956e89.18194798@news.cis.dfn.de>

On Thu, 22 Feb 2001 14:11:01 -0000, "O'Rourke Clodagh-corour01" :
>
>    import os
>    os.environ['VARIABLE'] = 'value'
>
>

Works for me.... but if you change it to paren then you get the error
below.

>I'm trying to set environment variables in a GSM network environment I tried 
>
>os.environ['VARIABLE'] = 'value'
>
>And I got.....
>
>SyntaxError: can't assign to function call 
>



From simon at joyful.com  Thu Feb 22 11:19:01 2001
From: simon at joyful.com (Simon Michael)
Date: 22 Feb 2001 10:19:01 +0000
Subject: nice articles (Re: The price of fame...)
References: <mailman.982834675.5030.python-list@python.org>
Message-ID: <87vgq358bu.fsf@joyful.com>

I found eg "Text Processing in Python: Tips for Beginners" a very
useful overview. Thanks!

If I may, an attempt to summarize for my fellow newbies:

  "Charming Python" columns can be found at, eg
  http://gnosis.cx/publish/tech_index.html .
  To read, click the title (not the eye logo).

Best regards


From Norman_Shelley-RRDN60 at email.sps.mot.com  Thu Feb 22 17:47:14 2001
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Thu, 22 Feb 2001 09:47:14 -0700
Subject: nice articles (Re: The price of fame...)
References: <mailman.982834675.5030.python-list@python.org> <87vgq358bu.fsf@joyful.com>
Message-ID: <3A954292.FEED9271@email.sps.mot.com>

Simon Michael wrote:

> I found eg "Text Processing in Python: Tips for Beginners" a very
> useful overview. Thanks!
>
> If I may, an attempt to summarize for my fellow newbies:
>
>   "Charming Python" columns can be found at, eg
>   http://gnosis.cx/publish/tech_index.html .
>   To read, click the title (not the eye logo).
>
> Best regards

Funny, I still get this when I click on the above URL

Not Found
The requested URL /publish/tech_index.html was not found on this server.

Additionally, a 404 Not Found error was encountered while trying to use
an ErrorDocument to handle the request.



From fig at monitor.net  Fri Feb 23 20:18:56 2001
From: fig at monitor.net (Stephen R. Figgins)
Date: Fri, 23 Feb 2001 11:18:56 -0800
Subject: nice articles (Re: The price of fame...)
References: <mailman.982834675.5030.python-list@python.org> <87vgq358bu.fsf@joyful.com> <3A954292.FEED9271@email.sps.mot.com>
Message-ID: <3A96B7A0.BFBE6A47@monitor.net>

> Funny, I still get this when I click on the above URL
> 
> Not Found
> The requested URL /publish/tech_index.html was not found on this server.

The web site was down for some reason.  It is back now.  You can find
Dr. Mertz's writings at:

  http://gnosis.cx/publish/tech_index.html

Stephen


From mertz at gnosis.cx  Thu Feb 22 19:49:54 2001
From: mertz at gnosis.cx (Dr. David Mertz)
Date: Thu, 22 Feb 2001 13:49:54 -0500
Subject: nice articles (Re: The price of fame...)
Message-ID: <S9Vl6kKkXYgU092yn@bellatlantic.net>                         

Simon Michael wrote:
> I found eg "Text Processing in Python: Tips for Beginners" a very
> useful overview. Thanks!
> If I may, an attempt to summarize for my fellow newbies:
>   "Charming Python" columns can be found at, eg
>   http://gnosis.cx/publish/tech_index.html .
>   To read, click the title (not the eye logo).

Norman Shelley
| Funny, I still get this when I click on the above URL
| Not Found
| The requested URL /publish/tech_index.html was not found on this server.
| Additionally, a 404 Not Found error was encountered while trying to use
| an ErrorDocument to handle the request.

Apparently my old host is AGAIN cycling though various unhelpful
messages at the site.  Let me repeat my recommendation against
Csoft.net.  Terrible customer support:  contemptuous, rude, unhelpful.

So I apologize for burdening the group with more matter that might seem
egotistical.  However, for now, you can get to the articles (on the new
host), at:

  http://64.41.64.172/publish/tech_index.html

In a day or two, the DNS updates should be distributed, and you can use
the domain name again.

Thanks, David...



From tuttledon at hotmail.com  Thu Feb 22 20:25:27 2001
From: tuttledon at hotmail.com (Don Tuttle)
Date: Thu, 22 Feb 2001 19:25:27 GMT
Subject: nice articles (Re: The price of fame...)
References: <mailman.982868404.12234.python-list@python.org>
Message-ID: <HIdl6.5971$h52.1486991@typhoon.southeast.rr.com>

David...
> Apparently my old host is AGAIN cycling though various unhelpful
> messages at the site.  Let me repeat my recommendation against
> Csoft.net.  Terrible customer support:  contemptuous, rude, unhelpful.
>
> So I apologize for burdening the group with more matter that might seem
> egotistical.  However, for now, you can get to the articles (on the new
> host), at:
>
>   http://64.41.64.172/publish/tech_index.html
>
> In a day or two, the DNS updates should be distributed, and you can use
> the domain name again.

Wow, what a gold mine!.  Hate to say it but I'm glad you've had problems,
otherwise I might not have found your work.  ;-)

Don




From djc at object-craft.com.au  Thu Feb 22 11:52:29 2001
From: djc at object-craft.com.au (Dave Cole)
Date: 22 Feb 2001 21:52:29 +1100
Subject: Sybase module 0.12 (Aki Yamashita release) released
Message-ID: <m3y9uz56s2.fsf@vole.object-craft.com.au>

What is it:

The Sybase module provides a Python interface to the Sybase relational
database system. The Sybase package supports almost all of the Python
Database API, version 2.0 with extensions.

The module works with Python versions 1.5.2 and later and Sybase
versions 11.0.3 and later.  It is based on the Sybase Client Library
(ct_* API), and the Bulk-Library Client (blk_* API) interfaces.

Changes:

- A bug has been fixed in the bulkcopy object which caused FLOAT
  columns to be corrupted on transfer to the server.

Where can you get it:

        http://www.object-craft.com.au/projects/sybase/

- Dave

-- 
http://www.object-craft.com.au


From vonWedel at lfpt.rwth-aachen.de  Thu Feb 22 13:42:59 2001
From: vonWedel at lfpt.rwth-aachen.de (vonWedel at lfpt.rwth-aachen.de)
Date: Thu, 22 Feb 2001 13:42:59 +0100
Subject: XML Module problems
Message-ID: <OFB6F6551C.884AED11-ONC12569FB.0044DC4F@lfpt.rwth-aachen.de>

Hello,

I'm trying to use an SAX-based parser to process XML files in Python 2.0.

First I tried Python on a Solaris machine, finding out that the python 
distribution
installed does not contain _any_ XML parsers!? Hence I chose to install
PyXML -- it installs fine but running it gives me a  segmentation fault in
sax.make_parser(), so probably it doesn't succeed in loading the 
pyexpat.so
module. Are there any binary versions of the module?

Running the program on an NT 4.0 machine succeeds in creating a parser,
but gives me the following error which seems to be internal to the parser 
module!?
(The warning is output from my program, meaning it doesn't handle element
<start>).

Warning: Start of element start skipped
Traceback (most recent call last):
  File "run.py", line 369, in ?
    main(sys.stdin, sys.stdout, sys.stderr)
  File "run.py", line 366, in main
    h.run()
  File "run.py", line 86, in run
    p.parse('case.xml')
  File "d:\programme\python20\lib\xml\sax\expatreader.py", line 42, in 
parse
    xmlreader.IncrementalParser.parse(self, source) 
  File "d:\programme\python20\lib\xml\sax\xmlreader.py", line 120, in 
parse
    self.feed(buffer)
  File "d:\programme\python20\lib\xml\sax\expatreader.py", line 81, in 
feed
    self._parser.Parse(data, isFinal)
TypeError: too many arguments; expected 1, got 2

Any other ideas on how to use XML in Python?

TNX,
Lars
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010222/f22aec9e/attachment.htm>

From loewis at informatik.hu-berlin.de  Thu Feb 22 14:20:30 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 22 Feb 2001 14:20:30 +0100
Subject: XML Module problems
References: <mailman.982845604.24626.python-list@python.org>
Message-ID: <j47l2iua5d.fsf@informatik.hu-berlin.de>

vonWedel at lfpt.rwth-aachen.de writes:

> First I tried Python on a Solaris machine, finding out that the
> python distribution installed does not contain _any_ XML parsers!?

It sure does include pyexpat - you have to have the expat library to
compile it, though. Of course, I don't know how you had installed
Python, so it is certainly possible that there is no XML parser in
your installation.

> Hence I chose to install PyXML -- it installs fine but running it
> gives me a segmentation fault in sax.make_parser(), so probably it
> doesn't succeed in loading the pyexpat.so module. Are there any
> binary versions of the module?

What version of PyXML? It works fine for me. If you need binary
releases: what version of Solaris?

>     self.feed(buffer)
>   File "d:\programme\python20\lib\xml\sax\expatreader.py", line 81, in 
> feed
>     self._parser.Parse(data, isFinal)
> TypeError: too many arguments; expected 1, got 2

That is a confusing traceback. It is not an error in the call to
Parse, but in the callback invoked from pyexpat. It would be good if
you could post your ContentHandler.

Regards,
Martin


From vonWedel at lfpt.rwth-aachen.de  Thu Feb 22 15:43:58 2001
From: vonWedel at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Thu, 22 Feb 2001 15:43:58 +0100
Subject: XML Module problems
References: <mailman.982845604.24626.python-list@python.org> <j47l2iua5d.fsf@informatik.hu-berlin.de>
Message-ID: <3A9525AE.D92CBBC4@lfpt.rwth-aachen.de>

Hello Martin,

> > First I tried Python on a Solaris machine, finding out that the
> > python distribution installed does not contain _any_ XML parsers!?
> 
> It sure does include pyexpat - you have to have the expat library to
> compile it, though. Of course, I don't know how you had installed
> Python, so it is certainly possible that there is no XML parser in
> your installation.

As far as I know we installed a binary package from some site, don't
know
from where exactly, though. However, in xml/parsers there is expat.py
but no pyexpat.

> > Hence I chose to install PyXML -- it installs fine but running it
> > gives me a segmentation fault in sax.make_parser(), so probably it
> > doesn't succeed in loading the pyexpat.so module. Are there any
> > binary versions of the module?
> 
> What version of PyXML? It works fine for me. If you need binary
> releases: what version of Solaris?
It shows the same behavior for both 0.6.2 and 0.6.3. 'uname -a'
says 'SunOS kazul 5.7 Generic sun4m sparc'. I just tried to recompile
the package on a more recent machine (SOS 5.7, but type sun4u instead
of sun4m) and this one seems to work (on both machines!), I can create
an Expat after installing the package. Strange, but anyway.

> >     self.feed(buffer)
> >   File "d:\programme\python20\lib\xml\sax\expatreader.py", line 81, in
> > feed
> >     self._parser.Parse(data, isFinal)
> > TypeError: too many arguments; expected 1, got 2
 
> That is a confusing traceback. It is not an error in the call to
> Parse, but in the callback invoked from pyexpat. It would be good if
> you could post your ContentHandler.

Now, that's indeed confusing! With this hint I could quickly trace the
problem down.

Thanks for your quick help,
Lars



From loewis at informatik.hu-berlin.de  Mon Feb 26 09:33:48 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 26 Feb 2001 09:33:48 +0100
Subject: XML Module problems
References: <mailman.982845604.24626.python-list@python.org> <j47l2iua5d.fsf@informatik.hu-berlin.de> <mailman.982853044.8344.python-list@python.org>
Message-ID: <j4k86dlu6r.fsf@informatik.hu-berlin.de>

Lars von Wedel <vonWedel at lfpt.rwth-aachen.de> writes:

> As far as I know we installed a binary package from some site, don't
> know from where exactly, though. However, in xml/parsers there is
> expat.py but no pyexpat.

That makes it difficult to analyse the problem further. In Python
proper, pyexpat, if present, might be builting or a lib-dynload
module; it will not reside in xml.parsers. If you'd know where the
binary came from, you could complain to the packager...

> > > Hence I chose to install PyXML -- it installs fine but running it
> > > gives me a segmentation fault in sax.make_parser(), so probably it
> > > doesn't succeed in loading the pyexpat.so module. Are there any
> > > binary versions of the module?
> > 
> > What version of PyXML? It works fine for me. If you need binary
> > releases: what version of Solaris?
> It shows the same behavior for both 0.6.2 and 0.6.3. 'uname -a'
> says 'SunOS kazul 5.7 Generic sun4m sparc'. I just tried to recompile
> the package on a more recent machine (SOS 5.7, but type sun4u instead
> of sun4m) and this one seems to work (on both machines!), I can create
> an Expat after installing the package. Strange, but anyway.

Strange indeed. Broken compiler?

> Now, that's indeed confusing! With this hint I could quickly trace the
> problem down.

It's good to hear it works now.

Regards,
Martin


From gradha at iname.com  Thu Feb 22 14:57:08 2001
From: gradha at iname.com (gradha at iname.com)
Date: Thu, 22 Feb 2001 14:57:08 +0100
Subject: Looking for better Tkinter's listbox widget
Message-ID: <kr5379.9hl.ln@127.0.0.1>

Hello.

I am looking forward to writting a Phython GUI for a program, and I need
to show a listbox with at least multiple columns, where it would be just
nice (not really needed) to display an icon in one of them. Such listbox
items should be individually selectable.

I've read Tkinter's Fredrik Introduction, and there it says that Tkinter's
native listbox doesn't support multiple columns. Where can I find a
replacement widget providing the multicolumn feature and being portable
to Windows and X?

-- 

 Grzegorz Adam Hankiewicz   gradha at iname.com - http://gradha.infierno.org
 Other web pages:  http://glub.ehu.es/  -  http://welcome.to/gogosoftware/


From jp.roumian at mobileway.com  Tue Feb 27 19:19:09 2001
From: jp.roumian at mobileway.com (jean-paul roumian)
Date: Tue, 27 Feb 2001 19:19:09 +0100
Subject: Looking for better Tkinter's listbox widget
References: <kr5379.9hl.ln@127.0.0.1>
Message-ID: <3A9BEF9D.2050507@mobileway.com>


gradha at iname.com wrote:

> Hello.
> 
> I am looking forward to writting a Phython GUI for a program, and I need
> to show a listbox with at least multiple columns, where it would be just
> nice (not really needed) to display an icon in one of them. Such listbox
> items should be individually selectable.
> 
> I've read Tkinter's Fredrik Introduction, and there it says that Tkinter's
> native listbox doesn't support multiple columns. Where can I find a
> replacement widget providing the multicolumn feature and being portable
> to Windows and X?

did you tried GtkCTree ? (from pygtk)

> 

-- 
"The Pythagorean Theorem employed 24 words, the Lord's Prayer has 66 words,
Archimedes Principle has 67words, the 10 Commandments have 179 words, 
the Declaration of Independence 1,300 words,
and finally the European Commission's regulation on the sale of 
cabbage: 26,911 words."



From martin.franklin at westerngeco.com  Wed Feb 28 08:49:25 2001
From: martin.franklin at westerngeco.com (Martin Franklin)
Date: Wed, 28 Feb 2001 07:49:25 +0000
Subject: Looking for better Tkinter's listbox widget
References: <kr5379.9hl.ln@127.0.0.1> <3A9BEF9D.2050507@mobileway.com>
Message-ID: <3A9CAD85.35EA4597@westerngeco.com>

Try MCScrolledListBox it is a Pmw/PmwContribD by Doug Hellmann
 http://members.home.net/doughellmann/PmwContribD/index.html

jean-paul roumian wrote:
> 
> gradha at iname.com wrote:
> 
> > Hello.
> >
> > I am looking forward to writting a Phython GUI for a program, and I need
> > to show a listbox with at least multiple columns, where it would be just
> > nice (not really needed) to display an icon in one of them. Such listbox
> > items should be individually selectable.
> >
> > I've read Tkinter's Fredrik Introduction, and there it says that Tkinter's
> > native listbox doesn't support multiple columns. Where can I find a
> > replacement widget providing the multicolumn feature and being portable
> > to Windows and X?
> 
> did you tried GtkCTree ? (from pygtk)
> 
> >
> 
> --
> "The Pythagorean Theorem employed 24 words, the Lord's Prayer has 66 words,
> Archimedes Principle has 67words, the 10 Commandments have 179 words,
> the Declaration of Independence 1,300 words,
> and finally the European Commission's regulation on the sale of
> cabbage: 26,911 words."
> 
> --
> http://mail.python.org/mailman/listinfo/python-list



From chrisw at nipltd.com  Thu Feb 22 15:21:12 2001
From: chrisw at nipltd.com (Chris Withers)
Date: Thu, 22 Feb 2001 14:21:12 +0000
Subject: Newbie Error? Compiling _tkinter from Source on Win32
Message-ID: <3A952058.D2D63AFE@nipltd.com>

>From PCBuild/readme.txt:

> _tkinter
>     Python wrapper for the Tk windowing system.  Requires tcl832.exe from
>         http://dev.scriptics.com/software/tcltk/downloadnow83.html
>     Run the installer, forcing installation into dist\Tcl.
>     Be sure to install everything, including the Tcl/Tk header files.

So I installed tcl832 into E:\Python-2.0\Tcl and then tried to use VC++ 6.0 to
build the _tkinter module, but I got:

> --------------------Configuration: _tkinter - Win32 Release--------------------
> Compiling...
> _tkinter.c
> E:\Python-2.0\Modules\_tkinter.c(44) : fatal error C1083: Cannot open include file: 'tcl.h': No such file or directory
> tkappinit.c
> E:\Python-2.0\Modules\tkappinit.c(15) : fatal error C1083: Cannot open include file: 'tcl.h': No such file or directory
> Error executing cl.exe.
> 
> _tkinter.pyd - 2 error(s), 0 warning(s)

Any ideas what I'm doing wrong?

cheers,

Chris



From fredrik at pythonware.com  Thu Feb 22 15:57:52 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 14:57:52 GMT
Subject: Newbie Error? Compiling _tkinter from Source on Win32
References: <mailman.982851663.3862.python-list@python.org>
Message-ID: <QN9l6.19328$AH6.2619674@newsc.telia.net>

Chris Withers wrote:
> > --------------------Configuration: _tkinter - Win32 Release--------------------
> > Compiling...
> > _tkinter.c
> > E:\Python-2.0\Modules\_tkinter.c(44) : fatal error C1083:
> > Cannot open include file: 'tcl.h': No such file or directory
> > tkappinit.c
> > E:\Python-2.0\Modules\tkappinit.c(15) : fatal error C1083:
> > Cannot open include file: 'tcl.h': No such file or directory
> > Error executing cl.exe.
> >
> > _tkinter.pyd - 2 error(s), 0 warning(s)
>
> Any ideas what I'm doing wrong?

Does "newbie" mean C newbie?  The error message pretty much
means what it says: the compiler cannot find and open that file.

Make sure you have tcl.h somewhere, and check the compiler's
include settings.

(iirc, the distributed build file expects to find the Tcl/Tk include
files in "<python-sources>\..\tcl\include")

Cheers /F




From tim.one at home.com  Fri Feb 23 07:33:07 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 23 Feb 2001 01:33:07 -0500
Subject: Newbie Error? Compiling _tkinter from Source on Win32
In-Reply-To: <3A952058.D2D63AFE@nipltd.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEGMJBAA.tim.one@home.com>

[posted & mailed]

[Chris Withers]
> From PCBuild/readme.txt:

>> _tkinter
>>     Python wrapper for the Tk windowing system.  Requires tcl832.exe
>>     from
> >         http://dev.scriptics.com/software/tcltk/downloadnow83.html
> >     Run the installer, forcing installation into dist\Tcl.
> >     Be sure to install everything, including the Tcl/Tk header files.

> So I installed tcl832 into E:\Python-2.0\Tcl and then tried to
> use VC++ 6.0 to build the _tkinter module, but I got:
> ...
> > E:\Python-2.0\Modules\_tkinter.c(44) : fatal error C1083:
> Cannot open include file: 'tcl.h': No such file or directory
> ...

Also from PCBuild/readme.txt:

    The following subprojects will generally NOT build out of the box.
    They wrap code Python doesn't control, and you'll need to download
    the base packages first and unpack them into siblings of PCbuilds's
    parent directory; for example, if your PCbuild is  .......\dist\src
    \PCbuild\, unpack into new subdirectories of dist\.

That's why the part you quoted said "forcing installation into dist\Tcl".
Unless your Python-2.0 is a sibling of PCbuild's parent directory (and,
trust me, it isn't <wink>), you're simply not following the instructions.

Bit o' advice:  the first thing a newbie should not do, and especially not
on Windows, is get creative about where they install things.  If you follow
instructions to the letter, they'll work; *then* you can get creative.




From rich_somerfield at tertio.com  Thu Feb 22 16:00:52 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Thu, 22 Feb 2001 15:00:52 -0000
Subject: Tkinter problems
Message-ID: <BP9l6.9665$MN.255212@news2-win.server.ntlworld.com>

I get the following exception when trying to load a jpeg image.

My system is Linux SuSE6.4 (however I have downloaded and recompiled Python
and PIL - the original versions had the same problem, i thought it would go
away if i upgraded, it didnt.) Python 2.0 / Pil 1.1.1 / Tk 8.0.4

I have been using my code successfully using the py15 installation from
PythonWare.

BTW, what is the problem with my Python interpreter?? The Warning msg about
the API version.


Regards
Rich


rsomerfield at FullCycle:~/SGE/sge > python sge_2.0.1.py
WARNING: Python C API version mismatch for module _tkinter:
  This Python has API version 1009, module _tkinter has version 1007.
WARNING: Python C API version mismatch for module _imaging:
  This Python has API version 1009, module _imaging has version 1007.
Traceback (most recent call last):
  File "sge_2.0.1.py", line 4585, in ?
    sge = SimpleGameEngine()
  File "sge_2.0.1.py", line 141, in __init__
    exec("self.main_function()")
  File "<string>", line 1, in ?
  File "sge_2.0.1.py", line 123, in main_function
    self.TG.produce_board_locations()
  File "sge_2.0.1.py", line 1807, in produce_board_locations
    self.vars.boardlocationimages[i] = self.DI.process_individual_image(i,
self.vars.theimageblank1)
  File "sge_2.0.1.py", line 1870, in process_individual_image
    theimagetemp = ImageTk.PhotoImage(theimagetemp)
  File "/usr/lib/python/site-packages/PIL/ImageTk.py", line 83, in __init__
    self.paste(image)
  File "/usr/lib/python/site-packages/PIL/ImageTk.py", line 113, in paste
    self.__photo.tk.call("PyImagingPhoto", self.__photo, block.id)
TclError: invalid command name "PyImagingPhoto"
rsomerfield at FullCycle:~/SGE/sge >




From whisper at oz.net  Fri Feb 23 20:09:19 2001
From: whisper at oz.net (Dave LeBlanc)
Date: 23 Feb 2001 19:09:19 GMT
Subject: Tkinter problems
References: <BP9l6.9665$MN.255212@news2-win.server.ntlworld.com>
Message-ID: <976cgv$4ii$0@216.39.170.247>

Oh, it's just a guess, but i'd say the version of tkinter you're
trying to use is incompatible with the version of python you're trying
to use.

Possible causes:
	Python 2.0 is finding an older version of tkinter (perhaps
associated with a Python 1.5.2 or 1.6.0 install?).

	You downloaded tkinter and python at different times.

	You recompiled Python but not tkinter.

My suggestion would be to get a standard python 2.0 distro with
tkinter included (does it actually come any other way?) and build and
install the whole mass at once.

BTW, Tcl 8.0.4 is _hopelessly_ out of date, having been superceded by
Tcl 8.0.5 long ago (not to mention many further versions not used with
the Python 1.5.2 version of tkinter).  AFAIK from reading, tkinter for
Python 2.0 now uses Tcl 8.3.2 which, as always, is available through
www.scriptics.com at sourceforge.

HTH

On Thu, 22 Feb 2001 15:00:52 -0000, "Rich Somerfield"
<rich_somerfield at tertio.com> wrote:

>I get the following exception when trying to load a jpeg image.
>
>My system is Linux SuSE6.4 (however I have downloaded and recompiled Python
>and PIL - the original versions had the same problem, i thought it would go
>away if i upgraded, it didnt.) Python 2.0 / Pil 1.1.1 / Tk 8.0.4
>
>I have been using my code successfully using the py15 installation from
>PythonWare.
>
>BTW, what is the problem with my Python interpreter?? The Warning msg about
>the API version.
>
>
>Regards
>Rich
>
>
>rsomerfield at FullCycle:~/SGE/sge > python sge_2.0.1.py
>WARNING: Python C API version mismatch for module _tkinter:
>  This Python has API version 1009, module _tkinter has version 1007.
>WARNING: Python C API version mismatch for module _imaging:
>  This Python has API version 1009, module _imaging has version 1007.
>Traceback (most recent call last):
>  File "sge_2.0.1.py", line 4585, in ?
>    sge = SimpleGameEngine()
>  File "sge_2.0.1.py", line 141, in __init__
>    exec("self.main_function()")
>  File "<string>", line 1, in ?
>  File "sge_2.0.1.py", line 123, in main_function
>    self.TG.produce_board_locations()
>  File "sge_2.0.1.py", line 1807, in produce_board_locations
>    self.vars.boardlocationimages[i] = self.DI.process_individual_image(i,
>self.vars.theimageblank1)
>  File "sge_2.0.1.py", line 1870, in process_individual_image
>    theimagetemp = ImageTk.PhotoImage(theimagetemp)
>  File "/usr/lib/python/site-packages/PIL/ImageTk.py", line 83, in __init__
>    self.paste(image)
>  File "/usr/lib/python/site-packages/PIL/ImageTk.py", line 113, in paste
>    self.__photo.tk.call("PyImagingPhoto", self.__photo, block.id)
>TclError: invalid command name "PyImagingPhoto"
>rsomerfield at FullCycle:~/SGE/sge >
>
>



From rich_somerfield at tertio.com  Mon Feb 26 10:17:14 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Mon, 26 Feb 2001 09:17:14 -0000
Subject: Tkinter problems
References: <BP9l6.9665$MN.255212@news2-win.server.ntlworld.com> <976cgv$4ii$0@216.39.170.247>
Message-ID: <vapm6.27678$5n4.595310@news6-win.server.ntlworld.com>

Sounds like u might have hit the nail on the head.  I was having problems
recompiling Tkinter [path problem!!]:

bash-2.03# python setup.py install
running install
Traceback (most recent call last):
  File "setup.py", line 25, in ?
    long_description = __doc__
  File "/usr/bin/python-2.0/Lib/distutils/core.py", line 138, in setup
    dist.run_commands()
  File "/usr/bin/python-2.0/Lib/distutils/dist.py", line 829, in
run_commands
    self.run_command(cmd)
  File "/usr/bin/python-2.0/Lib/distutils/dist.py", line 848, in run_command
    cmd_obj.ensure_finalized()
  File "/usr/bin/python-2.0/Lib/distutils/cmd.py", line 112, in
ensure_finalized
    self.finalize_options()
  File "/usr/bin/python-2.0/Lib/distutils/command/install.py", line 255, in
fina
lize_options
    (prefix, exec_prefix) = get_config_vars('prefix', 'exec_prefix')
  File "/usr/bin/python-2.0/Lib/distutils/sysconfig.py", line 368, in
get_config
_vars
    func()
  File "/usr/bin/python-2.0/Lib/distutils/sysconfig.py", line 280, in
_init_posi
x
    raise DistutilsPlatformError, my_msg
distutils.errors.DistutilsPlatformError: invalid Python installation: unable
to
open /usr/local/lib/python2.0/config/Makefile (No such file or directory)
bash-2.03#

I dont suppose u know the correct command line arguments to build Tkinter?
I couldnt find any documentation with the tar that i downloaded and the help
from the setup.py that was included didnt really suggest the best / exact
method.
My Python distro is in "/usr/bin/python-2.0", i cant find anything that
specifies where the Python install resides [ the problem with the above
command is that it is looking in /usr/local/lib/python2.0 !!!! - WRONG
PATH].

If i do download a later version of Tk, do i need to recompile Python2.0 AND
Tkinter??  Or will Python pick up the libraries dynamically??

Regards
Rich



Dave LeBlanc <whisper at oz.net> wrote in message
news:976cgv$4ii$0 at 216.39.170.247...
> Oh, it's just a guess, but i'd say the version of tkinter you're
> trying to use is incompatible with the version of python you're trying
> to use.
>
> Possible causes:
> Python 2.0 is finding an older version of tkinter (perhaps
> associated with a Python 1.5.2 or 1.6.0 install?).
>
> You downloaded tkinter and python at different times.
>
> You recompiled Python but not tkinter.
>
> My suggestion would be to get a standard python 2.0 distro with
> tkinter included (does it actually come any other way?) and build and
> install the whole mass at once.
>
> BTW, Tcl 8.0.4 is _hopelessly_ out of date, having been superceded by
> Tcl 8.0.5 long ago (not to mention many further versions not used with
> the Python 1.5.2 version of tkinter).  AFAIK from reading, tkinter for
> Python 2.0 now uses Tcl 8.3.2 which, as always, is available through
> www.scriptics.com at sourceforge.
>
> HTH
>
> On Thu, 22 Feb 2001 15:00:52 -0000, "Rich Somerfield"
> <rich_somerfield at tertio.com> wrote:
>
> >I get the following exception when trying to load a jpeg image.
> >
> >My system is Linux SuSE6.4 (however I have downloaded and recompiled
Python
> >and PIL - the original versions had the same problem, i thought it would
go
> >away if i upgraded, it didnt.) Python 2.0 / Pil 1.1.1 / Tk 8.0.4
> >
> >I have been using my code successfully using the py15 installation from
> >PythonWare.
> >
> >BTW, what is the problem with my Python interpreter?? The Warning msg
about
> >the API version.
> >
> >
> >Regards
> >Rich
> >
> >
> >rsomerfield at FullCycle:~/SGE/sge > python sge_2.0.1.py
> >WARNING: Python C API version mismatch for module _tkinter:
> >  This Python has API version 1009, module _tkinter has version 1007.
> >WARNING: Python C API version mismatch for module _imaging:
> >  This Python has API version 1009, module _imaging has version 1007.
> >Traceback (most recent call last):
> >  File "sge_2.0.1.py", line 4585, in ?
> >    sge = SimpleGameEngine()
> >  File "sge_2.0.1.py", line 141, in __init__
> >    exec("self.main_function()")
> >  File "<string>", line 1, in ?
> >  File "sge_2.0.1.py", line 123, in main_function
> >    self.TG.produce_board_locations()
> >  File "sge_2.0.1.py", line 1807, in produce_board_locations
> >    self.vars.boardlocationimages[i] =
self.DI.process_individual_image(i,
> >self.vars.theimageblank1)
> >  File "sge_2.0.1.py", line 1870, in process_individual_image
> >    theimagetemp = ImageTk.PhotoImage(theimagetemp)
> >  File "/usr/lib/python/site-packages/PIL/ImageTk.py", line 83, in
__init__
> >    self.paste(image)
> >  File "/usr/lib/python/site-packages/PIL/ImageTk.py", line 113, in paste
> >    self.__photo.tk.call("PyImagingPhoto", self.__photo, block.id)
> >TclError: invalid command name "PyImagingPhoto"
> >rsomerfield at FullCycle:~/SGE/sge >
> >
> >
>




From NoSpam at NoSpam.com  Thu Feb 22 17:59:08 2001
From: NoSpam at NoSpam.com (Tom)
Date: Thu, 22 Feb 2001 16:59:08 GMT
Subject: icons for python types??
Message-ID: <wzbl6.3855$XZ6.2212228@news4.rdc1.on.home.com>

I'm looking for icons to represent the standard python types (eg. function,
integer, frame).

Thanks,
Tom.




From cook at maya.com  Thu Feb 22 18:07:21 2001
From: cook at maya.com (Nicole Cook)
Date: Thu, 22 Feb 2001 12:07:21 -0500
Subject: another newboe mxodbc quest. -- update set
Message-ID: <000d01c09cf1$ebd79740$21fe46c0@maya.com>

I'm running python 1.5.2 on windows98 and using mxODBC to talk to an Access
database.   I'm having no trouble reading information from the database, but
now I want to write something to a column.

So I do :
db = ODBC.Windoes.connect("accessdb")
cs = db.cursor()

cs.execute("UPDATE Project SET Item='"+item+"' WHERE Name='"+name+"'")

When I do this from  the interpreter it returns 1, when I use it as a line
in a function Access freezes until I quit python.  Either way, nothing
happens to the column in the db that I'm trying to write.  Is there
something else I have to do to get mxODBC to execute an update?

Nicole




From db3l at fitlinxx.com  Sat Feb 24 02:08:21 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 23 Feb 2001 20:08:21 -0500
Subject: another newboe mxodbc quest. -- update set
References: <mailman.982861683.28194.python-list@python.org>
Message-ID: <uelwo99bu.fsf@ctwd0143.fitlinxx.com>

"Nicole Cook" <cook at maya.com> writes:

> I'm running python 1.5.2 on windows98 and using mxODBC to talk to an Access
> database.   I'm having no trouble reading information from the database, but
> now I want to write something to a column.
> 
> So I do :
> db = ODBC.Windoes.connect("accessdb")
> cs = db.cursor()
> 
> cs.execute("UPDATE Project SET Item='"+item+"' WHERE Name='"+name+"'")
> 
> When I do this from  the interpreter it returns 1, when I use it as a line
> in a function Access freezes until I quit python.  Either way, nothing
> happens to the column in the db that I'm trying to write.  Is there
> something else I have to do to get mxODBC to execute an update?

As far as an actual command goes, it looks ok (barring bad values in
either item or name that might mess up the quoting - you might want to
check that for debugging purposes).

Could you provide additional samples of the code - particularly how it
is used within the function?

As for seeing the change - mxODBC (conforming to the DB-API 2.0) sets
any database that supports transactions to use them by default.  I'm
not familiar enough with Access (which I'm guessing your using with
the default Jet engine?) to know if it supports them, but if it does,
then you won't see your change until you db.commit().  So if you just
exit your script (or drop the database connection object), your
changes will be rolled back.

mxODBC has an mxODBC specific named parameter on the connect() call
"clear_auto_commit" which can be used to defeat this default behavior,
as in:

    db = ODBC.Windows.connect(<database>,clear_auto_commit=0)

which will let the database's default transaction behavior take over.
That may or may not still require you to commit, depending on the
database and your ODBC definition for that database.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From embed at geocities.com  Thu Feb 22 18:22:40 2001
From: embed at geocities.com (Warren Postma)
Date: Thu, 22 Feb 2001 12:22:40 -0500
Subject: Randomize list order
Message-ID: <hXbl6.799$TW.6086@tor-nn1.netcom.ca>

You know, an in-place list.scramble() method would be nice; It could also be
useful card games and self test scripts. ;-)

I even have another use for it....

I have a BSDDB B+Tree file that when created all in one fell swoop, likes to
have its primary keys inserted in Chaotic order.  So, I wonder, is there a
better way to scramble my keys, or is this the most efficient way (lambda
would be less efficient even if on one line, since it would be generating
the necessary functions on demand, more or less, right?).

Warren

----

# scramble a list:

import random

def _scr1(a):
    "turns item a into (random#,a) pairs"
    return (random.randint(0,1000000),a)

def _scr2(a):
    "gets x from (random#,x) pairs"
    return a[1]

def _scramble_list( l ):
    "return a scrambled version of the list l. used by lwdb_btree files when
packing, to insert the new rows in random order."
    n = map(  _scr1, l) # create [ ( random1, keyvalue1), .... ] pairs
    n.sort() # sort by random numbers (scrambles list)
    return map( _scr2, n) # return back just the actual key values





From fredrik at pythonware.com  Thu Feb 22 18:43:17 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 17:43:17 GMT
Subject: Randomize list order
References: <hXbl6.799$TW.6086@tor-nn1.netcom.ca>
Message-ID: <Vccl6.19349$AH6.2625384@newsc.telia.net>

Warren Postma write:
> You know, an in-place list.scramble() method would be nice;
> It could also be useful card games and self test scripts. ;-)

random.shuffle(list)

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->






From embed at geocities.com  Thu Feb 22 19:02:03 2001
From: embed at geocities.com (Warren Postma)
Date: Thu, 22 Feb 2001 13:02:03 -0500
Subject: Randomize list order
References: <hXbl6.799$TW.6086@tor-nn1.netcom.ca> <Vccl6.19349$AH6.2625384@newsc.telia.net>
Message-ID: <bwcl6.816$TW.6179@tor-nn1.netcom.ca>

> random.shuffle(list)
>
> Cheers /F

Doh!  ;-)




From tim.one at home.com  Fri Feb 23 05:19:40 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 22 Feb 2001 23:19:40 -0500
Subject: Randomize list order
In-Reply-To: <bwcl6.816$TW.6179@tor-nn1.netcom.ca>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEGFJBAA.tim.one@home.com>

[Warren Postma, wants to shuffle]

[/F]
> random.shuffle(list)

[Warren]
> Doh!  ;-)

Not your fault!  For some reason (or utter lack of reason ...),
random.shuffle() never made it into the docs.  The docs for random were
extensively reworked for 2.1, and several new features added.  And shuffle()
is documented now.




From worm at gdp-group.com  Thu Feb 22 18:30:11 2001
From: worm at gdp-group.com (Thomas Volkmar Worm)
Date: Thu, 22 Feb 2001 18:30:11 +0100
Subject: Module rfc822 - uses file-objects - why?
Message-ID: <3A954CA3.72A6D51C@gdp-group.com>

Hi,

I started to write a little programm which reads mails from a POP3
account. To handle the message I thought, I use the rfc822 module or one
of its subclasses (mimetools). But the methods I want to use always
expect file-objects as parameters. Do I need to store every list/string
onto disk when I want to use these classes or methods on them?

To solve the problem I started to write a helper class which implements
methods like readline for my objects, but I really do not want to write
an emulation of the file-class again and again for each object I create.

It seems to me very odd to do this, so I guess there is another much
more clever way to use these classes without having files at all.

Regards
Thomas


-------------- next part --------------
A non-text attachment was scrubbed...
Name: worm.vcf
Type: text/x-vcard
Size: 366 bytes
Desc: Visitenkarte f?r Thomas Volkmar Worm
URL: <http://mail.python.org/pipermail/python-list/attachments/20010222/48ef2e04/attachment.vcf>

From fredrik at pythonware.com  Thu Feb 22 18:43:15 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 17:43:15 GMT
Subject: Module rfc822 - uses file-objects - why?
References: <3A954CA3.72A6D51C@gdp-group.com>
Message-ID: <Tccl6.19348$AH6.2625576@newsc.telia.net>

Thomas Volkmar Worm wrote:
> I started to write a little programm which reads mails from a POP3
> account. To handle the message I thought, I use the rfc822 module or one
> of its subclasses (mimetools). But the methods I want to use always
> expect file-objects as parameters. Do I need to store every list/string
> onto disk when I want to use these classes or methods on them?

import StringIO
file = StringIO.StringIO(data)

see the library reference for details.

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From aahz at panix.com  Thu Feb 22 18:50:01 2001
From: aahz at panix.com (Aahz Maruch)
Date: 22 Feb 2001 09:50:01 -0800
Subject: Module rfc822 - uses file-objects - why?
References: <3A954CA3.72A6D51C@gdp-group.com> <Tccl6.19348$AH6.2625576@newsc.telia.net>
Message-ID: <973jg9$dcj$1@panix3.panix.com>

In article <Tccl6.19348$AH6.2625576 at newsc.telia.net>,
Fredrik Lundh <fredrik at effbot.org> wrote:
>Thomas Volkmar Worm wrote:
>>
>> I started to write a little programm which reads mails from a POP3
>> account. To handle the message I thought, I use the rfc822 module or one
>> of its subclasses (mimetools). But the methods I want to use always
>> expect file-objects as parameters. Do I need to store every list/string
>> onto disk when I want to use these classes or methods on them?
>
>import StringIO
>file = StringIO.StringIO(data)

For performance:

try:
    from StringIO import StringIO
except ImportError:
    from cStringIO import StringIO

(I learned that idiom back when cStringIO wasn't automatically compiled;
I'm not sure it's necessary anymore.)
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From barry at digicool.com  Thu Feb 22 22:46:13 2001
From: barry at digicool.com (Barry A. Warsaw)
Date: Thu, 22 Feb 2001 16:46:13 -0500
Subject: Module rfc822 - uses file-objects - why?
References: <3A954CA3.72A6D51C@gdp-group.com>
	<Tccl6.19348$AH6.2625576@newsc.telia.net>
	<973jg9$dcj$1@panix3.panix.com>
Message-ID: <14997.34981.147019.748162@anthem.wooz.org>

>>>>> "AM" == Aahz Maruch <aahz at panix.com> writes:

    AM> (I learned that idiom back when cStringIO wasn't automatically
    AM> compiled; I'm not sure it's necessary anymore.)

I'm not sure either, but I still do it that way too.  Old habits 'n'
all...

-Barry



From stephen_purcell at yahoo.com  Fri Feb 23 08:07:51 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 23 Feb 2001 08:07:51 +0100
Subject: Module rfc822 - uses file-objects - why?
In-Reply-To: <973jg9$dcj$1@panix3.panix.com>; from aahz@panix.com on Thu, Feb 22, 2001 at 09:50:01AM -0800
References: <3A954CA3.72A6D51C@gdp-group.com> <Tccl6.19348$AH6.2625576@newsc.telia.net> <973jg9$dcj$1@panix3.panix.com>
Message-ID: <20010223080751.A7076@freedom.puma-ag.com>

Aahz Maruch wrote:
> For performance:
> 
> try:
>     from StringIO import StringIO
> except ImportError:
>     from cStringIO import StringIO


Or for even better performance:   <wink>

 try:
     from cStringIO import StringIO
 except ImportError:
     from StringIO import StringIO

Pedantic, but someone might try it the other way round and post a bunch of
messages complaining that it's too slow...  :-)

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From annis at biostat.wisc.edu  Thu Feb 22 18:53:03 2001
From: annis at biostat.wisc.edu (William Annis)
Date: 22 Feb 2001 11:53:03 -0600
Subject: Module rfc822 - uses file-objects - why?
References: <3A954CA3.72A6D51C@gdp-group.com>
Message-ID: <jtvgq2zjsw.fsf@wazor.biostat.wisc.edu>

Thomas Volkmar Worm <worm at gdp-group.com> writes:

>                                Do I need to store every list/string
> onto disk when I want to use these classes or methods on them?
> 
> To solve the problem I started to write a helper class which implements
> methods like readline for my objects, but I really do not want to write
> an emulation of the file-class again and again for each object I create.

        You have solved a problem other people have had and solved
already.  The StringIO class -- part of standard Python -- can be used
to make strings act like files:

        http://www.python.org/doc/current/lib/module-StringIO.html

-- 
William Annis - System Administrator - Biomedical Computing Group
annis at biostat.wisc.edu                       PGP ID:1024/FBF64031
Mi parolas Esperanton - La Internacian Lingvon  www.esperanto.org


From echuck at mindspring.com  Thu Feb 22 19:10:48 2001
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Thu, 22 Feb 2001 13:10:48 -0500
Subject: Duplicate modules problem
Message-ID: <5.0.2.1.0.20010222130757.04aaeb80@mail.mindspring.com>

You can find a tarball with this description and accompanying source code at:
     ftp://webware.sourceforge.net/pub/webware/ModulesProb01.tar.gz

I have a problem where Python creates duplicate modules in memory rather 
than reuse the same one. For example, a module in Pkg/Mod.py ends up in 
sys.modules under keys "Pkg.Mod" and "Mod" **pointing to two distinct 
modules**.

This creates further problems: Suppose a module Foo.py contains a class 
named Foo. If the module is loaded twice as two separate instances (of 
ModuleType), then there are 2 separate Foo classes. This causes confusion 
including the failure of an assertion such as:
     assert issubclass(foo, Foo)

The Foo in that code may be pointing to the first class, while the instance 
foo may have been created from the second class. The assertion then fails!

The problem stems from the fact that Python tracks modules by a relative, 
rather than absolute, path. A simple os.chdir() or a subtley in packages 
can cause this problem.

This problem is easiest to see in an os.chdir() situation:

C:\>mkdir foo
C:\>cd foo
C:\foo>mkdir bar
C:\foo>cd bar
C:\foo\bar>echo class baz: pass > baz.py
C:\foo\bar>echo ### > __init__.py
C:\foo\bar>cd ..
C:\foo>python
ActivePython 2.0, build 202 (ActiveState Tool Corp.)
based on Python 2.0 (#8, Oct 19 2000, 11:30:05) [MSC 32 bit (Intel)] on win32
 >>> from bar.baz import baz as baz1
 >>> import os
 >>> os.chdir('bar')
 >>> from baz import baz as baz2
 >>> baz1
<class bar.baz.baz at 007908BC>
 >>> baz2
<class baz.baz at 0079075C>
 >>> baz1 == baz2
0
 >>> import sys
 >>> [mod for mod in sys.modules.items() if mod[0].count('baz')]
[('baz', <module 'baz' from 'baz.pyc'>),
  ('bar.baz', <module 'bar.baz' from 'bar\baz.pyc'>)]
 >>> mod1 = sys.modules['bar.baz']
 >>> mod2 = sys.modules['baz']
 >>> id(mod1)
136147728
 >>> id(mod2)
136147744


This problem can also be seen without any use of os.chdir(). See 
ManufactureWare/ which contains the "assert issubclass(foo, Foo)" problem 
described above.

This all applies to Python 2.0 on Windows & UNIX. I haven't tried 2.1 yet.

I believe the solution is for Python to track modules by their absolute 
path. I don't know of any other resolution to the situation other than 
modifying Python in this manner. I also don't know of any disadvantage for 
Python to track modules by absolute path.

- Does anyone know of any workarounds?
- Does anyone know why it would be bad for Python to track modules by 
absolute path?
- Is there any chance Python will fix this in the future?
- If so, by 2.1?


-Chuck

ftp://webware.sourceforge.net/pub/webware/ModulesProb01.tar.gz




From ben.hutchings at roundpoint.com  Fri Feb 23 03:51:08 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 22 Feb 2001 18:51:08 -0800
Subject: Duplicate modules problem
References: <mailman.982865645.5169.python-list@python.org>
Message-ID: <u66i2kt7n.fsf@roundpoint.com>

Chuck Esterbrook <echuck at mindspring.com> writes:
<snip>
> I have a problem where Python creates duplicate modules in memory rather 
> than reuse the same one. For example, a module in Pkg/Mod.py ends up in 
> sys.modules under keys "Pkg.Mod" and "Mod" **pointing to two distinct 
> modules**.
<snip>
> The problem stems from the fact that Python tracks modules by a relative, 
> rather than absolute, path. A simple os.chdir() or a subtley in packages 
> can cause this problem.

Python doesn't track paths at all.  It has a name-indexed cache of
modules, accessible as sys.modules, which the import statement checks
and updates.  It's entirely possible to load modules without checking
the cache and/or without updating the cache.  It's also possible to
create a Python environment with all modules pre-loaded from some
location which might not be a file at all.

<snip example>

> I believe the solution is for Python to track modules by their
> absolute path. I don't know of any other resolution to the situation
> other than modifying Python in this manner.  I also don't know of
> any disadvantage for Python to track modules by absolute path.
<snip>

File/directory linking in Unix means that a single file can have
multiple paths, and as I said above some environments don't load
modules from files at all.  So your 'solution' would just move the
problem (a problem which you've created for yourself by abusing the
import statement, IMHO).  I think the real solution is to set the
Python module path appropriately and to give consistent module names
to the import statement.  If the behaviour of the import statement
really isn't what you want, use the lower-level module import
functions.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From tshumway at transafari.com  Fri Feb 23 04:05:46 2001
From: tshumway at transafari.com (Terrel Shumway)
Date: Thu, 22 Feb 2001 22:05:46 -0500
Subject: [Webware-discuss] Duplicate modules problem
References: <5.0.2.1.0.20010222130757.04aaeb80@mail.mindspring.com>
Message-ID: <3A95D38A.84F455AD@transafari.com>

> I believe the solution is for Python to track modules by their absolute
> path. I don't know of any other resolution to the situation other than
> modifying Python in this manner. I also don't know of any disadvantage for
> Python to track modules by absolute path.
>
> - Does anyone know of any workarounds?

The workaround is surprisingly easy:
    1) replace all non-absolute paths in sys.path (especially "").
    2) make sure that no module is accessible from more than one of these roots
(sufficent condition: no name in sys.path is a prefix of any other.)

This will make some things slightly less convenient (you will have to think more
carefully about your directory structure), but it will prevent you from shooting
yourself in the foot by importing a module from two different places.

Python 2.0 (#5, Dec  5 2000, 17:59:47)
[GCC 2.96 20000731 (Red Hat Linux 7.0)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import sys
>>> from pprint import pprint as pp
>>> import os
>>> pp(sys.path)
['',
 '/usr/local/lib/python2.0',
 '/usr/local/lib/python2.0/plat-linux2',
 '/usr/local/lib/python2.0/lib-tk',
 '/usr/local/lib/python2.0/lib-dynload',
 '/usr/local/lib/python2.0/site-packages']
>>> os.getcwd()
'/home/tshumway/projects/current/ModulesProb/foo'
>>> sys.path[0]=os.getcwd()
>>> from bar.baz import baz as baz1
>>> os.chdir("bar")
>>> from baz import baz as baz2
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: No module named baz
>>> from bar.baz import baz as baz2
>>> baz1==baz2
1
>>> pp(sys.path)
['/home/tshumway/projects/current/ModulesProb/foo',
 '/usr/local/lib/python2.0',
 '/usr/local/lib/python2.0/plat-linux2',
 '/usr/local/lib/python2.0/lib-tk',
 '/usr/local/lib/python2.0/lib-dynload',
 '/usr/local/lib/python2.0/site-packages']
>>> pp(sys.modules)
{'UserDict': <module 'UserDict' from '/usr/local/lib/python2.0/UserDict.pyc'>,
 '__builtin__': <module '__builtin__' (built-in)>,
 '__main__': <module '__main__' (built-in)>,
 'bar': <module 'bar' from
'/home/tshumway/projects/current/ModulesProb/foo/bar/__init__.pyc'>,
 'bar.baz': <module 'bar.baz' from
'/home/tshumway/projects/current/ModulesProb/foo/bar/baz.pyc'>,
 'cStringIO': <module 'cStringIO' from
'/usr/local/lib/python2.0/lib-dynload/cStringIO.so'>,
....

Notice that although "/usr/local/lib/python2.0" is a prefix of four other names,
those other names are carefully choosen to be illegal as package names. (e.g.
import plat-linux.foo is a syntax error.)  I think the default sys.path under
Windows is not quite as clean.

>
> - Does anyone know why it would be bad for Python to track modules by
> absolute path?

It would make it more difficult to move modules around. (Java demonstrates
this.)  Tool support would help: it is easy to search for import statements.

>
> - Is there any chance Python will fix this in the future?

Hmmm.  Seems like it does not really need a fix from On High: only a
widely-published HOWTO documenting the problem and the solution.


    HTH
    -- Terrel

FactoryKit/main.py:
-------------------
# main.py

# Make sure FactoryKit is reachable, since widgets
# will want to import FactoryKit.Widget.
try:
        import FactoryKit
except ImportError:
        import os, sys
        # XXX dangerous: should make sure it is really replacing ""
        sys.path[0] = os.path.abspath(os.pardir)
        import FactoryKit

from FactoryKit import Factory
# A simple "import Factory" raises an exception, as it should.
# Since __main__ does not belong to the FactoryKit package,
# it cannot use the package-relative import shortcuts

Factory.LoadAndMakeWidget('Examples/ExampleWidget.py')






From echuck at mindspring.com  Fri Feb 23 04:12:11 2001
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Thu, 22 Feb 2001 22:12:11 -0500
Subject: [Webware-discuss] Duplicate modules problem
In-Reply-To: <3A95D38A.84F455AD@transafari.com>
References: <5.0.2.1.0.20010222130757.04aaeb80@mail.mindspring.com>
Message-ID: <5.0.2.1.0.20010222220319.04758a60@mail.mindspring.com>

At 10:05 PM 2/22/2001 -0500, Terrel Shumway wrote:
> > - Does anyone know why it would be bad for Python to track modules by
> > absolute path?
>
>It would make it more difficult to move modules around. (Java demonstrates
>this.)  Tool support would help: it is easy to search for import statements.

I don't follow that at all. This is strictly a run time phenomena. Who 
wants to move modules once the program starts?


> > - Is there any chance Python will fix this in the future?
>
>Hmmm.  Seems like it does not really need a fix from On High: only a
>widely-published HOWTO documenting the problem and the solution.

Well unless your previous comment pans out, there doesn't seem to be a 
reason to track them by relative path. And since doing so leads to 
problems, it seems that an On High fix would be more appropriate than 
considering the situation to be OK simply because it's documented.


Also, I'm not really sure if I want to replace '' in sys.path. I thought 
that was there as a relative path to the current module (not the current 
directory) and therefore helped a given file Foo.py say "import Bar from 
Bar" where Bar.py resided in the same package. Particularly if you didn't 
even import Foo directly.  Am I wrong about that?

As Geoff pointed out, we most likely encountered this problem because we're 
running a program out of a package, which is uncommon. I tried Geoff's fix 
on the example code I wrote and it worked like a charm. I'll try it on 
Webware next.

If the fix pans out for Webware, that will be great, but unless there is a 
concrete advantage to tracking packages by relative path, I still recommend 
a change in Python to track modules by absolute path. That would eliminate 
accidently getting 2 distinct copies of the same module for any Python program.


Thanks for the input,
-Chuck




From tshumway at transafari.com  Fri Feb 23 17:29:03 2001
From: tshumway at transafari.com (Terrel Shumway)
Date: Fri, 23 Feb 2001 11:29:03 -0500
Subject: [Webware-discuss] Duplicate modules problem
References: <5.0.2.1.0.20010222130757.04aaeb80@mail.mindspring.com> <5.0.2.1.0.20010222220319.04758a60@mail.mindspring.com>
Message-ID: <3A968FCF.31ABEE10@transafari.com>

Chuck Esterbrook wrote:

> At 10:05 PM 2/22/2001 -0500, Terrel Shumway wrote:
> > > - Does anyone know why it would be bad for Python to track modules by
> > > absolute path?
> >
> >It would make it more difficult to move modules around. (Java demonstrates
> >this.)  Tool support would help: it is easy to search for import statements.
>
> I don't follow that at all. This is strictly a run time phenomena. Who
> wants to move modules once the program starts?

No, this is not a runtime problem, it is a refactoring-time problem.  In Java,
just try renaming a package and see how long it takes you to get a clean compile
without using a tool like WoodenChair.
The people who designed Python's hierarchical package system wanted people to be
able to gather modules and packages into packages without breaking them.  The
package relative import is a fairly clean solution to this fairly nasty problem.
Tool support would be helpful for external code that uses the moved modules.
Consider this very very simple scenario:

ham.py
---
class Ham:...

eggs.py
---
class Eggs...
class Dozen...
spam.py
---
from ham import Ham
from eggs import EggMixin
class Spam(Ham,EggMixin):...
class Can...

spam_eater.py
---------
from spam import Spam
import eggs

factory = Spam()
breakfast = factory.getCan()
breakfast.eat()
lunch = eggs.Dozen()
lunch.eat()
supper = Eggs.leftovers((breakfast,lunch))
supper.eat("Yum,yum!")


spam_eater is an application that uses ham, eggs, and spam as library modules.
Suppose now that we wanted to serve Spam and Eggs over HTTP.  We could just dump
all of these modules into the same directory with Request,Response,Servlet,
et.al., (or, equivalently, add the directory to sys.path) but that would not be
clean.  Lets put them together into a package: FoodKit.

FoodKit
    +- __init__.py
    +- spam.py
    +- ham.py
    +- eggs.py

bin
    +- spam_eater.py

Now in our web server we can say:
import FoodKit
serveit = FoodKit.spam.Can()

spam,eggs,and ham will work without modification. spam_eater would too, if we had
put it in the package, but we decided that we shouldn't mix applications and
libraries.
Since FoodKit is not in sys.path (If it were, we would get the nasty duplicate
module problems that started this thread.)  spam_eater gets an ImportError because
it cannot find spam or eggs.  The solution is to add the package name, just as the
web server does.

from FoodKit.spam import Spam
from FoodKit import eggs

Finding and fixing all of these broken import statements is what tools can do.
Note that the problem would be much larger without the package relative imports.
(cf. Java)  Also note that this example is extremely simple.  A more realistic
scenario would probably involve multiple packages and dozens of modules. Without
tool support, this type of refactoring tends to get put off until it is really
nasty.

> Also, I'm not really sure if I want to replace '' in sys.path. I thought
> that was there as a relative path to the current module (not the current
> directory) and therefore helped a given file Foo.py say "import Bar from
> Bar" where Bar.py resided in the same package. Particularly if you didn't
> even import Foo directly.  Am I wrong about that?

No, the package relative import works without "" in sys.path.

> As Geoff pointed out, we most likely encountered this problem because we're
> running a program out of a package, which is uncommon. I tried Geoff's fix
> on the example code I wrote and it worked like a charm. I'll try it on
> Webware next.

A program (script) running within a package, if it tries to use the package
relative imports, is broken.  __main__ is in the default package.  Trying to use
relative imports from __main__ is breaking the package encapsulation.

> If the fix pans out for Webware, that will be great, but unless there is a
> concrete advantage to tracking packages by relative path,

There is a concrete advantage: a library module does not need to know where it is
in the package hierarchy. Renaming the package does not require any code changes,
as it does for example in Java, which uses absolute names.  Only external clients
need code changes.

> I still recommend
> a change in Python to track modules by absolute path. That would eliminate
> accidently getting 2 distinct copies of the same module for any Python program.

Python does track modules by *filename*. If all of the names in sys.path are
absolute, then you can never get a filename that is not absolute.

-- Terrel




From echuck at mindspring.com  Fri Feb 23 20:31:29 2001
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Fri, 23 Feb 2001 14:31:29 -0500
Subject: Duplicate modules problem
Message-ID: <5.0.2.1.0.20010223142751.047993f0@mail.mindspring.com>

In summary, the problem is that when launching a program from within a 
Python package, confusion can result about paths of modules resulting in 
duplicate modules residing in memory (and therefore duplication classes 
(and therefore problems with isinstance())).

The solution was to provide a Launch.py program that essentially does an 
"os.chdir(os.pardir))" before importing the package, its
"starter" module and invoking main().

Also, a "if '' not in sys.path: sys.path.insert(0, '')" was needed to fix 
problems when starting with "Launch.py" rather than "python Launch.py".

Overall this solution is pretty small and keeps things straight.

Thanks to Geoff Talvola for the fix and everyone else for their input.


-Chuck



"Chuck Esterbrook" <echuck at mindspring.com> wrote in message 
news:<mailman.982865645.5169.python-list at python.org>...
>You can find a tarball with this description and accompanying source code at:
>      ftp://webware.sourceforge.net/pub/webware/ModulesProb01.tar.gz
>I have a problem where Python creates duplicate modules in memory rather 
>than reuse the same one. For example, a module in Pkg/Mod.py ends up in 
>sys.modules under keys "Pkg.Mod" and "Mod" **pointing to two distinct 
>modules**.
>This creates further problems: Suppose a module Foo.py contains a class 
>named Foo. If the module is loaded twice as two separate instances (of 
>ModuleType), then there are 2 separate Foo classes. This causes confusion 
>including the failure of an assertion such as:
>      assert issubclass(foo, Foo)
>The Foo in that code may be pointing to the first class, while the 
>instance foo may have been created from the second class. The assertion 
>then fails!
>The problem stems from the fact that Python tracks modules by a relative, 
>rather than absolute, path. A simple os.chdir() or a subtley in packages 
>can cause this problem.
>This problem is easiest to see in an os.chdir() situation:
>C:\>mkdir foo
>C:\>cd foo
>C:\foo>mkdir bar
>C:\foo>cd bar
>C:\foo\bar>echo class baz: pass > baz.py
>C:\foo\bar>echo ### > __init__.py
>C:\foo\bar>cd ..
>C:\foo>python
>ActivePython 2.0, build 202 (ActiveState Tool Corp.)
>based on Python 2.0 (#8, Oct 19 2000, 11:30:05) [MSC 32 bit (Intel)] on win32
>  >>> from bar.baz import baz as baz1
>  >>> import os
>  >>> os.chdir('bar')
>  >>> from baz import baz as baz2
>  >>> baz1
><class bar.baz.baz at 007908BC>
>  >>> baz2
><class baz.baz at 0079075C>
>  >>> baz1 == baz2
>0
>  >>> import sys
>  >>> [mod for mod in sys.modules.items() if mod[0].count('baz')]
>[('baz', <module 'baz' from 'baz.pyc'>),
>   ('bar.baz', <module 'bar.baz' from 'bar\baz.pyc'>)]
>  >>> mod1 = sys.modules['bar.baz']
>  >>> mod2 = sys.modules['baz']
>  >>> id(mod1)
>136147728
>  >>> id(mod2)
>136147744
>
>This problem can also be seen without any use of os.chdir(). See 
>ManufactureWare/ which contains the "assert issubclass(foo, Foo)" problem 
>described above.
>This all applies to Python 2.0 on Windows & UNIX. I haven't tried 2.1 yet.
>I believe the solution is for Python to track modules by their absolute 
>path. I don't know of any other resolution to the situation other than 
>modifying Python in this manner. I also don't know of any disadvantage for 
>Python to track modules by absolute path.
>- Does anyone know of any workarounds?
>- Does anyone know why it would be bad for Python to track modules by 
>absolute path?
>- Is there any chance Python will fix this in the future?
>- If so, by 2.1?
>
>-Chuck
>ftp://webware.sourceforge.net/pub/webware/ModulesProb01.tar.gz




From echuck at mindspring.com  Fri Feb 23 21:58:02 2001
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Fri, 23 Feb 2001 15:58:02 -0500
Subject: Duplicate modules problem
In-Reply-To: <3A96C967.4D3420EE@transafari.com>
References: <5.0.2.1.0.20010223142751.047993f0@mail.mindspring.com>
Message-ID: <5.0.2.1.0.20010223153906.047404f0@mail.mindspring.com>

At 03:34 PM 2/23/2001 -0500, Terrel Shumway wrote:
>Chuck Esterbrook wrote:
>
> > In summary, the problem is that when launching a program from within a
> > Python package, confusion can result about paths of modules resulting in
> > duplicate modules residing in memory (and therefore duplication classes
> > (and therefore problems with isinstance())).
> >
> > The solution was to provide a Launch.py program that essentially does an
> > "os.chdir(os.pardir))" before importing the package, its
> > "starter" module and invoking main().
>
>I don't understand why this is necessary.  It seems much cleaner to me to 
>just be careful about sys.path and use the correct import statements.

Um, I never had any incorrect or abusive use of import statements.

When a file sat in the same directory, I said:
   from Foo import Foo

When it sat in an external package, I said:
   from Pkg.Foo import Foo

I learned these conventions from Python. I didn't invent them.

Perhaps I had an "incorrect" situation if it was never intended that 
programs start out of their own packages.


> > Also, a "if '' not in sys.path: sys.path.insert(0, '')" was needed to fix
> > problems when starting with "Launch.py" rather than "python Launch.py".
> >
>
>Please explain this problem. Can you show us Lauch.py and examples of its use?

You can get it from CVS:
https://sourceforge.net/cvs/?group_id=4866

See WebKit/Launch.py.


Again, the problem is when running a program out of a package. And if you 
don't mind duplicate modules, you don't need the fix. But duplicates 
eventually lead to subtle, arcane bugs. We need the fix in WebKit.

See more below.


>I think having "" in sys.path is just asking for trouble, but oh well -- 
>leave it in if you like.  The ManufactureWare/FactoryKit/main.py works 
>fine if you correct the broken import statement. __main__ is not part of 
>the package and cannot use the relative import shortcut:  "from FactoryKit 
>import Factory" is the correct import.  Removing "" from sys.path just 
>makes the breakage more visible.

Regarding the import, the fix does what you describe: Makes the package 
explicit to the program.


Regarding '' in sys.path, if you say:
   > python someprog.py

you get '' as the first element in the path. Again that's Python, not me.

However, if you say:
   > someprog.py

then you get an absolute path in place of ''. Geoff ran into this first and 
it caused problems, hence the fix. Since the fix puts you back to the 
sys.path you would have gotten from Python anyway (by launching 
differently), it doesn't feel so evil to me.

Regarding my choice of:
   os.chdir(os.pardir)

I could done something like this instead:
   sys.path.insert(1, os.path.abspath(os.pardir))

The first seemed a tad simpler.

Also, regarding the use of Launch.py, I might have chosen to bake this 
directly into the app server, BUT the problem is that we have at least 2 of 
them and will be adding another in the future. Also, Launch.py is a very 
small, very clean file, making it easier to deal with independent of all 
that goes on inside the app server source files.


I have been flexing the plug-ins, examples and test cases for WebKit and 
things are working good. At the very least, they're working much better 
than before, even if Launch.py doesn't suit your taste.

The standard way to launch the app server is still:

 > cd Webware/WebKit
 > ./AppServer


-Chuck
--
Get web dev at http://webware.sourceforge.net/




From db3l at fitlinxx.com  Sat Feb 24 02:34:26 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 23 Feb 2001 20:34:26 -0500
Subject: Duplicate modules problem
References: <5.0.2.1.0.20010223142751.047993f0@mail.mindspring.com> <mailman.982962124.8879.python-list@python.org>
Message-ID: <uae7c984d.fsf@ctwd0143.fitlinxx.com>

Chuck Esterbrook <echuck at mindspring.com> writes:

> Regarding '' in sys.path, if you say:
>    > python someprog.py
> 
> you get '' as the first element in the path. Again that's Python, not me.
> 
> However, if you say:
>    > someprog.py
> 
> then you get an absolute path in place of ''. Geoff ran into this first and 
> it caused problems, hence the fix. Since the fix puts you back to the 
> sys.path you would have gotten from Python anyway (by launching 
> differently), it doesn't feel so evil to me.

I expect that in both cases Python is just consistently taking the
dirname off of the script it is supplied on the command line and
inserting it as the first search entry (to include the script location
for imports).

When you run the script directly in the second case, I expect the
surrounding environment (Windows?) is using the full path when it
calls python.

> Regarding my choice of:
>    os.chdir(os.pardir)
> 
> I could done something like this instead:
>    sys.path.insert(1, os.path.abspath(os.pardir))

You'd want to use 0 to put it at the front of the path, right?

> The first seemed a tad simpler.

Simpler, but also environmentally more damaging since it affects an OS
environmental item that can affect non-Python code that might
eventually run in the process.

I'm not that familiar with how your application might be launched, but
if there's any chance of this code path being imported and run that
way (rather than directly from the command line), then I'd definitely
vote for the latter, since the former changes the current directory
for the existing process which might be unexpected by someone
importing your launch module.  Better to adjust a Python internal item
such as sys.path than an OS-environment feature, at least IMHO.

Even if it isn't generally expected to start that way, if there's any
good reason someone might wrap that module in their own code it might
prove unexpected.

I seem to have missed a post or two in this thread on my server, so
perhaps this violates some other requirement, but is there a reason
you just can't remove the first entry from sys.path in your launcher?

Since that's the location of the script, removing it would force
Python to use the normal search path, which your package should
already be on in order to be a package to Python?

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From patfourc at libertysurf.fr  Thu Feb 22 19:41:23 2001
From: patfourc at libertysurf.fr (Liberty)
Date: Thu, 22 Feb 2001 19:41:23 +0100
Subject: File browser
Message-ID: <973mgo$l3o$1@news2.isdnet.net>

Hello, I'm a french student looking for a file browser written in Python to
include it in my study project.

Does anyone got it ?


Thanks





From akuchlin at mems-exchange.org  Thu Feb 22 20:02:50 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: 22 Feb 2001 14:02:50 -0500
Subject: File browser
References: <973mgo$l3o$1@news2.isdnet.net>
Message-ID: <3dzofe35id.fsf@ute.cnri.reston.va.us>

"Liberty" <patfourc at libertysurf.fr> writes:
> Hello, I'm a french student looking for a file browser written in Python to
> include it in my study project.

cfm (cfm.sourceforge.net) is a cute little file browser that uses curses.

--amk


From rickylee at americasm01.nt.com  Thu Feb 22 20:32:50 2001
From: rickylee at americasm01.nt.com (Lee, Rick)
Date: Thu, 22 Feb 2001 14:32:50 -0500
Subject: Any way for a function to refer to itself?
Message-ID: <3A956962.3BBFED86@americasm01.nt.com>

I can't find any other way for a function or method to refer to itself
than something like the following:

def foo():
    myself = foo   # only if foo is global
    mydoc = myself.__doc__
    myname = myself.__name__

So it can be done this way, but:

- only if the function name can actually be accessed from the current
name space
- if the function name changes, that first line inside this function
also has to change

Seems to me there should be a more "Pythonic" way of doing this.  Is
there?



From jmarshal at mathworks.com  Thu Feb 22 21:17:23 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 22 Feb 2001 20:17:23 GMT
Subject: Any way for a function to refer to itself?
References: <3A956962.3BBFED86@americasm01.nt.com>
Message-ID: <973s4j$2h9$1@news.mathworks.com>

Lee, Rick <rickylee at americasm01.nt.com> wrote:
> I can't find any other way for a function or method to refer to itself
> than something like the following:

> def foo():
>     myself = foo   # only if foo is global
>     mydoc = myself.__doc__
>     myname = myself.__name__

> So it can be done this way, but:

> - only if the function name can actually be accessed from the current
> name space
> - if the function name changes, that first line inside this function
> also has to change

> Seems to me there should be a more "Pythonic" way of doing this.  Is
> there?

It might be unpleasant, but you can do something like:

  def fib(f, x):
     if x < 2: return 1
     return f(f, x-1) + f(f, x-2)

  print fib(fib, 10)


From jmarshal at mathworks.com  Thu Feb 22 21:22:16 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 22 Feb 2001 20:22:16 GMT
Subject: Any way for a function to refer to itself?
References: <3A956962.3BBFED86@americasm01.nt.com> <973s4j$2h9$1@news.mathworks.com>
Message-ID: <973sdo$2no$1@news.mathworks.com>

Joshua Marshall <jmarshal at mathworks.com> wrote:
> Lee, Rick <rickylee at americasm01.nt.com> wrote:
>> I can't find any other way for a function or method to refer to itself
>> than something like the following:

>> def foo():
>>     myself = foo   # only if foo is global
>>     mydoc = myself.__doc__
>>     myname = myself.__name__

>> So it can be done this way, but:

>> - only if the function name can actually be accessed from the current
>> name space
>> - if the function name changes, that first line inside this function
>> also has to change

>> Seems to me there should be a more "Pythonic" way of doing this.  Is
>> there?

> It might be unpleasant, but you can do something like:

>   def fib(f, x):
>      if x < 2: return 1
>      return f(f, x-1) + f(f, x-2)

>   print fib(fib, 10)

And I guess if you don't like the idea of always having to pass your
function in to itself, you can wrap it:

  def fib(x):
     def _fib(f, x):
        if x < 2: return 1
        return f(f, x-1) + f(f, x-2)
     return _fib(_fib, x)

  print fib(10)


From emile at fenx.com  Fri Feb 23 05:44:34 2001
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 22 Feb 2001 20:44:34 -0800
Subject: Any way for a function to refer to itself?
References: <3A956962.3BBFED86@americasm01.nt.com>
Message-ID: <974ptk$nsq9e$1@ID-11957.news.dfncis.de>

I'm not sure if it always plays nice, but you may try
starting with:

>>> def test():
 "test function description"
 import traceback
 me = eval( traceback.extract_stack()[-1][2])
 print "DocString: %s  functionName: %s " % (me.__doc__,
me.__name__)


>>> test()
DocString: test function description  functionName: test

HTH

--

Emile van Sebille
emile at fenx.com
-------------------


"Lee, Rick" <rickylee at americasm01.nt.com> wrote in message
news:3A956962.3BBFED86 at americasm01.nt.com...
> I can't find any other way for a function or method to
refer to itself
> than something like the following:
>
> def foo():
>     myself = foo   # only if foo is global
>     mydoc = myself.__doc__
>     myname = myself.__name__
>
> So it can be done this way, but:
>
> - only if the function name can actually be accessed from
the current
> name space
> - if the function name changes, that first line inside
this function
> also has to change
>
> Seems to me there should be a more "Pythonic" way of doing
this.  Is
> there?
>




From jmarshal at mathworks.com  Fri Feb 23 15:24:28 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 23 Feb 2001 14:24:28 GMT
Subject: Any way for a function to refer to itself?
References: <3A956962.3BBFED86@americasm01.nt.com> <974ptk$nsq9e$1@ID-11957.news.dfncis.de>
Message-ID: <975rqs$59o$1@news.mathworks.com>

Emile van Sebille <emile at fenx.com> wrote:
> I'm not sure if it always plays nice, but you may try
> starting with:

>>>> def test():
>  "test function description"
>  import traceback
>  me = eval( traceback.extract_stack()[-1][2])
>  print "DocString: %s  functionName: %s " % (me.__doc__,
> me.__name__)


>>>> test()
> DocString: test function description  functionName: test

This doesn't do what was asked - it only works if test is defined in
the global namespace:

  def f():
      def test():
          "test function description"
          import traceback
          me = eval( traceback.extract_stack()[-1][2])
          print "DocString: %s  functionName: %s " % (me.__doc__, me.__name__)
  
      test()
  
  f()

Yields:

  Traceback (innermost last):
    File "fu.py", line 10, in ?
      f()
    File "fu.py", line 8, in f
      test()
    File "fu.py", line 5, in test
      me = eval( traceback.extract_stack()[-1][2])
    File "<string>", line 0, in ?
  NameError: test


From emile at fenx.com  Fri Feb 23 17:37:15 2001
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 23 Feb 2001 08:37:15 -0800
Subject: Any way for a function to refer to itself?
References: <3A956962.3BBFED86@americasm01.nt.com> <974ptk$nsq9e$1@ID-11957.news.dfncis.de> <975rqs$59o$1@news.mathworks.com>
Message-ID: <9763lt$o4vka$1@ID-11957.news.dfncis.de>

Well then, it would need to be modified for additional
namespaces.  You could create a function that retrieves the
traceback for the caller, uses linecache to determine the
callers indentation level, walk up to the enclosing
indentation level, (unless already global), examine globals
for that name, verify it includes the right target name, and
then you'd have it.

See you,

--

Emile van Sebille
emile at fenx.com
-------------------


"Joshua Marshall" <jmarshal at mathworks.com> wrote in message
news:975rqs$59o$1 at news.mathworks.com...
> Emile van Sebille <emile at fenx.com> wrote:
> > I'm not sure if it always plays nice, but you may try
> > starting with:
>
> >>>> def test():
> >  "test function description"
> >  import traceback
> >  me = eval( traceback.extract_stack()[-1][2])
> >  print "DocString: %s  functionName: %s " % (me.__doc__,
> > me.__name__)
>
>
> >>>> test()
> > DocString: test function description  functionName: test
>
> This doesn't do what was asked - it only works if test is
defined in
> the global namespace:
>
>   def f():
>       def test():
>           "test function description"
>           import traceback
>           me = eval( traceback.extract_stack()[-1][2])
>           print "DocString: %s  functionName: %s " %
(me.__doc__, me.__name__)
>
>       test()
>
>   f()
>
> Yields:
>
>   Traceback (innermost last):
>     File "fu.py", line 10, in ?
>       f()
>     File "fu.py", line 8, in f
>       test()
>     File "fu.py", line 5, in test
>       me = eval( traceback.extract_stack()[-1][2])
>     File "<string>", line 0, in ?
>   NameError: test




From will at nospam.misconception.org.uk  Thu Feb 22 20:43:00 2001
From: will at nospam.misconception.org.uk (Will Newton)
Date: Thu, 22 Feb 2001 19:43:00 +0000
Subject: Python, Access and win32com
Message-ID: <973pvb$ms8$1@newsg1.svr.pol.co.uk>

I have been struggling to get MS Access 97 and the JET database engine to 
work with Active State Python and the win32com extensions.

If I run daodump.py I get the following error:

(Unfortuantely ASPython doesn't seem to have saved the log, why, I have no 
idea)

com_error(-(a large number, INT_MAX?),CLASS_E_NOTLICENSED...)



From gtalvola at nameconnector.com  Thu Feb 22 21:23:30 2001
From: gtalvola at nameconnector.com (Geoff Talvola)
Date: Thu, 22 Feb 2001 15:23:30 -0500
Subject: Python, Access and win32com
References: <973pvb$ms8$1@newsg1.svr.pol.co.uk>
Message-ID: <3A957542.476A501F@NameConnector.com>

Will Newton wrote:

> I have been struggling to get MS Access 97 and the JET database engine to
> work with Active State Python and the win32com extensions.
>
> If I run daodump.py I get the following error:
>
> (Unfortuantely ASPython doesn't seem to have saved the log, why, I have no
> idea)
>
> com_error(-(a large number, INT_MAX?),CLASS_E_NOTLICENSED...)

It might be the problem described in http://support.microsoft.com/support/kb/articles/q189/6/07.asp in which case you can use the FixDAO function in the attached module to fix the problem.

--


- Geoff Talvola
  Parlance Corporation
  gtalvola at NameConnector.com


-------------- next part --------------
import win32api, win32con, win32com.client, win32com.client.gencache

def FixDAO():
    """
    If DAO.DBEngine.35 cannot be created, this function will fix up the
    licensing Registry variables necessary so that it can be created.
    """
    try:
        key = win32api.RegOpenKeyEx(win32con.HKEY_CLASSES_ROOT,'LICENSES\\F4FC596D-DFFE-11CF-9551-00AA00A3DC45',0,win32con.KEY_SET_VALUE|win32con.KEY_READ)
    except:
        key = win32api.RegCreateKey(win32con.HKEY_CLASSES_ROOT,'LICENSES\\F4FC596D-DFFE-11CF-9551-00AA00A3DC45')
    if win32api.RegQueryValue(key,'') != 'mbmabptebkjcdlgtjmskjwtsdhjbmkmwtrak':
        win32api.RegSetValue(key, '', win32con.REG_SZ, 'mbmabptebkjcdlgtjmskjwtsdhjbmkmwtrak')

def DAO():
    """
    Create and return a DAO.DBEngine.35 object using early-bound automation
    """
    # make sure that the makepy imformation has been generated
    win32com.client.gencache.EnsureModule('{00025E01-0000-0000-C000-000000000046}', 0, 4, 0)
    # Try to create the DAO object; if it fails, fix the licensing and try again.
    try:
        return win32com.client.Dispatch('DAO.DBEngine.35')
    except:
        FixDAO()
        return win32com.client.Dispatch('DAO.DBEngine.35')


From timr at probo.com  Fri Feb 23 07:11:02 2001
From: timr at probo.com (Tim Roberts)
Date: Thu, 22 Feb 2001 22:11:02 -0800
Subject: Python, Access and win32com
References: <973pvb$ms8$1@newsg1.svr.pol.co.uk>
Message-ID: <qcvb9tckiuh1s64jtm7tp781pd37qgt0i8@4ax.com>

Will Newton <will at nospam.misconception.org.uk> wrote:

>
>I have been struggling to get MS Access 97 and the JET database engine to 
>work with Active State Python and the win32com extensions.
>
>If I run daodump.py I get the following error:
>
>(Unfortuantely ASPython doesn't seem to have saved the log, why, I have no 
>idea)
>
>com_error(-(a large number, INT_MAX?),CLASS_E_NOTLICENSED...)

It's not INT_MAX; CLASS_E_NOTLICENSED is Win32 error code 0x80040112.

This error says that either Access was not installed on your machine, or
the DAO DLLs were not properly registered.  This can happen if you do a
brain transplant to copy one hard disk to another, and you don't get the
registry quite right.

Does Access run standalone?
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From aleaxit at yahoo.com  Fri Feb 23 12:16:57 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 23 Feb 2001 12:16:57 +0100
Subject: Python, Access and win32com
References: <973pvb$ms8$1@newsg1.svr.pol.co.uk>
Message-ID: <975gr702df@news1.newsguy.com>

"Will Newton" <will at nospam.misconception.org.uk> wrote in message
news:973pvb$ms8$1 at newsg1.svr.pol.co.uk...
>
> I have been struggling to get MS Access 97 and the JET database engine to
> work with Active State Python and the win32com extensions.
>
> If I run daodump.py I get the following error:

I suggest not using DAO (an obsolete technology) but rather ADO (which
works just fine with many kinds of data sources, fully including Jet).
Get the latest MSDAC from the Microsoft site, makepy it, & be happy...


Alex





From arinagour at yahoo.co.uk  Thu Feb 22 20:47:20 2001
From: arinagour at yahoo.co.uk (chris lamb)
Date: Thu, 22 Feb 2001 19:47:20 -0000
Subject: executing shell commands
Message-ID: <R%dl6.77662$Nw6.251029@NewsReader>

I am very new to Python and having to learn quickly! I am enjoying it,
however. It is a flexible and intuitive environment.

Can someone advise me if there is an equivalent to running a system process
from inside a script as there is in Perl? I mean in perl if I use backticks
to enclose a system command, then the command is executed. Is there a
comparable way of running these commands from Python?

I want to build dynamic javadoc commands and module names and run the
commands inside the script.

best wishes

chris




From adam at deprince.net  Wed Feb 21 21:33:14 2001
From: adam at deprince.net (Adam DePrince)
Date: Wed, 21 Feb 2001 15:33:14 -0500
Subject: executing shell commands
References: <R%dl6.77662$Nw6.251029@NewsReader>
Message-ID: <3A94260A.CCEDC918@deprince.net>

chris lamb wrote:
> I am very new to Python and having to learn quickly! I am enjoying it,
> however. It is a flexible and intuitive environment.
> Can someone advise me if there is an equivalent to running a system process
> from inside a script as there is in Perl? I mean in perl if I use backticks
> to enclose a system command, then the command is executed. Is there a
> comparable way of running these commands from Python?
> I want to build dynamic javadoc commands and module names and run the
> commands inside the script.
> best wishes
> chris

http://www.python.org/doc/current/lib/module-commands.html

Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From phlip_cpp at my-deja.com  Thu Feb 22 21:37:30 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: Thu, 22 Feb 2001 12:37:30 -0800
Subject: executing shell commands
References: <R%dl6.77662$Nw6.251029@NewsReader>
Message-ID: <rOel6.32$Db4.41250@news.pacbell.net>

chris lamb wrote:

> I am very new to Python and having to learn quickly! I am enjoying it,
> however. It is a flexible and intuitive environment.
> 
> Can someone advise me if there is an equivalent to running a system
> process from inside a script as there is in Perl? I mean in perl if I use
> backticks to enclose a system command, then the command is executed. Is
> there a comparable way of running these commands from Python?

        os.system ("ls -l")

> I want to build dynamic javadoc commands and module names and run the
> commands inside the script.

May I suggest you use 'make'?

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  My opinions are those of your employer  --


From arinagour at yahoo.co.uk  Fri Feb 23 00:00:58 2001
From: arinagour at yahoo.co.uk (chris lamb)
Date: Thu, 22 Feb 2001 23:00:58 -0000
Subject: executing shell commands
References: <R%dl6.77662$Nw6.251029@NewsReader> <rOel6.32$Db4.41250@news.pacbell.net>
Message-ID: <oRgl6.4735$5U3.26506@NewsReader>

>         os.system ("ls -l")

Gulp! I have been trying to make it far too hard for myself!!

Thank you for your help, both replies have been very useful. My scripts are
running nicely. A very productive 1 days development under my belt!

Chris




From shaleh at valinux.com  Thu Feb 22 21:39:52 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Thu, 22 Feb 2001 12:39:52 -0800 (PST)
Subject: executing shell commands
In-Reply-To: <R%dl6.77662$Nw6.251029@NewsReader>
Message-ID: <XFMail.20010222123952.shaleh@valinux.com>

On 22-Feb-2001 chris lamb wrote:
> I am very new to Python and having to learn quickly! I am enjoying it,
> however. It is a flexible and intuitive environment.
> 
> Can someone advise me if there is an equivalent to running a system process
> from inside a script as there is in Perl? I mean in perl if I use backticks
> to enclose a system command, then the command is executed. Is there a
> comparable way of running these commands from Python?
> 

import os
os.system() or os.popen()

system will run the command, but you do not get anything back other than an
exit status.  popen() can launch and app and connect its output to a Python
file handler and let you read from it.



From matthias at book.mteege.de  Thu Feb 22 22:16:52 2001
From: matthias at book.mteege.de (Matthias Teege)
Date: Thu, 22 Feb 2001 22:16:52 +0100
Subject: Python and curses
Message-ID: <slrn99b0gq.ap.matthias@book.mteege.de>

Moin,

I want to use python to build some simple forms for database input. Is
there any python module for simple curses menus and forms out there?

Thanks for any hint
Matthias
 


From flognat at flognat.myip.org  Fri Feb 23 09:12:44 2001
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Fri, 23 Feb 2001 08:12:44 GMT
Subject: Python and curses
References: <slrn99b0gq.ap.matthias@book.mteege.de>
Message-ID: <m3lmqx95tj.fsf@master.athome>

/ matthias at book.mteege.de (Matthias Teege) wrote:
| [...] 
| I want to use python to build some simple forms for database input. Is
| there any python module for simple curses menus and forms out there?

Yes there are a ncurses/curses(?) module included in the distribution,
requires curses/ncurses on the machine to be run.

        /Andy


From jfontain at free.fr  Thu Feb 22 23:03:03 2001
From: jfontain at free.fr (Jean-Luc Fontaine)
Date: Thu, 22 Feb 2001 22:03:03 GMT
Subject: ANNOUNCE: tclpython-1.0
Message-ID: <3A958C9A.130C7838@free.fr>

This is the first release: please try it.
Comments and bug reports are obviously welcomed...


### README ###

tclpython version 1.0: a Python package for Tcl

This package allows the execution of Python code from a Tcl
interpreter, as in:


    package require tclpython
    set interpreter [python::interp new]
    $interpreter eval {x = 3/2.0}
    set result [$interpreter eval {print x}]
    puts $result


Sending data from the Python interpreter to the Tcl interpreter
requires the print Python command. This will hopefully no longer be
required in future versions.


This library is free software; you can redistribute it and/or modify
it under the terms of the GNU Library General Public License as
published by the Free Software Foundation; either version 2 of the
License, or (at your option) any later version.

This library is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

###

you may find it now at my homepage:

http://jfontain.free.fr/tclpython-1.0.tar.gz
http://jfontain.free.fr/tclpython-1.0-1.i386.rpm
http://jfontain.free.fr/tclpython-1.0-1.spec
http://jfontain.free.fr/tclpython.htm

Enjoy and please let me know what you think.


--
Jean-Luc Fontaine  mailto:jfontain at free.fr  http://jfontain.free.fr/





From robin at jessikat.fsnet.co.uk  Fri Feb 23 10:32:52 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 23 Feb 2001 09:32:52 +0000
Subject: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr>
Message-ID: <m1NVERAE5il6Ew$D@jessikat.fsnet.co.uk>

In article <3A958C9A.130C7838 at free.fr>, Jean-Luc Fontaine
<jfontain at free.fr> writes
>This is the first release: please try it.
>Comments and bug reports are obviously welcomed...
>

How does this compare with the primitive python for tcl package I wrote
long ago? 


http://www.jessikat.demon.co.uk/pyserver.zip

The main deficiency of my approach was that it was difficult to get
reverse communication ie callbacks from the python side and that
everything was done via sockets.


>
>### README ###
>
>tclpython version 1.0: a Python package for Tcl
>
>This package allows the execution of Python code from a Tcl
>interpreter, as in:
>
>
>    package require tclpython
>    set interpreter [python::interp new]
>    $interpreter eval {x = 3/2.0}
>    set result [$interpreter eval {print x}]
>    puts $result
>
>
>Sending data from the Python interpreter to the Tcl interpreter
>requires the print Python command. This will hopefully no longer be
>required in future versions.
>
>
>This library is free software; you can redistribute it and/or modify
>it under the terms of the GNU Library General Public License as
>published by the Free Software Foundation; either version 2 of the
>License, or (at your option) any later version.
>
>This library is distributed in the hope that it will be useful, but
>WITHOUT ANY WARRANTY; without even the implied warranty of
>MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>
>###
>
>you may find it now at my homepage:
>
>http://jfontain.free.fr/tclpython-1.0.tar.gz
>http://jfontain.free.fr/tclpython-1.0-1.i386.rpm
>http://jfontain.free.fr/tclpython-1.0-1.spec
>http://jfontain.free.fr/tclpython.htm
>
>Enjoy and please let me know what you think.
>
>
>--
>Jean-Luc Fontaine  mailto:jfontain at free.fr  http://jfontain.free.fr/
>
>
>

-- 
Robin Becker


From jfontain at winealley.com  Fri Feb 23 12:37:57 2001
From: jfontain at winealley.com (Jean-Luc Fontaine)
Date: Fri, 23 Feb 2001 12:37:57 +0100
Subject: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr> <m1NVERAE5il6Ew$D@jessikat.fsnet.co.uk>
Message-ID: <975i2e$pa8$1@s1.read.news.oleane.net>

Robin Becker wrote:

---
How does this compare with the primitive python for tcl package I wrote
long ago?

http://www.jessikat.demon.co.uk/pyserver.zip

The main deficiency of my approach was that it was difficult to get
reverse communication ie callbacks from the python side and that
everything was done via sockets.
---

I did not know that existed: I will check it out, thank you.

The difference is that this is using an embedded python interpreter 
(several I hope in a future version), so communication is done through the 
C APIs.

-- 
Jean-Luc Fontaine mailto:jfontain at winealley.com http://www.winealley.com


From robin at jessikat.fsnet.co.uk  Fri Feb 23 13:52:44 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 23 Feb 2001 12:52:44 +0000
Subject: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr>
 <m1NVERAE5il6Ew$D@jessikat.fsnet.co.uk>
 <975i2e$pa8$1@s1.read.news.oleane.net>
Message-ID: <vRP6jvAc0ll6Ewss@jessikat.demon.co.uk>

In article <975i2e$pa8$1 at s1.read.news.oleane.net>, Jean-Luc Fontaine
<jfontain at winealley.com> writes
...
>I did not know that existed: I will check it out, thank you.
>
>The difference is that this is using an embedded python interpreter 
>(several I hope in a future version), so communication is done through the 
>C APIs.
>
Wow, I downloaded my thing and after changing the Python version value
to 2.0 it ran the Excel demo perfectly. I guess there are both
advantages and disadvantages to having a combined interpreter version.
-- 
Robin Becker


From lvirden at cas.org  Fri Feb 23 12:51:08 2001
From: lvirden at cas.org (lvirden at cas.org)
Date: 23 Feb 2001 11:51:08 GMT
Subject: Languages from within Tcl [Was: Re: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr>
Message-ID: <975irc$16g$1@srv38.cas.org>

Okay, we now have 

o	tclblend
o	tclperl
o	tclpython
o	embedded c

What other programming languages have been integrated into Tcl?
-- 
-- 
"See, he's not just anyone ... he's my son."  Mark Schultz
<URL: mailto:lvirden at cas.org> <URL: http://www.purl.org/NET/lvirden/>
Even if explicitly stated to the contrary, nothing in this posting


From claird at starbase.neosoft.com  Fri Feb 23 15:35:38 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 23 Feb 2001 08:35:38 -0600
Subject: Languages from within Tcl [Was: Re: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr> <975irc$16g$1@srv38.cas.org>
Message-ID: <E43EEA3E68CAF21B.712DA4A67EDE6509.58FFB479E492D2F2@lp.airnews.net>

In article <975irc$16g$1 at srv38.cas.org>,  <lvirden at cas.org> wrote:
>
>Okay, we now have 
>
>o	tclblend
>o	tclperl
>o	tclpython
>o	embedded c
>
>What other programming languages have been integrated into Tcl?
			.
			.
			.
Ada, Forth, Prolog, ...  Occasionally I convulse into an effort to catalogue 'em
<URL: http://starbase.neosoft.com/~claird/comp.lang.misc/polyglot.html >
I welcome help.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From snowzone at home.com  Sun Feb 25 04:33:26 2001
From: snowzone at home.com (tony summerfelt)
Date: Sun, 25 Feb 2001 03:33:26 GMT
Subject: Languages from within Tcl [Was: Re: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr> <975irc$16g$1@srv38.cas.org> <E43EEA3E68CAF21B.712DA4A67EDE6509.58FFB479E492D2F2@lp.airnews.net>
Message-ID: <slrn99gv7u.u6s.snowzone@co146453-a.kico1.on.wave.home.com>

On 23 Feb 2001 08:35:38 -0600, Cameron Laird wrote:
 >
> >What other programming languages have been integrated into Tcl?
 			.
> Ada, Forth, Prolog, ...  Occasionally I convulse into an effort to catalogue

and the ada one is pretty good. it allows you write tcl extensions in ada
(and of course you get all the tcl/tk goodies for ada) ada being rock
solid, tcl/tk/ada is a great combination...

-- 
*----------------------------------------
| 
|http://www.biosys.net/snowzone
|
|remove myspleen to email
|
*-----------------------


From davygrvy at pobox.com  Sun Feb 25 05:08:24 2001
From: davygrvy at pobox.com (David Gravereaux)
Date: Sat, 24 Feb 2001 20:08:24 -0800
Subject: Languages from within Tcl [Was: Re: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr> <975irc$16g$1@srv38.cas.org> <E43EEA3E68CAF21B.712DA4A67EDE6509.58FFB479E492D2F2@lp.airnews.net>
Message-ID: <kv0h9tki4774kajje102uprdkf4seqqnga@4ax.com>

claird at starbase.neosoft.com (Cameron Laird) wrote:

>In article <975irc$16g$1 at srv38.cas.org>,  <lvirden at cas.org> wrote:
>>
>>Okay, we now have 
>>
>>o	tclblend
>>o	tclperl
>>o	tclpython
>>o	embedded c
>>
>>What other programming languages have been integrated into Tcl?
>			.
>			.
>			.
>Ada, Forth, Prolog, ...  Occasionally I convulse into an effort to catalogue 'em
><URL: http://starbase.neosoft.com/~claird/comp.lang.misc/polyglot.html >
>I welcome help.

>From the embedded side we could add delphi,{nasm,tasm,masm}, and VB if we only
had an h2lang convertor thing for tcl.h .  The Stubs issue might be kinda odd in
VB.

I tried once, but stopped.  Yet another Tcl project started and stopped.  I need
more free time :)

--
                  \\ ~ ~ //
                  (  @ @  )
----------------oOOo-(_)-oOOo--------------
  David Gravereaux <davygrvy at pobox.com>
  Just some Tcl hacker dude.
  http://dev.scriptics.com/doc/integration.html#Tcl
------------------------Oooo.--------------
               .oooO    (   )
               (   )     ) /
                \ (     (_/
                 \_)


From changl at neatware.com  Fri Feb 23 17:11:52 2001
From: changl at neatware.com (Chang LI)
Date: Fri, 23 Feb 2001 11:11:52 -0500
Subject: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr>
Message-ID: <pXvl6.1821$zi6.54134@newscontent-01.sprint.ca>

How could you call a function in Python or Perl from Tcl?
How to use the library in these languages?
I do not think it is very useful to write Python program
inside the Tcl.  However it is useful to use the Python
or Perl library.

Chang

Jean-Luc Fontaine wrote in message <3A958C9A.130C7838 at free.fr>...
>This is the first release: please try it.
>Comments and bug reports are obviously welcomed...
>
>
>### README ###
>
>tclpython version 1.0: a Python package for Tcl
>
>This package allows the execution of Python code from a Tcl
>interpreter, as in:
>
>
>    package require tclpython
>    set interpreter [python::interp new]
>    $interpreter eval {x = 3/2.0}
>    set result [$interpreter eval {print x}]
>    puts $result
>
>
>Sending data from the Python interpreter to the Tcl interpreter
>requires the print Python command. This will hopefully no longer be
>required in future versions.
>
>
>This library is free software; you can redistribute it and/or modify
>it under the terms of the GNU Library General Public License as
>published by the Free Software Foundation; either version 2 of the
>License, or (at your option) any later version.
>
>This library is distributed in the hope that it will be useful, but
>WITHOUT ANY WARRANTY; without even the implied warranty of
>MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>
>###
>
>you may find it now at my homepage:
>
>http://jfontain.free.fr/tclpython-1.0.tar.gz
>http://jfontain.free.fr/tclpython-1.0-1.i386.rpm
>http://jfontain.free.fr/tclpython-1.0-1.spec
>http://jfontain.free.fr/tclpython.htm
>
>Enjoy and please let me know what you think.
>
>
>--
>Jean-Luc Fontaine  mailto:jfontain at free.fr  http://jfontain.free.fr/
>
>
>




From mal at lemburg.com  Thu Feb 22 23:04:29 2001
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 22 Feb 2001 23:04:29 +0100
Subject: ANN: eGenix.com mx Extensions -- Version 2.0.0 (mxODBC, mxDateTime, ...)
Message-ID: <3A958CED.F7B95C33@lemburg.com>

________________________________________________________________________

ANNOUNCING:

               eGenix.com mx Extension Series for Python
                            Version 2.0.0

      Full Source Python extensions providing important and useful
                  services for Python programmers.


________________________________________________________________________

WHAT IS IT ?

The eGenix.com mx Extensions for Python are a collection of
professional quality software tools which enhance Python's usability
in many important areas such as ODBC database connectivity, fast text
processing, date/time processing and web site programming.

Python is an open-oriented Open Source programming language which runs
on all modern platforms. By integrating ease-of-use, clarity in
coding, enterprise application connectivity and rapid application
design, Python establishes an ideal programming platform for todays IT
challenges.

The tools have a proven record of being portable across many Unix and
Windows platforms, e.g. you can write applications which use an ODBC
database on Windows which then run on Unix platforms without change
due to the consistent platforms independent interfaces.

All available packages have proven their stability and usefulness in
many mission critical applications and various commercial settings all
around the world.

The two most well-known subpackages from the eGenix.com mx Extension
Series are mxDateTime and mxODBC providing date/time services and
professional ODBC database connectivity on practically all supported
Python platforms. These two packages enable database software which is
portable not only across platforms, but also across database backends.

________________________________________________________________________

WHAT'S NEW ?

mxODBC, mxDateTime and all other mx packages are now maintained by my
new company eGenix.com Software GmbH, Langenfeld in Germany
(http://www.egenix.com/). This move will provide a better base for
future development and also assures the availability of commercial
support, which is important for companies building applications based
on the eGenix.com mx Extensions.

The new versions of the eGenix.com mx Extensions use the Python 2.0
distutils packaging technology to simplify installation and use of the
many different subpackages. This is expected to greatly enhance the
installation and usage experience of the software.

In additions to many subpackage enhancements, this new release
repackages all mx Extensions under a new top-level package named
'mx'. This will allow you to use the subpackages side-by-side with
other Python tools using similar names.

The software is now delivered in two download archives: the eGenix.com
mx BASE package which holds all the Open Source tools from the series
and provides the basis for the other add-ons and the eGenix.com mx
COMMERCIAL package which currently only contains the mxODBC package.
The two packages are covered by the eGenix.com Public License and the
eGenix.com Commercial License respectively. Details about the licenses
can be found at:

       http://www.lemburg.com/files/python/mxLicense.html
	  
________________________________________________________________________

SPECIAL OFFER

theKompany.com has licensed the COMMERCIAL package (which includes
mxODBC) for inclusion in their brand new Qt-based Python IDE
BackAdder. It allows developing portable GUI-based database
applications which run on Windows and Linux platforms without any
change to the source code.

BlackAdder includes a 1 CPU license for this package at no extra cost,
so you may want to check out their great new product. See

http://www.lemburg.com/files/python/eGenix-mx-Extensions.html#BlackAdder

for details.

________________________________________________________________________

EGENIX.COM MX BASE PACKAGE OVERVIEW:

mxDateTime - Generic Date/Time Types

     mxDateTime is an extension package that provides three new object
     types, DateTime, DateTimeDelta and RelativeDateTime, which let
     you store and handle date/time values in a much more natural way
     than by using ticks (seconds since 1.1.70 0:00 UTC; the encoding
     used by the time module).

     You can add, subtract and even multiply instances, pickle and
     copy them and convert the results to strings, COM dates, ticks
     and some other more esoteric values. In addition, there are
     several convenient constructors and formatters at hand to greatly
     simplify dealing with dates and times in real-world applications.

     In addition to providing an easy-to-use Python interface the
     package also exports a comfortable C API interface for other
     extensions to build upon. This is especially interesting for
     database applications which often have to deal with date/time
     values (the mxODBC package is one example of an extension using
     this interface).

mxTextTools - Fast Text Processing Tools

     mxTextTools is an extension package for Python that provides
     several useful functions and types that implement
     high-performance text manipulation and searching algorithms in
     addition to a very flexible and extendable state machine, the
     Tagging Engine, that allows scanning and processing text based on
     low-level byte-code "programs" written using Python tuples. It
     gives you access to the speed of C without the need to do any
     compile and link steps every time you change the parsing
     description.

     Applications include parsing structured text, finding and
     extracting text (either exact or using translation tables) and
     recombining strings to form new text.

mxStack - Fast and Memory-Efficient Stack Type

     mxStack is an extension package that provides a new object type
     called Stack. It works much like what you would expect from such
     a type, having .push() and .pop() methods and focusses on
     obtaining maximum speed at low memory costs.

mxTools - Collection of Additional Builtins

     mxTools is an extension package that includes a collection of
     handy functions and objects giving additional functionality in
     form of new builtins to the Python programmer.

     The package auto-installs the new functions and objects as
     builtins upon first import. This means that they become
     instantely available to all other modules without any further
     action on your part. Add the line import NewBuiltins to your
     site.py script and they will be available to all users at your
     site as if they were installed in the Python interpreter itself.

mxProxy - Generic Proxy Wrapper Type

     mxProxy is an extension package that provides a new type that is
     suitable to implement Bastion like features without the need to
     use restricted execution environments.

     The type's main features are secure data encapsulation (the
     hidden objects are not accessible from Python since they are
     stored in internal C structures), customizable attribute lookup
     methods and a cleanup protocol that helps in breaking circular
     references prior to object deletion.

     The latest version adds a very interesting new feature: weak
     references which help you work with circular references in a way
     that doesn't cause memory leakage in a Python system.

mxBeeBase - On-disk B+Tree Based Database Kit

     mxBeeBase is a high performance construction kit for disk based
     indexed databases. It offers components which you can plug
     together to easily build your own custom mid-sized databases (the
     current size limit is <CODE>sizeof(long)</CODE> which gives you
     an address range of around 2GB on 32-bit platforms).
 
     The two basic building blocks in mxBeeBase are storage and
     index. Storage is implemented as variable record length data
     storage with integrated data protection features, automatic data
     recovery and locking for multi process access. Indexes use a high
     performance optimized B+Tree implementation built on top of
     Thomas Niemann's Cookbook B+Tree implementation
     (http://epaperpress.com/).

     Note: mxBeeBase is new in this release and still in beta.

________________________________________________________________________

EGENIX.COM MX COMMERCIAL PACKAGE OVERVIEW:

mxODBC - Generic ODBC 2.0 interface

     mxODBC is an extension package that provides a Python Database
     API compliant interface to ODBC 2.0 capable database drivers and
     managers.

     In addition to the capabilities provided through the standard DB
     API it also gives access to a rich set of catalog methods which
     allow you to scan the database for tables, procedures,
     etc. Furthermore, it uses the mxDateTime package for date/time
     value interfacing eliminating most of the problems these types
     normally introduce (other in/output formats are available too).

     The latest version of the interface allows you to interface to
     more than one database from one process. It includes a varity of
     preconfigured setups for many commonly used databases such as
     MySQL, Oracle, Informix, Solid and many more. A precompiled
     version of the extension for use with the Windows ODBC manager is
     also included.

________________________________________________________________________

WHERE CAN I GET IT ?

The download archives and instructions for installing the packages can
be found at:

     http://www.lemburg.com/files/python/

________________________________________________________________________

WHAT DOES IT COST ?

The BASE packages come with a Python 2.0 style license, which means
that you can use them in both commercial and non-commercial settings
without fee or charge. All packages come with full source code.

mxODBC comes with a new licenses which allows non-commercial use
at no charge, but costs a moderate fee for commercial use. Please see 

http://www.lemburg.com/files/python/eGenix-mx-Extensions.html#mxCOMMERCIAL

for details.

________________________________________________________________________

WHERE CAN I GET SUPPORT ?

Commercial quality support for these packages is available from
eGenix.com Software GmbH. Please see

http://www.lemburg.com/files/python/eGenix-mx-Extensions.html#Support 

for details about the eGenix support offerings.

________________________________________________________________________

REFERENCE:

<P><A HREF="http://www.lemburg.com/files/python/">eGenix.com mx
Extension Series</A> - eGenix.com mx Extension Series with distutils
support and precompiled binaries for Windows and Linux. (22-Feb-2001)

________________________________________________________________________


...thanks for reading this far ;-)

In the spirit of Python software announcements, I am leaving off for
the Python Conference in Long Beach, California tomorrow. 

If you have any problems with these packages or the downloads, please
email me at mal at egenix.com. I will read your mail once I get back from
the conference and will do my best make the eGenix.com mx Extensions a
success for you.

See you at the conference,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Pages:                           http://www.lemburg.com/python/




From gustav at morpheus.demon.co.uk  Sat Feb 24 23:15:05 2001
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Sat, 24 Feb 2001 22:15:05 +0000
Subject: ANN: eGenix.com mx Extensions -- Version 2.0.0 (mxODBC, mxDateTime, ...)
References: <3A958CED.F7B95C33@lemburg.com>
Message-ID: <r7cg9toqrkgklmk6egokarju1g3mc9mei2@4ax.com>

On Thu, 22 Feb 2001 23:04:29 +0100, "M.-A. Lemburg" <mal at lemburg.com>
wrote:

>The download archives and instructions for installing the packages can
>be found at:
>
>     http://www.lemburg.com/files/python/

One thing which would be nice would be a Win32 binary download, which
was NOT packaged using an installer. Just a zip file to be unzipped in
an appropriate place...

At the very least, an option to change the install directory (to a
"local modules" directory of my choosing) is pretty much a
necessity...

Paul.


From thomas.heller at ion-tof.com  Tue Feb 27 10:26:21 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Tue, 27 Feb 2001 10:26:21 +0100
Subject: ANN: eGenix.com mx Extensions -- Version 2.0.0 (mxODBC, mxDateTime, ...)
References: <3A958CED.F7B95C33@lemburg.com> <r7cg9toqrkgklmk6egokarju1g3mc9mei2@4ax.com>
Message-ID: <97frrt$p1a5r$1@ID-59885.news.dfncis.de>

"Paul Moore" <gustav at morpheus.demon.co.uk> wrote in message news:r7cg9toqrkgklmk6egokarju1g3mc9mei2 at 4ax.com...
> On Thu, 22 Feb 2001 23:04:29 +0100, "M.-A. Lemburg" <mal at lemburg.com>
> wrote:
>
> >The download archives and instructions for installing the packages can
> >be found at:
> >
> >     http://www.lemburg.com/files/python/
>
> One thing which would be nice would be a Win32 binary download, which
> was NOT packaged using an installer. Just a zip file to be unzipped in
> an appropriate place...
>

The windows installer (egenix-mx-base-2.0.0.win32-py2.0.exe) is more or
less a self-extracting zip.file, which also can be opened with WinZip
or other tools...

Thomas




From matt at mondoinfo.com  Thu Feb 22 23:12:04 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Thu, 22 Feb 2001 22:12:04 GMT
Subject: os.path.join("/a","/b")
Message-ID: <slrn99b3lk.vh.matt@happy-hour.mondoinfo.com>

I was a little surprised today to notice that

os.path.join("/a","/b") returns '/b'

(os.path is posixpath in this case of course) while

os.path.join("/a/","b") returns "/a/b"

which I expected from both. Poking back into my dusty archives, I find
that it has been that way since at least 1.4 so I presume that it's a
feature but I'll be darned if I can figure out the logic. It sure
doesn't seem to have much to do with joining. The module's docstring
just says "Join two or more pathname components, inserting '/' as
needed".

If someone could explain the logic to me, I'd be grateful.

Regards,
Matt


From jmarshal at mathworks.com  Thu Feb 22 23:25:35 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 22 Feb 2001 22:25:35 GMT
Subject: os.path.join("/a","/b")
References: <slrn99b3lk.vh.matt@happy-hour.mondoinfo.com>
Message-ID: <9743kv$7ur$1@news.mathworks.com>

Matthew Dixon Cowles <matt at mondoinfo.com> wrote:
> I was a little surprised today to notice that

> os.path.join("/a","/b") returns '/b'

> (os.path is posixpath in this case of course) while

> os.path.join("/a/","b") returns "/a/b"

> which I expected from both. Poking back into my dusty archives, I find
> that it has been that way since at least 1.4 so I presume that it's a
> feature but I'll be darned if I can figure out the logic. It sure
> doesn't seem to have much to do with joining. The module's docstring
> just says "Join two or more pathname components, inserting '/' as
> needed".

> If someone could explain the logic to me, I'd be grateful.

> Regards,
> Matt

It's as if you cd to directory "/a" and reference "/b".  Since "/b" is
an absolute path, the current directory is ignored.


From fredrik at pythonware.com  Thu Feb 22 23:36:09 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 22:36:09 GMT
Subject: os.path.join("/a","/b")
References: <slrn99b3lk.vh.matt@happy-hour.mondoinfo.com>
Message-ID: <tvgl6.19417$AH6.2643616@newsc.telia.net>

Matthew Dixon Cowles wrote:
> I was a little surprised today to notice that
>
> os.path.join("/a","/b") returns '/b'
>
> (os.path is posixpath in this case of course) while
>
> os.path.join("/a/","b") returns "/a/b"
>
> which I expected from both. Poking back into my dusty archives, I find
> that it has been that way since at least 1.4 so I presume that it's a
> feature but I'll be darned if I can figure out the logic.

http://www.python.org/doc/current/lib/module-os.path.html

    join (path1[, path2[, ...]])

    Joins one or more path components intelligently. If any
    component is an absolute path, all previous components
    are thrown away, and joining continues. The return value
    is the concatenation of path1, and optionally path2, etc.,
    with exactly one slash ('/') inserted between components,
    unless path is empty.

the rationale is probably that it's a good idea if

    os.path.join(os.getcwd(), filename)

refers to the same file as:

    filename

for any filename.

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->






From matt at mondoinfo.com  Fri Feb 23 00:36:55 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Thu, 22 Feb 2001 23:36:55 GMT
Subject: os.path.join("/a","/b")
References: <slrn99b3lk.vh.matt@happy-hour.mondoinfo.com> <tvgl6.19417$AH6.2643616@newsc.telia.net>
Message-ID: <slrn99b8kn.lg.matt@happy-hour.mondoinfo.com>

On Thu, 22 Feb 2001 22:36:09 GMT, Fredrik Lundh <fredrik at pythonware.com> wrote:

>the rationale is probably that it's a good idea if
>
>    os.path.join(os.getcwd(), filename)
>
>refers to the same file as:
>
>    filename
>
>for any filename.

Thanks to you and Joshua for the clues. I'm sure that you and he are
right about the rationale. Still, I'd have chosen a name other than
join for that behavior.

Regards,
Matt


From bsb at winnegan.de  Fri Feb 23 00:01:29 2001
From: bsb at winnegan.de (Siggy Brentrup)
Date: 23 Feb 2001 00:01:29 +0100
Subject: os.path.join("/a","/b")
In-Reply-To: <slrn99b3lk.vh.matt@happy-hour.mondoinfo.com>
References: <slrn99b3lk.vh.matt@happy-hour.mondoinfo.com>
Message-ID: <871ysqqq46.fsf@winnegan.de>

matt at mondoinfo.com (Matthew Dixon Cowles) writes:

> I was a little surprised today to notice that
> 
> os.path.join("/a","/b") returns '/b'
> 
> (os.path is posixpath in this case of course) while
> 
> os.path.join("/a/","b") returns "/a/b"
> 
> which I expected from both. Poking back into my dusty archives, I find
> that it has been that way since at least 1.4 so I presume that it's a
> feature but I'll be darned if I can figure out the logic. It sure
> doesn't seem to have much to do with joining. The module's docstring
> just says "Join two or more pathname components, inserting '/' as
> needed".

> If someone could explain the logic to me, I'd be grateful.

Consult the library reference, it says "if any component is an
absolute path ..."

Thx
  Siggy




From lreilly at cs.strath.ac.uk  Thu Feb 22 23:40:10 2001
From: lreilly at cs.strath.ac.uk (Lee Reilly CS1997)
Date: Thu, 22 Feb 2001 22:40:10 +0000
Subject: which IDE?
Message-ID: <3A95954A.72621882@cs.strath.ac.uk>

Hi,

I'm relatively new to Python and have been developing a product using
Zope and experimenting using IDLE. I was wodnering if anyone could tell
me the name of a good IDE - something similar to Symantec Visual Cafe
perhaps, where I could develop my code and it can spice it up
(text/variable highlighting, etc.).

WingIDE looks good http://archaeopteryx.com/wingide but AFAIK it's not
available on Win32.

Anyone


From tino at wildenhain.de  Fri Feb 23 00:08:08 2001
From: tino at wildenhain.de (Tino Wildenhain)
Date: Fri, 23 Feb 2001 00:08:08 +0100
Subject: which IDE?
In-Reply-To: <3A95954A.72621882@cs.strath.ac.uk>
References: <3A95954A.72621882@cs.strath.ac.uk>
Message-ID: <15423962.982886887@[10.0.0.101]>

Hi Lee,

> Hi,
>
> I'm relatively new to Python and have been developing a product using
> Zope and experimenting using IDLE. I was wodnering if anyone could tell
> me the name of a good IDE - something similar to Symantec Visual Cafe
> perhaps, where I could develop my code and it can spice it up
> (text/variable highlighting, etc.).
>
> WingIDE looks good http://archaeopteryx.com/wingide but AFAIK it's not
> available on Win32.

You will for sure like the pythonwin IDE that comes with the win32extensions
by Mark Hammond. Its near the place where you downloaded the windows version
of python. I recommend the couple python2.0/win32 135 iirc. You get both
with one installer from activestate.com

HTH
Tino Wildenhain





From wayne.izatt at _nospam_myself.com  Sat Feb 24 07:37:34 2001
From: wayne.izatt at _nospam_myself.com (Wayne Izatt)
Date: Sat, 24 Feb 2001 06:37:34 GMT
Subject: which IDE?
References: <3A95954A.72621882@cs.strath.ac.uk>
Message-ID: <OEIl6.170259$df5.5070164@news1.crdva1.bc.home.com>

Activestate are also pushing something called Komodo, an ide for both Perl
and Python. I haven't tried it, so I can't rate it.

ciao

"Lee Reilly CS1997" <lreilly at cs.strath.ac.uk> wrote in message
news:3A95954A.72621882 at cs.strath.ac.uk...
> Hi,
>
> I'm relatively new to Python and have been developing a product using
> Zope and experimenting using IDLE. I was wodnering if anyone could tell
> me the name of a good IDE - something similar to Symantec Visual Cafe
> perhaps, where I could develop my code and it can spice it up
> (text/variable highlighting, etc.).
>
> WingIDE looks good http://archaeopteryx.com/wingide but AFAIK it's not
> available on Win32.
>
> Anyone




From dsavitsk at e-coli.net  Fri Feb 23 00:32:34 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 22 Feb 2001 17:32:34 -0600
Subject: httplib and buffered data
Message-ID: <mkhl6.124395$Ch.22717962@newsrump.sjc.telocity.net>

i am capturing the data being returned from a script running on a web
server.  The script takes several minutes to run, and trickles back data a
little at a time.
Using httplib (or anything else) is there a way to capture the data line by
line instead of as a lump sum with getfile()? i would like to alert the user
as to the % of the data (a known quantity) that has been processed.

thanks,

doug




From dwig at advancedmp.net  Sat Feb 24 01:33:31 2001
From: dwig at advancedmp.net (Don Dwiggins)
Date: 23 Feb 2001 16:33:31 -0800
Subject: httplib and buffered data
In-Reply-To: "dsavitsk"'s message of "Thu, 22 Feb 2001 17:32:34 -0600"
References: <mkhl6.124395$Ch.22717962@newsrump.sjc.telocity.net>
Message-ID: <uhf1kapic.fsf@advancedmp.net>

dsavitsk  writes:
> i am capturing the data being returned from a script running on a web
> server.  The script takes several minutes to run, and trickles back data a
> little at a time.
> Using httplib (or anything else) is there a way to capture the data line by
> line instead of as a lump sum with getfile()? i would like to alert the user
> as to the % of the data (a known quantity) that has been processed.

You could use asyncore/asynchat, if you're willing to wrap your mind around
that style of operation.  The library documentation for asyncore shows a
little http client as an example.  Better documentation for both is in the
Medusa pages on www.nightmare.com.

Happy hunting,

-- 
Don Dwiggins                    "Solvitur Ambulando"
Advanced MP Technology
dwig at advancedmp.net




From dhjung7 at orgio.net  Fri Feb 23 03:18:35 2001
From: dhjung7 at orgio.net (iact21)
Date: Fri, 23 Feb 2001 11:18:35 +0900
Subject: How do I make damon process in Python ?
Message-ID: <974hap$4r4$1@b5nntp2.channeli.net>

How can I make damon process in Python..?

I'd like to see example code...






From cg at schlund.de  Fri Feb 23 10:39:03 2001
From: cg at schlund.de (Carsten Gaebler)
Date: Fri, 23 Feb 2001 10:39:03 +0100
Subject: How do I make damon process in Python ?
References: <974hap$4r4$1@b5nntp2.channeli.net>
Message-ID: <3A962FB7.161814B1@schlund.de>

iact21 wrote:
> How can I make damon process in Python..?
> 
> I'd like to see example code...

#!/usr/bin/python 

import os

pid = os.fork()
if pid == 0:
        # Child process (daemon)
        while 1:
                pass
else:
        # Parent process:
        # Do something useful, or nothing, then exit
        # and leave the child alone.
        pass



cg.
-- 
Carsten Gaebler                    Computerhomoeopathie und
Schlund + Partner AG                  Schlangenbeschwoerung
Erbprinzenstr. 4 - 12             voice://49.721.91374.543/
76133 Karlsruhe                     fax://49.721.91374.228/


From berning at teuto.de  Sat Feb 24 01:33:40 2001
From: berning at teuto.de (Ulrich Berning)
Date: Sat, 24 Feb 2001 01:33:40 +0100
Subject: How do I make damon process in Python ?
References: <974hap$4r4$1@b5nntp2.channeli.net>
Message-ID: <3A970164.BBA246FF@teuto.de>

iact21 wrote:

> How can I make damon process in Python..?
>
> I'd like to see example code...

We used something like the attached example daemon.

The daemon ensures, that only one instance can run on a single machine
(using a pid file).
You can also use this daemon to stop a running instance or to query if
there is a running instance.

This daemon puts itself into background and disconnects completely from
the controlling terminal,
so it can not write to stdout or stderr after is has initialized.
For this reason, stdout and stderr are closed and then reasigned to a
message log file and an error log file.

I hope its useful.

Ulli

-------------- next part --------------
#!/usr/bin/env python

#============================================================================#
# Import the necessary modules                                               #
#----------------------------------------------------------------------------#
import sys, os, string, time, signal, errno

#============================================================================#
# ----- Contants ----------------------------------------------------------- #
#----------------------------------------------------------------------------#
LOG_DIR  = '/tmp/my_daemon'

#============================================================================#
# ----- Signal Handlers ---------------------------------------------------- #
#----------------------------------------------------------------------------#
#============================================================================#
# terminate() - Signal handler that is called on SIGTERM                     #
#----------------------------------------------------------------------------#
def terminate(signal, param):
    try:
        # Do necessary cleanup handling
        #...
        #...
        #...

        # Remove the pid file
        os.remove(os.path.join(LOG_DIR, 'run.pid'))
    except:
        pass                            # Ignore any errors
    sys.stdout.write("........terminating\n")
    sys.exit(0)

#============================================================================#
# ----- Main Routine ------------------------------------------------------- #
#----------------------------------------------------------------------------#
# Store the name of the programm
call_name = os.path.split(sys.argv[0])[1]

# Get the call arguments (stop, status)
if len(sys.argv) > 2:
    sys.stderr.write("Usage: %s [stop|status]\n" % call_name)
    sys.exit(1)
if len(sys.argv) == 2:
    if sys.argv[1] == 'stop':
        if os.path.isfile(os.path.join(LOG_DIR, 'run.pid')):
            f = open(os.path.join(LOG_DIR, 'run.pid'), 'r')
            pid = string.atoi(string.strip(f.readline()))
            f.close()
            try:
                os.kill(pid, 0)
            except os.error, args:
                if args[0] != errno.ESRCH: # NO SUCH PROCESS
                    raise os.error, args
            else:
                os.kill(pid, signal.SIGTERM)
                sys.exit(0)            
        sys.stdout.write("%s: daemon is not running\n" % call_name)
        sys.exit(1)
    elif sys.argv[1] == 'status':
        if os.path.isfile(os.path.join(LOG_DIR, 'run.pid')):
            f = open(os.path.join(LOG_DIR, 'run.pid'), 'r')
            pid = string.atoi(string.strip(f.readline()))
            f.close()
            try:
                os.kill(pid, 0)
            except os.error, args:
                if args[0] != errno.ESRCH: # NO SUCH PROCESS
                    raise os.error, args
            else:
                sys.stdout.write("%s: daemon [%d] is running\n" % (call_name, pid))
                sys.exit(0)            
        sys.stdout.write("%s: daemon is not running\n" % call_name)
        sys.exit(0)
    else:
        sys.stderr.write("Usage: %s [stop|status]\n" % call_name)
        sys.exit(1)

# Check, if the log directory is already there. If not, create it
if not os.path.isdir(LOG_DIR):
    os.mkdir(LOG_DIR, 0755)

# Check, if there is already a running instance
if os.path.isfile(os.path.join(LOG_DIR, 'run.pid')):
    f = open(os.path.join(LOG_DIR, 'run.pid'), 'r')
    pid = string.atoi(string.strip(f.readline()))
    f.close()
    try:
        os.kill(pid, 0)
    except os.error, args:
        if args[0] != errno.ESRCH: # NO SUCH PROCESS
            raise os.error, args
    else:
        sys.stderr.write("%s: there is already a running instance\n" % call_name)
        sys.exit(1)
        
# Initialize error logging
f_stderr = open(os.path.join(LOG_DIR, 'error.log'), 'w')
    
# Initialize message logging
f_stdout = open(os.path.join(LOG_DIR, 'message.log'), 'w')
    
# Print a startup message
sys.stdout.write("%s: started at %s\n" % (call_name, time.strftime("%d.%m.%Y / %H:%M:%S", time.localtime(time.time()))))
sys.stdout.write("%s: log directory is is '%s'\n" % (call_name, LOG_DIR))
sys.stdout.write("%s: going into background........\n" % call_name)

#============================================================================#
# Initialize a daemon process                                                #
#----------------------------------------------------------------------------#
# Change to the log directory, this permits an umount on this directory
os.chdir(LOG_DIR)

# Do the following only, if we were not started by the init process
if os.getppid() != 1:
    # Ignore a terminal output signal
    signal.signal(signal.SIGTTOU, signal.SIG_IGN)
    # Ignore a terminal input signal
    signal.signal(signal.SIGTTIN, signal.SIG_IGN)
    # Ignore a terminal stop signal
    signal.signal(signal.SIGTSTP, signal.SIG_IGN)
    # Fork the first child
    pid = os.fork()
    # The parent ends here
    if pid > 0:
        sys.exit(0)
    # Make ourself a leader of the process group
    os.setpgrp()
    # Ignore a hangup signal
    signal.signal(signal.SIGHUP, signal.SIG_IGN)
    # Fork the second child
    pid = os.fork()
    # The parent ends here
    if pid > 0:
        sys.exit(0)

# Close sys.stdin
sys.stdin.close()
# Close sys.stdout and redirect the message log to sys.stdout
sys.stdout.close()
sys.stdout = f_stdout
# Close sys.stderr and redirect the error log to sys.stderr
sys.stderr.close()
sys.stderr = f_stderr
# Reset the umask
os.umask(000)
# Ignore a dead of child signal
signal.signal(signal.SIGCLD, signal.SIG_IGN)
# Install a hander for ther terminate signal
signal.signal(signal.SIGTERM, terminate)
# Install a handler for the interrupt signal
signal.signal(signal.SIGINT, terminate)
# Install a handler for the quit signal
signal.signal(signal.SIGQUIT, terminate)

# Write our process id into the pid file
f = open(os.path.join(LOG_DIR, 'run.pid'), 'w')
f.write("%d" % os.getpid())
f.close()

# Print a startup message
sys.stdout.write("%s: started at %s\n" % (call_name, time.strftime("%d.%m.%Y / %H:%M:%S", time.localtime(time.time()))))
sys.stdout.write("%s: log directory is is '%s'\n" % (call_name, LOG_DIR))
sys.stdout.write("%s: going into background........\n" % call_name)
sys.stdout.flush()

# Do other necessary initializations before entering the endless loop
#...
#...
#...

#============================================================================#
# Enter the endles loop                                                      #
#----------------------------------------------------------------------------#
while(1):
    # Replace the following sleep() by a real worker
    time.sleep(10)

#============================================================================#
# EOF                                                                        #
#----------------------------------------------------------------------------#

From sai at dft.chem.cuhk.edu.hk  Fri Feb 23 03:36:49 2001
From: sai at dft.chem.cuhk.edu.hk (SC Zhong)
Date: Fri, 23 Feb 2001 10:36:49 +0800
Subject: Togl cannot be loaded
Message-ID: <Pine.SGI.3.91.1010223102242.20695A-100000@dft.chem.cuhk.edu.hk>

Dear all,

I have an application which use the togl extension
of Tk through python.  I tried to compile python
again with the activation of the togl extension
in the Modules/Setup file.  But somehow the togl
cannot be loaded:

>>> from Tkinter import * 
>>> mst = Tk() 
>>> mst.tk.call('package','require', 'Togl') 
Traceback (innermost last): 
  File "<stdin>", line 1, in ?  
TclError: can't find package Togl
>>> 

There are no compiling errors.  I also searched a lot but
couldn't find some useful documentation of this 'module'.
I have python1.5.2 and on SGI.  The relevant part of the Setup file is: 

# *** Always uncomment this (leave the leading underscore in!):
 _tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
# *** Uncomment and edit to reflect where your Tcl/Tk headers are:
        -I/usr/freeware/include \
# *** Uncomment and edit to reflect where your X11 header files are:
        -I/usr/include/X11 \ 
# *** Or uncomment this for Solaris: 
#       -I/usr/openwin/include \
# *** Uncomment and edit for Tix extension only:
#       -DWITH_TIX -ltix4.1.8.0 \
# *** Uncomment and edit for BLT extension only:
        -DWITH_BLT -I/usr/local/blt2.4u -I/usr/local/blt2.4u/src -lBLT24 \
# *** Uncomment and edit for PIL (TkImaging) extension only:
#       -DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
# *** Uncomment and edit for TOGL extension only:
        -DWITH_TOGL togl.c \
# *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
        -L/usr/freeware/lib \
# *** Uncomment and edit to reflect your Tcl/Tk versions:
        -ltk8.0 -ltcl8.0 \
# *** Uncomment and edit to reflect where your X11 libraries are:
        -L/usr/lib \
# *** Or uncomment this for Solaris:
#       -L/usr/openwin/lib \
# *** Uncomment these for TOGL extension only:
        -lGL -lGLU -lXext -lXmu \
# *** Uncomment for AIX:
#       -lld \
# *** Always uncomment this; X11 libraries to link with:
        -lX11

thanks for your help
sc zhong



From guido at digicool.com  Fri Feb 23 03:59:26 2001
From: guido at digicool.com (Guido van Rossum)
Date: Thu, 22 Feb 2001 21:59:26 -0500
Subject: Nested scopes resolution -- you can breathe again!
In-Reply-To: Your message of "Thu, 22 Feb 2001 16:45:00 EST."
             <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> 
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> 
Message-ID: <200102230259.VAA19238@cj20424-a.reston1.va.home.com>

We (PythonLabs) have received a lot of flak over our plan to introduce
nested scopes despite the fact that it appears to break a small but
significant amount of working code.  We discussed this at an
PythonLabs group meeting today.  After the meeting, Tim posted this
teaser:

> PS: At the internal PythonLabs mtg today, I voted against nested
> scopes.  But also for them.  Leaving that to Jeremy to explain.

After the meeting Jeremy had a four hour commute home due to bad
weather, so let me do the honors for him.  (Jeremy will update the
PEP, implement the feature, and update the documentation, in that
order.)

We have clearly underestimated how much code the nested scopes would
break, but more importantly we have underestimated how much value our
community places on stability.  At the same time we really like nested
scopes, and we would like to see the feature introduced at some point.

So here's the deal: we'll make nested scopes an optional feature in
2.1, default off, selectable on a per-module basis using a mechanism
that's slightly hackish but is guaranteed to be safe.  (See below.)

At the same time, we'll augment the compiler to detect all situations
that will break when nested scopes are introduced in the future, and
issue warnings for those situations.  The idea here is that warnings
don't break code, but encourage folks to fix their code so we can
introduce nested scopes in 2.2.  Given our current pace of releases
that should be about 6 months warning.

These warnings are *not* optional -- they are issued regardless of
whether you select to use nested scopes.  However there is a command
line option (crudest form: -Wi) to disable warnings; there are also
ways to disable them programmatically.  If you want to make sure that
you don't ignore the warnings, there's also a way to turn warnings
into errors (-We from the command line).

How do you select nested scopes?  Tim suggested a mechanism that is
used by the ANSI C committee to enable language features that are
backwards incompatible: they trigger on the import of a specific
previously non-existant header file.  (E.g. after #include
<imaginary.h>, "imaginary" becomes a reserved word.)

The Python equivalent of this is a magical import that is recognized
by the compiler; this was also proposed by David Scherer for making
integer division yield a float.  (See
http://mail.python.org/pipermail/edu-sig/2000-May/000499.html)  You
could say that Perl's "use" statement is similar.

We haven't decided yet which magical import; two proposals are:

    import __nested_scopes__
    from __future__ import nested_scopes

The magical import only affects the source file in which it occurs.
It is recognized by the compiler as it is scanning the source code.
It must appear at the top-level (no "if" or "try" or "def" or anything
else around it) and before any code that could be affected.

We realize that PEP 5 specifies a one-year transition period.  We
believe that that is excessive in this case, and would like to change
the PEP to be more flexible.  (The PEP has questionable status -- it
was never formally discussed.)

We also believe that the magical import mechanism is useful enough to
be reused for other situations like this; Tim will draft a PEP to
describe in excruciating detail.

I thank everybody who gave feedback on this issue.  And thanks to
Jeremy for implementing nested scopes!

--Guido van Rossum (home page: http://www.python.org/~guido/)



From greg at cosc.canterbury.ac.nz  Fri Feb 23 05:25:25 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 23 Feb 2001 17:25:25 +1300 (NZDT)
Subject: [Python-Dev] Nested scopes resolution -- you can breathe again!
In-Reply-To: <200102230259.VAA19238@cj20424-a.reston1.va.home.com>
Message-ID: <200102230425.RAA01507@s454.cosc.canterbury.ac.nz>

> from __future__ import nested_scopes

Hey, neat, now we've got programmatic access to
Guido's time machine from the Python interpreter!

What else will I be able to import from the
future? I'd quite like

  from __future__ import mips_100000

too.

Greg Ewing, Computer Science Dept, +--------------------------------------+
University of Canterbury,	   | A citizen of NewZealandCorp, a	  |
Christchurch, New Zealand	   | wholly-owned subsidiary of USA Inc.  |
greg at cosc.canterbury.ac.nz	   +--------------------------------------+



From tanzer at swing.co.at  Fri Feb 23 08:04:56 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Fri, 23 Feb 2001 08:04:56 +0100
Subject: Nested scopes resolution -- you can breathe again! 
In-Reply-To: Your message of "Thu, 22 Feb 2001 21:59:26 EST."
             <200102230259.VAA19238@cj20424-a.reston1.va.home.com> 
Message-ID: <m14WCHg-000wcEC@swing.co.at>

Guido van Rossum <guido at digicool.com> wrote:

> We (PythonLabs) have received a lot of flak over our plan to introduce
> nested scopes despite the fact that it appears to break a small but
> significant amount of working code.

There were lots of complaints about newly introduced Python features
recently. While I understand concerns about code breakage (myself being
worried about breakage to my code), far more complaints concerned
style issues than breakage.

As I normally don't want to waste time over religious issues -- like
`print >>' being cool or the worst wart ever -- I don't participate in
such discussions on c.l.py.

But IMO Python still is an excellent language and it got even better in the
last year. Thank you and your team for the great work you are doing.

> How do you select nested scopes?  Tim suggested a mechanism that is
> used by the ANSI C committee to enable language features that are
> backwards incompatible: they trigger on the import of a specific
> previously non-existant header file.  (E.g. after #include
> <imaginary.h>, "imaginary" becomes a reserved word.)
> 
> The Python equivalent of this is a magical import that is recognized
> by the compiler; this was also proposed by David Scherer for making
> integer division yield a float.  (See
> http://mail.python.org/pipermail/edu-sig/2000-May/000499.html)  You
> could say that Perl's "use" statement is similar.
> 
> We haven't decided yet which magical import; two proposals are:
> 
>     import __nested_scopes__
>     from __future__ import nested_scopes

Cool.

I hope that `from __future__ import <cool-feature>' will win this
contest -- it defines a common pattern for introducing backwards
incompatible features. 

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From mikael at isy.liu.se  Fri Feb 23 10:41:12 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 23 Feb 2001 10:41:12 +0100 (MET)
Subject: Nested scopes resolution -- you can breathe again!
In-Reply-To: <200102230259.VAA19238@cj20424-a.reston1.va.home.com>
Message-ID: <XFMail.010223104112.mikael@isy.liu.se>

On 23-Feb-01 Guido van Rossum wrote:
 >      from __future__ import nested_scopes

There really is a time machine. So I guess I can get the full Python 3k
functionality by doing

from __future__ import *

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    23-Feb-01
Time:    10:39:52

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From moshez at zadka.site.co.il  Fri Feb 23 10:52:45 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Fri, 23 Feb 2001 11:52:45 +0200 (IST)
Subject: [Python-Dev] RE: Nested scopes resolution -- you can breathe again!
In-Reply-To: <XFMail.010223104112.mikael@isy.liu.se>
References: <XFMail.010223104112.mikael@isy.liu.se>
Message-ID: <20010223095245.A69E2A840@darjeeling.zadka.site.co.il>

On Fri, 23 Feb 2001, Mikael Olofsson <mikael at isy.liu.se> wrote:
 
> There really is a time machine. So I guess I can get the full Python 3k
> functionality by doing
> 
> from __future__ import *

In Py3K from import * will be illegal, so this will unfortunately blow
up the minute the "import_star_bad" is imported.
You'll just have to try them one by one...
-- 
"I'll be ex-DPL soon anyway so I'm        |LUKE: Is Perl better than Python?
looking for someplace else to grab power."|YODA: No...no... no. Quicker,
   -- Wichert Akkerman (on debian-private)|      easier, more seductive.
For public key, finger moshez at debian.org  |http://www.{python,debian,gnu}.org



From news at dorb.com  Sat Feb 24 22:58:13 2001
From: news at dorb.com (Darrell)
Date: Sat, 24 Feb 2001 21:58:13 GMT
Subject: [Python-Dev] RE: Nested scopes resolution -- you can breathe again!
References: <XFMail.010223104112.mikael@isy.liu.se> <mailman.982922104.24606.python-list@python.org>
Message-ID: <V7Wl6.166281$o91.20750508@typhoon.nyroc.rr.com>

"Moshe Zadka" wrote:
> In Py3K from import * will be illegal, so this will unfortunately blow
> up the minute the "import_star_bad" is imported.

Has it been decided that "from x import *" will become illegal?
While dangerous, it can be useful.

--Darrell






From tim.one at home.com  Sun Feb 25 05:58:01 2001
From: tim.one at home.com (Tim Peters)
Date: Sat, 24 Feb 2001 23:58:01 -0500
Subject: [Python-Dev] RE: Nested scopes resolution -- you can breathe again!
In-Reply-To: <V7Wl6.166281$o91.20750508@typhoon.nyroc.rr.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEMOJBAA.tim.one@home.com>

[Moshe Zadka]
> In Py3K from import * will be illegal, so this will unfortunately blow
> up the minute the "import_star_bad" is imported.

[Darrell]
> Has it been decided that "from x import *" will become illegal?

No.

> While dangerous, it can be useful.

Guido agrees.  There are likely to be more restrictions imposed on its use
over time, though.  For example, the Language Reference Manual has always
said "import *" is not to be used except at module scope (but that the
current implementation didn't enforce that).  2.2 *will* enforce that, but
only in some cases when it's combined with nested functions containing free
variables (see PEP 227 (Statically Nested Scopes) for details), and 2.1 will
start *warning* about such uses.




From mikael at isy.liu.se  Fri Feb 23 11:21:06 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 23 Feb 2001 11:21:06 +0100 (MET)
Subject: [Python-Dev] RE: Nested scopes resolution -- you can breathe
In-Reply-To: <01c301c5198d$c6bcc3f0$0900a8c0@SPIFF>
Message-ID: <XFMail.010223112106.mikael@isy.liu.se>

On 23-Feb-05 Fredrik Lundh wrote:
 >  Mikael Olofsson wrote:
 > > from __future__ import *
 >  
 >  I wouldn't do that: it imports both "warnings_are_errors" and
 >  "from_import_star_is_evil", and we've found that it's impossible
 >  to catch ParadoxErrors in a platform independent way.

Naturally. More seriously though, I like 

    from __future__ import something

as an idiom. It gives us a clear reusable syntax to incorporate new 
features before they are included in the standard distribution. It is
not obvious to me that the proposed alternative

    import __something__

is a way to incorporate something new.

Perhaps Py3k should allow

    from __past__ import something

to give us a way to keep some functionality from 2.* that has been 
(will be) changed in Py3k.

explicit-is-better-than-implicit-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    23-Feb-01
Time:    11:07:11

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------



From tim.one at home.com  Fri Feb 23 17:34:58 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 23 Feb 2001 11:34:58 -0500
Subject: [Python-Dev] RE: Nested scopes resolution -- you can breathe
In-Reply-To: <XFMail.010223112106.mikael@isy.liu.se>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEICJBAA.tim.one@home.com>

[Mikael Olofsson]
> Naturally. More seriously though, I like
>
>     from __future__ import something
>
> as an idiom. It gives us a clear reusable syntax to incorporate new
> features before they are included in the standard distribution. It is
> not obvious to me that the proposed alternative
>
>     import __something__
>
> is a way to incorporate something new.

Bingo.  That's why I'm pushing for it.  Also means we only have to create
one artificial module (__future__.py) for this; and besides the doc value,
it occurs to me we *do* have to create a real module anyway so that masses
of tools don't get confused searching for things that look like modules but
don't actually exist.

> Perhaps Py3k should allow
>
>     from __past__ import something
>
> to give us a way to keep some functionality from 2.* that has been
> (will be) changed in Py3k.

Actually, I thought that's something PythonWare could implement as an
extension, to seize the market opportunity created by mean old Guido
breaking all the code he can on a whim <wink>.  Except they'll probably have
to extend the syntax a bit, to make that

    from __past__ import not something

Maybe we should add

    from __future__ import __past__with_not

now to make that easier for them.

> explicit-is-better-than-implicit-ly y'rs

otoh-implicit-manages-to-hide-explicit-suckiness-a-bit-longer-ly
    y'rs  - tim




From nas at arctrix.com  Fri Feb 23 15:36:51 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Fri, 23 Feb 2001 06:36:51 -0800
Subject: [Python-Dev] Nested scopes resolution -- you can breathe again!
In-Reply-To: <200102230259.VAA19238@cj20424-a.reston1.va.home.com>; from guido@digicool.com on Thu, Feb 22, 2001 at 09:59:26PM -0500
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <200102230259.VAA19238@cj20424-a.reston1.va.home.com>
Message-ID: <20010223063651.B23270@glacier.fnational.com>

On Thu, Feb 22, 2001 at 09:59:26PM -0500, Guido van Rossum wrote:
> from __future__ import nested_scopes

I this this alternative better since there is only one "reserved"
module name.  I still think releasing 2.0.1 with warnings is a
good idea.  OTOH, maybe its hard for that compiler to detect
questionable code.

  Neil



From guido at digicool.com  Fri Feb 23 15:42:12 2001
From: guido at digicool.com (Guido van Rossum)
Date: Fri, 23 Feb 2001 09:42:12 -0500
Subject: [Python-Dev] Nested scopes resolution -- you can breathe again!
In-Reply-To: Your message of "Fri, 23 Feb 2001 06:36:51 PST."
             <20010223063651.B23270@glacier.fnational.com> 
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <200102230259.VAA19238@cj20424-a.reston1.va.home.com>  
            <20010223063651.B23270@glacier.fnational.com> 
Message-ID: <200102231442.JAA24227@cj20424-a.reston1.va.home.com>

> > from __future__ import nested_scopes
> 
> I this this alternative better since there is only one "reserved"
> module name.

Noted.

> I still think releasing 2.0.1 with warnings is a
> good idea.  OTOH, maybe its hard for that compiler to detect
> questionable code.

The problem is that in order to do a decent job of compile-time
warnings, not only the warnings module and API would have to be
retrofitted in 2.0.1, but also the entire new compiler, which has the
symbol table needed to be able to detect the situations we want to
warn about.

--Guido van Rossum (home page: http://www.python.org/~guido/)




From jeremy at alum.mit.edu  Wed Feb 28 18:58:58 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Wed, 28 Feb 2001 12:58:58 -0500 (EST)
Subject: PEP 227 (was Re: Nested scopes resolution -- you can breathe again!)
In-Reply-To: <200102230259.VAA19238@cj20424-a.reston1.va.home.com>
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com>
	<200102230259.VAA19238@cj20424-a.reston1.va.home.com>
Message-ID: <15005.15458.703037.373890@w221.z064000254.bwi-md.dsl.cnc.net>

Last week Guido sent a message about our decisions to make nested
scopes an optional feature for 2.1 in advance of their mandatory
introduction in Python 2.2.

I've included an ndiff of the PEP for reference.  The beta release on
Friday will contain the features as described in the PEP.

Jeremy

-: old-pep-0227.txt
+: pep-0227.txt
  PEP: 227
  Title: Statically Nested Scopes
- Version: $Revision: 1.6 $
?                       ^   
+ Version: $Revision: 1.7 $
?                       ^   
  Author: jeremy at digicool.com (Jeremy Hylton)
  Status: Draft
  Type: Standards Track
  Python-Version: 2.1
  Created: 01-Nov-2000
  Post-History:
  
  Abstract
  
      This PEP proposes the addition of statically nested scoping
      (lexical scoping) for Python 2.1.  The current language definition
      defines exactly three namespaces that are used to resolve names --
      the local, global, and built-in namespaces.  The addition of
      nested scopes would allow resolution of unbound local names in
      enclosing functions' namespaces.
  
      One consequence of this change that will be most visible to Python
      programs is that lambda statements could reference variables in
      the namespaces where the lambda is defined.  Currently, a lambda
      statement uses default arguments to explicitly creating bindings
      in the lambda's namespace.
  
  Introduction
  
      This proposal changes the rules for resolving free variables in
-     Python functions.  The Python 2.0 definition specifies exactly
-     three namespaces to check for each name -- the local namespace,
-     the global namespace, and the builtin namespace.  According to
-     this defintion, if a function A is defined within a function B,
-     the names bound in B are not visible in A.  The proposal changes
-     the rules so that names bound in B are visible in A (unless A
+     Python functions.  The new name resolution semantics will take
+     effect with Python 2.2.  These semantics will also be available in
+     Python 2.1 by adding "from __future__ import nested_scopes" to the
+     top of a module.
+ 
+     The Python 2.0 definition specifies exactly three namespaces to
+     check for each name -- the local namespace, the global namespace,
+     and the builtin namespace.  According to this definition, if a
+     function A is defined within a function B, the names bound in B
+     are not visible in A.  The proposal changes the rules so that
+     names bound in B are visible in A (unless A contains a name
-     contains a name binding that hides the binding in B).
?    ----------------                                       
+     binding that hides the binding in B).
  
      The specification introduces rules for lexical scoping that are
      common in Algol-like languages.  The combination of lexical
      scoping and existing support for first-class functions is
      reminiscent of Scheme.
  
      The changed scoping rules address two problems -- the limited
-     utility of lambda statements and the frequent confusion of new
+     utility of lagmbda statements and the frequent confusion of new
?                  +                                                  
      users familiar with other languages that support lexical scoping,
      e.g. the inability to define recursive functions except at the
      module level.
+ 
+ XXX Konrad Hinsen suggests that this section be expanded
  
      The lambda statement introduces an unnamed function that contains
      a single statement.  It is often used for callback functions.  In
      the example below (written using the Python 2.0 rules), any name
      used in the body of the lambda must be explicitly passed as a
      default argument to the lambda.
  
      from Tkinter import *
      root = Tk()
      Button(root, text="Click here",
             command=lambda root=root: root.test.configure(text="..."))
  
      This approach is cumbersome, particularly when there are several
      names used in the body of the lambda.  The long list of default
      arguments obscure the purpose of the code.  The proposed solution,
      in crude terms, implements the default argument approach
      automatically.  The "root=root" argument can be omitted.
  
+     The new name resolution semantics will cause some programs to
+     behave differently than they did under Python 2.0.  In some cases,
+     programs will fail to compile.  In other cases, names that were
+     previously resolved using the global namespace will be resolved
+     using the local namespace of an enclosing function.  In Python
+     2.1, warnings will be issued for all program statement that will
+     behave differently.
+ 
  Specification
  
      Python is a statically scoped language with block structure, in
      the traditional of Algol.  A code block or region, such as a
-     module, class defintion, or function body, is the basic unit of a
+     module, class definition, or function body, is the basic unit of a
?                        +                                               
      program.
  
      Names refer to objects.  Names are introduced by name binding
      operations.  Each occurrence of a name in the program text refers
      to the binding of that name established in the innermost function
      block containing the use.
  
      The name binding operations are assignment, class and function
      definition, and import statements.  Each assignment or import
      statement occurs within a block defined by a class or function
      definition or at the module level (the top-level code block).
  
      If a name binding operation occurs anywhere within a code block,
      all uses of the name within the block are treated as references to
      the current block.  (Note: This can lead to errors when a name is
      used within a block before it is bound.)
  
      If the global statement occurs within a block, all uses of the
      name specified in the statement refer to the binding of that name
      in the top-level namespace.  Names are resolved in the top-level
      namespace by searching the global namespace, the namespace of the
      module containing the code block, and the builtin namespace, the
      namespace of the module __builtin__.  The global namespace is
      searched first.  If the name is not found there, the builtin
-     namespace is searched.
+     namespace is searched.  The global statement must precede all uses
+     of the name.
  
      If a name is used within a code block, but it is not bound there
      and is not declared global, the use is treated as a reference to
      the nearest enclosing function region.  (Note: If a region is
      contained within a class definition, the name bindings that occur
      in the class block are not visible to enclosed functions.)
  
      A class definition is an executable statement that may uses and
      definitions of names.  These references follow the normal rules
      for name resolution.  The namespace of the class definition
      becomes the attribute dictionary of the class.
  
      The following operations are name binding operations.  If they
      occur within a block, they introduce new local names in the
      current block unless there is also a global declaration.
  
-     Function defintion: def name ...
+     Function definition: def name ...
?                   +                   
      Class definition: class name ...
      Assignment statement: name = ...    
      Import statement: import name, import module as name,
          from module import name
      Implicit assignment: names are bound by for statements and except
          clauses
  
      The arguments of a function are also local.
  
      There are several cases where Python statements are illegal when
      used in conjunction with nested scopes that contain free
      variables.
  
      If a variable is referenced in an enclosing scope, it is an error
      to delete the name.  The compiler will raise a SyntaxError for
      'del name'.
  
-     If the wildcard form of import (import *) is used in a function
+     If the wild card form of import (import *) is used in a function
?                +                                                     
      and the function contains a nested block with free variables, the
      compiler will raise a SyntaxError.
  
      If exec is used in a function and the function contains a nested
      block with free variables, the compiler will raise a SyntaxError
-     unless the exec explicit specifies the local namespace for the
+     unless the exec explicitly specifies the local namespace for the
?                             ++                                       
      exec.  (In other words, "exec obj" would be illegal, but 
      "exec obj in ns" would be legal.)
  
+     If a name bound in a function scope is also the name of a module
+     global name or a standard builtin name and the function contains a
+     nested function scope that references the name, the compiler will
+     issue a warning.  The name resolution rules will result in
+     different bindings under Python 2.0 than under Python 2.2.  The
+     warning indicates that the program may not run correctly with all
+     versions of Python.
+ 
  Discussion
  
      The specified rules allow names defined in a function to be
      referenced in any nested function defined with that function.  The
      name resolution rules are typical for statically scoped languages,
      with three primary exceptions:
  
          - Names in class scope are not accessible.
          - The global statement short-circuits the normal rules.
          - Variables are not declared.
  
      Names in class scope are not accessible.  Names are resolved in
-     the innermost enclosing function scope.  If a class defintion
+     the innermost enclosing function scope.  If a class definition
?                                                              +     
      occurs in a chain of nested scopes, the resolution process skips
      class definitions.  This rule prevents odd interactions between
      class attributes and local variable access.  If a name binding
-     operation occurs in a class defintion, it creates an attribute on
+     operation occurs in a class definition, it creates an attribute on
?                                      +                                 
      the resulting class object.  To access this variable in a method,
      or in a function nested within a method, an attribute reference
      must be used, either via self or via the class name.
  
      An alternative would have been to allow name binding in class
      scope to behave exactly like name binding in function scope.  This
      rule would allow class attributes to be referenced either via
      attribute reference or simple name.  This option was ruled out
      because it would have been inconsistent with all other forms of
      class and instance attribute access, which always use attribute
      references.  Code that used simple names would have been obscure.
  
      The global statement short-circuits the normal rules.  Under the
      proposal, the global statement has exactly the same effect that it
-     does for Python 2.0.  It's behavior is preserved for backwards
?                             -                                      
+     does for Python 2.0.  Its behavior is preserved for backwards
      compatibility.  It is also noteworthy because it allows name
      binding operations performed in one block to change bindings in
      another block (the module).
  
      Variables are not declared.  If a name binding operation occurs
      anywhere in a function, then that name is treated as local to the
      function and all references refer to the local binding.  If a
      reference occurs before the name is bound, a NameError is raised.
      The only kind of declaration is the global statement, which allows
      programs to be written using mutable global variables.  As a
      consequence, it is not possible to rebind a name defined in an
      enclosing scope.  An assignment operation can only bind a name in
      the current scope or in the global scope.  The lack of
      declarations and the inability to rebind names in enclosing scopes
      are unusual for lexically scoped languages; there is typically a
      mechanism to create name bindings (e.g. lambda and let in Scheme)
      and a mechanism to change the bindings (set! in Scheme).
  
      XXX Alex Martelli suggests comparison with Java, which does not
      allow name bindings to hide earlier bindings.  
  
  Examples
  
      A few examples are included to illustrate the way the rules work.
  
      XXX Explain the examples
  
      >>> def make_adder(base):
      ...     def adder(x):
      ...         return base + x
      ...     return adder
      >>> add5 = make_adder(5)
      >>> add5(6)
      11
  
      >>> def make_fact():
      ...     def fact(n):
      ...         if n == 1:
      ...             return 1L
      ...         else:
      ...             return n * fact(n - 1)
      ...     return fact
      >>> fact = make_fact()
      >>> fact(7)    
      5040L
  
      >>> def make_wrapper(obj):
      ...     class Wrapper:
      ...         def __getattr__(self, attr):
      ...             if attr[0] != '_':
      ...                 return getattr(obj, attr)
      ...             else:
      ...                 raise AttributeError, attr
      ...     return Wrapper()
      >>> class Test:
      ...     public = 2
      ...     _private = 3
      >>> w = make_wrapper(Test())
      >>> w.public
      2
      >>> w._private
      Traceback (most recent call last):
        File "<stdin>", line 1, in ?
      AttributeError: _private
  
-     An example from Tim Peters of the potential pitfalls of nested scopes
?                                 ^                          -------------- 
+     An example from Tim Peters demonstrates the potential pitfalls of
?                                +++ ^^^^^^^^                           
-     in the absence of declarations:
+     nested scopes in the absence of declarations:
?    ++++++++++++++                                 
  
      i = 6
      def f(x):
          def g():
              print i
          # ...
          # skip to the next page
          # ...
          for i in x:  # ah, i *is* local to f, so this is what g sees
              pass
          g()
  
      The call to g() will refer to the variable i bound in f() by the for
      loop.  If g() is called before the loop is executed, a NameError will
      be raised.
  
      XXX need some counterexamples
  
  Backwards compatibility
  
      There are two kinds of compatibility problems caused by nested
      scopes.  In one case, code that behaved one way in earlier
-     versions, behaves differently because of nested scopes.  In the
?             -                                                       
+     versions behaves differently because of nested scopes.  In the
      other cases, certain constructs interact badly with nested scopes
      and will trigger SyntaxErrors at compile time.
  
      The following example from Skip Montanaro illustrates the first
      kind of problem:
  
      x = 1
      def f1():
          x = 2
          def inner():
              print x
          inner()
  
      Under the Python 2.0 rules, the print statement inside inner()
      refers to the global variable x and will print 1 if f1() is
      called.  Under the new rules, it refers to the f1()'s namespace,
      the nearest enclosing scope with a binding.
  
      The problem occurs only when a global variable and a local
      variable share the same name and a nested function uses that name
      to refer to the global variable.  This is poor programming
      practice, because readers will easily confuse the two different
      variables.  One example of this problem was found in the Python
      standard library during the implementation of nested scopes.
  
      To address this problem, which is unlikely to occur often, a
      static analysis tool that detects affected code will be written.
-     The detection problem is straightfoward.
+     The detection problem is straightforward.
?                                        +      
  
-     The other compatibility problem is casued by the use of 'import *'
?                                           -                            
+     The other compatibility problem is caused by the use of 'import *'
?                                          +                             
      and 'exec' in a function body, when that function contains a
      nested scope and the contained scope has free variables.  For
      example:
  
      y = 1
      def f():
          exec "y = 'gotcha'" # or from module import *
          def g():
              return y
          ...
  
      At compile-time, the compiler cannot tell whether an exec that
-     operators on the local namespace or an import * will introduce
?           ^^                                                       
+     operates on the local namespace or an import * will introduce
?           ^                                                       
      name bindings that shadow the global y.  Thus, it is not possible
      to tell whether the reference to y in g() should refer to the
      global or to a local name in f().
  
      In discussion of the python-list, people argued for both possible
      interpretations.  On the one hand, some thought that the reference
      in g() should be bound to a local y if one exists.  One problem
      with this interpretation is that it is impossible for a human
      reader of the code to determine the binding of y by local
      inspection.  It seems likely to introduce subtle bugs.  The other
      interpretation is to treat exec and import * as dynamic features
      that do not effect static scoping.  Under this interpretation, the
      exec and import * would introduce local names, but those names
      would never be visible to nested scopes.  In the specific example
      above, the code would behave exactly as it did in earlier versions
      of Python.
  
-     Since each interpretation is problemtatic and the exact meaning
?                                         -                           
+     Since each interpretation is problematic and the exact meaning
      ambiguous, the compiler raises an exception.
  
      A brief review of three Python projects (the standard library,
      Zope, and a beta version of PyXPCOM) found four backwards
      compatibility issues in approximately 200,000 lines of code.
      There was one example of case #1 (subtle behavior change) and two
      examples of import * problems in the standard library.
  
      (The interpretation of the import * and exec restriction that was
      implemented in Python 2.1a2 was much more restrictive, based on
      language that in the reference manual that had never been
      enforced.  These restrictions were relaxed following the release.)
  
+ Compatibility of C API
+ 
+     The implementation causes several Python C API functions to
+     change, including PyCode_New().  As a result, C extensions may
+     need to be updated to work correctly with Python 2.1.  
+ 
  locals() / vars()
  
      These functions return a dictionary containing the current scope's
      local variables.  Modifications to the dictionary do not affect
      the values of variables.  Under the current rules, the use of
      locals() and globals() allows the program to gain access to all
      the namespaces in which names are resolved.
  
      An analogous function will not be provided for nested scopes.
      Under this proposal, it will not be possible to gain
      dictionary-style access to all visible scopes.
  
+ Warnings and Errors
+ 
+     The compiler will issue warnings in Python 2.1 to help identify
+     programs that may not compile or run correctly under future
+     versions of Python.  Under Python 2.2 or Python 2.1 if the
+     nested_scopes future statement is used, which are collectively
+     referred to as "future semantics" in this section, the compiler
+     will issue SyntaxErrors in some cases.
+ 
+     The warnings typically apply when a function that contains a
+     nested function that has free variables.  For example, if function
+     F contains a function G and G uses the builtin len(), then F is a
+     function that contains a nested function (G) with a free variable
+     (len).  The label "free-in-nested" will be used to describe these
+     functions. 
+ 
+     import * used in function scope
+ 
+         The language reference specifies that import * may only occur
+         in a module scope.  (Sec. 6.11)  The implementation of C
+         Python has supported import * at the function scope.
+ 
+         If import * is used in the body of a free-in-nested function,
+         the compiler will issue a warning.  Under future semantics,
+         the compiler will raise a SyntaxError.
+ 
+     bare exec in function scope
+ 
+         The exec statement allows two optional expressions following
+         the keyword "in" that specify the namespaces used for locals
+         and globals.  An exec statement that omits both of these
+         namespaces is a bare exec.
+ 
+         If a bare exec is used in the body of a free-in-nested
+         function, the compiler will issue a warning.  Under future
+         semantics, the compiler will raise a SyntaxError.
+ 
+     local shadows global
+ 
+         If a free-in-nested function has a binding for a local
+         variable that (1) is used in a nested function and (2) is the
+         same as a global variable, the compiler will issue a warning.
+ 
  Rebinding names in enclosing scopes
  
      There are technical issues that make it difficult to support
      rebinding of names in enclosing scopes, but the primary reason
      that it is not allowed in the current proposal is that Guido is
      opposed to it.  It is difficult to support, because it would
      require a new mechanism that would allow the programmer to specify
      that an assignment in a block is supposed to rebind the name in an
      enclosing block; presumably a keyword or special syntax (x := 3)
      would make this possible.
  
      The proposed rules allow programmers to achieve the effect of
      rebinding, albeit awkwardly.  The name that will be effectively
      rebound by enclosed functions is bound to a container object.  In
      place of assignment, the program uses modification of the
      container to achieve the desired effect:
  
      def bank_account(initial_balance):
          balance = [initial_balance]
          def deposit(amount):
              balance[0] = balance[0] + amount
              return balance
          def withdraw(amount):
              balance[0] = balance[0] - amount
              return balance
          return deposit, withdraw
  
      Support for rebinding in nested scopes would make this code
      clearer.  A class that defines deposit() and withdraw() methods
      and the balance as an instance variable would be clearer still.
      Since classes seem to achieve the same effect in a more
      straightforward manner, they are preferred.
  
  Implementation
  
      The implementation for C Python uses flat closures [1].  Each def
      or lambda statement that is executed will create a closure if the
      body of the function or any contained function has free
      variables.  Using flat closures, the creation of closures is
      somewhat expensive but lookup is cheap.
  
      The implementation adds several new opcodes and two new kinds of
      names in code objects.  A variable can be either a cell variable
      or a free variable for a particular code object.  A cell variable
      is referenced by containing scopes; as a result, the function
      where it is defined must allocate separate storage for it on each
-     invocation.  A free variable is reference via a function's closure.
?                                                               --------- 
+     invocation.  A free variable is referenced via a function's
?                                              +                  
+     closure. 
+ 
+     The choice of free closures was made based on three factors.
+     First, nested functions are presumed to be used infrequently,
+     deeply nested (several levels of nesting) still less frequently.
+     Second, lookup of names in a nested scope should be fast.
+     Third, the use of nested scopes, particularly where a function
+     that access an enclosing scope is returned, should not prevent
+     unreferenced objects from being reclaimed by the garbage
+     collector. 
  
      XXX Much more to say here
  
  References
  
      [1] Luca Cardelli.  Compiling a functional language.  In Proc. of
      the 1984 ACM Conference on Lisp and Functional Programming,
      pp. 208-217, Aug. 1984
          http://citeseer.nj.nec.com/cardelli84compiling.html
  



From robin at jessikat.fsnet.co.uk  Fri Feb 23 10:45:28 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 23 Feb 2001 09:45:28 +0000
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com>
 <mailman.982897324.9109.python-list@python.org>
Message-ID: <V0yJ0qA4Ejl6EwIt@jessikat.fsnet.co.uk>

In article <mailman.982897324.9109.python-list at python.org>, Guido van
Rossum <guido at digicool.com> writes
.....
>We have clearly underestimated how much code the nested scopes would
>break, but more importantly we have underestimated how much value our
>community places on stability.  At the same time we really like nested
>scopes, and we would like to see the feature introduced at some point.
>
>So here's the deal: we'll make nested scopes an optional feature in
>2.1, default off, selectable on a per-module basis using a mechanism
>that's slightly hackish but is guaranteed to be safe.  (See below.)
>
....
The problem here is that of stability; introducing new versions which
break old extensions at the rate of two a year is the problem. Does this
2.1 release make all the changes to the API which are needed and if so
why do we need to turn python21.dll into python22.dll. The unixers might
get away with smart loaders, but win32 will require a complete rebuild
for all .pyds. The confusion caused by multiple versions is an issue. I
would prefer it if 2.1 moved to 2.11 instead of 2.2.

Whilst helping my colleague migrate from 1.5.2 to 2.0 we came up against
the dreaded python15.dll not found message. The irritating thing about
the message is that it didn't tell us which .pyd it was attempting to
load.
-- 
Robin Becker


From db3l at fitlinxx.com  Mon Feb 26 22:52:58 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 26 Feb 2001 16:52:58 -0500
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <V0yJ0qA4Ejl6EwIt@jessikat.fsnet.co.uk>
Message-ID: <ulmqt158l.fsf@ctwd0143.fitlinxx.com>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> Whilst helping my colleague migrate from 1.5.2 to 2.0 we came up against
> the dreaded python15.dll not found message. The irritating thing about
> the message is that it didn't tell us which .pyd it was attempting to
> load.

Note that if you actually have a python15.dll on your system (so that
the Windows system loader can resolve it), then the Python import
mechanism will complain about the mismatch as an ImportError - at
least if the dependency is directly from the module being imported -
which may be of more help.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From robin at jessikat.fsnet.co.uk  Tue Feb 27 11:43:55 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Tue, 27 Feb 2001 10:43:55 +0000
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com>
 <mailman.982897324.9109.python-list@python.org>
 <V0yJ0qA4Ejl6EwIt@jessikat.fsnet.co.uk> <ulmqt158l.fsf@ctwd0143.fitlinxx.com>
Message-ID: <B8hhIIArT4m6EwC7@jessikat.demon.co.uk>

In article <ulmqt158l.fsf at ctwd0143.fitlinxx.com>, David Bolen
<db3l at fitlinxx.com> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>
>> Whilst helping my colleague migrate from 1.5.2 to 2.0 we came up against
>> the dreaded python15.dll not found message. The irritating thing about
>> the message is that it didn't tell us which .pyd it was attempting to
>> load.
>
>Note that if you actually have a python15.dll on your system (so that
>the Windows system loader can resolve it), then the Python import
>mechanism will complain about the mismatch as an ImportError - at
>least if the dependency is directly from the module being imported -
>which may be of more help.
>
>--
>-- David
yes, but it would be just as easy to get the failure to do that. The
message may come from windows, but the extension (in our case zlib)
still wasn't loaded and that should have been enough, but we had a
try/except around the particular import. Finding a matching pyd/dll
which then fails to load is surely not 'just an import error' even if it
eventually gets raised as one.
-- 
Robin Becker


From db3l at fitlinxx.com  Wed Feb 28 02:26:13 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 20:26:13 -0500
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <V0yJ0qA4Ejl6EwIt@jessikat.fsnet.co.uk> <ulmqt158l.fsf@ctwd0143.fitlinxx.com> <B8hhIIArT4m6EwC7@jessikat.demon.co.uk>
Message-ID: <uu25fmwcq.fsf@ctwd0143.fitlinxx.com>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> yes, but it would be just as easy to get the failure to do that. The
> message may come from windows, but the extension (in our case zlib)

The reason I suggested placing a python15.dll on the system is to get
rid of the Windows message, leaving it just to the Python ImportError.

> still wasn't loaded and that should have been enough, but we had a
> try/except around the particular import. Finding a matching pyd/dll
> which then fails to load is surely not 'just an import error' even if it
> eventually gets raised as one.

Well, but it doesn't have to be too much work.  I'm thinking of the
specific ImportError of:

 "Module use of python15.dll conflicts with this version of Python."

in which case you know that you have an old extension (which has to be
on Python's sys.path somewhere), and that one of three things are
true:

1. The imported module directly depends on python15.dll (probably most
   likely for C extension modules being directly imported, like zlib).
2. The module you imported itself imports some other module with the
   dependency - in that case you have the traceback showing which import
   of which module failed, which reverts to the prior 1.
3. The module has a Windows DLL dependency which eventually pulls in
   python15.dll.  In this case, you know the top level module, so use
   a standard windows tool like depends to check the dependencies.

It's really only case (3) where you need something more than Python
itself provides, but even in (3) you know the module to check, and
it's a relatively straight forward path to fully identifying the root
cause.  It should also be reasonably rare since most extension modules
directly use python15.dll rather than linking to another DLL which itself
is the extension.

Of course, none of this is to excuse the mess that is Windows DLL-land :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From robin at jessikat.fsnet.co.uk  Wed Feb 28 11:27:55 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Wed, 28 Feb 2001 10:27:55 +0000
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com>
 <mailman.982897324.9109.python-list@python.org>
 <V0yJ0qA4Ejl6EwIt@jessikat.fsnet.co.uk> <ulmqt158l.fsf@ctwd0143.fitlinxx.com>
 <B8hhIIArT4m6EwC7@jessikat.demon.co.uk> <uu25fmwcq.fsf@ctwd0143.fitlinxx.com>
Message-ID: <YqkqTnArKNn6Ew4O@jessikat.fsnet.co.uk>

In article <uu25fmwcq.fsf at ctwd0143.fitlinxx.com>, David Bolen
<db3l at fitlinxx.com> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
....
>Well, but it doesn't have to be too much work.  I'm thinking of the
>specific ImportError of:
>
> "Module use of python15.dll conflicts with this version of Python."
>
>in which case you know that you have an old extension (which has to be
>on Python's sys.path somewhere), and that one of three things are
>true:
>
>1. The imported module directly depends on python15.dll (probably most
>   likely for C extension modules being directly imported, like zlib).
>2. The module you imported itself imports some other module with the
>   dependency - in that case you have the traceback showing which import
>   of which module failed, which reverts to the prior 1.
>3. The module has a Windows DLL dependency which eventually pulls in
>   python15.dll.  In this case, you know the top level module, so use
>   a standard windows tool like depends to check the dependencies.
>
....

I guess the 'import' of this is that I should be more specific in
trapping my import errors. Because zlib is often absent we have

try:
        import zlib
except ImportError:
        #other stuff
        ...

The catchall exception is my problem. I reckon the specific error trap
should be

except ImportError, 'No module named zlib'

but this dependency on the format of the message seems a bit fragile.

>Of course, none of this is to excuse the mess that is Windows DLL-land :-)

that is of course true ;(

>-- David

-- 
Robin Becker


From owen at astrono.spamwashington.emu  Fri Feb 23 18:01:19 2001
From: owen at astrono.spamwashington.emu (Russell E. Owen)
Date: Fri, 23 Feb 2001 09:01:19 -0800
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org>
Message-ID: <owen-2302010901210001@cs205-86.fsmodem.washington.edu>

In article <mailman.982897324.9109.python-list at python.org>, Guido van Rossum <guido at digicool.com> wrote:

>We (PythonLabs) have received a lot of flak over our plan to introduce
>nested scopes despite the fact that it appears to break a small but
>significant amount of working code.

I'm unhappy about the plan not because of potentially broken old code, but because it's not as clean as it might be. There are some unfortunate exceptions (such as classes) and it seems a bit more dangerous in Python than in other languages because of Python's lack of requirement to declare variables. A mis-typed variable name may accidentally refer to a variable instead of just creating a new local variable -- a bad bug becomes a worse bug.

I think I am not the only one.

So what I'm saying, at the risk of giving offense, is I'd rather wait and have a cleaner implementation. In a posting yesterday I suggested considering scaling back the proposal to allow it to be simpler and more regular, for example have a keyword or other explicit notation to say "this variable is not in local scope". One person replied with a notation that sounded interesting.

Thank you for the many wonderful enhancements in 2.0. Lots of nice changes in there to make a programmer's life easier. An outstanding upgrade to a great language!

-- Russell


From qrczak at knm.org.pl  Sat Feb 24 15:21:29 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 24 Feb 2001 14:21:29 GMT
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <owen-2302010901210001@cs205-86.fsmodem.washington.edu>
Message-ID: <slrn99fgr9.etd.qrczak@qrnik.zagroda>

Fri, 23 Feb 2001 09:01:19 -0800, Russell E. Owen <owen at astrono.spamwashington.emu> pisze:

> In a posting yesterday I suggested considering scaling back the
> proposal to allow it to be simpler and more regular, for example
> have a keyword or other explicit notation to say "this variable is
> not in local scope".

I would certainly not call it simpler. Maybe for the language
implementor, but not for the programmer, which must search the
expression for its free variables to list them at the top, obscuring
the program.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From rturpin at do.not.use  Sat Feb 24 18:06:09 2001
From: rturpin at do.not.use (Russell Turpin)
Date: Sat, 24 Feb 2001 11:06:09 -0600
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <owen-2302010901210001@cs205-86.fsmodem.washington.edu> <slrn99fgr9.etd.qrczak@qrnik.zagroda>
Message-ID: <3A97EA01.C83F5082@do.not.use>

Fri, 23 Feb 2001 09:01:19 -0800, Russell E. Owen
<owen at astrono.spamwashington.emu> pisze:
>> I suggested .. scaling back the proposal to allow it 
>> to be simpler and more regular, for example have a 
>> keyword or other explicit notation to say "this 
>> variable is not in local scope".

The best argument for this is that it is more EXPLICIT.
Currently, the "global" keyword is required to set a 
variable outside local scope. Simple reference, though,
is automatic. This is consistent with automatic 
declaration, though in both cases, it risks program
errors. I suspect an option that requires explicit 
declaration is one of the more frequent requests by
Python programmers. 

Marcin 'Qrczak' Kowalczyk wrote:
> I would certainly not call it simpler. Maybe for 
> the language implementor, but not for the programmer, 
> which must search the expression for its free variables 
> to list them at the top, obscuring the program.

???

When a programmer writes the variable "x," they had 
better know, in that context, to what it refers. If
they don't, there is pretty low chance that the code
will work. And far from obscuring the intent, perhaps
the SINGLE, BEST practice a programmer can employ, to 
help those who come later understand their work, is 
to put a manifest of all variables at the top of each 
function, class, and module. 

Russell


From qrczak at knm.org.pl  Sat Feb 24 18:55:03 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 24 Feb 2001 17:55:03 GMT
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <owen-2302010901210001@cs205-86.fsmodem.washington.edu> <slrn99fgr9.etd.qrczak@qrnik.zagroda> <3A97EA01.C83F5082@do.not.use>
Message-ID: <slrn99ftbm.59i.qrczak@qrnik.zagroda>

Sat, 24 Feb 2001 11:06:09 -0600, Russell Turpin <rturpin at do.not.use> pisze:

> When a programmer writes the variable "x," they had better know,
> in that context, to what it refers.

How do you describe the "what"? It's just the one that is bound
in the nearest enclosing function. There are no explicit variable
declarations in Python, so nested scopes make it exactly as explicit
as needed and practical under this assumption.

A requirement to qualify each variable with the full path to its
definition (__builtin__, global, functions etc.) would be impractical.

When a variable is not defined but referenced in some functions, what
would you want it to refer to? An obvious candidate is to make it
the same variable as one level up - I didn't tell Python that I want
this name to refer to anything else, so it should stay the same in the
nested function. Making it silently refer to a toplevel definition is
evil, error-prone, counter-intuitive, and I am glad that it's changing.

> And far from obscuring the intent, perhaps the SINGLE, BEST practice
> a programmer can employ, to help those who come later understand
> their work, is to put a manifest of all variables at the top of
> each function, class, and module.

Not at all. You don't manifest names you refer to which are defined
at the module level, nor builtin names, nor you don't explicitly state
that they are local. Why would you make an exception for names bound
in outer but not global scopes?

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From rturpin at do.not.use  Sun Feb 25 17:54:21 2001
From: rturpin at do.not.use (Russell Turpin)
Date: Sun, 25 Feb 2001 10:54:21 -0600
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <owen-2302010901210001@cs205-86.fsmodem.washington.edu> <slrn99fgr9.etd.qrczak@qrnik.zagroda> <3A97EA01.C83F5082@do.not.use> <slrn99ftbm.59i.qrczak@qrnik.zagroda>
Message-ID: <3A9938BD.AE5F0B2B@do.not.use>

Turpin:
>> And far from obscuring the intent, perhaps the SINGLE, 
>> BEST practice a programmer can employ, to help those 
>> who come later understand their work, is to put a 
>> manifest of all variables at the top of each function, 
>> class, and module.

Marcin 'Qrczak' Kowalczyk wrote:
> Not at all. You don't manifest names you refer to which 
> are defined at the module level, nor builtin names, nor 
> you don't explicitly state that they are local. Why would
> you make an exception for names bound in outer but not
> global scopes?

If you're using good programming practices, module names
either (a) explicitly refer to the module everywhere they
appear, e.g., "x = someMod.SomeClass()," or (b) are 
explicitly manifested where they are imported, e.g., 
"from someMod import SomeClass". The wildcard feature on 
the "from" statement is a mistake, and should be avoided. 
Using either (a) or (b), anyone reading the code can 
easily determine which variables are non-local, and where 
they originate.  

Built-in names are are simply part of the language. You
can't extrapolate from how they work to how names generally 
should work.

Even though Python doesn't implement data encapsulation,
it is good practice to lexically differentiate between 
private names and public ones. Where I work, we use naming 
conventions for this purpose. One of the problems with 
wildcard imports is the risk of accidental private name 
collision, where an imported module uses the same private
name as the one that imports it. 

Nested scoping essentially is a wildcard import from all 
containing scopes. There are a VERY FEW cases in OOP where 
this is genuinely useful. For the most part, its use 
should be avoided like the plague. 99% of the time it is
used, there is a more robust and more understandable way
to solve the problem without nested scoping. I understand
the desire for nested scoping with respect to lambda
definitions. Frankly, though, that seems a weak reason
for changing the name resolution mechanism. The only other
argument I've heard is that it makes Python more like
languages that borrowed naming rules from Algol. Python 
is a much better language than Algol, or most of its
successors, so I don't see why it should be jammed into 
that mold.

Python is a great language. Implicit declaration works 
well for programmers who have developed good programming
practices. It can be rope for those who haven't. The 
wildcard on import is a mistake. Someone should submit a
PEP to spit out a warning everytime someone uses it: 
"Hey STUPID! DON'T use wildcard on import!" Then, remove
it from Python3K.

Russell


From qrczak at knm.org.pl  Sun Feb 25 19:11:20 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 25 Feb 2001 18:11:20 GMT
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <owen-2302010901210001@cs205-86.fsmodem.washington.edu> <slrn99fgr9.etd.qrczak@qrnik.zagroda> <3A97EA01.C83F5082@do.not.use> <slrn99ftbm.59i.qrczak@qrnik.zagroda> <3A9938BD.AE5F0B2B@do.not.use>
Message-ID: <slrn99iim8.4f4.qrczak@qrnik.zagroda>

Sun, 25 Feb 2001 10:54:21 -0600, Russell Turpin <rturpin at do.not.use> pisze:

> The wildcard feature on the "from" statement is a mistake, and
> should be avoided.

I agree. This doesn't conflict with my point.

Look this way: a name has not been explicitly hidden in an inner scope,
so it should not be silently made unavailable. Explicit is better than
implicit, so please don't hide names which were not explicitly shadowed
(i.e. assigned to).

Hiding as in older versions of Python leads to obscure bugs when the
default argument hack is forgotten, and more ugly code.

This hack is also not general, because you can't have a variable
number of arguments:
    lambda x, y, var=var, *rest: ...reference var here...
is not a replacement for
    lambda x, y, *rest: ...reference var here...

> Using either (a) or (b), anyone reading the code can easily determine
> which variables are non-local, and where they originate.

A good practice is to not shadow names from outer scopes explicitly.
In this case nested scopes make it easier to find where a name is
coming from: just spot *any* assignment to that name, and it is
defined there.

Python's practice does not create deeply nested scopes anyway,
so this is the smallest problem in finding variable's definition.

> One of the problems with wildcard imports is the risk of accidental
> private name collision, where an imported module uses the same
> private name as the one that imports it.
> 
> Nested scoping essentially is a wildcard import from all containing
> scopes.

You shouldn't show a disadvantage of wildcard imports and tell that
nested scopes are like wildcard imports, because nested scopes don't
have this problem! Names which you define are not shadowed by names
"imported" from nested scopes.

PS. A construction similar to wildcard import is not so bad where the
language protects from using ambiguous names (imported from different
modules or imported from a module and defined locally), e.g. as in
Haskell. They are useful especially when the language uses toplevel
functions where Python uses methods. I agree that it's bad in Python.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From wtanksle at dolphin.openprojects.net  Fri Feb 23 20:57:34 2001
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Fri, 23 Feb 2001 19:57:34 GMT
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org>
Message-ID: <slrn99dg5d.p6b.wtanksle@dolphin.openprojects.net>

On Thu, 22 Feb 2001 21:59:26 -0500, Guido van Rossum wrote:

>    import __nested_scopes__
>    from __future__ import nested_scopes

I like the second version.  That's quite cool.

Can I import stock market quotes from the future, too?  I'm sure that's
going to be a FAQ.  Come to think of it, perhaps __features__ is a better
name than __future__.

And perhaps _old_ code should bear the brunt of the change, with one of
the features being old_unnested_scoping. This way we don't have to keep a
default that's deprecated forever.

So we'd never see the line "from __features__ import nested_scoping"; that
would become an anachonism, and would quickly get irritating.  Instead,
we'd see "from __features__ import old_unnested_scoping" at the head of
modules which won't compile without warning.

>We also believe that the magical import mechanism is useful enough to
>be reused for other situations like this; Tim will draft a PEP to
>describe in excruciating detail.

>I thank everybody who gave feedback on this issue.  And thanks to
>Jeremy for implementing nested scopes!

Absolutely.  I'm very glad that nested scopes are here at last.

>--Guido van Rossum (home page: http://www.python.org/~guido/)

-- 
-William "Billy" Tanksley


From tim.one at home.com  Fri Feb 23 22:17:40 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 23 Feb 2001 16:17:40 -0500
Subject: Nested scopes resolution -- you can breathe again!
In-Reply-To: <slrn99dg5d.p6b.wtanksle@dolphin.openprojects.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEJHJBAA.tim.one@home.com>

[Guido]
>    import __nested_scopes__
>    from __future__ import nested_scopes

[William Tanksley]
> I like the second version.  That's quite cool.

I'd have a hard time guessing that from the rest of this msg <wink>.

> ...
> And perhaps _old_ code should bear the brunt of the change, with one of
> the features being old_unnested_scoping. This way we don't have to keep a
> default that's deprecated forever.

No.  The point of the scheme is to let motivated people use release N+1
features in release N without anyone else needing to change any of their
code.  To the extent possible, release N will instead produce warnings about
code that *will* break in release N+1.  Nobody is required to endure N+1
semantics before N+1; but people who want to experiment with N+1 features in
N can do so by explicitly asking for them.

> So we'd never see the line "from __features__ import nested_scoping";
> that would become an anachonism, and would quickly get irritating.

Good!  That means people will be motivated to get rid of it when it becomes
an anachronism (in N+1).  This puts the burden where it should be (i.e., on
the pioneers, not on the laggards <0.9 wink>).  BTW, for reasons to be
explained in an upcoming PEP, it will be very easy to find unneeded
__future__ imports as the millennia roll by.

> Instead, we'd see "from __features__ import old_unnested_scoping" at
> the head of modules which won't compile without warning.

In release N+1, release N semantics are *gone*.  There's no going back, else
the Python implementation would become an unboundedly complicated mess as
the years go by, and so would Python programs.  __future__ is for
incompatible changes, not for optional behaviors.  If, at release N+1, you
refuse to move to release N+1 semantics, fine:  stay with release N forever.
That's your choice.




From grante at visi.com  Sat Feb 24 03:58:27 2001
From: grante at visi.com (Grant Edwards)
Date: Sat, 24 Feb 2001 02:58:27 GMT
Subject: Nested scopes resolution -- you can breathe again!
References: <mailman.982921325.22962.python-list@python.org>
Message-ID: <nrFl6.937$V6.94615@ruti.visi.com>

In article <mailman.982921325.22962.python-list at python.org>, Mikael Olofsson wrote:
>
>On 23-Feb-01 Guido van Rossum wrote:
> >      from __future__ import nested_scopes
>
>There really is a time machine. So I guess I can get the full Python 3k
>functionality by doing
>
>from __future__ import *

Careful, you never know what the __future__ holds...

-- 
Grant Edwards                   grante             Yow!  Okay... I'm going
                                  at               home to write the "I HATE
                               visi.com            RUBIK's CUBE HANDBOOK FOR
                                                   DEAD CAT LOVERS"...


From wtanksle at dolphin.openprojects.net  Tue Feb 27 00:08:37 2001
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Mon, 26 Feb 2001 23:08:37 GMT
Subject: Nested scopes resolution -- you can breathe again!
References: <mailman.982963084.10423.python-list@python.org>
Message-ID: <slrn99lofm.gv4.wtanksle@dolphin.openprojects.net>

On Fri, 23 Feb 2001 16:17:40 -0500, Tim Peters wrote:
>[Guido]
>>    import __nested_scopes__
>>    from __future__ import nested_scopes

>> And perhaps _old_ code should bear the brunt of the change, with one of
>> the features being old_unnested_scoping. This way we don't have to keep a
>> default that's deprecated forever.

>No.  The point of the scheme is to let motivated people use release N+1
>features in release N without anyone else needing to change any of their
>code.  To the extent possible, release N will instead produce warnings about
>code that *will* break in release N+1.  Nobody is required to endure N+1
>semantics before N+1; but people who want to experiment with N+1 features in
>N can do so by explicitly asking for them.

Ah, so this is only intended for experimental code.  Sorry, I just have a
phobia of "temporary" language features.  Obviously, I've seen too many of
them become permanent.

I say skip the warning stuff unless it's really easy.  In release N,
provide a simple way for people to downgrade to the old semantics when
their code doesn't run right.  In release N+1, try to remove that
mechanism.

>> So we'd never see the line "from __features__ import nested_scoping";
>> that would become an anachonism, and would quickly get irritating.

>Good!  That means people will be motivated to get rid of it when it
>becomes an anachronism (in N+1).

That is a relief; I was worried that it would never be removed.  Looks
like I misread the post.  I was envisioning a world where all newbies
would be taught to start their Python scripts with five or six lines of
boilerplate to get the useful features :-).

>This puts the burden where it should be (i.e., on the pioneers, not on the
>laggards <0.9 wink>).

The *pioneers* are using CVS, never release versions.  They have enough
troubles in life.  The laggards will never be affected: they're still
using 1.5 (my company is!).  The middle-of-the-road people will upgrade to
a new version to get compelling new features, and will be discouraged to
hear that one of the compelling new features has to be accessed in a
manner which will become obsolete shortly.

>BTW, for reasons to be explained in an upcoming PEP, it will be very easy
>to find unneeded __future__ imports as the millennia roll by.

That's cool.

-- 
-William "Billy" Tanksley


From tim.one at home.com  Tue Feb 27 01:24:12 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 26 Feb 2001 19:24:12 -0500
Subject: Nested scopes resolution -- you can breathe again!
In-Reply-To: <slrn99lofm.gv4.wtanksle@dolphin.openprojects.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEECNJCAA.tim.one@home.com>

[Tim, on __future__]
>> The point of the scheme is to let motivated people use release N+1
>> features in release N without anyone else needing to change any of
>> their code.  To the extent possible, release N will instead produce
>> warnings about code that *will* break in release N+1.  Nobody is
>> required to endure N+1 semantics before N+1; but people who want to
>> experiment with N+1 features in N can do so by explicitly asking for
>> them.

[William Tanksley]
> Ah, so this is only intended for experimental code.

No, it's to allow people to migrate incrementally, on a module-by-module
basis, to release N+1 semantics while they're still happily running release
N under release N-1 semantics.

> Sorry, I just have a phobia of "temporary" language features.

Nothing temporary about it.  For example, nested scopes will be THE rule in
2.2.  2.1 will let you get ready for that change, via __future__, without
yet forcing the issue, should you happen to have old code that may break.

> ...
> I say skip the warning stuff unless it's really easy.

The warning stuff is essential.

> In release N, provide a simple way for people to downgrade to the
> old semantics when their code doesn't run right.

Why?  The default is not to break old code, which is overwhelmingly more
often requested.

> In release N+1, try to remove that mechanism.

Six of one, half dozen of the other to me, except that there's no "try" in
N+1, the old behavior is history then.

> ...
> The middle-of-the-road people will upgrade to a new version to get
> compelling new features, and will be discouraged to hear that one
> of the compelling new features has to be accessed in a manner which
> will become obsolete shortly.

1. This has nothing to do with new features per se.  It's for incompatible
   changes.  "incompatible" is key there.  You'll never need a __future__
   statement just to get a new feature, unless that feature *also* has
   the potential to *break* old code.  So, for example, had we done this
   earlier in Python's life, you would NOT have needed a __future__
   statement to get at list comprehensions or print>>.  Those were wholly
   backward compatible.  You WOULD have needed it long ago when "lambda"
   was introduced, because that added a new keyword and so broke any old
   code that used "lambda" as a vrbl, function, class or attribute name.

2. Not obsolete, simply unnecessary -- at N+1 they don't *need* the old
   __future__s anymore, but they're still allowed.

Note too that incompatible changes typically break few lines of code,
although that will vary by user.  For example, the nested scopes change
breaks perhaps three lines of Python in the standard distribution.  So we
normally anticipate that __future__ will get used in only a small percentage
of modules, and usually where people are playing on the edge of the
language, or even in direct violation of what the Language Reference Manual
has said all along (which appears to be the case with real-life nested scope
breakage so far).

it's-not-a-dessert-topping-it's-a-floor-wax-ly y'rs  - tim




From wtanksle at dolphin.openprojects.net  Wed Feb 28 01:09:58 2001
From: wtanksle at dolphin.openprojects.net (William Tanksley)
Date: Wed, 28 Feb 2001 00:09:58 GMT
Subject: Nested scopes resolution -- you can breathe again!
References: <mailman.983233507.13255.python-list@python.org>
Message-ID: <slrn99ogem.2r5.wtanksle@dolphin.openprojects.net>

On Mon, 26 Feb 2001 19:24:12 -0500, Tim Peters wrote:
>1. This has nothing to do with new features per se.  It's for incompatible
>   changes.  "incompatible" is key there.

Whoops!  I didn't get that key distinction.  I see your point, and have to
concede that I have no experience with introducing incompatible changes
into a language.

I'm glad that we're getting the ability to do that, though.  There *are*
some oddities in Python, and it'll be good to be able to make them go away.

>it's-not-a-dessert-topping-it's-a-floor-wax-ly y'rs  - tim

Hmm...  Tastes like foot to me.

-- 
-William "Billy" Tanksley


From jayhawks at netsgo.com  Fri Feb 23 06:22:45 2001
From: jayhawks at netsgo.com (jayhawks at netsgo.com)
Date: Fri, 23 Feb 2001 05:22:45 -0000
Subject: may i get sample source ?
Message-ID: <974s35+jk6u@eGroups.com>

HI?

may i get sample source about python + informix ?
or 
where get i sample ?

i don't know wel about using python+informix .

i am new to python .

thanks.





From adam at deprince.net  Fri Feb 23 07:34:04 2001
From: adam at deprince.net (Adam DePrince)
Date: Fri, 23 Feb 2001 01:34:04 -0500
Subject: may i get sample source ?
References: <mailman.982905784.23883.python-list@python.org>
Message-ID: <3A96045C.FC182C5D@deprince.net>

jayhawks at netsgo.com wrote:
> 
> HI?
> 
> may i get sample source about python + informix ?
> or
> where get i sample ?
> 
> i don't know wel about using python+informix .
> 
> i am new to python .
> 
> thank

You can look up python addons here: http://www.vex.net/parnassus/

Two interfaces to informix are here: 
http://www.vex.net/parnassus/apyllo.py?find=informix

Informixdb claims to support DB1.0; documentation on DB1.0 can be found
here: http://www.python.org/topics/database/DatabaseAPI-1.0.html

Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From rolney at pcug.org.au  Fri Feb 23 06:44:27 2001
From: rolney at pcug.org.au (Robert Olney)
Date: Fri, 23 Feb 2001 15:44:27 +1000
Subject: Can't redirect output
Message-ID: <3A95F8BA.9CD22C6F@pcug.org.au>

I am running Python 2.0 on Windows 98. When I try to redirect output
from a command prompt, e.g.

C:\> python hello.py > out.txt

the output is printed on the screen and not to the file. Why? The FAQ
mentions that this is a problem under Linux.

Thanks,
Robert.




From chris.gonnerman at usa.net  Fri Feb 23 14:44:39 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Fri, 23 Feb 2001 07:44:39 -0600
Subject: Can't redirect output
References: <3A95F8BA.9CD22C6F@pcug.org.au>
Message-ID: <002401c09d9e$c91eb3c0$a100000a@local>

From: "Robert Olney" <rolney at pcug.org.au>
Subject: Can't redirect output
> I am running Python 2.0 on Windows 98. When I try to redirect output
> from a command prompt, e.g.
> 
> C:\> python hello.py > out.txt
> 
> the output is printed on the screen and not to the file. Why? 

Is the output you are talking about from print commands and/or 
sys.stdout.write(...) calls, or is it error messages?  Unix and Windows 
processes have two standard channels for writing messages, stdout and
stderr, which Python maps to sys.stdout and sys.stderr.  The redirection
you show only redirects stdout/sys.stdout.

On Unix (and reportedly Windows NT) you can write:

    python hello.py >out.txt 2>errors.txt

which separates the output channels, or

    python hello.py >out.txt 2>&1

to combine them in the file.  Unfortunately this does not work on Win9X,
but there is a program available called stderr which combines the two 
channels.  Get it from:

    http://www.teaser.fr/~amajorel/stderr/

IF NONE OF THIS HELPS YOU... please post your hello.py so we can
inspect it.

>The FAQ
> mentions that this is a problem under Linux.

I'll have to look at the FAQ, I'm not sure why Linux would have a problem 
with it.





From rolney at pcug.org.au  Sat Feb 24 01:50:33 2001
From: rolney at pcug.org.au (Robert Olney)
Date: Sat, 24 Feb 2001 00:50:33 GMT
Subject: Can't redirect output
References: <3A95F8BA.9CD22C6F@pcug.org.au> <mailman.982935605.18416.python-list@python.org>
Message-ID: <3a9703b0.4598430@newshost.pcug.org.au>

"Chris Gonnerman" <chris.gonnerman at usa.net> wrote:

>From: "Robert Olney" <rolney at pcug.org.au>
>Subject: Can't redirect output
>> I am running Python 2.0 on Windows 98. When I try to redirect output
>> from a command prompt, e.g.
>> 
>> C:\> python hello.py > out.txt
>> 
>> the output is printed on the screen and not to the file. Why? 
>
>Is the output you are talking about from print commands and/or 
>sys.stdout.write(...) calls, or is it error messages?  Unix and Windows 
>processes have two standard channels for writing messages, stdout and
>stderr, which Python maps to sys.stdout and sys.stderr.  The redirection
>you show only redirects stdout/sys.stdout.
>
>On Unix (and reportedly Windows NT) you can write:
>
>    python hello.py >out.txt 2>errors.txt
>
>which separates the output channels, or
>
>    python hello.py >out.txt 2>&1
>
>to combine them in the file.  Unfortunately this does not work on Win9X,
>but there is a program available called stderr which combines the two 
>channels.  Get it from:
>
>    http://www.teaser.fr/~amajorel/stderr/
>
>IF NONE OF THIS HELPS YOU... please post your hello.py so we can
>inspect it.
>
>>The FAQ
>> mentions that this is a problem under Linux.
>
>I'll have to look at the FAQ, I'm not sure why Linux would have a problem 
>with it.
>
>
>
I am using normal print commands.

Hello.py:
print "Hello"

I have now found I can't redirect this under DOS (python 1.5.2)
either.

Robert.


From chris.gonnerman at usa.net  Sat Feb 24 07:03:31 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 24 Feb 2001 00:03:31 -0600
Subject: Can't redirect output
References: <3A95F8BA.9CD22C6F@pcug.org.au> <mailman.982935605.18416.python-list@python.org> <3a9703b0.4598430@newshost.pcug.org.au>
Message-ID: <00e201c09e27$8929b920$a100000a@local>

----- Original Message -----
From: "Robert Olney" <rolney at pcug.org.au>
Subject: Re: Can't redirect output
> >> I am running Python 2.0 on Windows 98. When I try to redirect output
> >> from a command prompt, e.g.
> >>
> >> C:\> python hello.py > out.txt
> >>
> >> the output is printed on the screen and not to the file. Why?
> >
> I am using normal print commands.
>
> Hello.py:
> print "Hello"

Huh.  I just put that single line (print "Hello") into a script file and did

    python Hello.py > out.txt

as you specified, and it worked find.  I'm using BeOpen Python 2.0 on
Windows 98.
The same code works fine under Python 1.5.2 on Linux 2.2.

I'm puzzled.  Can you try this on a different Win98 computer?  Do you have a
keyboard
or console redirector running (not DOSKEY but a clone)?  Are you running
COMMAND.COM, 4DOS, Win95CMD.exe, or what?  Can you think of anything
unusual about your console configuration?

> I have now found I can't redirect this under DOS (python 1.5.2)
> either.

I've never tried the DOS version, so I can't comment here.




From timo at lightdog.com  Mon Feb 26 18:00:49 2001
From: timo at lightdog.com (timo at lightdog.com)
Date: Mon, 26 Feb 2001 12:00:49 -0500
Subject: Can't redirect output
In-Reply-To: <00e201c09e27$8929b920$a100000a@local>
Message-ID: <3A9A4571.25103.BF62D3@localhost>

On 24 Feb 2001, at 0:03, Chris Gonnerman wrote:
> > I have now found I can't redirect this under DOS (python 1.5.2)
> > either.
> 
> I've never tried the DOS version, so I can't comment here.
> 

I guess this is a known problem on NT? I have to work with NT, and 
I never found out exactly why it won't redirect. I found that if I create 
a .bat file that does nothing but call the .py file, I can redirect that 
just fine. Sounds like something evil going on under the covers.

Tim (no not that one)



From daniel at dittmar.net  Mon Feb 26 20:48:42 2001
From: daniel at dittmar.net (Daniel Dittmar)
Date: Mon, 26 Feb 2001 20:48:42 +0100
Subject: Can't redirect output
References: <mailman.983207049.352.python-list@python.org>
Message-ID: <3A9AB31A.31DC2202@dittmar.net>

> I guess this is a known problem on NT? I have to work with NT, and
> I never found out exactly why it won't redirect. I found that if I create
> a .bat file that does nothing but call the .py file, I can redirect that
> just fine. Sounds like something evil going on under the covers.

There is a known Problem on NT that starting a program through it's
extension will make the redirection impossible.

So 'python hello.py > somefile.txt' should work. But 'hello.py >
somefile.txt' does not. Solutions: don't know, I'm using 4NT as a shell,
so I don't have this problem. Maybe auto-creating batch files for every
Python program is the simplest method.

Daniel


From adam at deprince.net  Fri Feb 23 07:44:50 2001
From: adam at deprince.net (Adam DePrince)
Date: Fri, 23 Feb 2001 01:44:50 -0500
Subject: interactive programming
References: <mailman.982954873.25844.python-list@python.org>
Message-ID: <3A9606E2.9EEDE70A@deprince.net>

Chris Richard Adams wrote:
> 
> How can I create a loop - interactively.  I'm just trying to learn
> syntax for example, so Id like to create conditional if statements like
> a = 'feb'
> 
> if a == 'jan':
>         mo = 1
> elif a == 'feb':
>         mo = 2
> elif a == 'mar':
>         mo = 3
> ...
> 
> I tried to type this interactively, but I get a mess of error about
> syntax or mis placed whitespace...
> 
> 1.) After the return after mo = 1 - how do I make sure theelif is
> aligned with if?
> 
> 2.) Is there any refernces for interactive programming - the tutorial is
> too short?
> 
> Thanks,Chris

Ah.  To end a block interactively you must hit enter twice.

$ python
Python 2.0 (#1, Dec 18 2000, 00:59:58) 
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> for x in range(10):
...     print x <- Enter
...             <- Enter
0
1
2
3
4
5
6
7
8
9
>>> 


Your code:

> if a == 'jan':
>         mo = 1
> elif a == 'feb':
>         mo = 2
> elif a == 'mar':
>         mo = 3
> ...
> 

would require an extra enter after mo=12.

Instead of using nested ifs, I would do this:

months = {"jan":1,"feb":2,"mar":3 ... :12}


b = months[a]

Of course, you don't want to create months each and every time you do
the lookup, otherwise you havn't really gained anything.

Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From adam at deprince.net  Fri Feb 23 07:48:15 2001
From: adam at deprince.net (Adam DePrince)
Date: Fri, 23 Feb 2001 01:48:15 -0500
Subject: Can I leverage Java classes from a python program
References: <mailman.982955345.26200.python-list@python.org>
Message-ID: <3A9607AF.347A5BCB@deprince.net>

Chris Richard Adams wrote:
> 
> If so, questions:
> 
> I am working with Java Oracle Intermedia classes (just a Java API that
> helps you manipulate images (Blobs) stored in a database). ANyway - I
> have to work with the java oracle classes, but wanted to do this from a
> python program. Can I use inheritance to access these types of classes,
> even though they are in Java. If so could you point to a reference that
> might show some examples???? AND does this have anything to do with
> JPython - would I want to use that rather than my standard Linux 2.0
> version.
> 
> thanks!

You will have to use Jython for this.  

Alternatively, you can connect to Oracle via C-Python to do the work and
forget Java altogether.  Take a look at DCOracle:
http://www.zope.org/Products/DCOracle


-- 
Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From juergen0509 at web.de  Sat Feb 24 00:00:40 2001
From: juergen0509 at web.de (Juergen Bocklage)
Date: Sat, 24 Feb 2001 00:00:40 +0100
Subject: Can I leverage Java classes from a python program
References: <mailman.982955345.26200.python-list@python.org>
Message-ID: <3A96EB98.9CAC1121@web.de>

Yes, I think JPython or now Jython is the right choice.See
http://www.jython.org/  and at http://www.jython.org/docs/index.html you
should find everything what is needed to subclass existing java classes.

Linux 2.0?? or do you mean Python 2.0 ;-)

- JB

Chris Richard Adams wrote:

> If so, questions:
>
> I am working with Java Oracle Intermedia classes (just a Java API that
> helps you manipulate images (Blobs) stored in a database). ANyway - I
> have to work with the java oracle classes, but wanted to do this from a
> python program. Can I use inheritance to access these types of classes,
> even though they are in Java. If so could you point to a reference that
> might show some examples???? AND does this have anything to do with
> JPython - would I want to use that rather than my standard Linux 2.0
> version.
>
> thanks!

--
Juergen Bocklage
================================
mailto:juergen.bocklage at gmx.net




From dkuhlman at rexx.com  Sun Feb 25 23:18:03 2001
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Sun, 25 Feb 2001 22:18:03 -0000
Subject: Can I leverage Java classes from a python program
References: <mailman.982955345.26200.python-list@python.org>
Message-ID: <983139494.154967@rexx.com>

Others have given you pointers to JPython.

What about JPE -- Java-Python Extension.  See

    http://sourceforge.net/projects/jpe


I'd be interested in seeing an evaluation of JPE.

  - Dave


Chris Richard Adams <chrisa at aspatech.com.br> wrote:
> 
> If so, questions:
> 
> I am working with Java Oracle Intermedia classes (just a Java API that
> helps you manipulate images (Blobs) stored in a database). ANyway - I
> have to work with the java oracle classes, but wanted to do this from a
> python program. Can I use inheritance to access these types of classes,
> even though they are in Java. If so could you point to a reference that
> might show some examples???? AND does this have anything to do with
> JPython - would I want to use that rather than my standard Linux 2.0
> version.
> 
> thanks!
> 
> 

-- 
Dave Kuhlman
dkuhlman at rexx.com


From shaleh at valinux.com  Fri Feb 23 08:54:17 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Thu, 22 Feb 2001 23:54:17 -0800
Subject: parsing based on BNF?
Message-ID: <20010222235417.A12624@valinux.com>

Anyone have any pointers to Python tools for parsing based on BNF's?

I am looking to verify fields in files and to point out mistakes in these
fields.




From johnw at gnu.org  Fri Feb 23 09:18:53 2001
From: johnw at gnu.org (John Wiegley)
Date: 23 Feb 2001 01:18:53 -0700
Subject: parsing based on BNF?
References: <mailman.982914844.7621.python-list@python.org>
Message-ID: <m3y9ux95hu.fsf@gnu.org>

>>>>> On Fri Feb 23, Sean writes:

> Anyone have any pointers to Python tools for parsing based on BNF's?

On the Vaults of Parnassus (http://www.vex.net/parnassus), you can
find a FlexBison module in the parsing section.  Also, there's
mxTextTools, which is kind of a mix between BNF and assembly.  It's
killer fast, though.


From uche at ogbuji.net  Sat Feb 24 23:24:58 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Sat, 24 Feb 2001 22:24:58 GMT
Subject: parsing based on BNF?
References: <mailman.982914844.7621.python-list@python.org> <m3y9ux95hu.fsf@gnu.org>
Message-ID: <3A9834BB.664DE48E@ogbuji.net>

John Wiegley wrote:
> 
> >>>>> On Fri Feb 23, Sean writes:
> 
> > Anyone have any pointers to Python tools for parsing based on BNF's?
> 
> On the Vaults of Parnassus (http://www.vex.net/parnassus), you can
> find a FlexBison module in the parsing section.  Also, there's
> mxTextTools, which is kind of a mix between BNF and assembly.  It's
> killer fast, though.

Problem is, Sean asked for parsing based on BNFs.  I'm assuming he has a
huge grammar lying about in BNF (or EBNF) format, which we doesn't want
to translate to some other format if he can avoid it.

I'm also curious to know if there is such a beast for Python or C/C++.


-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From johnw at gnu.org  Sun Feb 25 01:33:58 2001
From: johnw at gnu.org (John Wiegley)
Date: 24 Feb 2001 17:33:58 -0700
Subject: parsing based on BNF?
References: <mailman.982914844.7621.python-list@python.org> <m3y9ux95hu.fsf@gnu.org> <3A9834BB.664DE48E@ogbuji.net>
Message-ID: <m34rxjd2ix.fsf@gnu.org>

>>>>> On Sat Feb 24, Uche writes:

> Problem is, Sean asked for parsing based on BNFs.  I'm assuming he
> has a huge grammar lying about in BNF (or EBNF) format, which we
> doesn't want to translate to some other format if he can avoid it.

Well, somebody did write a BNF->mxTextTools converter.  I forget the
name of it, but it's on the mxTextTools web page.


From dsh8290 at rit.edu  Mon Feb 26 04:03:23 2001
From: dsh8290 at rit.edu (D-Man)
Date: Sun, 25 Feb 2001 22:03:23 -0500
Subject: parsing based on BNF?
In-Reply-To: <3A9834BB.664DE48E@ogbuji.net>; from uche@ogbuji.net on Sat, Feb 24, 2001 at 10:24:58PM +0000
References: <mailman.982914844.7621.python-list@python.org> <m3y9ux95hu.fsf@gnu.org> <3A9834BB.664DE48E@ogbuji.net>
Message-ID: <20010225220323.A13230@harmony.cs.rit.edu>

On Sat, Feb 24, 2001 at 10:24:58PM +0000, Uche Ogbuji wrote:
| 
| I'm also curious to know if there is such a beast for Python or C/C++.
| 

If you want to get C code as the result, lex + yacc is a good
combination.  With lex (flex) you specify regexes that describe the
various tokens in your language.  With yacc (bison) you specify the BNF
in terms of the lex tokens and what actions the parser should take
when that expression is found.

-D




From uche at ogbuji.net  Tue Feb 27 06:04:49 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Tue, 27 Feb 2001 05:04:49 GMT
Subject: parsing based on BNF?
References: <mailman.982914844.7621.python-list@python.org> <m3y9ux95hu.fsf@gnu.org> <3A9834BB.664DE48E@ogbuji.net> <mailman.983156647.18494.python-list@python.org>
Message-ID: <3A9B356F.6A4350FA@ogbuji.net>

D-Man wrote:
> 
> On Sat, Feb 24, 2001 at 10:24:58PM +0000, Uche Ogbuji wrote:
> |
> | I'm also curious to know if there is such a beast for Python or C/C++.
> |
> 
> If you want to get C code as the result, lex + yacc is a good
> combination.  With lex (flex) you specify regexes that describe the
> various tokens in your language.  With yacc (bison) you specify the BNF
> in terms of the lex tokens and what actions the parser should take
> when that expression is found.

Not what I was talking about.

Lex uses regular expressions and a proprietary glue language.

Yacc is based on BNF, but not quite.

I'm quite familiar with all the various scanners and parsers, which is
why I asked my question.  I do not know of one that implements plain BNF
(or EBNF to provide some lexical specifications).  This would have been
useful in implementing the many little languages I have: XPath, OQL,
ODL, etc. which are specified in BNF.

So far I've either had to convert them to lex or yacc, a tedious
process, or to our XML-based BisonGen format.


-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From aycock at csc.uvic.ca  Wed Feb 28 01:05:08 2001
From: aycock at csc.uvic.ca (John Aycock)
Date: 28 Feb 2001 00:05:08 GMT
Subject: parsing based on BNF?
References: <mailman.982914844.7621.python-list@python.org> <m3y9ux95hu.fsf@gnu.org> <3A9834BB.664DE48E@ogbuji.net> <mailman.983156647.18494.python-list@python.org> <3A9B356F.6A4350FA@ogbuji.net>
Message-ID: <97hfbk$uno$1@casper.UVic.CA>

Uche Ogbuji <uche at ogbuji.net> wrote:
> I'm quite familiar with all the various scanners and parsers, which is
> why I asked my question.  I do not know of one that implements plain BNF
> (or EBNF to provide some lexical specifications).  This would have been
> useful in implementing the many little languages I have: XPath, OQL,
> ODL, etc. which are specified in BNF.

Well, I guess I'll add my shameless plug.  SPARK's parser can handle
arbitrary BNF grammars, even ambiguous ones.  It was originally
intended for little languages.  You can find SPARK at

	http://csr.uvic.ca/~aycock/python

John


From jdries at mail.com  Wed Feb 28 17:46:58 2001
From: jdries at mail.com (Jan Dries)
Date: Wed, 28 Feb 2001 17:46:58 +0100
Subject: parsing based on BNF?
References: <mailman.982914844.7621.python-list@python.org> <m3y9ux95hu.fsf@gnu.org> <3A9834BB.664DE48E@ogbuji.net> <mailman.983156647.18494.python-list@python.org> <3A9B356F.6A4350FA@ogbuji.net> <97hfbk$uno$1@casper.UVic.CA>
Message-ID: <3A9D2B82.9ECF6391@mail.com>

John Aycock wrote:
> 
> Uche Ogbuji <uche at ogbuji.net> wrote:
> > I'm quite familiar with all the various scanners and parsers, which is
> > why I asked my question.  I do not know of one that implements plain BNF
> > (or EBNF to provide some lexical specifications).  This would have been
> > useful in implementing the many little languages I have: XPath, OQL,
> > ODL, etc. which are specified in BNF.
> 
> Well, I guess I'll add my shameless plug.  SPARK's parser can handle
> arbitrary BNF grammars, even ambiguous ones.  It was originally
> intended for little languages.  

I've been using the SPARK parser, and I like it a lot because of the
flexibility is has in dealing with grammars. In fact, I used it to
implement XPath, something the previous poster was refering too. 
However, there's one disadvantage in SPARK: it's slow. Parsing a few not
even very complex XPath expression takes something in the order of a
second.

Regards,
Jan



From mfletch at tpresence.com  Sun Feb 25 05:08:27 2001
From: mfletch at tpresence.com (Mike Fletcher)
Date: Sat, 24 Feb 2001 23:08:27 -0500
Subject: parsing based on BNF?
Message-ID: <CC59868A0A76D311B1F50090277B321C1E14C0@VRTMAIL>

Hmm, that would be me, I think.  But it's a particular EBNF (there are lots
of EBNF formats), so some translation might still be required.  At present
it (SimpleParse) only allows for... trying to recall... I think it's LL(1)?
languages.  I don't work with SimpleParse much directly anymore (there are
problems with defining new languages related to reporting of failed groups.
I was given a patch for that but never had the time to thoroughly test and
revise the generator to take advantage of the fix.), I just use the VRML97
parser that it generates for me.

Might be useful to "Sean" (didn't see the original post) or not, see the
homepage at:

http://members.home.com/mcfletch/programming/simpleparse/simpleparse.html

Enjoy,
Mike

-----Original Message-----
From: John Wiegley [mailto:johnw at gnu.org]
Sent: Saturday, February 24, 2001 7:34 PM
To: python-list at python.org
Subject: Re: parsing based on BNF?



>>>>> On Sat Feb 24, Uche writes:

> Problem is, Sean asked for parsing based on BNFs.  I'm assuming he
> has a huge grammar lying about in BNF (or EBNF) format, which we
> doesn't want to translate to some other format if he can avoid it.

Well, somebody did write a BNF->mxTextTools converter.  I forget the
name of it, but it's on the mxTextTools web page.
-- 
http://mail.python.org/mailman/listinfo/python-list



From robin at alldunn.com  Fri Feb 23 09:00:32 2001
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 23 Feb 2001 00:00:32 -0800
Subject: ANNOUNCE: Release 3.0b3 of bsddb3 is available
Message-ID: <008901c09d6e$b317d1e0$0100a8c0@Rogue>

The Python wrappers for BerkeleyDB 3.1.17 (http://www.sleepycat.com/) are
available at http://pybsddb.sourceforge.net/.

Berkeley DB is a programmatic toolkit that provides high-performance
built-in database support for desktop and server applications.
The Berkeley DB access methods include B+tree, Extended Linear Hashing,
Fixed and Variable-length records, and Queues. Berkeley DB provides full
transactional support, database recovery, online backups, multi-threaded and
multi-process access, etc.

The Python wrappers allow you to store Python string objects of any length,
keyed either by strings or integers depending on the database access method.
With the use of another module in the package standard shelve-like
functionality is provided allowing you to store any picklable Python object!

Changes in this release include the following:

* Lots of bugs fixed.

* General code cleanup and some refactoring.

* Better support for building on many different platforms.

* Better build/install docs.

* More and better unit tests.

* Added DBLock type and methods to DBEnv for acquiring and releasing
arbitrary locks.

* Partial record reads/writes, and class for doing it in a file-like way.

* Classes added for doing lightweight table-like access using a DB for
storage.

* Various other methods added.


The focus of the next release will be support for the recently released
BerkeleyDB 3.2.9.

--
Robin Dunn
Software Craftsman
robin at AllDunn.com       Java give you jitters?
http://wxPython.org      Relax with wxPython!






From ycheng at sinica.edu.tw  Fri Feb 23 09:01:22 2001
From: ycheng at sinica.edu.tw (Y. Cheng)
Date: Fri, 23 Feb 2001 08:01:22 +0000 (UTC)
Subject: How python output a unicode string.
Message-ID: <9755ci$1t4m$1@news1.sinica.edu.tw>

Hi,

I try to install a big5.py on /usr/lib/python2.0/encodings
and it seem working fine. Then I try to output an string
which is u"\u9DFD" to stdout by using print. Before that,
I call

	sys.setdefaultencoding("big5")

without any error message.

As I try to output the string, I got error message:

-----------------------
  File "/usr/lib/python2.0/encodings/big5.py", line 18, in encode
    return codecs.charmap_encode(input,errors,encoding_map)
TypeError: character mapping must be in range(256)
-----------------------

is this means that python can't output multi-byte character
with unicode internal encoding ?

Yuan-Chen Cheng


From footech at get2net.dk  Fri Feb 23 10:17:52 2001
From: footech at get2net.dk (Mikkel Rasmussen)
Date: Fri, 23 Feb 2001 10:17:52 +0100
Subject: Newbie: Large dictionaries
Message-ID: <WUpl6.43$Gv2.2355@news.get2net.dk>

Why can't there be more than about 65.000 entries in a dictionary, or more
importantly: how can I make a large dictionary with room for at least
500.000
elements?

I have looked in "Programming Python", "Python in 24 hours" (don't laugh!)
and "Python Pocket Reference" but I can't even find that there is a limit.


Mikkel Rasmussen
www.mellon.dk




From cg at schlund.de  Fri Feb 23 10:29:44 2001
From: cg at schlund.de (Carsten Gaebler)
Date: Fri, 23 Feb 2001 10:29:44 +0100
Subject: Newbie: Large dictionaries
References: <WUpl6.43$Gv2.2355@news.get2net.dk>
Message-ID: <3A962D88.BB3C0282@schlund.de>

Mikkel Rasmussen wrote:
> 
> Why can't there be more than about 65.000 entries in a dictionary, or more
> importantly: how can I make a large dictionary with room for at least
> 500.000
> elements?

Don't know where your problem is:

#!/usr/bin/python

d = {}
for i in xrange(100000):
	d[i] = i

for (k, v) in d.items():
	print k, v



cg.


From stephen_purcell at yahoo.com  Fri Feb 23 11:36:54 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 23 Feb 2001 11:36:54 +0100
Subject: Newbie: Large dictionaries
In-Reply-To: <WUpl6.43$Gv2.2355@news.get2net.dk>; from footech@get2net.dk on Fri, Feb 23, 2001 at 10:17:52AM +0100
References: <WUpl6.43$Gv2.2355@news.get2net.dk>
Message-ID: <20010223113654.A1286@freedom.puma-ag.com>

Mikkel Rasmussen wrote:
> Why can't there be more than about 65.000 entries in a dictionary, or more
> importantly: how can I make a large dictionary with room for at least
> 500.000
> elements?

It's probably a memory issue: how big are the keys and/or values?

On my machine, I can easily get a million entries into a dictionary:-

>>> dict = {}
>>> try:
...   for i in xrange(1000000):
...     dict[i] = None
... except:
...   print "stopped at", i
...   raise
... 
>>> 

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From footech at get2net.dk  Fri Feb 23 13:09:08 2001
From: footech at get2net.dk (Mikkel Rasmussen)
Date: Fri, 23 Feb 2001 13:09:08 +0100
Subject: Newbie: Large dictionaries
References: <WUpl6.43$Gv2.2355@news.get2net.dk> <mailman.982924567.27809.python-list@python.org>
Message-ID: <vpsl6.134$Gv2.4218@news.get2net.dk>

Hi!

I can also put more than 4 million entries *into* the dictionary. But I can
only get less than 65.000 back out!

What do you get with: len(dict.keys()) ?

And with a loop like

count = 0
for elem in dict:
    count = count + 1
print count

The keys are ordinary words like an ordinary dictionary :-) with an average
length of about 6 characters.


Mikkel Rasmussen


Steve Purcell <stephen_purcell at yahoo.com> wrote in message
news:mailman.982924567.27809.python-list at python.org...
> Mikkel Rasmussen wrote:
> > Why can't there be more than about 65.000 entries in a dictionary, or
more
> > importantly: how can I make a large dictionary with room for at least
> > 500.000
> > elements?
>
> It's probably a memory issue: how big are the keys and/or values?
>
> On my machine, I can easily get a million entries into a dictionary:-
>
> >>> dict = {}
> >>> try:
> ...   for i in xrange(1000000):
> ...     dict[i] = None
> ... except:
> ...   print "stopped at", i
> ...   raise
> ...
> >>>
>
> -Steve
>
> --
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Get servlets at http://pyserv.sourceforge.net/
> "Even snakes are afraid of snakes." -- Steven Wright
>




From fredrik at pythonware.com  Fri Feb 23 13:32:03 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 23 Feb 2001 12:32:03 GMT
Subject: Newbie: Large dictionaries
References: <WUpl6.43$Gv2.2355@news.get2net.dk> <mailman.982924567.27809.python-list@python.org> <vpsl6.134$Gv2.4218@news.get2net.dk>
Message-ID: <7Lsl6.15130$Qb7.2431128@newsb.telia.net>

Mikkel Rasmussen wrote:
> I can also put more than 4 million entries *into* the dictionary. But I can
> only get less than 65.000 back out!
>
> What do you get with: len(dict.keys()) ?

if that reports 65,000, you probably don't have more
than 65,000 *unique* entries.

>>> dict = {}
>>> dict["a"] = 1
>>> dict["a"] = 2
>>> dict["a"] = 3
>>> len(dict)
1
>>> len(dict.keys())
1

> And with a loop like
>
> count = 0
> for elem in dict:
>     count = count + 1
> print count

>>> count = 0
>>> for elem in dict:
...     count = count + 1
...
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: loop over non-sequence

are you sure you're using Python?

> The keys are ordinary words like an ordinary dictionary :-) with an average
> length of about 6 characters.

and what language has four million *unique* words
having six characters and less?

Cheers /F




From footech at get2net.dk  Fri Feb 23 13:52:58 2001
From: footech at get2net.dk (Mikkel Rasmussen)
Date: Fri, 23 Feb 2001 13:52:58 +0100
Subject: Newbie: Large dictionaries
References: <WUpl6.43$Gv2.2355@news.get2net.dk> <mailman.982924567.27809.python-list@python.org> <vpsl6.134$Gv2.4218@news.get2net.dk> <7Lsl6.15130$Qb7.2431128@newsb.telia.net>
Message-ID: <B2tl6.144$Gv2.4637@news.get2net.dk>

>
> >>> count = 0
> >>> for elem in dict:
> ...     count = count + 1
> ...
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> TypeError: loop over non-sequence
>
> are you sure you're using Python?
>
> > The keys are ordinary words like an ordinary dictionary :-) with an
average
> > length of about 6 characters.
>
> and what language has four million *unique* words
> having six characters and less?
>
> Cheers /F
>
>
Yes, I know you can't loop over dict - dict.keys() is much better :-)

Languages containing spelling errors can contain a lot of variations of each
word. I'm using spelling misstakes as keys. Spelling misstakes also occur in
programming languages. :-)

Mikkel Rasmussen




From mikael at isy.liu.se  Fri Feb 23 14:59:13 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 23 Feb 2001 14:59:13 +0100 (MET)
Subject: Newbie: Large dictionaries
In-Reply-To: <7Lsl6.15130$Qb7.2431128@newsb.telia.net>
Message-ID: <XFMail.010223145913.mikael@isy.liu.se>

On 23-Feb-01 Fredrik Lundh wrote:
 >  and what language has four million *unique* words
 >  having six characters and less?

Well, there are approximately 300 million combinations of 6 letters
from an alphabet of size 26. I wouldn't be surprized if there actually 
are 4 million unique words of length 6 in some language.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    23-Feb-01
Time:    14:44:44

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From max at alcyone.com  Fri Feb 23 17:46:59 2001
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 23 Feb 2001 08:46:59 -0800
Subject: Newbie: Large dictionaries
References: <XFMail.010223145913.mikael@isy.liu.se>
Message-ID: <3A969403.59F2114@alcyone.com>

Mikael Olofsson wrote:

> Well, there are approximately 300 million combinations of 6 letters
> from an alphabet of size 26. I wouldn't be surprized if there actually
> are 4 million unique words of length 6 in some language.

That tells you very little, because only a tiny proportion of those
three hundred million letter combinations represent reasonable
combinations of sounds that are legal in a given language, much less
actual words in that language.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ You and I / We've seen it all / Chasing our hearts' desire
\__/ The Russian and Florence, _Chess_
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From footech at get2net.dk  Sat Feb 24 19:41:56 2001
From: footech at get2net.dk (Mikkel Rasmussen)
Date: Sat, 24 Feb 2001 19:41:56 +0100
Subject: Newbie: Large dictionaries (OT)
References: <XFMail.010223145913.mikael@isy.liu.se> <3A969403.59F2114@alcyone.com>
Message-ID: <MI6m6.85$Bi4.4881@news.get2net.dk>

An example of non-use of combinations:
In Danish there are about 10.000 single syllable words (not word forms) out
of about 50.000 valid combinations. A single syllable word can be anything
from one letter to seven letters.

Question 1: How many percent of the possible combinations are are used?
Question 2: How many percent of the possible combinations are pronounciable?


Mikkel Rasmussen
www.mellon.dk


Erik Max Francis <max at alcyone.com> wrote in message
news:3A969403.59F2114 at alcyone.com...
> Mikael Olofsson wrote:
>
> > Well, there are approximately 300 million combinations of 6 letters
> > from an alphabet of size 26. I wouldn't be surprized if there actually
> > are 4 million unique words of length 6 in some language.
>
> That tells you very little, because only a tiny proportion of those
> three hundred million letter combinations represent reasonable
> combinations of sounds that are legal in a given language, much less
> actual words in that language.
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ You and I / We've seen it all / Chasing our hearts' desire
> \__/ The Russian and Florence, _Chess_
>     Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
>  An Esperanto reference for English speakers.




From timr at probo.com  Mon Feb 26 03:59:50 2001
From: timr at probo.com (Tim Roberts)
Date: Sun, 25 Feb 2001 18:59:50 -0800
Subject: Newbie: Large dictionaries
References: <XFMail.010223145913.mikael@isy.liu.se> <3A969403.59F2114@alcyone.com>
Message-ID: <5ihj9tsbtmcjp9rqmlnkevohu2kjkml9u3@4ax.com>

Erik Max Francis <max at alcyone.com> wrote:
>
>That tells you very little, because only a tiny proportion of those
>three hundred million letter combinations represent reasonable
>combinations of sounds that are legal in a given language, much less
>actual words in that language.

I don't know what you're qquxxj zxooxx yfdtty rrrrpo mnlkaaing about.

Actually, I'll bet a substantial fraction of those 300,000,000 sequences
now exist as the middle token in the sequence www.??????.com.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From footech at get2net.dk  Fri Feb 23 13:46:00 2001
From: footech at get2net.dk (Mikkel Rasmussen)
Date: Fri, 23 Feb 2001 13:46:00 +0100
Subject: Newbie: Large dictionaries
References: <WUpl6.43$Gv2.2355@news.get2net.dk> <mailman.982924567.27809.python-list@python.org> <vpsl6.134$Gv2.4218@news.get2net.dk>
Message-ID: <S_sl6.143$Gv2.4564@news.get2net.dk>

I found the error, that is: my error.

I tried building a dictionary of 65.531 words held in a list. That worked
fine. Then I wanted to test Python by building a *large* dictionary of four
million words. I obtained my test data by concatenating the list with
itself. I poured the four million words *into* the dictionary, but only got
65.531 words out - because there was only as many different words and each
word was its own key.

Always beware of your test data :-)

Now I tested again with better data and got a fine working dictionary. It
only took 90 seconds to delete it after the test :-)

Thanks for your answers!


Mikkel Rasmussen


Mikkel Rasmussen <footech at get2net.dk> wrote in message
news:vpsl6.134$Gv2.4218 at news.get2net.dk...
> Hi!
>
> I can also put more than 4 million entries *into* the dictionary. But I
can
> only get less than 65.000 back out!
>
> What do you get with: len(dict.keys()) ?
>
> And with a loop like
>
> count = 0
> for elem in dict:
>     count = count + 1
> print count
>
> The keys are ordinary words like an ordinary dictionary :-) with an
average
> length of about 6 characters.
>
>
> Mikkel Rasmussen
>
>
> Steve Purcell <stephen_purcell at yahoo.com> wrote in message
> news:mailman.982924567.27809.python-list at python.org...
> > Mikkel Rasmussen wrote:
> > > Why can't there be more than about 65.000 entries in a dictionary, or
> more
> > > importantly: how can I make a large dictionary with room for at least
> > > 500.000
> > > elements?
> >
> > It's probably a memory issue: how big are the keys and/or values?
> >
> > On my machine, I can easily get a million entries into a dictionary:-
> >
> > >>> dict = {}
> > >>> try:
> > ...   for i in xrange(1000000):
> > ...     dict[i] = None
> > ... except:
> > ...   print "stopped at", i
> > ...   raise
> > ...
> > >>>
> >
> > -Steve
> >
> > --
> > Steve Purcell, Pythangelist
> > Get testing at http://pyunit.sourceforge.net/
> > Get servlets at http://pyserv.sourceforge.net/
> > "Even snakes are afraid of snakes." -- Steven Wright
> >
>
>




From bencr at bigpond.com  Fri Feb 23 13:01:37 2001
From: bencr at bigpond.com (Ben)
Date: Fri, 23 Feb 2001 23:01:37 +1100
Subject: Suffering For Your Art
Message-ID: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>

# Just A Quick Observation

Is it just me or is there a lot of animosity towards Python as a serious
programming language ... it seems to be an old school attitude that  if you
do not have to struggle and suffer with an idiosyncratic, unintuitive
programming syntax then it cannot possibly be a powerful language. All these
ascetics who would like to see everyone go back to the 'halcyon days' of
assembly ... IMHO I think Python is a marvellously eloquant and beautiful
language because it does not get in the road of programming. It pisses me
off if I have to consult a tome of a help file/manual to find some obscure
reference to do something that should have been quite simple ...

# Would Anyone Like To Comment?




From ws-news at gmx.at  Fri Feb 23 13:08:09 2001
From: ws-news at gmx.at (Werner Schiendl)
Date: Fri, 23 Feb 2001 13:08:09 +0100
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <982930354.223466@newsmaster-04.atnet.at>

Hi,

I think you will not find too many people in this NG complainig about
python's syntax.
You should probably try that one at comp.lang.c or comp.lang.c++
But don't mention "C/C++", or you'll be blamed an idiot ;-)

hth
werner

Ben <bencr at bigpond.com> wrote in message
news:n2rl6.1686$v5.5388 at newsfeeds.bigpond.com...
> # Just A Quick Observation
>
> Is it just me or is there a lot of animosity towards Python as a serious
> programming language ... it seems to be an old school attitude that  if
you
> do not have to struggle and suffer with an idiosyncratic, unintuitive
> programming syntax then it cannot possibly be a powerful language. All
these
> ascetics who would like to see everyone go back to the 'halcyon days' of
> assembly ... IMHO I think Python is a marvellously eloquant and beautiful
> language because it does not get in the road of programming. It pisses me
> off if I have to consult a tome of a help file/manual to find some obscure
> reference to do something that should have been quite simple ...
>
> # Would Anyone Like To Comment?
>
>




From bowman at montana.com  Fri Feb 23 14:16:30 2001
From: bowman at montana.com (bowman)
Date: Fri, 23 Feb 2001 06:16:30 -0700
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <vtul6.275$yU2.1291@newsfeed.slurp.net>

Ben <bencr at bigpond.com> wrote in message
news:n2rl6.1686$v5.5388 at newsfeeds.bigpond.com...
>
>  it seems to be an old school attitude that  if you
> do not have to struggle and suffer with an idiosyncratic, unintuitive
> programming syntax then it cannot possibly be a powerful language.

well known psychological quirk of the species. after suffering through X to
achieve goal Y, the experimental subject will value Y over Z. You can find
it everywhere, from Marines to fraternity members. In the amateur radio
community, morse code proficiency was stressed even after the rest of the
world abandoned it -- kept the riffraff off the air, y'know.





From danielk at aracnet.com  Fri Feb 23 14:49:11 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Fri, 23 Feb 2001 05:49:11 -0800
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <hfqc9tg00gfb1fnic7dor8gfd04qa14p5j@4ax.com>

On Fri, 23 Feb 2001 23:01:37 +1100, "Ben" <bencr at bigpond.com> wrote:

>Is it just me or is there a lot of animosity towards Python as a serious
>programming language ...
[snip]
># Would Anyone Like To Comment?

Just cos they don't understand us doesn't make us confused.

Dan



From stephen_purcell at yahoo.com  Fri Feb 23 15:15:50 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 23 Feb 2001 15:15:50 +0100
Subject: Suffering For Your Art
In-Reply-To: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>; from bencr@bigpond.com on Fri, Feb 23, 2001 at 11:01:37PM +1100
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <20010223151550.A3228@freedom.puma-ag.com>

Ben wrote:
> Is it just me or is there a lot of animosity towards Python as a serious
> programming language ... it seems to be an old school attitude that  if you
> do not have to struggle and suffer with an idiosyncratic, unintuitive
> programming syntax then it cannot possibly be a powerful language.

Many of those on this list have been through the wringer of those
idiosyncratic languages and have attained Pythonic enlightenment.

We should be patient with those that have yet to transcend the suffering
of \$, -> and BEGIN/END.

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From ljohnson at resgen.com  Fri Feb 23 15:23:04 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Fri, 23 Feb 2001 08:23:04 -0600
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <t9cse0s7mra90f@corp.supernews.com>

> Is it just me or is there a lot of animosity towards Python as a serious
> programming language ... it seems to be an old school attitude that  if
you
> do not have to struggle and suffer with an idiosyncratic, unintuitive
> programming syntax then it cannot possibly be a powerful language.

When I first started looking at Python a few years back I had to get past
the mental block that Python was a "scripting" language and was therefore
inherently less powerful than the more traditional "compiled" languages
(e.g. C/C++ or Fortran). So while it was OK for small or even moderately
sized programming tasks, there's no way you'd write a full-blown application
in Python. I believe that's still a big misconception for a lot of people.




From kevin at GamesandTheory.com  Fri Feb 23 16:05:49 2001
From: kevin at GamesandTheory.com (Kevin Douglas)
Date: Fri, 23 Feb 2001 10:05:49 -0500
Subject: Suffering For Your Art
In-Reply-To: <t9cse0s7mra90f@corp.supernews.com>
Message-ID: <DKEEJHGMIIOMNIJFOCAIEEFKDGAA.kevin@GamesandTheory.com>

> > Is it just me or is there a lot of animosity towards Python as
> > a serious programming language ...
>
> When I first started looking at Python a few years back I had to
> get past the mental block that Python was a "scripting" language
> and was therefore inherently less powerful than the more
> traditional "compiled" languages (e.g. C/C++ or Fortran).

This is the first time I have ever posted to this list, in over
a year of being subscribed.

I've built and/or worked on systems that are scalable to the level 
of billions of transactions per month with Python and C and ASM.

The premise is simple - Python is an AUTOMATION and PROTOTYPING
language.  You use it to automate either administrative tasks,
management tasks, monitoring and quality assurance tasks,
data aggregation or analysis tasks, or - more powerfully -
programming tasks (self replicating code, code management, etc).

It's quite easy to build an advertising exchange and collaborative
filtering system such as DoubleClick has (even better than theirs,
actually), or BlueFly or any other such company - in about one
to three months - then work for the next 6 to 12 months on filling
in the details, and native C functionality and inline ASM where
needed.  Such a system usually lasts for 3 to 4 years without
needing much change (from industry experience, including one of
the examples I just mentioned *cough cough*).

I've spoken with executives and venture capitalist that are
responsible for several million dollars worth of technology
expenditures - who re completely confident in the capacity
of Python for Enterprise Level developments of any size - ERP,
HIRS, CRM, Financial Systems, Accounting Systems, Billing
Systems, Trading Systems, data Mining/Warehousing Systems, you
name it.

I was about to build a Python system for Deutsche Bank, the
world's largest bank, with North American, European, and
Asia-Pacific operations, multilingual, HRIS, graphical charts
and analytics, database, security systems, the whole nine
(yes, thin-client Java Swing graphics with distributed load
balancing on the client end, plus ORACLE and encrypted
one-way anonymous data repository for survey data and aggregate
analysis).

Of course, we would never tell the general public that it was
written in Python - that's silly.  Of course, we would never
tell the general public if we were running Java either -
because that is equally as silly.  We run "heterogeneous
systems" - and that's all anyone needs to know - for Risk
Management purposes.

Unfortunately, people generally tend to suck.  So that project
never completed because someone unnamed wasn't willing to wake up
everyday and be their own boss.

And Python coders like myself are starving artists because of
such silly and trivial things.  Mostly because of "grown men"
being considerably more immature than most of my peers.

On an interesting side note, I've noticed that more money seems
to be spent on Welfare Administration than on the actual welfare
itself, given some of the rules and regulations that these
agencies claim are legal and strictly adhere to.  I was turned
down for Food Stamps even though I only have water left right
now - it's interesting how a CTO level programmer gets no respect
because he focused on cost effective realistic solutions instead
of Multi-Level Sales and Marketing tactics like the ones that
destroyed the US Economy.

EMC just announced they'd make $1B less than they thought -
because - yes - all those con artists who were selling "solutions"
based more on merchant partnerships than real solutions - ripped
everyone off and no one is left!

My 2 cents worth.  If I had 2 cents anyway.

Always available for consulting or freelance work or start-up
prototyping, resume available if anyone really cares.


Kevin Douglas
President / Chief Technology Officer
Games and Theory Supercomputing, Inc.






From dan at eevolved.com  Fri Feb 23 17:51:52 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 23 Feb 2001 16:51:52 GMT
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com> <t9cse0s7mra90f@corp.supernews.com>
Message-ID: <Iywl6.303914$JT5.10340026@news20.bellglobal.com>

Lyle Johnson wrote:

> When I first started looking at Python a few years back I had to get past
> the mental block that Python was a "scripting" language and was therefore
> inherently less powerful than the more traditional "compiled" languages
> (e.g. C/C++ or Fortran). So while it was OK for small or even moderately
> sized programming tasks, there's no way you'd write a full-blown
> application in Python. 

I am. A rather large project as well that hits consumers as well as 
businesses. Python has been the breath of fresh air I've always wished 
programming could be. Because of Python, we are going to be able to produce 
some incredible applications using
-Python Sockets and cPickle
-Shelve (Robin Dunn's bsddb3.dbshelve which is excellent btw)
-Dictionaries whose keys are arbitrary methods/functions as dynamic message 
handlers
- FXPy (which you coded :) which comes to a 1.5MB full featured, cross 
platform GUI toolkit
- RExec

We learned about Python in January, switched our entire project from c++ 
(losing a large amount of code), and we hope to have a beta out relatively 
soon (weeks instead of months). I always tell my programmer friends, to 
learn of python is to learn python.

Dan


From aahz at panix.com  Fri Feb 23 17:07:23 2001
From: aahz at panix.com (Aahz Maruch)
Date: 23 Feb 2001 08:07:23 -0800
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <9761rr$gal$1@panix2.panix.com>

In article <n2rl6.1686$v5.5388 at newsfeeds.bigpond.com>,
Ben <bencr at bigpond.com> wrote:
>
>Is it just me or is there a lot of animosity towards Python as a serious
>programming language ... 

Where I see such animosity, it's almost always also directed toward Perl
and Java.  The only people who have animosity toward Python in specific
are those for whom writing braces gives them a woodie.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From max at alcyone.com  Fri Feb 23 17:48:23 2001
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 23 Feb 2001 08:48:23 -0800
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <3A969457.44BC94C5@alcyone.com>

Ben wrote:

> Is it just me or is there a lot of animosity towards Python as a
> serious
> programming language ... it seems to be an old school attitude that 
> if you
> do not have to struggle and suffer with an idiosyncratic, unintuitive
> programming syntax then it cannot possibly be a powerful language.

Some people don't want to use a new language no matter what,
particularly if their repertoire is limited and they're inexperienced. 
You'll find that experienced programmers know a lot of languages, and
adding another one to their belt is a welcomed experience rather than a
feared one.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Winners are men who have dedicated their whole lives to winning.
\__/ Woody Hayes
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From mkx at excite.com  Fri Feb 23 19:42:00 2001
From: mkx at excite.com (mkx at excite.com)
Date: Fri, 23 Feb 2001 13:42:00 -0500
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <a8bd9tkvianmkqtg5c3qg33i8igq47n6kj@4ax.com>

On Fri, 23 Feb 2001 23:01:37 +1100, "Ben" <bencr at bigpond.com> wrote:

># Just A Quick Observation
>
>Is it just me or is there a lot of animosity towards Python as a serious
>programming language ...

Only among those who believe that the way they learned to do things,
is the one and only way that things should be done, and the only way
things should ever be done.

Many projects suffer from a lack of progress because of the belief by
many that "the way it has been done is the only way it will be done."

>it seems to be an old school attitude

That some consider a clinical condition.

>that if you do not have to struggle and suffer with an idiosyncratic, unintuitive
>programming syntax then it cannot possibly be a powerful language.

"In my day we walked to school, barefoot, up-hill, in the snow, BOTH
WAYS!"

>All these ascetics who would like to see everyone go back to the 'halcyon days' of
>assembly ...

"Real programmers write in binary, but only on punch cards."


>IMHO I think Python is a marvellously eloquant and beautiful
>language because it does not get in the road of programming. It pisses me
>off if I have to consult a tome of a help file/manual to find some obscure
>reference to do something that should have been quite simple ...

Other environments allow the true computer geeks to flex their brain
muscles, in an attempt to impress the intellectually-poor masses.




From aleaxit at yahoo.com  Sat Feb 24 17:37:34 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 24 Feb 2001 17:37:34 +0100
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com> <a8bd9tkvianmkqtg5c3qg33i8igq47n6kj@4ax.com>
Message-ID: <978oih02b74@news1.newsguy.com>

<mkx at excite.com> wrote in message
news:a8bd9tkvianmkqtg5c3qg33i8igq47n6kj at 4ax.com...
    [snip]
> >IMHO I think Python is a marvellously eloquant and beautiful
> >language because it does not get in the road of programming. It pisses me
> >off if I have to consult a tome of a help file/manual to find some
obscure
> >reference to do something that should have been quite simple ...
>
> Other environments allow the true computer geeks to flex their brain
> muscles, in an attempt to impress the intellectually-poor masses.

So does Python (I'm guilty of my share of "clever" solutions posted
to this group), but in Python's case it's clearer that "being clever" is
a temptation to be *wisely resisted* -- that "the simplest thing that
can possibly work" is the RIGHT way to go (and I try to recall to
point this out just about every time I come up with "cleverness"
that would likely be a very bad idea to use in production code...:-).


Alex





From mkx at excite.com  Sun Feb 25 10:02:26 2001
From: mkx at excite.com (mkx at excite.com)
Date: Sun, 25 Feb 2001 04:02:26 -0500
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com> <a8bd9tkvianmkqtg5c3qg33i8igq47n6kj@4ax.com> <978oih02b74@news1.newsguy.com>
Message-ID: <q1ih9t0qktq6ju8kjg7rsp5hk4bivvb66d@4ax.com>

On Sat, 24 Feb 2001 17:37:34 +0100, "Alex Martelli"
<aleaxit at yahoo.com> wrote:

>So does Python (I'm guilty of my share of "clever" solutions posted
>to this group), but in Python's case it's clearer that "being clever" is
>a temptation to be *wisely resisted* -- that "the simplest thing that
>can possibly work" is the RIGHT way to go (and I try to recall to
>point this out just about every time I come up with "cleverness"
>that would likely be a very bad idea to use in production code...:-).

I find that usually, if I have had enough sleep in the past week, I
can at lest _read_ most Python code posted by others with varying CQs
(clever quotients) {often reading is the 1st step in understanding}.
This was the first factor in swaying me to the language. I came from
another c.l.p* newsgroup, where more often than not, I could not tell
if the postings contained code or line-noise. And if I asked for
assistance in deciphering, I got hit with a big FAQ upside the head. 

I think that part of the art in Python is that clever code is most
often an elegant way to accomplish a task. In the past, I had spent
far too much time deciphering code where clever meant "difficult for
humans to comprehend."




From mkx at excite.com  Sun Feb 25 10:10:01 2001
From: mkx at excite.com (mkx at excite.com)
Date: Sun, 25 Feb 2001 04:10:01 -0500
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com>
Message-ID: <goih9tobrpss758tloblb8j2c56upu8r5f@4ax.com>

I only suffer when my coworkers make me work with their VBScript code,
because they don't know Python, and they have a TON of VB stuff built
over the past years.

The only thing worse (for me - no flame baiting of offense intended),
is when we find left over Perl code that the previous guy wrote. Then
we are pained _and_ confused.


On Fri, 23 Feb 2001 23:01:37 +1100, "Ben" <bencr at bigpond.com> wrote:

># Just A Quick Observation
>
>Is it just me or is there a lot of animosity towards Python as a serious
>programming language ... it seems to be an old school attitude that  if you
>do not have to struggle and suffer with an idiosyncratic, unintuitive
>programming syntax then it cannot possibly be a powerful language. All these
>ascetics who would like to see everyone go back to the 'halcyon days' of
>assembly ... IMHO I think Python is a marvellously eloquant and beautiful
>language because it does not get in the road of programming. It pisses me
>off if I have to consult a tome of a help file/manual to find some obscure
>reference to do something that should have been quite simple ...
>
># Would Anyone Like To Comment?
>




From nospam at diveintopython.org  Fri Feb 23 17:23:04 2001
From: nospam at diveintopython.org (nospam at diveintopython.org)
Date: 23 Feb 2001 16:23:04 GMT
Subject: ANN: Dive Into Python chapter 3 released
Message-ID: <9762p8$jig$1@news.netmar.com>

"Dive Into Python" (http://diveintopython.org/) is a free Python tutorial for
experienced programmers.  You can read the book online or download it in a
variety of formats.

Chapter 3 (http://diveintopython.org/fileinfo_divein.html) is finally
complete.  It covers all of these concepts:
- defining, subclassing, and instantiating Python classes
- special class methods
- data attributes and class attributes
- exception handling
- basic file management (opening, reading files)
- writing cross-platform libraries by encapsulating platform differences
(like
the getpass, os modules)
- assigning multiple values at once

If you're one of the 5,000 people who have already downloaded "Dive Into
Python", come back and see what's new:
- All examples are now color-coded.
- Each section now has "further reading" links to relevant sections of other
free resources, including the Python library reference, Python language
reference, Python tutorial, Python Knowledge Base, and other free Python
books available online.
- Chapter 2 now uses the getattr function instead of eval -- a much more
elegant solution suggested by an articulate reader.
(http://diveintopython.org/apihelper_getattr.html)
- Text edited, confusions clarified, bugs fixed. :)

-M
--
Mark Pilgrim, f8dy at diveintopython.org
You're smart; why haven't you learned Python yet?  http://diveintopython.org/


 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse at newsone.net


From aahz at panix.com  Fri Feb 23 17:54:03 2001
From: aahz at panix.com (Aahz Maruch)
Date: 23 Feb 2001 08:54:03 -0800
Subject: ANN: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com>
Message-ID: <9764jb$nt0$1@panix2.panix.com>

In article <9762p8$jig$1 at news.netmar.com>,  <nospam at diveintopython.org> wrote:
>
>"Dive Into Python" (http://diveintopython.org/) is a free Python tutorial for
>experienced programmers.  You can read the book online or download it in a
>variety of formats.

Nice!  Good work!
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From ecastro at cicei.ulpgc.es  Fri Feb 23 19:40:29 2001
From: ecastro at cicei.ulpgc.es (Enrique)
Date: 23 Feb 2001 17:40:29 -0100
Subject: ANN: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com>
Message-ID: <3A969F67.D5FA9EC4@cicei.ulpgc.es>

hi,
    I have downloaded all *.zip files (in any format), but no one unzip correctly.
They die due to a CRC error
    Where is the error?

nospam at diveintopython.org wrote:

> "Dive Into Python" (http://diveintopython.org/) is a free Python tutorial for
> experienced programmers.  You can read the book online or download it in a
> variety of formats.
>



From wayne.izatt at _nospam_myself.com  Sat Feb 24 07:31:17 2001
From: wayne.izatt at _nospam_myself.com (Wayne Izatt)
Date: Sat, 24 Feb 2001 06:31:17 GMT
Subject: ANN: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com> <3A969F67.D5FA9EC4@cicei.ulpgc.es>
Message-ID: <VyIl6.170258$df5.5070181@news1.crdva1.bc.home.com>

Sounds like you're downloading in ascii format (I'm not sure what ftp client
you're using). Specifying binary might help.

cheers

"Enrique" <ecastro at cicei.ulpgc.es> wrote in message
news:3A969F67.D5FA9EC4 at cicei.ulpgc.es...
> hi,
>     I have downloaded all *.zip files (in any format), but no one unzip
correctly.
> They die due to a CRC error
>     Where is the error?
>
> nospam at diveintopython.org wrote:
>
> > "Dive Into Python" (http://diveintopython.org/) is a free Python
tutorial for
> > experienced programmers.  You can read the book online or download it in
a
> > variety of formats.
> >
>




From rtrocca at libero.it  Sun Feb 25 16:38:55 2001
From: rtrocca at libero.it (Riccardo Trocca)
Date: Sun, 25 Feb 2001 15:38:55 GMT
Subject: ANN: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com> <3A969F67.D5FA9EC4@cicei.ulpgc.es> <VyIl6.170258$df5.5070181@news1.crdva1.bc.home.com>
Message-ID: <1epclmi.xa1xz81ny2o0sN%rtrocca@libero.it>

I had the same problem with my macintosh.
I dled the zip files and my mac couldn't expand them.
But as soon as I expanded the same files (no further dl) under Linux
everything worked.
any hint?

Riccardo

Wayne Izatt <wayne.izatt at _nospam_myself.com> wrote:

> Sounds like you're downloading in ascii format (I'm not sure what ftp client
> you're using). Specifying binary might help.
> 
> cheers
> 
> "Enrique" <ecastro at cicei.ulpgc.es> wrote in message
> news:3A969F67.D5FA9EC4 at cicei.ulpgc.es...
> > hi,
> >     I have downloaded all *.zip files (in any format), but no one unzip
> correctly.
> > They die due to a CRC error
> >     Where is the error?
> >
> > nospam at diveintopython.org wrote:
> >
> > > "Dive Into Python" (http://diveintopython.org/) is a free Python
> tutorial for
> > > experienced programmers.  You can read the book online or download it in
> a
> > > variety of formats.
> > >
> >


From francois.granger at free.fr  Sun Feb 25 23:46:54 2001
From: francois.granger at free.fr (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Sun, 25 Feb 2001 22:46:54 GMT
Subject: ANN: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com> <3A969F67.D5FA9EC4@cicei.ulpgc.es> <VyIl6.170258$df5.5070181@news1.crdva1.bc.home.com> <1epclmi.xa1xz81ny2o0sN%rtrocca@libero.it>
Message-ID: <1epedtk.1qhe3edokehmzN%francois.granger@free.fr>

Riccardo Trocca <rtrocca at libero.it> wrote:

> I had the same problem with my macintosh.
> I dled the zip files and my mac couldn't expand them.
> But as soon as I expanded the same files (no further dl) under Linux
> everything worked.

I downloaded the Unix files on my Mac, work like a charme.

-- 
La realit?,  y compris la realit? ?comonique,
est toujours complexe. - Noelle Adam


From f8dy at yahoo.com  Mon Feb 26 15:34:43 2001
From: f8dy at yahoo.com (Mark Pilgrim)
Date: Mon, 26 Feb 2001 09:34:43 -0500
Subject: ANN: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com> <3A969F67.D5FA9EC4@cicei.ulpgc.es> <VyIl6.170258$df5.5070181@news1.crdva1.bc.home.com> <1epclmi.xa1xz81ny2o0sN%rtrocca@libero.it>
Message-ID: <97dpf2$orked$1@ID-77331.news.dfncis.de>

I tested the Mac files on my iMac before releasing them, and they appeared
to work.  Perhaps they require Stuffit Expander 5?

This is not really a Python-related question, but what's the general
consensus on providing platform-specific archive formats?  I know there is
unzip on most (all?) UN*X platforms, WinZip can handle tar/gzipped files,
and Mac Stuffit Expander (at least version 5) can handle pretty much
whatever you throw at it.  I initially felt that it was friendlier to
provide separate archives for each platform, but perhaps it's more trouble
than it's worth?

-M
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


"Riccardo Trocca" <rtrocca at libero.it> wrote in message
news:1epclmi.xa1xz81ny2o0sN%rtrocca at libero.it...
> I had the same problem with my macintosh.
> I dled the zip files and my mac couldn't expand them.
> But as soon as I expanded the same files (no further dl) under Linux
> everything worked.
> any hint?
>
> Riccardo
>
> Wayne Izatt <wayne.izatt at _nospam_myself.com> wrote:
>
> > Sounds like you're downloading in ascii format (I'm not sure what ftp
client
> > you're using). Specifying binary might help.
> >
> > cheers
> >
> > "Enrique" <ecastro at cicei.ulpgc.es> wrote in message
> > news:3A969F67.D5FA9EC4 at cicei.ulpgc.es...
> > > hi,
> > >     I have downloaded all *.zip files (in any format), but no one
unzip
> > correctly.
> > > They die due to a CRC error
> > >     Where is the error?
> > >
> > > nospam at diveintopython.org wrote:
> > >
> > > > "Dive Into Python" (http://diveintopython.org/) is a free Python
> > tutorial for
> > > > experienced programmers.  You can read the book online or download
it in
> > a
> > > > variety of formats.
> > > >
> > >




From glenfant at equod.com.nospam  Sat Feb 24 17:09:55 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Sat, 24 Feb 2001 17:09:55 +0100
Subject: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com>
Message-ID: <978m16$fvi$1@reader1.imaginet.fr>

Good job...

Don't need to be an "experienced programmer" to read this tutorial.
You're an experienced programmer after reading this book.

<nospam at diveintopython.org> a ?crit dans le message news:
9762p8$jig$1 at news.netmar.com...
> "Dive Into Python" (http://diveintopython.org/) is a free Python tutorial
for
> experienced programmers.  You can read the book online or download it in a
> variety of formats.




From sheila at spamcop.net  Mon Feb 26 04:58:16 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 26 Feb 2001 03:58:16 GMT
Subject: ANN: Dive Into Python chapter 3 released
References: <9762p8$jig$1@news.netmar.com>
Message-ID: <1ekj9t0lconntup78pqfs127ksv74b0arj@4ax.com>

On 23 Feb 2001 16:23:04 GMT, nospam at diveintopython.org wrote in
comp.lang.python in article <9762p8$jig$1 at news.netmar.com>:

:"Dive Into Python" (http://diveintopython.org/) is a free Python tutorial for
:experienced programmers.  You can read the book online or download it in a
:variety of formats.

I know this resource has been mentioned before, but I only just looked at it
this week, after this announcement was posted. I'm only working on Chapter 1
so far, and making slow progress due to lack of time to devote to Python right
now, but I want to say that:

I find this approach fantastic! This is just what I needed. It DOES just dive
right in, and I like it, and the explanations. I especially liked the
cr0ss-referenced web links in each section. Nice job!


(I do have a couple of Python books on order: Core Python Programming and
Quick Python...I would've ordered from O'Reilly, but the new books aren't
available yet, so I'm waiting...)

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From ransen_spam_me_not at nemo.it  Fri Feb 23 17:32:40 2001
From: ransen_spam_me_not at nemo.it (Owen F. Ransen)
Date: Fri, 23 Feb 2001 16:32:40 GMT
Subject: PyImport_AddModule vs PyImport_Import
Message-ID: <3a966ad6.4798499@news.newsguy.com>

What is the difference between
   
   PyImport_AddModule 
   PyImport_Import

?  

What does AddModule add the module to exactly?


--
Owen F. Ransen
http://www.ransen.com/
Home of Gliftic & Repligator Image Generators


From owen at astrono.spamwashington.emu  Fri Feb 23 17:41:26 2001
From: owen at astrono.spamwashington.emu (Russell E. Owen)
Date: Fri, 23 Feb 2001 08:41:26 -0800
Subject: package question: getting names one level deep?
Message-ID: <owen-2302010841550001@cs205-86.fsmodem.washington.edu>

I like to break my code up into lots of files -- typically one file contains a class and turn a related set of files/classes into a package. When a file contains a class, naturally I give the file the same name as the class.

However, I've not found a nice way to refer to the resulting classes. I hope I'm missing some easy, standard trick.

Say I have a directory structure such as:
apackage/
  __init__.py  which contains __all__ = [list of all files]
  classa.py
  classb.py
  ...

To use these classes I write:

import apackage
x = package.classa.classa(etc)

It's that duplicated file-name.identical-class-name that I'd like to eliminate, so I could say:

import apackage (or something)
x = apackage.classa(etc)

Basically I'd like to do an import * into the "apackage" namespace, and ideally I'd like "import apackage" to somehow magically do it. In other words, ideally I'd like to set up my packages so they act as if all their code (at least all intended-to-be-visible-to-the-outside-world code) was in a single module.

Any hints?

-- Russell


From stephen_purcell at yahoo.com  Fri Feb 23 19:01:45 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 23 Feb 2001 19:01:45 +0100
Subject: package question: getting names one level deep?
In-Reply-To: <owen-2302010841550001@cs205-86.fsmodem.washington.edu>; from owen@astrono.spamwashington.emu on Fri, Feb 23, 2001 at 08:41:26AM -0800
References: <owen-2302010841550001@cs205-86.fsmodem.washington.edu>
Message-ID: <20010223190145.D4461@freedom.puma-ag.com>

Russell E. Owen wrote:
> I like to break my code up into lots of files -- typically one file
> contains a class and turn a related set of files/classes into a package.
> When a file contains a class, naturally I give the file the same name as
> the class.
> 
> However, I've not found a nice way to refer to the resulting classes.
> I hope I'm missing some easy, standard trick.

Ah, well, the easy, standard trick is to group your classes in modules, not
put each class in a submodule of it's 'real' module.

All you're currently doing is breaking one module up into a number of
little files for no concrete reason (c.f. PIL's plugin scheme).

It's a C++/Java idiom that doesn't map too well to Python, without the
application of the voodoo invocations you're soliciting. :-)

Rule #2343: "Be suspicious when something seems complicated in Python."

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From robin at jessikat.fsnet.co.uk  Fri Feb 23 21:00:53 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Fri, 23 Feb 2001 20:00:53 +0000
Subject: package question: getting names one level deep?
References: <owen-2302010841550001@cs205-86.fsmodem.washington.edu>
Message-ID: <XIFkFJA1Fsl6EwaX@jessikat.fsnet.co.uk>

In article <owen-2302010841550001 at cs205-86.fsmodem.washington.edu>,
Russell E. Owen <owen at astrono.spamwashington.emu> writes

one way is to put code into the __init__.py

eg

from classa import classa
from classb import classb

then classa, classb become visible as attributes of the package and you
can do
>>> from apackage import classa
>>> classa
<class apackage.classa.classa at 014C5FCC>

>I like to break my code up into lots of files -- typically one file contains a 
>class and turn a related set of files/classes into a package. When a file 
>contains a class, naturally I give the file the same name as the class.
>
>However, I've not found a nice way to refer to the resulting classes. I hope I'm 
>missing some easy, standard trick.
>
>Say I have a directory structure such as:
>apackage/
>  __init__.py  which contains __all__ = [list of all files]
>  classa.py
>  classb.py
>  ...
>
>To use these classes I write:
>
>import apackage
>x = package.classa.classa(etc)
>
>It's that duplicated file-name.identical-class-name that I'd like to eliminate, 
>so I could say:
>
>import apackage (or something)
>x = apackage.classa(etc)
>
>Basically I'd like to do an import * into the "apackage" namespace, and ideally 
>I'd like "import apackage" to somehow magically do it. In other words, ideally 
>I'd like to set up my packages so they act as if all their code (at least all 
>intended-to-be-visible-to-the-outside-world code) was in a single module.
>
>Any hints?
>
>-- Russell

-- 
Robin Becker


From max at alcyone.com  Fri Feb 23 17:49:45 2001
From: max at alcyone.com (Erik Max Francis)
Date: Fri, 23 Feb 2001 08:49:45 -0800
Subject: Many Threads + Many Sockets == SegFault?
References: <Mgul6.18735$Sl.810976@iad-read.news.verio.net>
Message-ID: <3A9694A9.FC38B944@alcyone.com>

Doug Fort wrote:

> I've attached a little program that I'm testing.  It basically creates
> a
> number of threads (I've been using 100), each of which performs a
> simple
> HTTP 'GET' over and over.  This works fine when I run through our T1
> line
> against a server out in the cloud.

Are you aware that what you're doing is a denial-of-service attack?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Wars don't last forever, Henry.  Only war does.
\__/ Capt. Benjamin "Hawkeye" Pierce
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From stephen_purcell at yahoo.com  Sat Feb 24 11:00:28 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sat, 24 Feb 2001 11:00:28 +0100
Subject: Many Threads + Many Sockets == SegFault?
In-Reply-To: <3A9694A9.FC38B944@alcyone.com>; from max@alcyone.com on Fri, Feb 23, 2001 at 08:49:45AM -0800
References: <Mgul6.18735$Sl.810976@iad-read.news.verio.net> <3A9694A9.FC38B944@alcyone.com>
Message-ID: <20010224110028.B6290@freedom.puma-ag.com>

Erik Max Francis wrote:
> Doug Fort wrote:
> 
> > I've attached a little program that I'm testing.  It basically creates
> > a number of threads (I've been using 100), each of which performs a
> > simple HTTP 'GET' over and over.  This works fine when I run through our T1
> > line against a server out in the cloud.
> 
> Are you aware that what you're doing is a denial-of-service attack?

He is aware of it, but people pay him to do it:

   http://stressmy.com/


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From max at alcyone.com  Sat Feb 24 21:39:13 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 24 Feb 2001 12:39:13 -0800
Subject: Many Threads + Many Sockets == SegFault?
References: <Mgul6.18735$Sl.810976@iad-read.news.verio.net> <3A9694A9.FC38B944@alcyone.com> <mailman.983008806.2343.python-list@python.org>
Message-ID: <3A981BF1.4F0AF09C@alcyone.com>

Steve Purcell wrote:

> He is aware of it, but people pay him to do it:
> 
>    http://stressmy.com/

People pay him to do it, but he can't get a program he wrote to do it
working?  That's not a good sign.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I go out with actresses because I'm not apt to marry one.
\__/ Henry Kissinger
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From erno-news at erno.iki.fi  Sat Feb 24 22:36:44 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 24 Feb 2001 23:36:44 +0200
Subject: Many Threads + Many Sockets == SegFault?
References: <Mgul6.18735$Sl.810976@iad-read.news.verio.net> <3A9694A9.FC38B944@alcyone.com> <mailman.983008806.2343.python-list@python.org> <3A981BF1.4F0AF09C@alcyone.com>
Message-ID: <ku3dd3sqz7.fsf@lasipalatsi.fi>

In article <3A981BF1.4F0AF09C at alcyone.com>, Erik Max Francis
<max at alcyone.com> writes:

| Steve Purcell wrote:
|| He is aware of it, but people pay him to do it:
|| 
|| http://stressmy.com/

| People pay him to do it, but he can't get a program he wrote to do it
| working?  That's not a good sign.

he has hit a bug in python or in something else - otherwise
python would not segfault. it is indeed not a good sign, but
not a problem in the original poster's code.

  -- erno


From jkraska1 at san.rr.com  Sun Feb 25 20:48:29 2001
From: jkraska1 at san.rr.com (Courageous)
Date: Sun, 25 Feb 2001 19:48:29 GMT
Subject: Many Threads + Many Sockets == SegFault?
References: <Mgul6.18735$Sl.810976@iad-read.news.verio.net> <3A9694A9.FC38B944@alcyone.com> <mailman.983008806.2343.python-list@python.org> <3A981BF1.4F0AF09C@alcyone.com> <ku3dd3sqz7.fsf@lasipalatsi.fi>
Message-ID: <iloi9t8me12i595uu260t8m37e8u8jfrc8@4ax.com>

>he has hit a bug in python or in something else - otherwise
>python would not segfault. it is indeed not a good sign, but
>not a problem in the original poster's code.

This is sometimes caused by poorly-written Python extensions;
it's quite easy to cause Python to crash if you screw up your
reference counting, for example. Is he using any kind of dubious
extension modules?



C//



From dougfort at downright.com  Sat Feb 24 22:48:21 2001
From: dougfort at downright.com (Doug Fort)
Date: Sat, 24 Feb 2001 16:48:21 -0500
Subject: Many Threads + Many Sockets == SegFault?
References: <Mgul6.18735$Sl.810976@iad-read.news.verio.net> <3A9694A9.FC38B944@alcyone.com> <mailman.983008806.2343.python-list@python.org> <3A981BF1.4F0AF09C@alcyone.com>
Message-ID: <G_Vl6.18960$Sl.824995@iad-read.news.verio.net>

The code I posted is not the code we use on the site (we use httplib).  It
is a test program to investigate a specific problem that occurs at very high
volumes.

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3A981BF1.4F0AF09C at alcyone.com...
> Steve Purcell wrote:
>
> > He is aware of it, but people pay him to do it:
> >
> >    http://stressmy.com/
>
> People pay him to do it, but he can't get a program he wrote to do it
> working?  That's not a good sign.
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ I go out with actresses because I'm not apt to marry one.
> \__/ Henry Kissinger
>     Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
>  An Esperanto reference for English speakers.




From rich_somerfield at tertio.com  Fri Feb 23 17:55:31 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Fri, 23 Feb 2001 16:55:31 -0000
Subject: Help with finding how much memory a variable is using
Message-ID: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com>

Hi,

I am generating a huge, huge list of data (cant think of a decent way to
reduce the required storage size and have it in a usable form).  Everytime i
try to keep this list for future derivations I get a memory problem from
Windows (effectively terminating my python script).

I presume this is because of the huge list.  Is it possible to find out the
amount of memory a variable [not a type of the variable, the actual data
contained within the variable] is taking up?  This variable is local to a
class and it would appear that it is when i go out of scope of that class
that the memory error is being issued.  Would I be better making this
variable global (across all classes)? e.g. would this problem still happen
if the variable was not just local to 1 class?

Would a different python datatype solve this problem for me ?

If I cant handle this amount of data then I will have to think of a way of
better representing it, this will be unfortunate though as and compression /
reduction of data will [as far as i can think] vastly reduce performance -
which is already a concern.

Regards
Rich





From stephen_purcell at yahoo.com  Fri Feb 23 18:48:09 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 23 Feb 2001 18:48:09 +0100
Subject: Help with finding how much memory a variable is using
In-Reply-To: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com>; from rich_somerfield@tertio.com on Fri, Feb 23, 2001 at 04:55:31PM -0000
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com>
Message-ID: <20010223184809.C4461@freedom.puma-ag.com>

Rich Somerfield wrote:
> I am generating a huge, huge list of data (cant think of a decent way to
> reduce the required storage size and have it in a usable form).  Everytime i
> try to keep this list for future derivations I get a memory problem from
> Windows (effectively terminating my python script).
> 
> I presume this is because of the huge list.  Is it possible to find out the
> amount of memory a variable [not a type of the variable, the actual data
> contained within the variable] is taking up?

A huge, huge in-memory list is very likely to be the cause of such a memory
problem! <wink>

Compression isn't going to help you much if the list is so big that even
a tenth of its size would still be too big.

Post some more technical information and I'm sure you'll get more helpful
suggestions/alternatives than you could have wished for*.

-Steve

* or wanted

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From cjensen at bioeng.ucsd.edu  Sat Feb 24 01:09:52 2001
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Fri, 23 Feb 2001 16:09:52 -0800
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com>
Message-ID: <3A96FBD0.26CD1E62@bioeng.ucsd.edu>

Rich Somerfield wrote:
> 
> Hi,
> 
> I am generating a huge, huge list of data (cant think of a decent way to
> reduce the required storage size and have it in a usable form).  Everytime i
> try to keep this list for future derivations I get a memory problem from
> Windows (effectively terminating my python script).
> 
> I presume this is because of the huge list.  Is it possible to find out the
> amount of memory a variable [not a type of the variable, the actual data
> contained within the variable] is taking up?  This variable is local to a
> class and it would appear that it is when i go out of scope of that class
> that the memory error is being issued.  Would I be better making this
> variable global (across all classes)? e.g. would this problem still happen
> if the variable was not just local to 1 class?
> 
> Would a different python datatype solve this problem for me ?
> 
> If I cant handle this amount of data then I will have to think of a way of
> better representing it, this will be unfortunate though as and compression /
> reduction of data will [as far as i can think] vastly reduce performance -
> which is already a concern.
> 
> Regards
> Rich

If your lists are numeric, you could use the Numeric module. 
Lists/arrays created by Numeric have a member function "itemsize()"
which corresponds to the number of bytes an individual element in that
array is using.  If your lists are not numeric, you might be able to
hack Numeric to give you memory usage (ie. you could use it to find
memory usage of lists of strings; anything more complicated than strings
and it could be difficult).

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451


From s713221 at student.gu.edu.au  Sun Feb 25 02:52:37 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Sun, 25 Feb 2001 11:52:37 +1000
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com>
Message-ID: <3A986565.BAF4E05B@student.gu.edu.au>

Rich Somerfield wrote:
> 
> Hi,
> 
> I am generating a huge, huge list of data (cant think of a decent way to
> reduce the required storage size and have it in a usable form).  Everytime i
> try to keep this list for future derivations I get a memory problem from
> Windows (effectively terminating my python script).
> 
> I presume this is because of the huge list.  Is it possible to find out the
> amount of memory a variable [not a type of the variable, the actual data
> contained within the variable] is taking up?  This variable is local to a
> class and it would appear that it is when i go out of scope of that class
> that the memory error is being issued.  Would I be better making this
> variable global (across all classes)? e.g. would this problem still happen
> if the variable was not just local to 1 class?
> 
> Would a different python datatype solve this problem for me ?
> 
> If I cant handle this amount of data then I will have to think of a way of
> better representing it, this will be unfortunate though as and compression /
> reduction of data will [as far as i can think] vastly reduce performance -
> which is already a concern.
> 
> Regards
> Rich

Do you have to load all the list into memory, or is there a way you can
load in parts and work on that sequentially. I haven't had a chance to
use persistance modules yet, but their documentation may be a place to
start.

Joal Heagney/AncientHart


From rich_somerfield at tertio.com  Mon Feb 26 10:11:59 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Mon, 26 Feb 2001 09:11:59 -0000
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com> <3A986565.BAF4E05B@student.gu.edu.au>
Message-ID: <B5pm6.28048$MN.708724@news2-win.server.ntlworld.com>

<s713221 at student.gu.edu.au> wrote in message
news:3A986565.BAF4E05B at student.gu.edu.au...
> Rich Somerfield wrote:
> >
> > Hi,
> >
> > I am generating a huge, huge list of data (cant think of a decent way to
> > reduce the required storage size and have it in a usable form).
Everytime i
> > try to keep this list for future derivations I get a memory problem from
> > Windows (effectively terminating my python script).
> >
> > I presume this is because of the huge list.  Is it possible to find out
the
> > amount of memory a variable [not a type of the variable, the actual data
> > contained within the variable] is taking up?  This variable is local to
a
> > class and it would appear that it is when i go out of scope of that
class
> > that the memory error is being issued.  Would I be better making this
> > variable global (across all classes)? e.g. would this problem still
happen
> > if the variable was not just local to 1 class?
> >
> > Would a different python datatype solve this problem for me ?
> >
> > If I cant handle this amount of data then I will have to think of a way
of
> > better representing it, this will be unfortunate though as and
compression /
> > reduction of data will [as far as i can think] vastly reduce
performance -
> > which is already a concern.
> >
> > Regards
> > Rich
>
> Do you have to load all the list into memory, or is there a way you can
> load in parts and work on that sequentially. I haven't had a chance to
> use persistance modules yet, but their documentation may be a place to
> start.
>
> Joal Heagney/AncientHart

Unfortunately I need all of the data in memory at the same time.  The only
way I can see to solve this problem is to reduce the memory consumption of
my application.  And the only way I can see how to do that is by vastly
reducing the performance.

Do u know of any information that suggests how large a datatype can be in
Python (are there any upper limits?) ?

Regards
Rich





From rich_somerfield at tertio.com  Mon Feb 26 10:09:46 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Mon, 26 Feb 2001 09:09:46 -0000
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com> <mailman.982950492.15473.python-list@python.org>
Message-ID: <v3pm6.27666$5n4.596175@news6-win.server.ntlworld.com>

I dont think compression would be the way to go as this would cause massive
problems with performance.

Technical Info:
The reason for the huge list is basically the AI tree of board
representations for AI processing.  If the programme was specialised to a
particular game then I could implement a space saving method of storing the
data.  However as the code uses a config file to determine the rules of the
game (hence game is unknown until runtime) the AI engine has to be
completely generic - otherwise the code would be useless!!

Look at this URL for more detailed info (and a download) :
www.geocities.com/djterrier/SGEMain.htm

The format of the list is :
thegamelist  = [ [node, [board_representation] ] , [.....] , [.....],
[.....] , ..... ]
thenodelist = [ [node, [ children_nodes ] , [.....] , [.....] , [.....] ,
..... ]

If anybody can suggest a better method of storing the data [bear in mind
that many functions need access to the data at each node for processing /
checking].  Also bear in mind that due to the interpreted nature of Python,
and the performance demanding nature of the application, any suggested
implementation would need not adversely effect the performance.

Regards
Rich.



Steve Purcell <stephen_purcell at yahoo.com> wrote in message
news:mailman.982950492.15473.python-list at python.org...
> Rich Somerfield wrote:
> > I am generating a huge, huge list of data (cant think of a decent way to
> > reduce the required storage size and have it in a usable form).
Everytime i
> > try to keep this list for future derivations I get a memory problem from
> > Windows (effectively terminating my python script).
> >
> > I presume this is because of the huge list.  Is it possible to find out
the
> > amount of memory a variable [not a type of the variable, the actual data
> > contained within the variable] is taking up?
>
> A huge, huge in-memory list is very likely to be the cause of such a
memory
> problem! <wink>
>
> Compression isn't going to help you much if the list is so big that even
> a tenth of its size would still be too big.
>
> Post some more technical information and I'm sure you'll get more helpful
> suggestions/alternatives than you could have wished for*.
>
> -Steve
>
> * or wanted
>
> --
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Get servlets at http://pyserv.sourceforge.net/
> "Even snakes are afraid of snakes." -- Steven Wright
>




From emile at fenx.com  Mon Feb 26 13:16:35 2001
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 26 Feb 2001 04:16:35 -0800
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com> <mailman.982950492.15473.python-list@python.org> <v3pm6.27666$5n4.596175@news6-win.server.ntlworld.com>
Message-ID: <97dhh2$p9ul0$1@ID-11957.news.dfncis.de>

"Rich Somerfield" <rich_somerfield at tertio.com> wrote in
message
news:v3pm6.27666$5n4.596175 at news6-win.server.ntlworld.com...
> I dont think compression would be the way to go as this
would cause massive
> problems with performance.
>
> Technical Info:
> The reason for the huge list is basically the AI tree of
board
> representations for AI processing.  If the programme was
specialised to a
> particular game then I could implement a space saving
method of storing the
> data.  However as the code uses a config file to determine
the rules of the
> game (hence game is unknown until runtime) the AI engine
has to be
> completely generic - otherwise the code would be useless!!
>
> Look at this URL for more detailed info (and a download) :
> www.geocities.com/djterrier/SGEMain.htm
>
> The format of the list is :
> thegamelist  = [ [node, [board_representation] ] , [.....]
, [.....],
> [.....] , ..... ]
> thenodelist = [ [node, [ children_nodes ] , [.....] ,
[.....] , [.....] ,
> ..... ]
>

Can your data be structured using consistent data types?  I
recently had a situation where I needed to store a ~100,000
element list where each entry consists of 180 floats.  The
memory usage was dramatically reduced by converting the data
to use the array module.

HTH

--

Emile van Sebille
emile at fenx.com
-------------------






From djc at object-craft.com.au  Tue Feb 27 11:13:17 2001
From: djc at object-craft.com.au (Dave Cole)
Date: 27 Feb 2001 21:13:17 +1100
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com> <mailman.982950492.15473.python-list@python.org> <v3pm6.27666$5n4.596175@news6-win.server.ntlworld.com> <97dhh2$p9ul0$1@ID-11957.news.dfncis.de>
Message-ID: <m33dd0juwy.fsf@vole.object-craft.com.au>

>>>>> "Emile" == Emile van Sebille <emile at fenx.com> writes:

Emile> Can your data be structured using consistent data types?  I
Emile> recently had a situation where I needed to store a ~100,000
Emile> element list where each entry consists of 180 floats.  The
Emile> memory usage was dramatically reduced by converting the data to
Emile> use the array module.

Or could you implement a native C datatype on top of which Python
proxy objects were created on demand?  The C objects could be stored
in whatever organisation made the most sense, a tree or an array.

A while ago, I wrote some readonly lazy dictionary / list types which
could be made to sit on top of a parse tree constructed by a yacc
grammar (or whatever).

I still have the code if you are interested.

- Dave

-- 
http://www.object-craft.com.au


From rich_somerfield at tertio.com  Wed Feb 28 10:24:12 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Wed, 28 Feb 2001 09:24:12 -0000
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com> <mailman.982950492.15473.python-list@python.org> <v3pm6.27666$5n4.596175@news6-win.server.ntlworld.com> <97dhh2$p9ul0$1@ID-11957.news.dfncis.de> <m33dd0juwy.fsf@vole.object-craft.com.au>
Message-ID: <yw3n6.1025$sT1.130520@news2-win.server.ntlworld.com>

That code would be gratefully received.

rich_somerfield at nO-sPaM.tertio.com

Remove the obvious, cheers.

Regards
Rich

Dave Cole <djc at object-craft.com.au> wrote in message
news:m33dd0juwy.fsf at vole.object-craft.com.au...
> >>>>> "Emile" == Emile van Sebille <emile at fenx.com> writes:
>
> Emile> Can your data be structured using consistent data types?  I
> Emile> recently had a situation where I needed to store a ~100,000
> Emile> element list where each entry consists of 180 floats.  The
> Emile> memory usage was dramatically reduced by converting the data to
> Emile> use the array module.
>
> Or could you implement a native C datatype on top of which Python
> proxy objects were created on demand?  The C objects could be stored
> in whatever organisation made the most sense, a tree or an array.
>
> A while ago, I wrote some readonly lazy dictionary / list types which
> could be made to sit on top of a parse tree constructed by a yacc
> grammar (or whatever).
>
> I still have the code if you are interested.
>
> - Dave
>
> --
> http://www.object-craft.com.au




From gnana at mips.biochem.mpg.de  Fri Feb 23 18:41:59 2001
From: gnana at mips.biochem.mpg.de (Gnanasekaran Thoppae)
Date: Fri, 23 Feb 2001 17:41:59 -0000
Subject: data parsing
Message-ID: <9767d7+43k5@eGroups.com>

Hi,

I have some data in a file 'test', which contains:

Joe|25|30|49|40|
|28|39|71||
|30|29|||
Malcolm|43|60|56||
|28|37|||
Amy||70|45||
|40|30||40
|40||30||

This is basically a multi line (record) values that belong
to the first line that starts with a filled field.

This is part belongs to joe 'Joe'

Joe|25|30|49|40|
|28|39|71||
|30|29|||

This to 'Malcolm'
Malcolm|43|60|56||
|28|37|||

and the rest to Amy.

I want to parse this data and format it in this way:

Joe|25;28;30|30;39;29|49;71|40|
Malcolm|43;28|60;37|56||
Amy|40;40|70;30|45;;30|;40;|

Basically speaking, I am trying to cluster multi record
data into one field, each seperated by a delimiter ';' and
if the field is empty, an empty ; will enable later on to
decode the field as empty field ''.

Thanks.

-gnana





From mfletch at tpresence.com  Fri Feb 23 23:23:57 2001
From: mfletch at tpresence.com (Mike Fletcher)
Date: Fri, 23 Feb 2001 17:23:57 -0500
Subject: data parsing
Message-ID: <CC59868A0A76D311B1F50090277B321C1E14BD@VRTMAIL>

Here's some ideas (untested, of course):

import string
lines = open( filename ).readlines()
records = map( string.split, lines, ['|']*len(lines))
results = []
while records:
	currentresult = [ string.join(records[0], ';')]
	del records[0]
	while records and not records[0][0]:
		currentresult.append( string.join(records[0], ';') )
		del records[0]
	results.append( string.join( currentresult, '|'))
open( newfilename, 'w').write( string.join( results, '\n'))

Try that in the interactive interpreter to get a feel for what's happening.
Hopefully it'll get you started.  Good luck,
Mike

-----Original Message-----
From: Gnanasekaran Thoppae [mailto:gnana at mips.biochem.mpg.de]
Sent: Friday, February 23, 2001 12:42 PM
To: python-list at cwi.nl
Subject: data parsing


Hi,

I have some data in a file 'test', which contains:

Joe|25|30|49|40|
|28|39|71||
|30|29|||
Malcolm|43|60|56||
|28|37|||
Amy||70|45||
|40|30||40
|40||30||

This is basically a multi line (record) values that belong
to the first line that starts with a filled field.

This is part belongs to joe 'Joe'

Joe|25|30|49|40|
|28|39|71||
|30|29|||

This to 'Malcolm'
Malcolm|43|60|56||
|28|37|||

and the rest to Amy.

I want to parse this data and format it in this way:

Joe|25;28;30|30;39;29|49;71|40|
Malcolm|43;28|60;37|56||
Amy|40;40|70;30|45;;30|;40;|

Basically speaking, I am trying to cluster multi record
data into one field, each seperated by a delimiter ';' and
if the field is empty, an empty ; will enable later on to
decode the field as empty field ''.

Thanks.

-gnana



-- 
http://mail.python.org/mailman/listinfo/python-list



From aleaxit at yahoo.com  Sat Feb 24 18:36:51 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 24 Feb 2001 18:36:51 +0100
Subject: data parsing
References: <mailman.982950184.14469.python-list@python.org>
Message-ID: <978rdf02e8b@news1.newsguy.com>

"Gnanasekaran Thoppae" <gnana at mips.biochem.mpg.de> wrote in message
news:mailman.982950184.14469.python-list at python.org...
> Hi,
>
> I have some data in a file 'test', which contains:
>
> Joe|25|30|49|40|
> |28|39|71||
> |30|29|||
> Malcolm|43|60|56||
> |28|37|||
    [snip]
> I want to parse this data and format it in this way:
>
> Joe|25;28;30|30;39;29|49;71|40|
> Malcolm|43;28|60;37|56||
> Amy|40;40|70;30|45;;30|;40;|
>
> Basically speaking, I am trying to cluster multi record
> data into one field, each seperated by a delimiter ';' and
> if the field is empty, an empty ; will enable later on to
> decode the field as empty field ''.

Problems of data parsing and reformatting are often quite
interesting.  Here, of course, the key is the intermediate,
internal format, in which a 'record' will include the name
and the list-of-lists of data items; all we want is a way to
form such structures from parsing the specified input file
format, and a way to output them into the requested form.

Output of a record can only be performed when it's all in,
which means either that the next record has started, or
that the whole file is over.  A top-down design, then, could
start with the outline:

def reformat(infileob, oufileob):
    data = None
    for line in infileob.lines():
        if line.startswith('|'):
            add_line(data, line)
        else:
            if data is not None:
                output_data(data, oufileob)
            data = new_data(line)
    output_data(data, oufileob)

All we need to do, then, is define more specifically what
we want to do in functions new_data, add_line, and
output_data.  This could be a good occasion to switch to
object-oriented design, making these into the constructor
and two methods of an appropriate class; the outline
would change only slightly:

def reformat(infileob, oufileob):
    data = None
    for line in infileob.lines():
        if line.startswith('|'):
            data.add_line(line)
        else:
            if data is not None:
                data.emit_to(oufileob)
            data = Data(line)
    data.emit_to(oufileob)

It's just a matter of style. In the *implementation* of
the constructor, mutator, and emittor, designing the
data object as a class instance would let us have named
fields as object attributes; since, here, we only really
need two fields (the name, and the list-of-lists of data
items), the advantage is not very substantial -- we may
as well use a two-items list.

Going back to the first outline, then, we could have...:

def new_data(line):
    fields = line.split('|')
    return ( fields[0], [ [field] for field in fields[1:-1] ] )

def add_line(data, line):
    fields = line.split('|')
    for fieldlist, newfield in zip(data[1], fields[1:-1]):
        fieldlist.append(newfield)

def output_data(data, oufileob):
    oufileob.write(data[0]+'|')
    for fieldlist in data[1]:
        oufileob.write(';'.join(fieldlist)+'|')
    oufileob.write('\n')

This will give us more regular output than your example
implicitly specifies, with leading and trailing semicolons
in the same numbers for each case (it's not clear to me
according to which rule you have them in some cases and
not in others), but when you test and adjust this code you
can no doubt implement the exact rules you desire, too.

The split and join method of string objects are what we
are mainly using here, of course; plus a few methods
(append) and operations on tuples (the data object
itself), lists (data[1] being a list of lists of strings),
strings (basically just + to concatenate them) and file
objects (just the write method).  Oh, and, a list
comprehension in new_data, and a little indexing and
slicing.  If any of these Python constructs and idioms
is not fully clear to you, it's possible to rephrase quite
a few of them in other ways (not quite as concise, and
thus, maybe, easier to understand), and anyway, of
course, we're always here for explanations...!


Alex





From news at dorb.com  Sat Feb 24 22:52:03 2001
From: news at dorb.com (Darrell)
Date: Sat, 24 Feb 2001 21:52:03 GMT
Subject: data parsing
References: <mailman.982950184.14469.python-list@python.org>
Message-ID: <72Wl6.166279$o91.20747427@typhoon.nyroc.rr.com>

import re
s="""
Joe|25|30|49|40|
|28|39|71||
|30|29|||
Malcolm|43|60|56||
|28|37|||
Amy||70|45||
|40|30||40
|40||30||
"""
s=s.replace("\012", "")
    # Assumes records start with text and consist of numbers
    # Match the name with \w+
    # Match as many characters as possible until a character not belonging
to a-z or A-Z
    # This should return your records as I understand them.
res1=re.findall("\w+[^a-zA-Z]+", s)
    # Split each record on "|" and join them back with ";"
    # This only works with Python2.0 or higher
res2=[";".join(i.split("|")) for i in res1]
for i in res2:
    print i

######### output
# Not the same as you requested because I don't understand your request.
Joe;25;30;49;40;;28;39;71;;;30;29;;;
Malcolm;43;60;56;;;28;37;;;
Amy;;70;45;;;40;30;;40;40;;30;;

--Darrell

"Gnanasekaran Thoppae" <gnana at mips.biochem.mpg.de> wrote in message
news:mailman.982950184.14469.python-list at python.org...
> Hi,
>
> I have some data in a file 'test', which contains:
>
> Joe|25|30|49|40|
> |28|39|71||
> |30|29|||
> Malcolm|43|60|56||
> |28|37|||
> Amy||70|45||
> |40|30||40
> |40||30||
>
> This is basically a multi line (record) values that belong
> to the first line that starts with a filled field.
>
> This is part belongs to joe 'Joe'
>
> Joe|25|30|49|40|
> |28|39|71||
> |30|29|||
>
> This to 'Malcolm'
> Malcolm|43|60|56||
> |28|37|||
>
> and the rest to Amy.
>
> I want to parse this data and format it in this way:
>
> Joe|25;28;30|30;39;29|49;71|40|
> Malcolm|43;28|60;37|56||
> Amy|40;40|70;30|45;;30|;40;|
>
> Basically speaking, I am trying to cluster multi record
> data into one field, each seperated by a delimiter ';' and
> if the field is empty, an empty ; will enable later on to
> decode the field as empty field ''.
>
> Thanks.
>
> -gnana
>
>
>




From othello at javanet.com  Sun Feb 25 07:43:53 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Sun, 25 Feb 2001 01:43:53 -0500
Subject: data parsing
References: <mailman.982950184.14469.python-list@python.org>
Message-ID: <3A98A9A9.8789B127@javanet.com>

Gnanasekaran Thoppae wrote:

> I have some data in a file 'test', which contains:
>
> Joe|25|30|49|40|
> |28|39|71||
> |30|29|||
> Malcolm|43|60|56||
> |28|37|||
> Amy||70|45||
> |40|30||40
> |40||30||
>
> I want to parse this data and format it in this way:
>
> Joe|25;28;30|30;39;29|49;71|40|
> Malcolm|43;28|60;37|56||
> Amy|40;40|70;30|45;;30|;40;|
>

name = 'unnamed'        # just in-case the file doesn't start with a name
result = {}
for line in inp.split():
    field = line.split('|')
    while len(field) < 6:   # fix misformatted line |40|40||40
        field.append('')
    k = field.pop(0)        # grab name field
    field.pop()             # clear final blank
    if k != '':             # on name change
        name = k
    if not result.has_key(name):
        result[name] = []
    result[name].append(field)

for name in result.keys():
    ans = apply( zip, result[name] )  # tranpose matrix
    ans = map( ';'.join, ans )
    ans.insert(0,name)
    ans.append('')
    print '|'.join(ans)

Raymond



From fgeiger at datec.at  Fri Feb 23 18:45:30 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Fri, 23 Feb 2001 18:45:30 +0100
Subject: Want to build a shop using Zope, but how?
Message-ID: <9767le$82f$1@newsreaderg1.core.theplanet.net>

I want to do the next step in CGI programming and want to use Zope to build
a web shop.

Besides downloading Zope and installing it on my Windows box I printed the
Zope book and read thru it. There is an example all thru the book which is
not of great help for me (at least I think so). While it is good to teach
you the very basics of Zope it does not seem to provide advanced topics.
When it comes to advanced concepts you are very likely to encounter "But
this is beyond this book" (ZClasses e.g.).

I've got the impression, that if one wants to use Zope for anything advanced
he has to download some modules from www.zope.org. But this site really
confuses me. Every day there are new modules released and every few days new
versions of those modules. I don't want to spend hours and days to figure
out what to download so that Zope is of advanced help for me.

The same seems to apply to documentation. Zope still seems to move very
fast. Which of the docs and how-to's are still relevant and which are
outdated?

Yes, I saw one or two Zope products on e-commerce which could be
interesting. But how can I understand those products if I do not understand
Zope yet?

I must admit that I am looking for fast success. Since I know Python I know
fast success is possible ;-). How can I achieve fast success using Zope? My
hope was to find information somewhere on how Zope supports a Pythoneer in
CGI programming: How to deal with cookies in Zope, how to manage session
etc. I am less interested in DTML.

Or did I get something completely wrong here? Anyone here who was in the
same situation? How did you proceed?

Any hint's welcome!

Many thanks in advance and best regards
Franz GEIGER


P.S.: Don't get me wrong because of my clumsy English: I WANT to use Zope.
It's oo approach is too appealing.








From doughellmann at bigfoot.com  Sat Feb 24 14:37:25 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Sat, 24 Feb 2001 13:37:25 GMT
Subject: Want to build a shop using Zope, but how?
References: <9767le$82f$1@newsreaderg1.core.theplanet.net>
Message-ID: <pOOl6.75135$m5.4523734@news1.rdc1.ga.home.com>

In article <9767le$82f$1 at newsreaderg1.core.theplanet.net>, "Franz GEIGER"
<fgeiger at datec.at> wrote:

> I want to do the next step in CGI programming and want to use Zope to
> build a web shop.
> 
> Besides downloading Zope and installing it on my Windows box I printed
> the Zope book and read thru it. There is an example all thru the book
> which is not of great help for me (at least I think so). While it is
> good to teach you the very basics of Zope it does not seem to provide
> advanced topics. When it comes to advanced concepts you are very likely
> to encounter "But this is beyond this book" (ZClasses e.g.).

My understanding was that the Zope book was for beginners, and that there
will be another more advanced book.

You may want to check out the New User How-Tos at
http://www.zope.org/Documentation/nuser.  Reading the online DTML manual
may help as well.

> I've got the impression, that if one wants to use Zope for anything
> advanced he has to download some modules from www.zope.org. But this
> site really confuses me. Every day there are new modules released and
> every few days new versions of those modules. I don't want to spend
> hours and days to figure out what to download so that Zope is of
> advanced help for me.

The Products available on zope.org are components you can instantiate on
your server.  The best place to get help about the Products is to look at
the documentation distributed with them, or to instantiate one and see
what you get -- many include online help.

> Or did I get something completely wrong here? Anyone here who was in the
> same situation? How did you proceed?

The learning curve for Zope is a bit steep, but not terribly long.  There
are actually a few plateaus, so you end up with a stair step learning
curve.

> Any hint's welcome!

Check out the mailing lists on zope.org, too.  They are very high
traffic, but that's generally when Zope developers go for help.

Doug


From fgeiger at datec.at  Fri Feb 23 18:46:41 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Fri, 23 Feb 2001 18:46:41 +0100
Subject: Not all ntpath.py-methods do recognize shares, do they?
Message-ID: <9767nl$83h$1@newsreaderg1.core.theplanet.net>

I had to change  exists() and isdir() of ntpath.py so that they work for my
purposes (shares ARE dirs in my apps):

##### DDG\fge (20001122)
#def exists(path):
#    """Test whether a path exists"""
#    try:
#        st = os.stat(path)
#    except os.error:
#        return 0
#    return 1
def exists(path):
    """Test whether a path exists"""
    try:
        st = os.stat(path)
    except os.error:
        if ismount(path):
            return 1
        return 0
    return 1
#
##### DDG\fge (20001122)


##### DDG\fge (20010223)
#def isdir(path):
#    """Test whether a path is a directory"""
#    try:
#        st = os.stat(path)
#    except os.error:
#        return 0
#    return stat.S_ISDIR(st[stat.ST_MODE])
def isdir(path):
    """Test whether a path is a directory"""
    try:
        st = os.stat(path)
    except os.error:
        if ismount(path):
            return 1
        return 0
    return stat.S_ISDIR(st[stat.ST_MODE])
#
##### DDG\fge (20010223)


Any comments? Did I miss something?

Best regards
Franz GEIGER







From MarkH at ActiveState.com  Sat Feb 24 02:35:12 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 24 Feb 2001 01:35:12 GMT
Subject: Not all ntpath.py-methods do recognize shares, do they?
References: <9767nl$83h$1@newsreaderg1.core.theplanet.net>
Message-ID: <3A970EF4.80400@ActiveState.com>

Franz GEIGER wrote:

> I had to change  exists() and isdir() of ntpath.py so that they work for my
> purposes (shares ARE dirs in my apps):
...

> Any comments? Did I miss something?

Off the top of my head I can't see a problem with this.  Submit a patch at sourceforge.

Mark.



From db3l at fitlinxx.com  Mon Feb 26 23:07:52 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 26 Feb 2001 17:07:52 -0500
Subject: Not all ntpath.py-methods do recognize shares, do they?
References: <9767nl$83h$1@newsreaderg1.core.theplanet.net>
Message-ID: <uitlx14jr.fsf@ctwd0143.fitlinxx.com>

"Franz GEIGER" <fgeiger at datec.at> writes:

> Any comments? Did I miss something?

When I last used this (and I just checked again), ismount() returns 1
for any directory immediately below a remote machine, whether or not
it is really a share.  So this might not be safe as a general change:

E.g., A remote machine "ctwh01" has a share "sitedata" with a directory
"ni" beneath it:

    os.path.ismount(r'\\ctwh01\sitedata\ni')  --> 0  
    os.path.ismount(r'\\ctwh01\sitedata')     --> 1
    os.path.ismount(r'\\ctwh01\foo')          --> 1
    os.path.ismount(r'\\ctwh01\bar')          --> 1
    os.path.ismount(r'\\ctwh01\\')            --> 1

But "foo", "bar" and the root directory "\" are not shares.

This was with Python 1.5.2 and Python 2.0 under NT4SP4.

If you want to treat a directory beneath a share as a directory, then
I would probably suggest that you use your own function wrapping
os.path.isdir rather than modifying the libraries themselves with
ismount() since I think that could lead to confusion in other cases.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From gem at hsv.crc.com  Fri Feb 23 18:55:36 2001
From: gem at hsv.crc.com (Glen Mettler)
Date: Fri, 23 Feb 2001 11:55:36 -0600
Subject: Python with AMS Real Time Project
Message-ID: <9767p9$ear$1@hobbes2.crc.com>

Is anyone using Python with AMS Real Time Project?
I have just acquired both and I am wondering about the ease and practicality
of manipulating RTS from Python.

Glen





From BgPorter at NOartlogicSPAM.com  Fri Feb 23 19:35:43 2001
From: BgPorter at NOartlogicSPAM.com (Brett g Porter)
Date: Fri, 23 Feb 2001 18:35:43 GMT
Subject: Palm/Python
Message-ID: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com>

Seen this morning on the PHO mailing list. Anyone know anything about this
port?

-------------------------

Peer-to-Peer Everywhere?
By: Mike Darrah 2/23/01
Published exclusively on PHO mailing list.

Endeavors Technology has announced the successful deployment
of an open source port of Python to the Palm OS, allowing the
company to bring it's "Magi" P2P technology to the Palm OS platform,
and enabling the automation of applications on the Palm OS.

Python is an interpreted, interactive, object-oriented programming
language, which allows for a great deal of dynamism, ease of
use, power, and flexibility in a programming language, often
compared to Perl or Java.

Endeavors Technology has been focusing on building a common infrastructure
for all and all connected devices, such as PDAs, PCs, servers,
information appliances, and Internet-ready phones to allow the
development of a P2P system structure to enable  software applications
to communicate to one another regardless of what type of device
it is.

Endeavors stated it will work with device manufacturers and software
developers to implement this small footprint, interoperable virtual
machine (VM) across more then just the Palm OS platform, thus
creating the common infrastructure needed to connect the devices
seamlessly in a peer-to-peer network. The technology, named "The
Pippy VM" is an extension of the work pioneered in a joint venture
between The Aerospace Corp. and University of California, Irvine,
which was initially funded in part by the Defense Advanced Research
Project Agency (DARPA).

The company has also already developed similar P2P systems for
the Compaq's iPAQ Pocket PC, announced just last month.

With the announcement of the successful deployment of the Python
port to the Palm OS, Endeavors Technology has taken a bold step
forward into making their vision of a complete P2P network with
all connected devices into a reality.

Additional Information References:
http://www.python.org/
http://www.endeavors.com/
http://www.endeavors.com/tech.html

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Once again, anyone interested in syndicating this news, please
feel free to do so free of charge. All I request is to be named
as the author of this news article in the syndication, and be
notified of it's syndication.

Sincerely,
Mike Darrah

-X-

Digital Media Consulting / Management

XGeneration Network
webmaster at xgeneration.net

--
// Brett g Porter * Senior Engineer, Application Development
// BgPorter @ artlogic . com * http://www.artlogic.com
// Art & Logic * Custom software solutions for hardware products
// Windows * MacOS * Embedded Systems





From fwittenburg at myokay.net  Fri Feb 23 21:49:49 2001
From: fwittenburg at myokay.net (Florian W.)
Date: Fri, 23 Feb 2001 21:49:49 +0100
Subject: Palm/Python
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com>
Message-ID: <976ibc$2ca8$1@news.okay.net>

Brett g Porter <BgPorter at NOartlogicSPAM.com> schrieb in im Newsbeitrag:
34yl6.301124$w35.49575483 at news1.rdc1.nj.home.com...
> Seen this morning on the PHO mailing list. Anyone know anything about this
> port?

Is it this ?

http://www.isr.uci.edu/projects/sensos/python/software.html

(I didn't manage to do something usefull with it.)

Mfg Florian





From jcollins at endeavors.com  Fri Feb 23 22:09:39 2001
From: jcollins at endeavors.com (Jeff Collins)
Date: Fri, 23 Feb 2001 14:09:39 -0700
Subject: Palm/Python
In-Reply-To: <976ibc$2ca8$1@news.okay.net>; from fwittenburg@myokay.net on Fri, Feb 23, 2001 at 09:49:49PM +0100
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <976ibc$2ca8$1@news.okay.net>
Message-ID: <20010223140939.T13871@endeavors.com>

The new site can be found at http://www.endeavors.com/pippy

On Fri, Feb 23, 2001 at 09:49:49PM +0100, Florian W. wrote:
> 
> Brett g Porter <BgPorter at NOartlogicSPAM.com> schrieb in im Newsbeitrag:
> 34yl6.301124$w35.49575483 at news1.rdc1.nj.home.com...
> > Seen this morning on the PHO mailing list. Anyone know anything about this
> > port?
> 
> Is it this ?
> 
> http://www.isr.uci.edu/projects/sensos/python/software.html
> 
> (I didn't manage to do something usefull with it.)
> 
> Mfg Florian

-- 
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.
http://www.endeavors.com



From jcollins at endeavors.com  Fri Feb 23 22:09:03 2001
From: jcollins at endeavors.com (Jeff Collins)
Date: Fri, 23 Feb 2001 14:09:03 -0700
Subject: Palm/Python
In-Reply-To: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com>; from BgPorter@NOartlogicSPAM.com on Fri, Feb 23, 2001 at 06:35:43PM +0000
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com>
Message-ID: <20010223140903.S13871@endeavors.com>

You can find the new home page for the port of Python to the Palm at
http://www.endeavors.com/pippy.



On Fri, Feb 23, 2001 at 06:35:43PM +0000, Brett g Porter wrote:
> Seen this morning on the PHO mailing list. Anyone know anything about this
> port?
> 
> -------------------------
> 
> Peer-to-Peer Everywhere?
> By: Mike Darrah 2/23/01
> Published exclusively on PHO mailing list.
> 
> Endeavors Technology has announced the successful deployment
> of an open source port of Python to the Palm OS, allowing the
> company to bring it's "Magi" P2P technology to the Palm OS platform,
> and enabling the automation of applications on the Palm OS.
> 
> Python is an interpreted, interactive, object-oriented programming
> language, which allows for a great deal of dynamism, ease of
> use, power, and flexibility in a programming language, often
> compared to Perl or Java.
> 
> Endeavors Technology has been focusing on building a common infrastructure
> for all and all connected devices, such as PDAs, PCs, servers,
> information appliances, and Internet-ready phones to allow the
> development of a P2P system structure to enable  software applications
> to communicate to one another regardless of what type of device
> it is.
> 
> Endeavors stated it will work with device manufacturers and software
> developers to implement this small footprint, interoperable virtual
> machine (VM) across more then just the Palm OS platform, thus
> creating the common infrastructure needed to connect the devices
> seamlessly in a peer-to-peer network. The technology, named "The
> Pippy VM" is an extension of the work pioneered in a joint venture
> between The Aerospace Corp. and University of California, Irvine,
> which was initially funded in part by the Defense Advanced Research
> Project Agency (DARPA).
> 
> The company has also already developed similar P2P systems for
> the Compaq's iPAQ Pocket PC, announced just last month.
> 
> With the announcement of the successful deployment of the Python
> port to the Palm OS, Endeavors Technology has taken a bold step
> forward into making their vision of a complete P2P network with
> all connected devices into a reality.
> 
> Additional Information References:
> http://www.python.org/
> http://www.endeavors.com/
> http://www.endeavors.com/tech.html
> 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Once again, anyone interested in syndicating this news, please
> feel free to do so free of charge. All I request is to be named
> as the author of this news article in the syndication, and be
> notified of it's syndication.
> 
> Sincerely,
> Mike Darrah
> 
> -X-
> 
> Digital Media Consulting / Management
> 
> XGeneration Network
> webmaster at xgeneration.net
> 
> --
> // Brett g Porter * Senior Engineer, Application Development
> // BgPorter @ artlogic . com * http://www.artlogic.com
> // Art & Logic * Custom software solutions for hardware products
> // Windows * MacOS * Embedded Systems

-- 
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.
http://www.endeavors.com



From bryan at eevolved.com  Sat Feb 24 16:51:17 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Sat, 24 Feb 2001 15:51:17 GMT
Subject: Palm/Python
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.982962545.9330.python-list@python.org>
Message-ID: <VLQl6.217821$Pm2.3527896@news20.bellglobal.com>

Hi Jeff,

Congratulations to you and Micheal for the very impressive work with Pippy. 
Even more impressive is that your website seems to be shrugging off a full 
scale slashdotting! :)

I was disheartened though to see in the README that floating points, file 
objects and long integers are dropped (I need those!), although I 
appreciate some concessions must be made to reduce the footprint for Palm.

I notice in the goals that you intend to lobby for incorporation of these 
modifications into the core distribution... Good luck. :) Personally as a 
python PDA application developer, I require the full functionality of 
python and consequently cannot cater to a platform incapable of supporting 
it completely.

Although I'm sure pippy is of great use to your application and I applaud 
your efforts, don't you see PDA capabilities as soon being able to run the 
complete python? 

Good luck with the venture,
-- 
<=====================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
Website:    http://www.eevolved.com
Public key: http://eevolved.com/bcm.pk
<=====================================>

"The genes are the master programmers, and they are programming for their 
lives." -- Richard Dawkins



From jcollins at endeavors.com  Mon Feb 26 17:16:16 2001
From: jcollins at endeavors.com (Jeff Collins)
Date: Mon, 26 Feb 2001 09:16:16 -0700
Subject: Palm/Python
In-Reply-To: <VLQl6.217821$Pm2.3527896@news20.bellglobal.com>; from bryan@eevolved.com on Sat, Feb 24, 2001 at 03:51:17PM +0000
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.982962545.9330.python-list@python.org> <VLQl6.217821$Pm2.3527896@news20.bellglobal.com>
Message-ID: <20010226091615.E13871@endeavors.com>

On Sat, Feb 24, 2001 at 03:51:17PM +0000, Bryan Mongeau wrote:
> Hi Jeff,
> 
> Congratulations to you and Micheal for the very impressive work with Pippy. 
> Even more impressive is that your website seems to be shrugging off a full 
> scale slashdotting! :)

Thanks!  I was surprised at how fast the news traveled.

> 
> I was disheartened though to see in the README that floating points, file 
> objects and long integers are dropped (I need those!), although I 
> appreciate some concessions must be made to reduce the footprint for Palm.
> 
> I notice in the goals that you intend to lobby for incorporation of these 
> modifications into the core distribution... Good luck. :) Personally as a 
> python PDA application developer, I require the full functionality of 
> python and consequently cannot cater to a platform incapable of supporting 
> it completely.
> 
> Although I'm sure pippy is of great use to your application and I applaud 
> your efforts, don't you see PDA capabilities as soon being able to run the 
> complete python? 

I wish I could predict the future:-) The earlier Palm platforms had
much less dynamic heap, stack and storage RAM than the latest models.
I would expect this trend to continue.  Check out PythonCE, a port of
an older version of Python to WinCE devices.  It is quite capable.

-- 
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.
http://www.endeavors.com



From hinsen at cnrs-orleans.fr  Tue Feb 27 11:08:50 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 27 Feb 2001 11:08:50 +0100
Subject: Palm/Python
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.982962545.9330.python-list@python.org> <VLQl6.217821$Pm2.3527896@news20.bellglobal.com> <mailman.983204227.22840.python-list@python.org>
Message-ID: <m3hf1gl9ot.fsf@chinon.cnrs-orleans.fr>

"Jeff Collins" <jcollins at endeavors.com> writes:

> I wish I could predict the future:-) The earlier Palm platforms had
> much less dynamic heap, stack and storage RAM than the latest models.

I wonder if that is also the reason why Pippy needs PalmOS 3.5?
Of course I have only 3.3 :-(

Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From ajung-ml at andreas-jung.com  Tue Feb 27 13:58:15 2001
From: ajung-ml at andreas-jung.com (Andreas Jung)
Date: Tue, 27 Feb 2001 07:58:15 -0500
Subject: Palm/Python
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.982962545.9330.python-list@python.org> <VLQl6.217821$Pm2.3527896@news20.bellglobal.com> <mailman.983204227.22840.python-list@python.org> <m3hf1gl9ot.fsf@chinon.cnrs-orleans.fr>
Message-ID: <013001c0a0c0$010cb970$d0063bd0@SUXLAP>

Don't worry. I am running 2.04 *sigh too*

Andreas
----- Original Message -----
From: "Konrad Hinsen" <hinsen at cnrs-orleans.fr>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Tuesday, February 27, 2001 5:08 AM
Subject: Re: Palm/Python


> "Jeff Collins" <jcollins at endeavors.com> writes:
>
> > I wish I could predict the future:-) The earlier Palm platforms had
> > much less dynamic heap, stack and storage RAM than the latest models.
>
> I wonder if that is also the reason why Pippy needs PalmOS 3.5?
> Of course I have only 3.3 :-(
>
> Konrad.
> --
> --------------------------------------------------------------------------
-----
> Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
> Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
> Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
> 45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
> France                                   | Nederlands/Francais
> --------------------------------------------------------------------------
-----
> --
> http://mail.python.org/mailman/listinfo/python-list





From jcollins at endeavors.com  Tue Feb 27 16:11:08 2001
From: jcollins at endeavors.com (Jeff Collins)
Date: Tue, 27 Feb 2001 08:11:08 -0700
Subject: Palm/Python
In-Reply-To: <m3hf1gl9ot.fsf@chinon.cnrs-orleans.fr>; from hinsen@cnrs-orleans.fr on Tue, Feb 27, 2001 at 11:08:50AM +0100
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.982962545.9330.python-list@python.org> <VLQl6.217821$Pm2.3527896@news20.bellglobal.com> <mailman.983204227.22840.python-list@python.org> <m3hf1gl9ot.fsf@chinon.cnrs-orleans.fr>
Message-ID: <20010227081108.K13871@endeavors.com>

We are testing Pippy on a device with 256K of heap, which is provided
by PalmOS 3.5 with at least 4M of RAM.  The 96K of heap provided by
the PalmOS 3.0,3.3 isn't enough to load the parser accelerators.  We
are continuing the effort to reduce the heap usage and hope to be
running on these devices soon.

On Tue, Feb 27, 2001 at 11:08:50AM +0100, Konrad Hinsen wrote:
> "Jeff Collins" <jcollins at endeavors.com> writes:
> 
> > I wish I could predict the future:-) The earlier Palm platforms had
> > much less dynamic heap, stack and storage RAM than the latest models.
> 
> I wonder if that is also the reason why Pippy needs PalmOS 3.5?
> Of course I have only 3.3 :-(
> 
> Konrad.

-- 
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.
http://www.endeavors.com



From fpm at u.washington.edu  Tue Feb 27 18:30:04 2001
From: fpm at u.washington.edu (Frank Miles)
Date: 27 Feb 2001 17:30:04 GMT
Subject: Palm/Python
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.983204227.22840.python-list@python.org> <m3hf1gl9ot.fsf@chinon.cnrs-orleans.fr> <mailman.983286732.8187.python-list@python.org>
Message-ID: <97go6s$vn4$1@nntp6.u.washington.edu>

In article <mailman.983286732.8187.python-list at python.org>,
Jeff Collins <jcollins at endeavors.com> wrote:
>
>We are testing Pippy on a device with 256K of heap, which is provided
>by PalmOS 3.5 with at least 4M of RAM.  The 96K of heap provided by
>the PalmOS 3.0,3.3 isn't enough to load the parser accelerators.  We
>are continuing the effort to reduce the heap usage and hope to be
>running on these devices soon.

Any hope that floats might be supported in the future?  Using MathLib
should take most of the pain out of that, other than possible space
problems...

	-frank
-- 


From jcollins at endeavors.com  Tue Feb 27 19:49:24 2001
From: jcollins at endeavors.com (Jeff Collins)
Date: Tue, 27 Feb 2001 11:49:24 -0700
Subject: Palm/Python
In-Reply-To: <97go6s$vn4$1@nntp6.u.washington.edu>; from fpm@u.washington.edu on Tue, Feb 27, 2001 at 05:30:04PM +0000
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.983204227.22840.python-list@python.org> <m3hf1gl9ot.fsf@chinon.cnrs-orleans.fr> <mailman.983286732.8187.python-list@python.org> <97go6s$vn4$1@nntp6.u.washington.edu>
Message-ID: <20010227114924.T13871@endeavors.com>

I have no immediate need for floats, but if someone is interested in
pursuing this, I may be able to offer suggestions.  

On Tue, Feb 27, 2001 at 05:30:04PM +0000, Frank Miles wrote:
> In article <mailman.983286732.8187.python-list at python.org>,
> Jeff Collins <jcollins at endeavors.com> wrote:
> >
> >We are testing Pippy on a device with 256K of heap, which is provided
> >by PalmOS 3.5 with at least 4M of RAM.  The 96K of heap provided by
> >the PalmOS 3.0,3.3 isn't enough to load the parser accelerators.  We
> >are continuing the effort to reduce the heap usage and hope to be
> >running on these devices soon.
> 
> Any hope that floats might be supported in the future?  Using MathLib
> should take most of the pain out of that, other than possible space
> problems...
> 
> 	-frank

-- 
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.
http://www.endeavors.com



From bobhicks at adelphia.net  Tue Feb 27 22:38:04 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Tue, 27 Feb 2001 21:38:04 GMT
Subject: Palm/Python
References: <34yl6.301124$w35.49575483@news1.rdc1.nj.home.com> <mailman.982962545.9330.python-list@python.org> <VLQl6.217821$Pm2.3527896@news20.bellglobal.com> <mailman.983204227.22840.python-list@python.org> <m3hf1gl9ot.fsf@chinon.cnrs-orleans.fr> <mailman.983286732.8187.python-list@python.org>
Message-ID: <B6C18868.1500%bobhicks@adelphia.net>

My M100 runs the beta fine...

> From: "Jeff Collins" <jcollins at endeavors.com>
> Organization: Baymountain
> Newsgroups: comp.lang.python
> Date: Tue, 27 Feb 2001 08:11:08 -0700
> To: Konrad Hinsen <hinsen at cnrs-orleans.fr>
> Cc: python-list at python.org
> Subject: Re: Palm/Python
> 
> 
> We are testing Pippy on a device with 256K of heap, which is provided
> by PalmOS 3.5 with at least 4M of RAM.  The 96K of heap provided by
> the PalmOS 3.0,3.3 isn't enough to load the parser accelerators.  We
> are continuing the effort to reduce the heap usage and hope to be
> running on these devices soon.
> 
> On Tue, Feb 27, 2001 at 11:08:50AM +0100, Konrad Hinsen wrote:
>> "Jeff Collins" <jcollins at endeavors.com> writes:
>> 
>>> I wish I could predict the future:-) The earlier Palm platforms had
>>> much less dynamic heap, stack and storage RAM than the latest models.
>> 
>> I wonder if that is also the reason why Pippy needs PalmOS 3.5?
>> Of course I have only 3.3 :-(
>> 
>> Konrad.
> 
> -- 
> Jeffery D. Collins, Ph.D.
> Sr. Software Developer
> Endeavors Technology, Inc.
> http://www.endeavors.com
> 



From chrisa at ASPATECH.COM.BR  Fri Feb 23 20:04:33 2001
From: chrisa at ASPATECH.COM.BR (Chris Richard Adams)
Date: Fri, 23 Feb 2001 16:04:33 -0300
Subject: interactive programming
Message-ID: <BD674CDC0A68C642B60038986FBB56DF015C43@mailsrv.ASPATECH.COM.BR>

How can I create a loop - interactively.  I'm just trying to learn
syntax for example, so Id like to create conditional if statements like
a = 'feb'

if a == 'jan':
	mo = 1
elif a == 'feb':
	mo = 2
elif a == 'mar':
	mo = 3
...

I tried to type this interactively, but I get a mess of error about
syntax or mis placed whitespace...

1.) After the return after mo = 1 - how do I make sure theelif is
aligned with if?

2.) Is there any refernces for interactive programming - the tutorial is
too short?

Thanks,Chris




From shaleh at valinux.com  Fri Feb 23 20:20:28 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 23 Feb 2001 11:20:28 -0800 (PST)
Subject: interactive programming
In-Reply-To: <BD674CDC0A68C642B60038986FBB56DF015C43@mailsrv.ASPATECH.COM.BR>
Message-ID: <XFMail.20010223112028.shaleh@valinux.com>

On 23-Feb-2001 Chris Richard Adams wrote:
> How can I create a loop - interactively.  I'm just trying to learn
> syntax for example, so Id like to create conditional if statements like
> a = 'feb'
> 
> if a == 'jan':
>       mo = 1
> elif a == 'feb':
>       mo = 2
> elif a == 'mar':
>       mo = 3
> ...
> 
> I tried to type this interactively, but I get a mess of error about
> syntax or mis placed whitespace...
> 

$ python
Python 1.5.2 (#0, Feb 20 2001, 23:57:58)  [GCC 2.95.3 20010125 (prerelease)] on
linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 

the >>> is the 'start a command here prompt'

>>> if a == 'jan':
...     print a # the ... is the 'continue previous command', when you see this
...             # you must indent your code (just hit tab or a few spaces)
... elif a == 'feb': # still under the if
...     print a # again, indented
...
feb
>>> # back to start a command



From chrisa at ASPATECH.COM.BR  Fri Feb 23 20:12:15 2001
From: chrisa at ASPATECH.COM.BR (Chris Richard Adams)
Date: Fri, 23 Feb 2001 16:12:15 -0300
Subject: Can I leverage Java classes from a python program
Message-ID: <BD674CDC0A68C642B60038986FBB56DF015C44@mailsrv.ASPATECH.COM.BR>

If so, questions:

I am working with Java Oracle Intermedia classes (just a Java API that
helps you manipulate images (Blobs) stored in a database). ANyway - I
have to work with the java oracle classes, but wanted to do this from a
python program. Can I use inheritance to access these types of classes,
even though they are in Java. If so could you point to a reference that
might show some examples???? AND does this have anything to do with
JPython - would I want to use that rather than my standard Linux 2.0
version.

thanks!




From dsh8290 at rit.edu  Fri Feb 23 22:21:42 2001
From: dsh8290 at rit.edu (D-Man)
Date: Fri, 23 Feb 2001 16:21:42 -0500
Subject: Can I leverage Java classes from a python program
In-Reply-To: <BD674CDC0A68C642B60038986FBB56DF015C44@mailsrv.ASPATECH.COM.BR>; from chrisa@aspatech.com.br on Fri, Feb 23, 2001 at 04:12:15PM -0300
References: <BD674CDC0A68C642B60038986FBB56DF015C44@mailsrv.ASPATECH.COM.BR>
Message-ID: <20010223162142.A5639@harmony.cs.rit.edu>

You can certainly use any Java classes from Python.  Go to
jythnon.sourceforge.net.  Jython is the new name for JPython.  It is
current with version 2.0 of CPython.


from javax.swing import JDialog , JLabel

class MyDialog( JDialog ) :
    def __init__( self ) :
        self.getContentPane().add( JLabel( "Hello World" ) )
        self.setSize( 200 , 300 )
        self.setVisible( 1 )

dialog = MyDialog()



-D




From whisper at oz.net  Fri Feb 23 20:39:21 2001
From: whisper at oz.net (Dave LeBlanc)
Date: 23 Feb 2001 19:39:21 GMT
Subject: Protecting Python Application Distributions - URGENT
Message-ID: <976e99$7pj$0@216.39.170.247>

Hi;

I have the possible chance to do a consulting project in Python, but
the customer has raised concerns about (not) doing a source
distribution.

I've read some stuff about python->C etc. but I haven't been able to
find any concrete information.

Any information about creating "exe" or "shrouded" or some other form
of protected distribution on Windows would be greatly appreciated. (I
realize that 100% protection is impossible given sufficient
motivation; I'd just like to make it irritating enough that most
people wouldn't bother trying to decompile/reconstruct the code.)

Cross-emails to whisper at oz.net greatly appreciated!

Regards,

Dave LeBlanc

P.S. Any solution that created stand-alone .exe files (no separate
python.exe needed) would be of special interest.



From adam at deprince.net  Fri Feb 23 08:03:53 2001
From: adam at deprince.net (Adam DePrince)
Date: Fri, 23 Feb 2001 02:03:53 -0500
Subject: Protecting Python Application Distributions - URGENT
References: <976e99$7pj$0@216.39.170.247>
Message-ID: <3A960B59.B10FF235@deprince.net>

Dave LeBlanc wrote:
> 
> Hi;
> 
> I have the possible chance to do a consulting project in Python, but
> the customer has raised concerns about (not) doing a source
> distribution.
> 
> I've read some stuff about python->C etc. but I haven't been able to
> find any concrete information.
> 
> Any information about creating "exe" or "shrouded" or some other form
> of protected distribution on Windows would be greatly appreciated. (I
> realize that 100% protection is impossible given sufficient
> motivation; I'd just like to make it irritating enough that most
> people wouldn't bother trying to decompile/reconstruct the code.)
> 
> Cross-emails to whisper at oz.net greatly appreciated!
> 
> Regards,
> 
> Dave LeBlanc
> 
> P.S. Any solution that created stand-alone .exe files (no separate
> python.exe needed) would be of special interest.


You could just ship the compiled python files (.pyc).   

If worry about others coping your stuff, well, stealing others code is
actually rather hard, it often takes longer to wrap ones mind around the
way others think than it does to rewrite it.  In fact, I've casually
observed that many projects, open source and closed, expereince the
greatest structural changes following changes in stewardship.  Besides,
if your code is good and your clients are smart, they are more likely to
say "hey, your stuff looks good.  Write more" if they can see it.


-- 
Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From butcuz at yahoo.com  Fri Feb 23 20:59:08 2001
From: butcuz at yahoo.com (sa)
Date: Fri, 23 Feb 2001 14:59:08 -0500
Subject: Protecting Python Application Distributions - URGENT
References: <976e99$7pj$0@216.39.170.247>
Message-ID: <3a96c1be$1_4@corp.newsfeeds.com>

- you might want to check these out:

Thomas Heller's py2exe : http://starship.python.net/crew/theller/py2exe/ (an
easy way create 'standalone' applications using disutils)

Gordon McMillan's python distribution utilities :
http://www.mcmillan-inc.com/install1.html (more options, but steeper
learning curve)

good luck,

sa

"Dave LeBlanc" <whisper at oz.net> wrote in message
news:976e99$7pj$0 at 216.39.170.247...
> Hi;
>
> I have the possible chance to do a consulting project in Python, but
> the customer has raised concerns about (not) doing a source
> distribution.
>
> I've read some stuff about python->C etc. but I haven't been able to
> find any concrete information.
>
> Any information about creating "exe" or "shrouded" or some other form
> of protected distribution on Windows would be greatly appreciated. (I
> realize that 100% protection is impossible given sufficient
> motivation; I'd just like to make it irritating enough that most
> people wouldn't bother trying to decompile/reconstruct the code.)
>
> Cross-emails to whisper at oz.net greatly appreciated!
>
> Regards,
>
> Dave LeBlanc
>
> P.S. Any solution that created stand-alone .exe files (no separate
> python.exe needed) would be of special interest.
>




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From whisper at oz.net  Fri Feb 23 21:27:58 2001
From: whisper at oz.net (Dave LeBlanc)
Date: 23 Feb 2001 20:27:58 GMT
Subject: Protecting Python Application Distributions - URGENT
References: <976e99$7pj$0@216.39.170.247> <3a96c1be$1_4@corp.newsfeeds.com>
Message-ID: <976h4e$dbh$0@216.39.170.247>

sa!

Thanks! Most Excellent!

Dave LeBlanc

On Fri, 23 Feb 2001 14:59:08 -0500, "sa" <butcuz at yahoo.com> wrote:

>
>- you might want to check these out:
>
>Thomas Heller's py2exe : http://starship.python.net/crew/theller/py2exe/ (an
>easy way create 'standalone' applications using disutils)
>
>Gordon McMillan's python distribution utilities :
>http://www.mcmillan-inc.com/install1.html (more options, but steeper
>learning curve)
>
>good luck,
>
>sa
>
>"Dave LeBlanc" <whisper at oz.net> wrote in message
>news:976e99$7pj$0 at 216.39.170.247...
>> Hi;
>>
>> I have the possible chance to do a consulting project in Python, but
>> the customer has raised concerns about (not) doing a source
>> distribution.
>>
>> I've read some stuff about python->C etc. but I haven't been able to
>> find any concrete information.
>>
>> Any information about creating "exe" or "shrouded" or some other form
>> of protected distribution on Windows would be greatly appreciated. (I
>> realize that 100% protection is impossible given sufficient
>> motivation; I'd just like to make it irritating enough that most
>> people wouldn't bother trying to decompile/reconstruct the code.)
>>
>> Cross-emails to whisper at oz.net greatly appreciated!
>>
>> Regards,
>>
>> Dave LeBlanc
>>
>> P.S. Any solution that created stand-alone .exe files (no separate
>> python.exe needed) would be of special interest.
>>
>
>
>
>
>-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
>http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
>-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----



From tismer at tismer.com  Fri Feb 23 20:42:45 2001
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 23 Feb 2001 20:42:45 +0100
Subject: Stackless Python 2.0 Minor Update ("Korea Edition")
Message-ID: <3A96BD35.7D56E8C@tismer.com>

This is yet another bugfix release for Stackless Python 2.0 .

x = []
for (a, b) in ((1,2),):
    if 1:
        x.append(a)

no longer crashes.
The reason was an unitialized error flag.

The new Win32 distribution and source code
can be downloaded from http://www.stackless.com/

People who don't want to do a download for just a small
change, and who like to compile Stackless anyway,
can find a simpe patch as well.

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net/
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com/

-- 
http://mail.python.org/mailman/listinfo/python-list



From nb0klun71q5001 at sneakemail.com  Fri Feb 23 20:47:47 2001
From: nb0klun71q5001 at sneakemail.com (Ron Scinta)
Date: Fri, 23 Feb 2001 19:47:47 GMT
Subject: Newbie question.. how to load modules with a dynamic name?
Message-ID: <D7zl6.221$Gm2.5547@skycache.prestige.net>

I have a need to run various python script files from a master control
 script, however, I do not know the names of the scripts until runtime.

 I will be getting the script names from a database.. and was thinking I
 could just use the import call with the supplied name.

 This does not work alas....
 >>> mymod = "ptw"
 >>> import mymod
 Traceback (innermost last):
   File "<interactive input>", line 1, in ?
 ImportError: No module named mym
 >>>

 What is the prefered method of doing this type of operation?

 Any help would be greatly appreciated!

 Best Regards,
 Ron









From shaleh at valinux.com  Fri Feb 23 22:13:26 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 23 Feb 2001 13:13:26 -0800 (PST)
Subject: Newbie question.. how to load modules with a dynamic name?
In-Reply-To: <D7zl6.221$Gm2.5547@skycache.prestige.net>
Message-ID: <XFMail.20010223131326.shaleh@valinux.com>

On 23-Feb-2001 Ron Scinta wrote:
> I have a need to run various python script files from a master control
>  script, however, I do not know the names of the scripts until runtime.
> 
>  What is the prefered method of doing this type of operation?
> 

>>> foo = 'sys'
>>> exec "import %s" % foo
>>> print sys.argv
['']
>>> dir(sys)      
['__doc__', '__name__', '__stderr__', '__stdin__', '__stdout__', 'argv',
'builtin_module_names', 'copyright', 'exc_info', 'exc_type', 'exec_prefix',
'executable', 'exit', 'getrefcount', 'hexversion', 'maxint', 'modules', 'path',
'platform', 'prefix', 'ps1', 'ps2', 'setcheckinterval', 'setprofile',
'settrace', 'stderr', 'stdin', 'stdout', 'version']




From dsavitsk at e-coli.net  Fri Feb 23 22:20:25 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 23 Feb 2001 15:20:25 -0600
Subject: Newbie question.. how to load modules with a dynamic name?
References: <D7zl6.221$Gm2.5547@skycache.prestige.net>
Message-ID: <uuAl6.125974$Ch.23398786@newsrump.sjc.telocity.net>

"Ron Scinta" <nb0klun71q5001 at sneakemail.com> wrote in message
news:D7zl6.221$Gm2.5547 at skycache.prestige.net...
> I have a need to run various python script files from a master control
>  script, however, I do not know the names of the scripts until runtime.
>
>  I will be getting the script names from a database.. and was thinking I
>  could just use the import call with the supplied name.
>
>  This does not work alas....
>  >>> mymod = "ptw"

>>> exec('import ' + mymod)

ds




From nb0klun71q5001 at sneakemail.com  Fri Feb 23 21:40:05 2001
From: nb0klun71q5001 at sneakemail.com (Ron Scinta)
Date: Fri, 23 Feb 2001 20:40:05 GMT
Subject: Newbie question.. how to load modules with a dynamic name?
References: <D7zl6.221$Gm2.5547@skycache.prestige.net> <uuAl6.125974$Ch.23398786@newsrump.sjc.telocity.net>
Message-ID: <FUzl6.227$Gm2.5865@skycache.prestige.net>

> >  This does not work alas....
> >  >>> mymod = "ptw"
>
> >>> exec('import ' + mymod)
>
> ds

OK, but how then do I reference the dynamic module by name?

mymod. (The period does not help here, still thinks I am using the string
class instead of my imported module)....

 Is there a way to alias the module name at this point?




From gerhard.nospam at bigfoot.de  Sat Feb 24 00:45:03 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Sat, 24 Feb 2001 00:45:03 +0100
Subject: Newbie question.. how to load modules with a dynamic name?
References: <D7zl6.221$Gm2.5547@skycache.prestige.net> <uuAl6.125974$Ch.23398786@newsrump.sjc.telocity.net> <FUzl6.227$Gm2.5865@skycache.prestige.net>
Message-ID: <3A96F5FF.1DE715EB@bigfoot.de>

Ron Scinta wrote:
> 
> > >  This does not work alas....
> > >  >>> mymod = "ptw"
> >
> > >>> exec('import ' + mymod)
> >
> > ds
> 
> OK, but how then do I reference the dynamic module by name?

In Python 2.0, you can use the "import ... as ..." construct.

for example:

modules2test = [ 'module1', 'module2' ]

for mymodule in modules2test:
    exec( 'import ' + mymodule + ' as test_module' )
    test_module.test()


Gerhard
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From aleaxit at yahoo.com  Sat Feb 24 10:31:40 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 24 Feb 2001 10:31:40 +0100
Subject: Newbie question.. how to load modules with a dynamic name?
References: <D7zl6.221$Gm2.5547@skycache.prestige.net> <uuAl6.125974$Ch.23398786@newsrump.sjc.telocity.net> <FUzl6.227$Gm2.5865@skycache.prestige.net>
Message-ID: <977v830vmc@news1.newsguy.com>

"Ron Scinta" <nb0klun71q5001 at sneakemail.com> wrote in message
news:FUzl6.227$Gm2.5865 at skycache.prestige.net...
> > >  This does not work alas....
> > >  >>> mymod = "ptw"
> >
> > >>> exec('import ' + mymod)
> >
> > ds
>
> OK, but how then do I reference the dynamic module by name?
>
> mymod. (The period does not help here, still thinks I am using the string
> class instead of my imported module)....
>
>  Is there a way to alias the module name at this point?

sys.modules[mymod] will refer to the module object (you'll need to
import sys, of course).  You can assign that to whatever name you
like, e.g.,

    floop = sys.modules[mymod]
    floop.pleep('praak')


Alex





From fredrik at pythonware.com  Fri Feb 23 22:27:16 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 23 Feb 2001 21:27:16 GMT
Subject: Newbie question.. how to load modules with a dynamic name?
References: <D7zl6.221$Gm2.5547@skycache.prestige.net>
Message-ID: <UAAl6.15213$Qb7.2447299@newsb.telia.net>

Ron Scinta wrote:
> I have a need to run various python script files from a master control
>  script, however, I do not know the names of the scripts until runtime.
>
>  I will be getting the script names from a database.. and was thinking I
>  could just use the import call with the supplied name.
>
>  This does not work alas....
>  >>> mymod = "ptw"
>  >>> import mymod
>  Traceback (innermost last):
>    File "<interactive input>", line 1, in ?
>  ImportError: No module named mym
>  >>>
>
>  What is the prefered method of doing this type of operation?

how about:

    execfile("mymod.py")

or:

    mymod = __import__("mymod")

see the library reference (builtin functions) for more info.

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->





From wjones at usachoice.net  Fri Feb 23 21:40:14 2001
From: wjones at usachoice.net (Wyn Jones)
Date: Fri, 23 Feb 2001 15:40:14 -0500
Subject: take me off this god damn list
Message-ID: <3A96CAAE.4040304@usachoice.net>

I DO NOT WANT ANY PART OF THIS BULL- SHIT!!!!!!!!!!!!!!!!!!!




From tim at vegeta.ath.cx  Sat Feb 24 07:29:50 2001
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Sat, 24 Feb 2001 06:29:50 GMT
Subject: take me off this god damn list
References: <mailman.982960807.7110.python-list@python.org>
Message-ID: <slrn99ell6.8vu.tim@vegeta.ath.cx>

Wyn Jones <wjones at usachoice.net> wrote:
> I DO NOT WANT ANY PART OF THIS BULL- SHIT!!!!!!!!!!!!!!!!!!!

Tell ya what: instead of cussing, why don't you take a look at your own
message headers, which in my newsreader (slrn) include:


NNTP-Posting-Host: mail.python.org
{ should've been your first clue }

User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; m18) Gecko/2000...
{ explains a lot...at least about the poster }

List-Help: <mailto:python-list-request at python.org?subject=help>
List-Post: <mailto:python-list at python.org>
List-Subscribe: <http://mail.python.org/mailman/listinfo/python-list>...
List-Id: General discussion list for the Python programming language...
List-Unsubscribe: <http://mail.python.org/mailman/listinfo/python-list>
List-Archive: <http://mail.python.org/pipermail/python-list/>

...and this would be the bulk of your help right here.
I'm gonna guess that either you (or possibly someone else) went to
mail.python.org/mailman and signed up to have newsgroup posts sent to
your email.  With me so far?

Now, I've never used Gecko, nor did I find anything from Netscape to be
stable on my Win32 box, but shouldn't it be able to display complete
message headers?

Here's that URL one more time, in case your mail client is bad at
finding URLs...
	http://mail.python.org/mailman/listinfo/python-list

And please watch your temper.  One of my favorite things about this
group is that it's not comp.lang.perl.misc where any response you
would've gotten would be either far _more_ sarcastic, a furious rage, or
just plain ignored.

Have a nice day.

-- 
-Tim Hammerquist <timmy at cpan.org>

God made the Idiot for practice, and then He made the School Board.
	-- Mark Twain


From kc5tja at garnet.armored.net  Mon Feb 26 20:07:18 2001
From: kc5tja at garnet.armored.net (Samuel A. Falvo II)
Date: 26 Feb 2001 11:07:18 -0800
Subject: take me off this god damn list
References: <mailman.982960807.7110.python-list@python.org> <slrn99ell6.8vu.tim@vegeta.ath.cx>
Message-ID: <slrn99la75.ac8.kc5tja@garnet.armored.net>

On Sat, 24 Feb 2001 06:29:50 GMT, Tim Hammerquist wrote:
>And please watch your temper.  One of my favorite things about this
>group is that it's not comp.lang.perl.misc where any response you
>would've gotten would be either far _more_ sarcastic, a furious rage, or
>just plain ignored.

No, that title would have to go to comp.lang.lisp...

-- 
KC5TJA/6, DM13, QRP-L #1447 | Official Channel Saint, *Team Amiga*
Samuel A. Falvo II	    |
Oceanside, CA		    |


From slaurent at sonicfoundry.com  Fri Feb 23 21:54:41 2001
From: slaurent at sonicfoundry.com (Sean Laurent)
Date: Fri, 23 Feb 2001 14:54:41 -0600
Subject: win32com and makepy
Message-ID: <88id9tgj54l5q8rgvtss4tc9u9k91kd0kn@4ax.com>

Howdy  All,
I'm having a strange problem with win32com.client.Dispatch().  I used
win32com\client\makepy.py to successfully generate Python sources.
Unfortunately, after I call win32com.client.Dispatch(), the object I
get back is of type "COMObject" instead of one of the generated
classes.

In particular, I'm trying to interact with Microsoft Windows Installer
Service (known as MSI) - the typelibrary is named "Microsoft Windows
Installer Object Library" and the PROGID is
"WindowsInstaller.Installer".

Using win32com\client\makepy.py:
makepy -i "Microsoft Windows Installer Object Library"
Microsoft Windows Installer Object Library
{000C1092-0000-0000-C000-000000000046}, lcid=1033, major=1, minor=0
>>> # Use these commands in Python code to auto generate .py support
>>> from win32com.client import gencache
>>>gencache.EnsureModule('{000C1092-0000-0000-C000-000000000046}',1033, 1, 0)

Looking in win32com\gen-py, the appropriate file _does_ exist:
000C1092-0000-0000-C000-000000000046x1033x1x0.py
and it does contain the correct class definitions.

Then, in my code:
>>> i = win32com.client.Dispatch("WindowsInstaller.Installer")
>>> i
<COMObject WindowsInstaller.Installer>
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Voila!  There's the problem.  According to the docs, this really
should be something more like:
<win32com.gen_py.Microsoft Windows Installer.Installer>

Furthermore, if I call a method which should create another object,
things get even worse:
>>> db = i.OpenDatabase("d:\\temp\\test.msi", 0)
>>> db
<COMObject <unknown>>
^^^^^^^^^^^^^^^^^^^^^
Again, this should be something more like:
<win32com.gen_py.Microsoft Windows Installer.Database>


However, if I try this with Microsoft Word, everything works great!
makepy -i "Microsoft Word 9.0 Object Library"
Microsoft Word 9.0 Object Library
{00020905-0000-0000-C000-000000000046}, lcid=0, major=8, minor=1
>>> # Use these commands in Python code to auto generate .py support
>>> from win32com.client import gencache
>>> gencache.EnsureModule('{00020905-0000-0000-C000-000000000046}', 0, 8, 1)

>>> w = win32com.client.Dispatch("Word.Application")
>>> w
<win32com.gen_py.Microsoft Word 9.0 Object Library._Application>


*arrrgghhh*

Any bright ideas as to what's going wrong here?

----------
Sean Laurent
Engineer -Install
Sonic Foundry, Inc.
slaurentNOSPAM at sonicfoundryNOSPAM.com
----------


From MarkH at ActiveState.com  Sat Feb 24 02:44:27 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 24 Feb 2001 01:44:27 GMT
Subject: win32com and makepy
References: <88id9tgj54l5q8rgvtss4tc9u9k91kd0kn@4ax.com>
Message-ID: <3A97111F.9050705@ActiveState.com>

Sean Laurent wrote:

> Howdy  All,
> I'm having a strange problem with win32com.client.Dispatch().  I used
> win32com\client\makepy.py to successfully generate Python sources.
> Unfortunately, after I call win32com.client.Dispatch(), the object I
> get back is of type "COMObject" instead of one of the generated
> classes.

This comes up fairly regularly, and unfortunately it depends on the object in question.

The simplest work around is to use the classes in that generated module directly.  Eg:


mod = gencache.EnsureModule('{000C1092-0000-0000-C000-000000000046}',1033, 1, 0)

i = mod.WhateverTheClassNameIs()

And to make life more painful:

>>>> db = i.OpenDatabase("d:\\temp\\test.msi", 0)

Will need to done like:

db = i.OpenDatabase("d:\\temp\\test.msi", 0)
db = mod.Database(db) # Assuming "Database" is the class name

Mark.



From slaurent at sonicfoundry.com  Sat Feb 24 04:49:25 2001
From: slaurent at sonicfoundry.com (Sean Laurent)
Date: Fri, 23 Feb 2001 21:49:25 -0600
Subject: win32com and makepy
References: <88id9tgj54l5q8rgvtss4tc9u9k91kd0kn@4ax.com> <3A97111F.9050705@ActiveState.com>
Message-ID: <1r8e9tc3jn5pi734ldi4r3ikr3bhnudvgb@4ax.com>

On Sat, 24 Feb 2001 01:44:27 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>> I'm having a strange problem with win32com.client.Dispatch().  I used
>> win32com\client\makepy.py to successfully generate Python sources.
>> Unfortunately, after I call win32com.client.Dispatch(), the object I
>> get back is of type "COMObject" instead of one of the generated
>> classes.
>
>This comes up fairly regularly, and unfortunately it depends on the object in question.
>
>The simplest work around is to use the classes in that generated module directly.  Eg:
>
>mod = gencache.EnsureModule('{000C1092-0000-0000-C000-000000000046}',1033, 1, 0)
>
>i = mod.WhateverTheClassNameIs()
>
>And to make life more painful:
>
>>>>> db = i.OpenDatabase("d:\\temp\\test.msi", 0)
>
>Will need to done like:
>
>db = i.OpenDatabase("d:\\temp\\test.msi", 0)
>db = mod.Database(db) # Assuming "Database" is the class name


Ah hah!  *light bulb goes off*

Interestingly enough, I tried:
     >>>> i = mod.Installer()
then I tried:
     >>>> db = i.OpenDatabase( ... )
At this point, "db" _is_ the correct type - no need to do:
     >>>> db = mod.Database(db)

All things considered, this isn't any more painful than calling
client.Dispatch().

I finally took a look at your "Advanced Python and COM" presentation
from the 8th International Python Conference... and it seems that the
following is also a viable alternative:
     >>>> gencache.EnsureModule(blahblahblah)
     >>>> klass = gencache.GetClassForProgID(blah)
     >>>> i = klass()
     >>>> db = i.OpenDatabase(blahblah)

The only advantage to that technique (in my mind), is that it is
indepedent of the actual class name in the cached module.  In reality,
I suppose there isn't any need to be independent of the class name.
If Microsoft (in their infinite wisdom) changes the structure of the
type library such that the class name changes... they would probably
also change enough other stuff that the whole script would have to be
reworked anyway.


As a side note, I'm still puzzled as to why the original code didn't
work:
     >>>> gencache.EnsureModule(blahblah)
     >>>> i = win32com.client.Dispatch("WindowsInstaller.Installer")

I spent some time stepping through the code and I found that the code
was bailing in win32com.client.dynamic._GetGoodDispatch() at the line
reading:
     IDispatch = pythoncom.connect(IDispatch)

     com_error: (-2147221021, 'Operation unavailable', None, None)

So, I decided to spoonfeed the call to Dispatch and hand it the
classID of the interface in question.  I cut and paste the ClassID
directly from the script generated by makepy.  With this, I followed
the debugger to __WrapDispatch() in win32com.client.__init__, where it
called:
     klass = gencache.GetClassForCLSID(resultCLSID)

Looking at gencache.GetClassForCLSID(), it calls
CLSIDToClass.HasClass().  And this is where things suddenly became
interesting.  Through the debugger I looked at the entries in the
mapCLSIDToClass dictionary - indeed there were a _number_ of entries,
but not the CLSID I needed.

At this point, I sat down with OLEVIEW and I discovered something
interesting.  The generated code set a CLSID of:

{000C1092-0000-0000-C000-000000000046}

While this is the ID of the type library, it is neither the CLSIDnor
the AppID (which carry the same value) for the actual "Microsoft
Windows Installer" object.  Unfortunately, this turned out to be a
dead end - I tried changing the ClassID in the generated file, but
this didn't help.  (As you may have already guessed!)


The other bizarre thing I found was in __WrapDispatch() in
win32com.client.__init__.  In this function, you find the following
code:
     if resultCLSID is None:
          try:
               typeinfo = dispatch.GetTypeInfo()

In this case, we don't have a CLSID, so this code gets called.  Yet it
bails with:  "com_error: (-2147467262, 'No such interface supported',
None, None)"  which seems a bit bizarre, since dispatch is of type
PyIDispatch - which should support this method...

Any thoughts?  Am I beating a dead horse?  Or what's the word?

-Sean Laurent-


From MarkH at ActiveState.com  Sat Feb 24 15:58:33 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 24 Feb 2001 14:58:33 GMT
Subject: win32com and makepy
References: <88id9tgj54l5q8rgvtss4tc9u9k91kd0kn@4ax.com> <3A97111F.9050705@ActiveState.com> <1r8e9tc3jn5pi734ldi4r3ikr3bhnudvgb@4ax.com>
Message-ID: <3A97CB3E.30204@ActiveState.com>

Sean Laurent wrote:

> As a side note, I'm still puzzled as to why the original code didn't
> work:
>      >>>> gencache.EnsureModule(blahblah)
>      >>>> i = win32com.client.Dispatch("WindowsInstaller.Installer")
> 
> I spent some time stepping through the code and I found that the code
> was bailing in win32com.client.dynamic._GetGoodDispatch() at the line
> reading:
>      IDispatch = pythoncom.connect(IDispatch)
> 
>      com_error: (-2147221021, 'Operation unavailable', None, None)

This should be fine - we then go and try the CoCreateInstance route and end up down the same path you traced.

> The other bizarre thing I found was in __WrapDispatch() in
> win32com.client.__init__.  In this function, you find the following
> code:
>      if resultCLSID is None:
>           try:
>                typeinfo = dispatch.GetTypeInfo()
> 
> In this case, we don't have a CLSID, so this code gets called.  Yet it
> bails with:  "com_error: (-2147467262, 'No such interface supported',
> None, None)"  which seems a bit bizarre, since dispatch is of type
> PyIDispatch - which should support this method...

This is the crux of the whole problem.  The object implements GetTypeInfo, but just fails to provide it.  If this was available, then win32com would be able to find the CLSID of the object, and get the makepy generated class.

> Any thoughts?  Am I beating a dead horse?

Pretty much I fear - at least until MS upgrade their interfaces.

Mark.
-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From slaurent at sonicfoundry.com  Mon Feb 26 19:32:18 2001
From: slaurent at sonicfoundry.com (Sean Laurent)
Date: Mon, 26 Feb 2001 12:32:18 -0600
Subject: win32com and makepy
References: <88id9tgj54l5q8rgvtss4tc9u9k91kd0kn@4ax.com> <3A97111F.9050705@ActiveState.com> <1r8e9tc3jn5pi734ldi4r3ikr3bhnudvgb@4ax.com> <3A97CB3E.30204@ActiveState.com>
Message-ID: <338l9tsq0ti17s22aitjdh2e19lm6m3l0a@4ax.com>

On Sat, 24 Feb 2001 14:58:33 GMT, Mark Hammond <MarkH at ActiveState.com>
wrote:

>This is the crux of the whole problem.  The object implements GetTypeInfo, but just fails
>to provide it.  If this was available, then win32com would be able to find the CLSID of the
>object, and get the makepy generated class.
>
>> Any thoughts?  Am I beating a dead horse?
>
>Pretty much I fear - at least until MS upgrade their interfaces.


D'oh!  As a developer, why do I I always find myself boxed into some
corner or forced to come up with workarounds just because Micro$oft
doesn't do something correctly?  *ugh*

Well, your kludge (described in an earlier message) is very beneficial
and solves the problem.  I greatly appreciate the help!

----------
Sean Laurent
Engineer -Install
Sonic Foundry, Inc.



From lbates at syscon-computers.com  Sat Feb 24 00:42:38 2001
From: lbates at syscon-computers.com (Larry Bates)
Date: Fri, 23 Feb 2001 17:42:38 -0600
Subject: _winreg - Windows registry access
Message-ID: <3a96f575$1@news.dbtech.net>

I'm a newbie, so please forgive me.  Does anyone have any example code on
how to use the new _winreg class?  I've tried everything that makes sense,
but
nothing seems to work.  Examples sure go a long way when you are learning.

Thanks in advance for any assistance.

-Larry





From MarkH at ActiveState.com  Sat Feb 24 15:45:42 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 24 Feb 2001 14:45:42 GMT
Subject: _winreg - Windows registry access
References: <3a96f575$1@news.dbtech.net>
Message-ID: <3A97C83A.2050807@ActiveState.com>

Larry Bates wrote:

> I'm a newbie, so please forgive me.  Does anyone have any example code on
> how to use the new _winreg class?  I've tried everything that makes sense,
> but
> nothing seems to work.  Examples sure go a long way when you are learning.

Here is a really simple one:

>>> import _winreg, sys
>>> key = _winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, "Software\\Python\\PythonCore\\" + sys.winver + "\\PythonPath")
>>> _winreg.QueryValue(key, "")
'f:\\src\\python-cvs\\lib;f:\\src\\python-cvs\\pcbuild'
>>> _winreg.EnumKey(key, 0)
'Pythonwin'
>>> _winreg.EnumKey(key, 1)
'win32'
>>>

This shows the default PythonPath for my current version, and shows that I have at least 2 keys under this, called "Pythonwin" and "win32".  It doesnt show the values for these 2 subkeys.

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From jcollins at endeavors.com  Sat Feb 24 01:29:48 2001
From: jcollins at endeavors.com (Jeff Collins)
Date: Fri, 23 Feb 2001 16:29:48 -0800
Subject: [ANN] Pippy: Python for the Palm
Message-ID: <20010223162948.B9773@endeavors.com>

We are pleased to announce the 0.6b release of Pippy, a port of Python to
the Palm.  Those interested in creating Python-based applications for Palm
handheld computers and other devices that use the Palm operating system will
find this to be an excellent VM.  For more information, please see
http://www.endeavors.com/pippy.


--
Jeffery D. Collins, Ph.D.
Sr. Software Developer
Endeavors Technology, Inc.
http://www.endeavors.com



From noahnoah.org  Sat Feb 24 02:09:48 2001
From: noahnoah.org (Noah Spurrier (a))
Date: 24 Feb 2001 01:09:48 GMT
Subject: Need to unread or push back bytes to a file
Message-ID: <9771ks0ic2@news2.newsguy.com>


I need to push bytes back into a file object. 
Is there a buffered file wrapper for file objects?
(Yes, I looked, but I gave up after almost five minutes of searching.)

I'm reading from a pipe, but this could be the same for a file.
I call an "expect" method which reads until a pattern is found,
but read() with a pipe will read as much data as there is available
in the pipe. When I have found a pattern match I want to return 
the stream with the extra character pushed-back.

Another way to word this is that I want to the of a pattern, but
not past the pattern.

My only other option would be to read(1) character at a time
and then search for the pattern. This will work, but I think a
push-back would be more elegant and efficient.

Yours,
Noah

==================================
Posted via http://nodevice.com
Linux Programmer's Site


From aleaxit at yahoo.com  Sat Feb 24 10:25:27 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 24 Feb 2001 10:25:27 +0100
Subject: Need to unread or push back bytes to a file
References: <9771ks0ic2@news2.newsguy.com>
Message-ID: <977ukk0uvd@news1.newsguy.com>

"Noah Spurrier" <noah(a)noah.org> wrote in message
news:9771ks0ic2 at news2.newsguy.com...
>
>
> I need to push bytes back into a file object.
> Is there a buffered file wrapper for file objects?
> (Yes, I looked, but I gave up after almost five minutes of searching.)
>
> I'm reading from a pipe, but this could be the same for a file.
> I call an "expect" method which reads until a pattern is found,
> but read() with a pipe will read as much data as there is available
> in the pipe. When I have found a pattern match I want to return
> the stream with the extra character pushed-back.

what about something like:

class NSWrapper:
    def __init__(self, fileob):
        self.fileob = fileob
        self.buffer = ''
    def close(self):
        self.fileob.close()
        self.buffer = ''
    def push_back(self, piece):
        self.buffer = piece + self.buffer
    def readall(self):
        result = self.buffer + self.fileob.read()
        self.buffer = ''
        return result
    def read(self, N=None):
        if N is None: return self.readall()
        avail = len(self.buffer)
        if N>avail:
            result = self.buffer+file.read(N-avail)
            self.buffer = ''
        else:
            result = self.buffer[:N]
            self.buffer = self.buffer[N:]
        return result

needs testing, but this should be roughly what you want...?


Alex





From noahnoah.org  Sat Feb 24 21:03:34 2001
From: noahnoah.org (Noah Spurrier (a))
Date: 24 Feb 2001 20:03:34 GMT
Subject: Need to unread or push back bytes to a file
References: <977ukk0uvd@news1.newsguy.com>
Message-ID: <97942m02njh@news1.newsguy.com>

Yeah, this looks good. I was hoping there was a way to do it
using some included library.

Also, this shows why file should be a first class object!
This class would be so much easier if we could just inherit from file.
I want to make sure that my new object can be used anywhere a
file is normally expected. As it currently stands, I will have to
provide wrapper implementations for every method that file supports;
about 12 methods in all.

I don't know how to wrap the file objects read-only attributes.
Does seek() and tell() make sense in this type of file?
At the end I added an expanded version of the class wrapper.

Yours,
Noah

class NSWrapper:
	# I don't know how to wrap these read-only attributes:
	# closed 
	# mode 
	# name 
	# softspace 

	def __init__(self, fileob):
		self.fileob = fileob
		self.buffer = ''
	def close (self):
		self.fileob.close()
		self.buffer = ''
	def flush (self):
		self.fileob.flush()
	def isatty (self):
		return self.fileob.isatty()
	def fileno (self):
		return self.fileob.fileno()
	def push_back(self, piece):
		self.buffer = piece + self.buffer
	def readall(self):
		result = self.buffer + self.fileob.read()
		self.buffer = ''
		return result
	def read (self, size=-1):
		if size <= -1: 
			return self.readall()
		avail = len(self.buffer)
		if size > avail:
			result = self.buffer+file.read(size-avail)
			self.buffer = ''
		else:
			result = self.buffer[:size]
			self.buffer = self.buffer[size:]
		return result
	def readline (self, size=-1):
		return self.fileob.readline(size)
	def readlines (self, sizehint=-1):
		return self.fileob.readlines (sizehint)
	#!!! Some methods don't make sense on a stream.
	#def seek (self, offset, whence=0):
	#	self.fileob.seek (offset, whence)
	#def tell (self):
	#	return self.fileob.tell()
	#def truncate (self, size=-1):
	#	self.fileob.truncate(size)
	def write (self, str):
		self.fileob.write(str)
	def writelines (self, list):
		self.fileob.writelines(list)



> "Noah Spurrier" <noah(a)noah.org> wrote in message
> news:9771ks0ic2 at news2.newsguy.com...
> >
> >
> > I need to push bytes back into a file object.
> > Is there a buffered file wrapper for file objects?
> >  ...
> what about something like:
> 
> class NSWrapper:
>     def __init__(self, fileob):
>  ...
> needs testing, but this should be roughly what you want...?

> Alex



==================================
Posted via http://nodevice.com
Linux Programmer's Site


From duke at radix.net  Sat Feb 24 04:35:24 2001
From: duke at radix.net (Pastor Duke)
Date: Fri, 23 Feb 2001 22:35:24 -0500
Subject: where to learn python programing
Message-ID: <9778so$725$1@news1.Radix.Net>

Greetings,
I'm very interested in learning Python programming as I'm a 3D graphics
designer, and want to develop some games.

Thanks,
Duke




From parkw at better.net  Sat Feb 24 04:34:30 2001
From: parkw at better.net (William Park)
Date: Fri, 23 Feb 2001 22:34:30 -0500
Subject: where to learn python programing
In-Reply-To: <9778so$725$1@news1.Radix.Net>; from duke@radix.net on Fri, Feb 23, 2001 at 10:35:24PM -0500
References: <9778so$725$1@news1.Radix.Net>
Message-ID: <20010223223430.A599@better.net>

On Fri, Feb 23, 2001 at 10:35:24PM -0500, Pastor Duke wrote:
> Greetings,
> I'm very interested in learning Python programming as I'm a 3D graphics
> designer, and want to develop some games.

Standard answer is Tutorial which is part of standard docs.

---William Park, Open Geometry Consulting



From wware at world.std.com  Sat Feb 24 13:23:09 2001
From: wware at world.std.com (Will Ware)
Date: Sat, 24 Feb 2001 12:23:09 GMT
Subject: where to learn python programing
References: <9778so$725$1@news1.Radix.Net> <mailman.982985707.26846.python-list@python.org>
Message-ID: <G99H2L.A6s@world.std.com>

On Fri, Feb 23, 2001 at 10:35:24PM -0500, Pastor Duke wrote:
> I'm very interested in learning Python programming as I'm a 3D graphics
> designer, and want to develop some games.

William Park (parkw at better.net) wrote:
> Standard answer is Tutorial which is part of standard docs.

You can find this on-line at:
http://www.python.org/doc/current/tut/tut.html

Other useful stuff here, including the Library Reference:
http://www.python.org/doc/current/

-- 
-----------------------------------+---------------------
 22nd century: Esperanto, geodesic | Will Ware
 domes, hovercrafts, metric system | wware at world.std.com


From bencr at bigpond.com  Sat Feb 24 14:41:25 2001
From: bencr at bigpond.com (Ben)
Date: Sun, 25 Feb 2001 00:41:25 +1100
Subject: where to learn python programing
References: <9778so$725$1@news1.Radix.Net> <mailman.982985707.26846.python-list@python.org>
Message-ID: <kBNl6.2527$v5.8078@newsfeeds.bigpond.com>

Try www.andamooka.org for a intro to Programming in Python ...

Without being aware of your intentions ... are you sure that Python is a
suitable programming language for implementing 3D games???  ...  you should
also take a look at a couple of existing engines (quake 3, Unreal Tournament
and Lithtech to name the most common) ... as some of these allow a high
level API without needing to invest in a licensing agreement ....



"William Park" <parkw at better.net> wrote in message
news:mailman.982985707.26846.python-list at python.org...
> On Fri, Feb 23, 2001 at 10:35:24PM -0500, Pastor Duke wrote:
> > Greetings,
> > I'm very interested in learning Python programming as I'm a 3D graphics
> > designer, and want to develop some games.
>
> Standard answer is Tutorial which is part of standard docs.
>
> ---William Park, Open Geometry Consulting
>




From al053162 at anubis.uji.es  Sat Feb 24 14:07:13 2001
From: al053162 at anubis.uji.es (LANCELOT)
Date: Sat, 24 Feb 2001 14:07:13 +0100
Subject: =?ISO-8859-1?Q?Espa=F1ol_=2F_Spanish?=
In-Reply-To: <kBNl6.2527$v5.8078@newsfeeds.bigpond.com>
References: <9778so$725$1@news1.Radix.Net> <mailman.982985707.26846.python-list@python.org> <kBNl6.2527$v5.8078@newsfeeds.bigpond.com>
Message-ID: <Pine.A41.4.31.0102241402390.26180-100000@anubis.uji.es>

Este es un mensaje en espa?ol / This is a message in Spanish:

Buenas,

Acabo de conocer este grupo de noticias, que no se de donde es, pero me
he dado cuenta de que todos hablais en ingles. Hay alguien que hable
espa?ol? Me gustaria poder hablar de Python con alguien que hable en
espa?ol, porque necesito informacion y la busco en espa?ol. Gracias.

Saludos.

PD: Does anybody know anything of Python in Spanish? Thank you.





From alan.gauld at gssec.bt.co.uk  Mon Feb 26 17:47:46 2001
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Mon, 26 Feb 2001 16:47:46 +0000
Subject: =?iso-8859-1?Q?Espa=F1ol?= / Spanish
References: <9778so$725$1@news1.Radix.Net> <mailman.982985707.26846.python-list@python.org> <kBNl6.2527$v5.8078@newsfeeds.bigpond.com> <Pine.A41.4.31.0102241402390.26180-100000@anubis.uji.es>
Message-ID: <3A9A88B2.D9316AFC@gssec.bt.co.uk>


LANCELOT wrote:
> Acabo de conocer este grupo de noticias, que no se de donde es, pero me
> he dado cuenta de que todos hablais en ingles. Hay alguien que hable
> espa?ol? Me gustaria poder hablar de Python con alguien que hable en
> espa?ol, porque necesito informacion y la busco en espa?ol. Gracias.
> 
> Saludos.
> 
> PD: Does anybody know anything of Python in Spanish? Thank you.

Absolute beginners tutorial:

http://www.crosswinds.net/~agauld/spanish

Also in progress in Portuguese and German....
http://www.crosswinds.net/~agauld/port
http://www.crosswinds.net/~agauld/german

Alan G.


From piet at cs.uu.nl  Mon Feb 26 20:13:36 2001
From: piet at cs.uu.nl (Piet van Oostrum)
Date: 26 Feb 2001 20:13:36 +0100
Subject: =?iso-8859-1?q?Espa=F1ol?= / Spanish
References: <9778so$725$1@news1.Radix.Net> <mailman.982985707.26846.python-list@python.org> <kBNl6.2527$v5.8078@newsfeeds.bigpond.com> <Pine.A41.4.31.0102241402390.26180-100000@anubis.uji.es> <3A9A88B2.D9316AFC@gssec.bt.co.uk>
Message-ID: <wzitlxtfz3.fsf@sunshine.cs.uu.nl>

>>>>> Alan Gauld <alan.gauld at gssec.bt.co.uk> (AG) writes:

AG> LANCELOT wrote:
>> Acabo de conocer este grupo de noticias, que no se de donde es, pero me
>> he dado cuenta de que todos hablais en ingles. Hay alguien que hable
>> espa?ol? Me gustaria poder hablar de Python con alguien que hable en
>> espa?ol, porque necesito informacion y la busco en espa?ol. Gracias.
>> 
>> Saludos.
>> 
>> PD: Does anybody know anything of Python in Spanish? Thank you.

AG> Absolute beginners tutorial:

AG> http://www.crosswinds.net/~agauld/spanish

AG> Also in progress in Portuguese and German....
AG> http://www.crosswinds.net/~agauld/port
AG> http://www.crosswinds.net/~agauld/german

THE Python tutorial is also available in Spanish:
http://www.zope.org/Members/rapto/tut

But he wanted to discuss about python in Spanish.

-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From mfletch at tpresence.com  Sat Feb 24 05:14:33 2001
From: mfletch at tpresence.com (Mike Fletcher)
Date: Fri, 23 Feb 2001 23:14:33 -0500
Subject: Unix's Zoo of Languages (Eric Raymond)
Message-ID: <CC59868A0A76D311B1F50090277B321C1E14BE@VRTMAIL>

http://www.tuxedo.org/~esr/writings/taoup/chapter3.html

Just thought some of those who like jousting in the lists of language
tournaments might find that chapter interesting.  I'd have liked some deeper
coverage of Ruby (which I rather like the look of as I read the new book,
but feel no desire to learn when I actually sit down to it) and Squeak, but
overall it's a fairly balanced feeling overview of the languages available
to the Unix person.  Would have been nice to include something of IDE
comparisons (GUI development and the like), but oh well.

The PSU's bribe money is obviously well spent ;) , Sun should learn
something from them :) .

Enjoy yourselves,
Mike
__________________________________
 Mike C. Fletcher
 Designer, VR Plumber
 http://members.home.com/mcfletch



From thomas at xs4all.net  Sat Feb 24 05:20:46 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sat, 24 Feb 2001 05:20:46 +0100
Subject: Unix's Zoo of Languages (Eric Raymond)
In-Reply-To: <CC59868A0A76D311B1F50090277B321C1E14BE@VRTMAIL>; from mfletch@tpresence.com on Fri, Feb 23, 2001 at 11:14:33PM -0500
References: <CC59868A0A76D311B1F50090277B321C1E14BE@VRTMAIL>
Message-ID: <20010224052046.G16781@xs4all.nl>

On Fri, Feb 23, 2001 at 11:14:33PM -0500, Mike Fletcher wrote:
> http://www.tuxedo.org/~esr/writings/taoup/chapter3.html

> The PSU's bribe money is obviously well spent ;) , Sun should learn
> something from them :) .

Well, considering Eric is a valued member of the python-dev team, I'm not
sure if that organization was actually involved. I hear they screen quite
vigorously before you're allowed in python-dev, not to mention the screening
and full body cavity search you get when being considered for bribing by the
P




From mfletch at tpresence.com  Sun Feb 25 00:54:41 2001
From: mfletch at tpresence.com (Mike Fletcher)
Date: Sat, 24 Feb 2001 18:54:41 -0500
Subject: Unix's Zoo of Languages (Eric Raymond)
Message-ID: <CC59868A0A76D311B1F50090277B321C1E14BF@VRTMAIL>

Drat it!  How many times do I have to tell them, it's no use if you just buy
them outright, take them into the inner chambers and tell the world they're
there (it's not like anyone would know, after all, who really knows what
happens behind closed doors).  It wrecks their credibility as a hostile
witness.  They have to appear unbiased and free to make their own decisions.

When we take Torvalds and Gates** we'll need to make sure they don't get
seen hanging out in the Cabal's meeting rooms (-dev) so we can use them to
sink the other languages.

Though, come to think of it, Python in the kernal would be nice...
Mike

** Yes, Gates, we need to take over Windows too.

-----Original Message-----
From: Thomas Wouters [mailto:thomas at xs4all.net]
...
Well, considering Eric is a valued member of the python-dev team, I'm not
sure if that organization was actually involved. I hear they screen quite
vigorously before you're allowed in python-dev, not to mention the screening
and full body cavity search you get when being considered for bribing by the
P
...



From vasvir at iit.demokritos.gr  Sat Feb 24 06:54:56 2001
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Sat, 24 Feb 2001 07:54:56 +0200
Subject: abusing __builtins__
Message-ID: <3A974CB0.EA7B0356@iit.demokritos.gr>

In python 1.5.2 (and possible to any future python) is possible to set a
global variable shared by all modules simply by doing:

setattr(__builtins__, 'GlobalVariable', GlobalVariable)

without the need to import anything.

Not being an experienced python programmer I would like ot hear some
arguments on whether this is considered abuse of the language or not.
Should it be avoided?  Are there any hidden pitfalls? except from the
obvious one that I should not have global variables in the first place.

Thanks in advance...
	.Bill


From nas at arctrix.com  Sat Feb 24 07:17:42 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Fri, 23 Feb 2001 22:17:42 -0800
Subject: abusing __builtins__
In-Reply-To: <3A974CB0.EA7B0356@iit.demokritos.gr>; from vasvir@iit.demokritos.gr on Sat, Feb 24, 2001 at 07:54:56AM +0200
References: <3A974CB0.EA7B0356@iit.demokritos.gr>
Message-ID: <20010223221742.A24805@glacier.fnational.com>

On Sat, Feb 24, 2001 at 07:54:56AM +0200, Vassilis Virvilis wrote:
> setattr(__builtins__, 'GlobalVariable', GlobalVariable)

So simple yet so evil.  You should win a prize.

  Neil



From tim.one at home.com  Mon Feb 26 00:08:35 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 25 Feb 2001 18:08:35 -0500
Subject: abusing __builtins__
In-Reply-To: <3A974CB0.EA7B0356@iit.demokritos.gr>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEOJJBAA.tim.one@home.com>

[Vassilis Virvilis]
> In python 1.5.2 (and possible to any future python) is possible to set a
> global variable shared by all modules simply by doing:
>
> setattr(__builtins__, 'GlobalVariable', GlobalVariable)
>
> without the need to import anything.

Cute!

> Not being an experienced python programmer I would like ot hear some
> arguments on whether this is considered abuse of the language or not.
> Should it be avoided?  Are there any hidden pitfalls? except from the
> obvious one that I should not have global variables in the first place.

The language doesn't guarantee this will always work, so it's certainly
abuse.

Looking up builtins is slower than looking up module globals (a module's
global dict is always searched before the builtin dict).

The language may someday make "GlobalVariable" the name of its own builtin,
and then your code won't work correctly if you either try to use the new
builtin or call any other module that does (incl. without limitation std
library modules).

People other than you will be unable to understand your code regardless,
because GlobalVariable will look to them like an unbound variable (i.e.,
nobody else uses this trick, so nobody else will recognize the usage
pattern).

Introspective tools won't think to look in the builtins either (again
because nobody else etc).

i-wouldn't-let-any-of-that-stop-me-though<wink>-ly y'rs  - tim




From thomas at xs4all.net  Mon Feb 26 00:16:54 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 26 Feb 2001 00:16:54 +0100
Subject: abusing __builtins__
In-Reply-To: <LNBBLJKPBEHFEDALKOLCKEOJJBAA.tim.one@home.com>; from tim.one@home.com on Sun, Feb 25, 2001 at 06:08:35PM -0500
References: <3A974CB0.EA7B0356@iit.demokritos.gr> <LNBBLJKPBEHFEDALKOLCKEOJJBAA.tim.one@home.com>
Message-ID: <20010226001653.J16781@xs4all.nl>

On Sun, Feb 25, 2001 at 06:08:35PM -0500, Tim Peters wrote:

> People other than you will be unable to understand your code regardless,
> because GlobalVariable will look to them like an unbound variable (i.e.,
> nobody else uses this trick, so nobody else will recognize the usage
> pattern).

Actually, Mailman used this trick for a while. Not in any release, though,
and Barry removed the ugly hack a week or two ago. Not sure if it was
originally Barry's idea, or just part of the internationalization patch he
accepted.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From mfletch at tpresence.com  Mon Feb 26 02:47:27 2001
From: mfletch at tpresence.com (Mike Fletcher)
Date: Sun, 25 Feb 2001 20:47:27 -0500
Subject: abusing __builtins__
Message-ID: <CC59868A0A76D311B1F50090277B321C1F37CD@VRTMAIL>

I've used this extensively in one (very large) in-house demo I created, I
used builtins.APPLICATION to refer to a global application object.  Was very
simple and useful.  Yes, a hack, but I considered it a rather nice hack.
Apparently this is another of those "you're wrong, Mike, just plain wrong"
things :) .

I seem to get more and more of those these days,
Mike

-----Original Message-----
From: Thomas Wouters [mailto:thomas at xs4all.net]
Sent: Sunday, February 25, 2001 6:17 PM
To: python-list at python.org
Subject: Re: abusing __builtins__


On Sun, Feb 25, 2001 at 06:08:35PM -0500, Tim Peters wrote:

> People other than you will be unable to understand your code regardless,
> because GlobalVariable will look to them like an unbound variable (i.e.,
> nobody else uses this trick, so nobody else will recognize the usage
> pattern).

Actually, Mailman used this trick for a while. Not in any release, though,
and Barry removed the ugly hack a week or two ago. Not sure if it was
originally Barry's idea, or just part of the internationalization patch he
accepted.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me
spread!

-- 
http://mail.python.org/mailman/listinfo/python-list



From root at rainerdeyke.com  Mon Feb 26 03:48:30 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Mon, 26 Feb 2001 02:48:30 GMT
Subject: abusing __builtins__
References: <mailman.983152087.5110.python-list@python.org>
Message-ID: <2ujm6.305444$ge4.105940925@news2.rdc2.tx.home.com>

"Mike Fletcher" <mfletch at tpresence.com> wrote in message
news:mailman.983152087.5110.python-list at python.org...
> I've used this extensively in one (very large) in-house demo I created, I
> used builtins.APPLICATION to refer to a global application object.  Was
very
> simple and useful.  Yes, a hack, but I considered it a rather nice hack.
> Apparently this is another of those "you're wrong, Mike, just plain wrong"
> things :) .

For the record, I think this hack has its uses, especially in non-Python
programs that embed Python as a scripting language.  I don't use it myself,
though.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From vasvir at iit.demokritos.gr  Mon Feb 26 06:50:35 2001
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Mon, 26 Feb 2001 07:50:35 +0200
Subject: abusing __builtins__
References: <mailman.983142547.10845.python-list@python.org>
Message-ID: <3A99EEAB.C756064C@iit.demokritos.gr>

Tim Peters wrote:
> Cute!
> 
> 
> The language doesn't guarantee this will always work, so it's certainly
> abuse.

Oups! That's the killer argument here.

> 
> Looking up builtins is slower than looking up module globals (a module's
> global dict is always searched before the builtin dict).

certainly

 
> The language may someday make "GlobalVariable" the name of its own builtin,
> and then your code won't work correctly if you either try to use the new
> builtin or call any other module that does (incl. without limitation std
> library modules).

This I don't get it. Why is that? GlobalVariable is just a name to
indicate the purpose of my variable. If a newer version of python use it
in a later stage as a reversed word (__builtins__ variable or function),
then obviously my old program won't be able to use the new functionality
but it will overwrite it happily. So it would work even then.

Consider the following imaginary example. In a project where there is no
need to open files one replaces the open function. So this project will
never manage to open a file but aside from that it will continue to work
as designed.

 
> People other than you will be unable to understand your code regardless,
> because GlobalVariable will look to them like an unbound variable (i.e.,
> nobody else uses this trick, so nobody else will recognize the usage
> pattern).
> 
> Introspective tools won't think to look in the builtins either (again
> because nobody else etc).
> 

Valid points too.



Thank you and everybody else for your answers. I think I will try to
avoid that trick anyway.

	.Bill


From tim.one at home.com  Mon Feb 26 07:36:47 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 26 Feb 2001 01:36:47 -0500
Subject: abusing __builtins__
In-Reply-To: <3A99EEAB.C756064C@iit.demokritos.gr>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEPGJBAA.tim.one@home.com>

[Tim]
>> The language may someday make "GlobalVariable" the name of its
>> own builtin,> and then your code won't work correctly if you
>> either try to use the new builtin or call any other module that
>> does (incl. without limitation std library modules).

[Vassilis Virvilis]
> This I don't get it. Why is that?

I think because you're reading it the opposite of the way it's written
<wink>:  I'm not saying you can't get at *your* meaning of the name anymore,
but that the *new* meaning gets masked (because you've overwritten it).

> ...
> Consider the following imaginary example. In a project where there is no
> need to open files one replaces the open function. So this project will
> never manage to open a file but aside from that it will continue to work
> as designed.

Very dangerous.  It's virtually impossible to write a substantial Python
program without invoking at least some of the std libraries.  And you simply
can't know whether *those* will use "open" across releases.  If they do,
they certainly aren't expecting to get *your* meaning of the name, they're
expecting the builtin meaning.

There are uses for that, though.  For example, on rare occasions and for
debugging purposes, I've replaced the builtin open with my own version that
logs the name of the file being opened along with a timestamp.  In that
case, it's "a feature" that the std libraries also see my redefined version.
That was a lifesaver at least once, and it's way cool that Python let me get
away with it.  It's never meant to be a permanent thing, though.

That's OK by Python:  it generally supports all the sickness you can inflict
on yourself, provided only that you agree in advance that you're in a state
of sin and won't complain if you're later condemned for it <wink>.

guido-makes-a-nice-living-selling-indulgences-on-the-side-ly y'rs  - tim




From thomas at xs4all.net  Mon Feb 26 07:56:31 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 26 Feb 2001 07:56:31 +0100
Subject: abusing __builtins__
In-Reply-To: <LNBBLJKPBEHFEDALKOLCKEPGJBAA.tim.one@home.com>; from tim.one@home.com on Mon, Feb 26, 2001 at 01:36:47AM -0500
References: <3A99EEAB.C756064C@iit.demokritos.gr> <LNBBLJKPBEHFEDALKOLCKEPGJBAA.tim.one@home.com>
Message-ID: <20010226075631.L16781@xs4all.nl>

On Mon, Feb 26, 2001 at 01:36:47AM -0500, Tim Peters wrote:

> guido-makes-a-nice-living-selling-indulgences-on-the-side-ly y'rs  - tim

Actually, I thought he just gave them away for free. Free sampling, and all
that. And then when he takes them away, you have to get him (or one of his
cowboys <wink>) to fix it -- and that's the part that costs! :) 

I know it cost me two keyboards the past month <0.9 wink>.

Just-say-no-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From jkraska1 at san.rr.com  Mon Feb 26 07:48:08 2001
From: jkraska1 at san.rr.com (Courageous)
Date: Mon, 26 Feb 2001 06:48:08 GMT
Subject: abusing __builtins__
References: <mailman.983142547.10845.python-list@python.org> <3A99EEAB.C756064C@iit.demokritos.gr>
Message-ID: <bbvj9ts78mki9fcb5pa3ale4uij64bk3ih@4ax.com>

>> The language doesn't guarantee this will always work, so it's certainly
>> abuse.
>
>Oups! That's the killer argument here.

Well, not to worry. It works now, as you've noticed. :)

C//



From thomas at xs4all.net  Mon Feb 26 07:53:22 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 26 Feb 2001 07:53:22 +0100
Subject: abusing __builtins__
In-Reply-To: <3A99EEAB.C756064C@iit.demokritos.gr>; from vasvir@iit.demokritos.gr on Mon, Feb 26, 2001 at 07:50:35AM +0200
References: <mailman.983142547.10845.python-list@python.org> <3A99EEAB.C756064C@iit.demokritos.gr>
Message-ID: <20010226075322.K16781@xs4all.nl>

On Mon, Feb 26, 2001 at 07:50:35AM +0200, Vassilis Virvilis wrote:

> > The language may someday make "GlobalVariable" the name of its own builtin,
> > and then your code won't work correctly if you either try to use the new
> > builtin or call any other module that does (incl. without limitation std
> > library modules).

> This I don't get it. Why is that? GlobalVariable is just a name to
> indicate the purpose of my variable. If a newer version of python use it
> in a later stage as a reversed word (__builtins__ variable or function),
> then obviously my old program won't be able to use the new functionality
> but it will overwrite it happily. So it would work even then.

First off, entries in builtin aren't "reserved words" as such. This is quite
a subtle difference, but an important one none the less. Reserved words are
very limited in Python, and I believe none were added since Python 1.0, when
'lambda' was put in at the very last moment. In fact, in Python 1.4, a
formerly reserved word ('access') was removed ! :) Reserved words are
reserved specifically for statements and special syntax. Reserved words are
part of the grammar of the language, and can't be used as variable names or
(in CPython) other places that need a name (like def, class, attribute
setting/getting, etc.) This is why (in CPython) you can't even have a method
called 'print', 'def' or 'class'. (If you want to know why: it's a flaw in
the parser/tokenizer. Feel free to fix it and submit a patch!)

Builtin functions and variables are a different story, however. They are
just conveniency names, are searched as a 'last resort', and are perfectly
permissible as variable names wherever you want them. If you use a name that
is defined in the builtins namespace as a local or global variable, nothing
weird happens. You lose access to the builtin from that function or module
(respectively), but since you wrote the code yourself, you are probably not
using it anyway. Other functions and modules don't notice or care that
you're shadowing the builtin function in your function or module.

(The masking of globals is one of the reasons why 'from module import *' is
generally a bad idea even on the module level.)

But if you add an entry into builtins, and later a builtin of that name gets
added, you will replace the builtin function with your own, probably
completely different function. All functions and modules will notice this.
And whenever your code, which expects 'GlobalVariable' to be your specific
function/variable, imports/execs code that expects 'GlobalVariable' to be
the new builtin funtion, you end up with weird errors (at best) or very
weird behaviour (if your code doesn't throw an exception.)

And especially the standard library has a very good chance to already be
using the new 'GlobalVariable' builtin function, since it's released at the
same time as the source, and is expected to work only with the Python
version it's released with. 

> Consider the following imaginary example. In a project where there is no
> need to open files one replaces the open function. So this project will
> never manage to open a file but aside from that it will continue to work
> as designed.

This is a different situation entirely! As long as you provide the same
interface for the function you are replacing, and *intend* for it to replace
the normal 'open' call, there is no problem. You might run into a problem in
the future, though, when builtins becomes unwritable. (Which, for all we
know, might happen in Python 2.3 -- but not without warnings first <wink> :)

If you really want this, though, you should consider using something like
Bastion: run your code in a shielded environment with some builtins/modules
removed or wrapped in protective functions.

Readability-is-still-the-best-argument-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tanzer at swing.co.at  Mon Feb 26 08:58:42 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Mon, 26 Feb 2001 08:58:42 +0100
Subject: abusing __builtins__ 
In-Reply-To: Your message of "Mon, 26 Feb 2001 07:53:22 +0100."
             <20010226075322.K16781@xs4all.nl> 
Message-ID: <m14XIYM-000wcBC@swing.co.at>

Thomas Wouters <thomas at xs4all.net> wrote:

> > Consider the following imaginary example. In a project where there is no
> > need to open files one replaces the open function. So this project will
> > never manage to open a file but aside from that it will continue to work
> > as designed.
> 
> This is a different situation entirely! As long as you provide the same
> interface for the function you are replacing, and *intend* for it to replace
> the normal 'open' call, there is no problem. You might run into a problem in
> the future, though, when builtins becomes unwritable. (Which, for all we
> know, might happen in Python 2.3 -- but not without warnings first <wink> :)

I sure hope that won't happen.

Sometimes there are pretty good reasons to augment builtins.

For instance, I recently implemented a Class_Proxy. To make
Class_Proxy instances fully compatible with the classes they proxy, I
used:

###############################################################################
def _fix_builtins () :
    import __builtin__

    def isinstance (object, C, isinstance = __builtin__.isinstance) :
        if hasattr (C, "_Class") :
            C = C._Class
        return isinstance (object, C)
    isinstance.__doc__     = __builtin__.isinstance.__doc__
    __builtin__.isinstance = isinstance

    def issubclass (C, B, issubclass = __builtin__.issubclass) :
        if hasattr (C, "_Class") : C = C._Class
        if hasattr (B, "_Class") : B = B._Class
        return issubclass (C, B)
    issubclass.__doc__     = __builtin__.issubclass.__doc__
    __builtin__.issubclass = issubclass
    
_fix_builtins ()
del _fix_builtins
###############################################################################

Works like a charm. I could retrofit the class proxies into 
existing applications without any changes in the users of the classes
to be proxied.

If builtins were readonly, every place where a class proxy is used
would need to import Class_Proxy explicitly. This defeats the purpose
of the proxies. 

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From thomas at xs4all.net  Mon Feb 26 09:50:38 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 26 Feb 2001 09:50:38 +0100
Subject: abusing __builtins__
In-Reply-To: <m14XIYM-000wcBC@swing.co.at>; from tanzer@swing.co.at on Mon, Feb 26, 2001 at 08:58:42AM +0100
References: <20010226075322.K16781@xs4all.nl> <m14XIYM-000wcBC@swing.co.at>
Message-ID: <20010226095038.M16781@xs4all.nl>

On Mon, Feb 26, 2001 at 08:58:42AM +0100, Christian Tanzer wrote:

> > You might run into a problem in the future, though, when builtins
> > becomes unwritable. (Which, for all we know, might happen in Python 2.3
> > -- but not without warnings first <wink> :)

> I sure hope that won't happen.

I do, too. But lets not fight that battle until it comes up... I promise to
look out for the early signs ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From anton at home.ogre.nu  Sat Feb 24 07:06:16 2001
From: anton at home.ogre.nu (Anton Sherwood)
Date: 23 Feb 2001 22:06:16 -0800
Subject: Programming Habits in Python
References: <8visq6$p3o$1@news.nuri.net> <slrn91pu5o.82h.scarblac@pino.selwerd.nl> <yay1yw2hknt.fsf@ganglot.ifi.uio.no>
Message-ID: <977j0o$2hh$1@home.ogre.nu>

: Remco Gerlich wrote
: > If you had some idea of bad habits caused by Python,
: > please elaborate. Any examples you've noticed?

at 23 Nov 2000 14:39:50 +0100,
Vetle Roeim <vr at acm.org> writes

: I have some examples: [...]
:  2) space-oriented blocks.
:     it takes a while to discover that no matter how many spaces
:     you include or omit, your C++ code simply won't compile.

Heh.  My first programming language, circa 1973, was TUTOR,
which looks like this:

	unit	foo	$$ function label
	calc	y:=200	$$ assignment
	do	spam	$$ function call
	if	x < 150
	.	do	larch
	.	calc	x := 2x 	$$ implicit mult!
	else
	.	do	annelk
	endif

My second was FORTRAN, in which the first six columns (iirc)
are reserved for labels, and the next for a continuation mark.

I don't recall blundering by expecting indentation to be meaningful
in Pascal or C; nor, conversely, do I find it hard to do in Python.

-- 
Anton Sherwood  --  br0nt0 at p0b0x.com  --  http://ogre.nu/


From p.magwene at snet.net  Sat Feb 24 08:19:08 2001
From: p.magwene at snet.net (Paul Magwene)
Date: Sat, 24 Feb 2001 02:19:08 -0500
Subject: Lovin' list comprehensions
Message-ID: <CfJl6.600$GD.445288@typhoon.snet.net>

I was slow to upgrade from 1.5.2 to the 2.x series.  I was saying to myself,
"List comprehensions?! We don't need no steenkin' list comprehensions!"
(never having used list comprehensions before, I was sure they weren't
useful ;)

Lo and behold, I finally upgraded, and what's my favorite new feature? List
comprehensions of course!  They're so useful and compact - you gotta love
'em.

Thanks Greg Ewing, Skip Montanaro, and all the rest of the gang!

Ciao,
Paul





From francois.granger at free.fr  Sat Feb 24 14:15:06 2001
From: francois.granger at free.fr (=?ISO-8859-1?Q?Fran=E7ois_Granger?=)
Date: Sat, 24 Feb 2001 13:15:06 GMT
Subject: for french users of Python
Message-ID: <1epc72k.1sjxr9ga1ivzmN%francois.granger@free.fr>

[en]
There is currently a discussion for the creation of fr.comp.lang.python
in fr.usenet.forum.evolution

[fr]
Il yn a une discussion en cour sur la cr?ation d'un forum
fr.comp.lang.python dans fr.usenet.forum.evolution
-- 
"Ce qui est gratuit ne vaut pas cher" - inconnu


From gmcm at hypernet.com  Sat Feb 24 16:26:26 2001
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 24 Feb 2001 10:26:26 -0500
Subject: ANN: Win32 Installer - new release for Python 2
Message-ID: <3A978C52.10169.22C663ED@localhost>

With many, many thanks to Barry Scott, a new release of the 
Win32 Installer package is here:

http://www.mcmillan-inc.com/install1.html

Barry found my major boo-boo in the code that resolves a C 
extension inside a package. He also greatly improved the 
messages that come out of the runtime when something goes 
wrong at the top level. And if you're a user of Mark Hammond's 
stuff, he also made it so, for example, "import pywintypes" 
automatically finds PyWinTypes20.dll.

Also in this release: the fix for builtin modules, and a fix to the 
analysis of C extensions and dlls.

The documentation now has a page of how-tos for common 
extensions / packages (e.g., Pmw, PIL among others).

Background:
 The Win32 Installer is a sort of compilerless Freeze - a way 
of distributing Python apps with all the required support built 
in. It also understands binary resources (.dlls and .pyds) and 
can package up arbitrary files that your app might require.

Old-Python-style license.

Contact: gmcm at hypernet.com

<P><A 
HREF="http://www.mcmillan-inc.com/install1.html">
Win32 Installer release 3j</A> A compiler-less way 
of distributing Python 2.0 apps on Windows. (24-Feb-01)



- Gordon



From echapin at sympatico.ca  Sat Feb 24 17:04:51 2001
From: echapin at sympatico.ca (Elliott Chapin)
Date: Sat, 24 Feb 2001 16:04:51 GMT
Subject: grid details question
Message-ID: <DYQl6.217828$Pm2.3529476@news20.bellglobal.com>

I'm hoping there is a way to get the runtime values of measurements
determined by "grid."

Thanks ...

--

---------------------------------------------------------------

Elliott Chapin         http:// www3.sympatico.ca/echapin




From s2mdalle at titan.vcu.edu  Sat Feb 24 21:56:35 2001
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Sat, 24 Feb 2001 20:56:35 GMT
Subject: rm -rf in python
Message-ID: <7eVl6.680$3y6.446542@typhoon2.ba-dsg.net>

I need to write the equivalent of "rm -rf" in python,
so I wrote this, but I'm having a hard time understanding
why it's not working.  

The assumptions that I'm working off of is that
os.rmdir() will fail on a non-empty directory, and
that os.unlink() will fail on a directory.  Here's
the code:

import dircache
import os

def recursive_delete(dirname):
    """Recursively deletes everything under dirname.  This does NOT make any
    symlink distinctions or otherwise, it's functionally equivalent to an
    rm -rf dirname so use it with much care."""
    try:
        print "Removing directory:  \"%s\"" % dirname
        os.rmdir(dirname)                            
    except OSError:        # Directory isn't empty.
        files = dircache.listdir(dirname)

        for file in files:
            # Get full pathname...
            file = "%s%s%s" % (dirname, os.sep, file)
            if os.path.isdir(file):
                return recursive_delete(file)
            else:
                print "Removing file:  \"%s\"" % file
                retval = os.unlink(file)
                return file

        # We've nuked all the files, now kill the directory.
        try:
            os.rmdir(dirname)
        except OSError:
            print "Wuh?  Huh?"

This prints out the actual filenames such as...

Removing file "/home/user/blarg/snuge.txt"

and so on, but the file actually still exists after
the program runs.  The module documentation in python
doesn't say anything about any particular return
value signaling an error and not being able to delete
a file, and all of these calls seem to be returning
None.  I've programmed in C before, and these syscalls
should return 0 for success, and != 0 otherwise,
but does that hold in python?

Any help would be appreciated.
-- 
David Allen
http://opop.nols.com/
----------------------------------------
Alliance, n.: 
In international politics, the union of two thieves who have 
their hands so deeply inserted in each other's pocket that they cannot 
separately plunder a third. 
- Ambrose Bierce, "The Devil's Dictionary"


From mwh21 at cam.ac.uk  Sat Feb 24 22:15:30 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 24 Feb 2001 21:15:30 +0000
Subject: rm -rf in python
References: <7eVl6.680$3y6.446542@typhoon2.ba-dsg.net>
Message-ID: <m3hf1jsryl.fsf@atrus.jesus.cam.ac.uk>

"David Allen" <s2mdalle at titan.vcu.edu> writes:

> I need to write the equivalent of "rm -rf" in python,
> so I wrote this, but I'm having a hard time understanding
> why it's not working.  
[schnipp]
> Any help would be appreciated.

I haven't read your code, but do you know about shutil.rmtree?

>>> print shutil.rmtree.__doc__ 
Recursively delete a directory tree.

    If ignore_errors is set, errors are ignored; otherwise, if
    onerror is set, it is called to handle the error; otherwise, an
    exception is raised.

Hmm, that docstring isn't quite right...

Cheers,
M.

-- 
  ROOSTA:  Ever since you arrived on this planet last night you've
           been going round telling people that you're Zaphod
           Beeblebrox, but that they're not to tell anyone else.
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 7


From s2mdalle at titan.vcu.edu  Sat Feb 24 22:38:41 2001
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Sat, 24 Feb 2001 21:38:41 GMT
Subject: rm -rf in python
References: <7eVl6.680$3y6.446542@typhoon2.ba-dsg.net> <m3hf1jsryl.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <BRVl6.736$3y6.460417@typhoon2.ba-dsg.net>

In article <m3hf1jsryl.fsf at atrus.jesus.cam.ac.uk>, "Michael Hudson"
<mwh21 at cam.ac.uk> wrote:

> "David Allen" <s2mdalle at titan.vcu.edu> writes:
> 
>> I need to write the equivalent of "rm -rf" in python, so I wrote this, but
>> I'm having a hard time understanding why it's not working.  
> [schnipp]
>> Any help would be appreciated.
> 
> I haven't read your code, but do you know about shutil.rmtree?

I hadn't until now...

Searching yields:

http://www.python.org/doc/current/lib/module-shutil.html

Question:  I'm not seeing in the documentation for
this whether or not it's good on windows machines.
Is it?  In my original code I had to be careful to
use os.sep rather than just "/" because this code 
will have to run on microsoft OS's.

-- 
David Allen
http://opop.nols.com/
----------------------------------------
"The reasonable man adapts himself to the world; the unreasonable one
 persists in trying to adapt the world to himself.  Therefore, all
 progress depends on the unreasonable man."
                                - George Bernard Shaw


From mwh21 at cam.ac.uk  Sat Feb 24 22:54:19 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 24 Feb 2001 21:54:19 +0000
Subject: rm -rf in python
References: <7eVl6.680$3y6.446542@typhoon2.ba-dsg.net> <m3hf1jsryl.fsf@atrus.jesus.cam.ac.uk> <BRVl6.736$3y6.460417@typhoon2.ba-dsg.net>
Message-ID: <m3elwnsq5w.fsf@atrus.jesus.cam.ac.uk>

"David Allen" <s2mdalle at titan.vcu.edu> writes:

> In article <m3hf1jsryl.fsf at atrus.jesus.cam.ac.uk>, "Michael Hudson"
> <mwh21 at cam.ac.uk> wrote:
> 
> > "David Allen" <s2mdalle at titan.vcu.edu> writes:
> > 
> >> I need to write the equivalent of "rm -rf" in python, so I wrote this, but
> >> I'm having a hard time understanding why it's not working.  
> > [schnipp]
> >> Any help would be appreciated.
> > 
> > I haven't read your code, but do you know about shutil.rmtree?
> 
> I hadn't until now...
> 
> Searching yields:
> 
> http://www.python.org/doc/current/lib/module-shutil.html
> 
> Question:  I'm not seeing in the documentation for
> this whether or not it's good on windows machines.
> Is it?  In my original code I had to be careful to
> use os.sep rather than just "/" because this code 
> will have to run on microsoft OS's.

>From glancing at the source, I'd say it should work.  It does its
stuff by building up a list of tuples of the form

(os.remove, path_to_file)

or

(os.rmdir, path_to_dir)

(building up the paths using os.path.join) and then rattles through
doing essentially

for a, b in list_of_tuples:
    a(b)

I don't know how old the module is, but I think it was in 1.5.2 (and
if this turns out to be a problem you can just nick the code from
2.1).

Cheers,
M.

-- 
  I really hope there's a catastrophic bug insome future e-mail
  program where if you try and send an attachment it cancels your
  ISP account, deletes your harddrive, and pisses in your coffee
                                                         -- Adam Rixey


From noahnoah.org  Sun Feb 25 00:07:17 2001
From: noahnoah.org (Noah Spurrier (a))
Date: 24 Feb 2001 23:07:17 GMT
Subject: rm -rf in python
References: <7eVl6.680$3y6.446542@typhoon2.ba-dsg.net>
Message-ID: <979er50ij@news1.newsguy.com>

Your function couldn't work because it was returning early
from your 'for' loop (in both directory and file deletion). 
The fixed code is below. I also moved the os.rmdir AFTER the
listdir looping, because it makes the code cleaner. You don't
have to detect if the directory is empty or not. If the directory
is empty then listdir returns an empty list and the loop does nothing.
I also replaced:
     file = '%s%s%s' % (dirname, os.sep, file)
with:
     path = os.path.join (dirname, file)
which does pretty much the same thing...

Finally I added a dirty shell script to make a bunch of files 
and directories for testing. The root directory is called 'foodir'.

Also, consider experimenting with os.path.walk().

Yours,
Noah

------------ Version 1 ---------------------------------------------
<pre>
import os, dircache

def recursive_delete(dirname):
	files = dircache.listdir(dirname)
	for file in files:
		path = os.path.join (dirname, file)
		if os.path.isdir(path):
			recursive_delete(path)
		else:
			print 'Removing file: "%s"' % path
			retval = os.unlink(path)
			
	print 'Removing directory:', dirname
	os.rmdir(dirname)
</pre>

------- Script to make test files and directories -----------------
#!/bin/sh
mkdir foodir
cd foodir
echo foo > foofile1
echo foo > foofile2
echo foo > foofile3
mkdir foodir1
mkdir foodir2
cd foodir1
echo foo > foofile1
echo foo > foofile2
echo foo > foofile3
cd ..
cd foodir2
echo foo > foofile1
echo foo > foofile2
echo foo > foofile3
mkdir foodir2_1
mkdir foodir2_2
cd foodir2_2
echo foo > foofile1
echo foo > foofile2
echo foo > foofile3

> I need to write the equivalent of "rm -rf" in python,
> so I wrote this, but I'm having a hard time understanding
> why it's not working.  
> 
> The assumptions that I'm working off of is that
> os.rmdir() will fail on a non-empty directory, and
> that os.unlink() will fail on a directory.  Here's
> the code:
> 
> ...
> and so on, but the file actually still exists after
> the program runs.  The module documentation in python
> doesn't say anything about any particular return
> value signaling an error and not being able to delete
> a file, and all of these calls seem to be returning
> None.  I've programmed in C before, and these syscalls
> should return 0 for success, and != 0 otherwise,
> but does that hold in python?
> 
> Any help would be appreciated.
> -- 
> David Allen
> http://opop.nols.com/
> ----------------------------------------
> Alliance, n.: 
> In international politics, the union of two thieves who have 
> their hands so deeply inserted in each other's pocket that they cannot 
> separately plunder a third. 
> - Ambrose Bierce, "The Devil's Dictionary"


==================================
Posted via http://nodevice.com
Linux Programmer's Site


From s2mdalle at titan.vcu.edu  Sun Feb 25 04:39:35 2001
From: s2mdalle at titan.vcu.edu (David Allen)
Date: Sun, 25 Feb 2001 03:39:35 GMT
Subject: rm -rf in python
References: <7eVl6.680$3y6.446542@typhoon2.ba-dsg.net> <979er50ij@news1.newsguy.com>
Message-ID: <X7%l6.894$3y6.610945@typhoon2.ba-dsg.net>

In article <979er50ij at news1.newsguy.com>, "Noah Spurrier" <noah(a)noah.org>
wrote:

> 
> Your function couldn't work because it was returning early from your 'for'
> loop (in both directory and file deletion).  The fixed code is below. I also
> moved the os.rmdir AFTER the listdir looping, because it makes the code
> cleaner. You don't have to detect if the directory is empty or not. If the
> directory is empty then listdir returns an empty list and the loop does
> nothing. I also replaced:
>      file = '%s%s%s' % (dirname, os.sep, file)
> with:
>      path = os.path.join (dirname, file)
> which does pretty much the same thing...
> 
> Finally I added a dirty shell script to make a bunch of files  and
> directories for testing. The root directory is called 'foodir'.
> 
> Also, consider experimenting with os.path.walk().
> 
> Yours, Noah

Thanks a lot - this is a lot cleaner and shorter
than mine was too.  :)

I had briefly looked at os.path.walk() but I decided
against it because this recursive solution just
seemed cleaner to me than walking through each file
and having specific logic to see if it was a directory
or file, empty or not, etc. and whether or not 
the walk was passing them to me in the right order or
not.  (i.e. if the path /home/user/foo/bar/baz/foo.txt
exists, I'd want to get the paths in this order:

/home/user/foo/bar/baz/foo.txt
/home/user/foo/bar/baz/
/home/user/foo/bar/
/home/user/foo/
/home/user)

Also, I think I'm going to go from dircache.listdir()
to os.listdir() since I don't actually need the 
caching mechanism at all.
-- 
David Allen
http://opop.nols.com/
----------------------------------------
Math is like love - a simple idea but it can get complicated. 
- R. Drabek


From uche at ogbuji.net  Sat Feb 24 23:20:26 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Sat, 24 Feb 2001 22:20:26 GMT
Subject: Article: Transforming Python performance data
Message-ID: <3A9833A4.A4A279FA@ogbuji.net>

Just plugging my brother's work.

http://www-106.ibm.com/developerworks/library/x-transpy/?dwzone=xml

for XML/RDF or natural language processing geeks, I also suggest

http://www.unixinsider.com/swol-12-2000/swol-1229-ontology.html

-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From uche at ogbuji.net  Sat Feb 24 23:40:04 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Sat, 24 Feb 2001 22:40:04 GMT
Subject: Column: The Python Web services developer
Message-ID: <3A983844.CB5DE5E5@ogbuji.net>

Mike Olson and I have begun a monthly column at IBM developerWorks on
using Python to develop Web Services.

See

http://www-106.ibm.com/developerworks/webservices/

The first installment is at

http://www-106.ibm.com/developerworks/library/ws-pyth1.html?dwzone=ws

And is an overview of Python tools for Web Services.


-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From slhath at home.com  Sun Feb 25 00:30:17 2001
From: slhath at home.com (Scott Hathaway)
Date: Sat, 24 Feb 2001 23:30:17 GMT
Subject: placing widgets at x,y coordinates on a window with Tkinter
Message-ID: <duXl6.112018$bb.1303633@news1.rdc1.tx.home.com>

I have a VB program that I want to convert to Tkinter.  The VB program uses
"skins" to allow the user to createa graphics that are used for the window.
The window of the program resizes itself to be the size of the background
image.  Then, there are 6 buttons and a drop-down box that I allow the users
to tell me (in the theme.dat file for a theme) where they are to be placed
on the screen.  The users give me the top and left coordinates for the
images.

How can I do this with Tkinter.  From what I understand about .pack()
.place() and grid(), none of them would allow for me to specify the x,y
coords in pixels on the fly.  Is this correct?  How can I do this?

TIA,
Scott




From zzizz_ at notmail.com  Sun Feb 25 00:50:31 2001
From: zzizz_ at notmail.com (zzzzz)
Date: Sun, 25 Feb 2001 10:50:31 +1100
Subject: win98 spawnv vs execv
Message-ID: <46ig9tcbuu6g8qrk6pnthf32ugs1j5phqg@4ax.com>

Hi all,
I running be-open python 2.0 with pythonwin on win98 and get an
windows execption when I launch a program with os.execv causing
pythonwin to immediately close... even though the program is
succefully launched. However, when I use os.spawnv with a os.P_WAIT
option the function works fine.

Is this likely to be:
a) a platform limitation,
b) something to do with my computer (I run Nortons Anti-Virus 5.0 not
2000 and I don't think I'm launching popen*),
c) something else I've haven't thought of, or
d) stop making up other options.


From chris.gonnerman at usa.net  Sun Feb 25 06:39:25 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 24 Feb 2001 23:39:25 -0600
Subject: win98 spawnv vs execv
References: <46ig9tcbuu6g8qrk6pnthf32ugs1j5phqg@4ax.com>
Message-ID: <012501c09eed$54ba30e0$a100000a@local>

The exec family of functions terminate the Python interpreter and replace
it with the called program; spawnv starts an additional process.  

os.execv() is a Unixism, used together with os.fork() to start new
processes.  My best guess is, PythonWin simply isn't designed to 
"keel over" when os.execv() is called.

As the doctor says, if it hurts when you do that, don't do that.

BTW when you post to the list about error messages, you probably 
should post at least the first sentence or two of the actual error message
*verbatim* as Windows has thousands of messages that don't always
mean what they say.

----- Original Message ----- 
From: "zzzzz" <zzizz_ at notmail.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Saturday, February 24, 2001 5:50 PM
Subject: win98 spawnv vs execv


> Hi all,
> I running be-open python 2.0 with pythonwin on win98 and get an
> windows execption when I launch a program with os.execv causing
> pythonwin to immediately close... even though the program is
> succefully launched. However, when I use os.spawnv with a os.P_WAIT
> option the function works fine.
> 
> Is this likely to be:
> a) a platform limitation,
> b) something to do with my computer (I run Nortons Anti-Virus 5.0 not
> 2000 and I don't think I'm launching popen*),
> c) something else I've haven't thought of, or
> d) stop making up other options.





From zzizz_ at notmail.com  Sun Feb 25 23:11:05 2001
From: zzizz_ at notmail.com (zzzzz)
Date: Mon, 26 Feb 2001 09:11:05 +1100
Subject: win98 spawnv vs execv
References: <46ig9tcbuu6g8qrk6pnthf32ugs1j5phqg@4ax.com> <mailman.983080025.30146.python-list@python.org>
Message-ID: <hk0j9tkul58f9p2eo6gauk5se6nflna00a@4ax.com>

On Sat, 24 Feb 2001 23:39:25 -0600, "Chris Gonnerman"
<chris.gonnerman at usa.net> wrote:


>BTW when you post to the list about error messages, you probably 
>should post at least the first sentence or two of the actual error message
>*verbatim* as Windows has thousands of messages that don't always
>mean what they say.
>
Thanks Chris, I'll do that in future.

z.


>----- Original Message ----- 
>From: "zzzzz" <zzizz_ at notmail.com>
>Newsgroups: comp.lang.python
>To: <python-list at python.org>
>Sent: Saturday, February 24, 2001 5:50 PM
>Subject: win98 spawnv vs execv
>
>
>> Hi all,
>> I running be-open python 2.0 with pythonwin on win98 and get an
>> windows execption when I launch a program with os.execv causing
>> pythonwin to immediately close... even though the program is
>> succefully launched. However, when I use os.spawnv with a os.P_WAIT
>> option the function works fine.
>> 
>> Is this likely to be:
>> a) a platform limitation,
>> b) something to do with my computer (I run Nortons Anti-Virus 5.0 not
>> 2000 and I don't think I'm launching popen*),
>> c) something else I've haven't thought of, or
>> d) stop making up other options.
>
>

Regards,

zzzzz.
---------------------


From fimafeng at yahoo.com  Sun Feb 25 00:56:45 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Sat, 24 Feb 2001 23:56:45 GMT
Subject: Newbie Question---IDLE on Linux
Message-ID: <3A984A1F.4030801@yahoo.com>

Hi:
I'm running Red Hat 7.0 which comes with Python 1.5.2 installed by default.
I downloaded the source for 2.0 and Tkinter
The source for 2.0 seems to have finished installing and I can
get the 2.0 interpreter to come up but I have type in the full path
which is   /myinstalls/Python-2.0/python

I have tkinter 1.52 already installed from the inital build of the 
system. Do I still need to upgrade? How do I upgrade the TKinter? I'm 
guessing that it is the setup.py file in the TKiner directory.
When I try to run the program I get the following.....

[root at localhost bin]# python /myinstalls/Tkinter-2.0-8.0/setup.py
Traceback (innermost last):
   File "/myinstalls/Tkinter-2.0-8.0/setup.py", line 8, in ?
     from distutils.core import setup, Extension
ImportError: No module named distutils.core

What AM I doing wrong????

Please help. I'm trying to get IDLE to run....

Chris



From bryan at eevolved.com  Sun Feb 25 01:27:53 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Sun, 25 Feb 2001 00:27:53 GMT
Subject: Newbie Question---IDLE on Linux
References: <3A984A1F.4030801@yahoo.com>
Message-ID: <dkYl6.312580$JT5.10583064@news20.bellglobal.com>

Easiest solution:
http://www.python.org/ftp/python/2.0/BeOpen-Python-2.0-1.i386.rpm

When you compile from source, you have to set the environment variable 
$PYTHONPATH to the directory where your libs are.

Good luck.
-- 
<=====================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
Website:    http://www.eevolved.com
Public key: http://eevolved.com/bcm.pk
<=====================================>

"The genes are the master programmers, and they are programming for their 
lives." -- Richard Dawkins



From fimafeng at yahoo.com  Sun Feb 25 01:41:15 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Sun, 25 Feb 2001 00:41:15 GMT
Subject: Newbie Question---IDLE on Linux
References: <3A984A1F.4030801@yahoo.com> <dkYl6.312580$JT5.10583064@news20.bellglobal.com>
Message-ID: <3A98548D.7080207@yahoo.com>

I tried That--- Here's what the FAQ said at python.org


Q. I just downloaded version 2.0 (BeOpen-Python-2.0-1.i386.rpm) and 
there is no trace of IDLE in there; an "rpm -qpl 
BeOpen-Python-2.0-1.i386.rpm | grep -i idle" comes up empty. 
http://www.python.org/idle/ claims that IDLE 0.6 is distributed as part 
of Python 2.0. Checking further, I see that IDLE is included in 
ftp://ftp.python.org/pub/python/src/python-2.0c1.tar.gz in the tools 
directory. Is this an oversight? Is there a way to install IDLE without 
unpacking the source tarball? (bob at drzyzgula.org)

A. This is an oversight. None of the code in Tools is installed. I will 
fix this as soon as possible. (jeremy at alum.mit.edu)




Bryan Mongeau wrote:

> Easiest solution:
> 


> http://www.python.org/ftp/python/2.0/BeOpen-Python-2.0-1.i386.rpm
> 
> 
> 
> When you compile from source, you have to set the environment variable 
> 
> $PYTHONPATH to the directory where your libs are.
> 
> 
> 
> Good luck.

  Where's IDLE?

Q. I just downloaded version 2.0 (BeOpen-Python-2.0-1.i386.rpm) and 
there is no trace of IDLE in there; an "rpm -qpl 
BeOpen-Python-2.0-1.i386.rpm | grep -i idle" comes up empty. 
http://www.python.org/idle/ claims that IDLE 0.6 is distributed as part 
of Python 2.0. Checking further, I see that IDLE is included in 
ftp://ftp.python.org/pub/python/src/python-2.0c1.tar.gz in the tools 
directory. Is this an oversight? Is there a way to install IDLE without 
unpacking the source tarball? (bob at drzyzgula.org)

A. This is an oversight. None of the code in Tools is installed. I will 
fix this as soon as possible. (jeremy at alum.mit.edu)




From andymac at bullseye.apana.org.au  Sun Feb 25 01:02:04 2001
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Sun, 25 Feb 2001 11:02:04 +1100 (EDT)
Subject: Python 2.0 vs OS/2+EMX
Message-ID: <Pine.OS2.4.32.0102251028370.1907-100000@CENTRAL>

In attempting to build 2.0 on OS/2 (v4 + FP12) with the EMX gcc compiler
(0.9d fix03 w/gcc 2.8.1) I've run into a couple of oddities.

I've patched the source based on patches done by Andrew Zabolotny for his
EMX port of 1.5.2, and have succeeded in building both single and
multithreaded versions.

I have been able to run the regression test with the single threaded
interpreter and a fair proportion of the tests pass, and most of the rest
fail graciously although a number need to be followed up as to why they
fail.  However I'd appreciate thoughts about the issues causing the
following problems:-
1) when running the regression test, regrtest.py reports that it is unable
to create the output files for each test (EACCES is the C lib error).
However I can manually fire up the python binary and create the exact same
file no problems.  ????

2) the longexp test causes python to consume all available memory until
SWAPPER.DAT exhausts available disk space.  Reducing the number of REPS to
30000 (from 65580) doesn't help, but REPS = 10000 works fine.  On the face
of it a memory management problem of some sort, but where to start?

The multithreaded interpreter fails on startup with a message something
like "don't have interpreter lock".  Unfortunately I haven't kept the
multithreaded build, so I can't check this at the moment, and in any case
its a lower priority than getting the single threaded interpreter running.

Any thoughts or suggestions gratefully received.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From andymac at bullseye.apana.org.au  Sun Feb 25 01:24:48 2001
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Sun, 25 Feb 2001 11:24:48 +1100 (EDT)
Subject: Python 2.0 vs OS/2+EMX
In-Reply-To: <Pine.OS2.4.32.0102251028370.1907-100000@CENTRAL>
Message-ID: <Pine.OS2.4.32.0102251121560.1924-100000@CENTRAL>

On Sun, 25 Feb 2001, Andrew MacIntyre wrote:

> 2) the longexp test causes python to consume all available memory until
> SWAPPER.DAT exhausts available disk space.  Reducing the number of REPS to
> 30000 (from 65580) doesn't help, but REPS = 10000 works fine.  On the face
> of it a memory management problem of some sort, but where to start?

I forgot to clarify that it appears to be eval() which is having the
problem, as creating the string of the long expression interactively with
REPS = 65580 (as in test_longexp.py) works fine.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andrew.macintyre at aba.gov.au    (work) | Snail: PO Box 370
        andymac at bullseye.apana.org.au  (play) |        Belconnen  ACT  2616
        andymac at pcug.org.au           (play2) |        Australia




From dkstewart at hotmail.com  Sun Feb 25 01:24:36 2001
From: dkstewart at hotmail.com (David Stewart)
Date: Sun, 25 Feb 2001 10:24:36 +1000
Subject: Newbie: How do I run a script from IDLE (Win98)
Message-ID: <3A9850C4.AA6ED04B@hotmail.com>

I can import and run my ?Hello World? module/function in the immediate
window but if I then edit and save the source file I can not see how to
run the modified code without closing and restarting the shell.

If I try and do it all from the IDLE edit window both import module (f5)
and run script (ctrl+f5) open a shell with the cursor in the line below
the prompt and then nothing!  Ctrl+f5 in the shell gives me the message
?The buffer for Python Shell is not saved 
.?

What am I missing?

Thanks

Dave Stewart




From bernie at pacific.net.hk  Wed Feb 28 18:30:17 2001
From: bernie at pacific.net.hk (bernie)
Date: Thu, 01 Mar 2001 01:30:17 +0800
Subject: Newbie: How do I run a script from IDLE (Win98)
References: <3A9850C4.AA6ED04B@hotmail.com>
Message-ID: <3A9D35A8.70E80DE8@pacific.net.hk>

David Stewart wrote:

> I can import and run my ?Hello World? module/function in the immediate
> window but if I then edit and save the source file I can not see how to
> run the modified code without closing and restarting the shell.
>

1.    Open <Hello World.py>
2.    Make your change
3.    Save the changes <Ctrl-s>
4.    Run the script <Ctrl-F5>

If you are dealing with mutliple module files, then you should:

A.    Repeat step 1 to 3 above
B.    Go to the main module
C.    Repeat Step 4

It works most of the time, unless you use thread modules.

Hope it helps!


Bernie




From bernie at pacific.net.hk  Wed Feb 28 18:30:38 2001
From: bernie at pacific.net.hk (bernie)
Date: Thu, 01 Mar 2001 01:30:38 +0800
Subject: Newbie: How do I run a script from IDLE (Win98)
References: <3A9850C4.AA6ED04B@hotmail.com>
Message-ID: <3A9D35BD.E731FC63@pacific.net.hk>

David Stewart wrote:

> I can import and run my ?Hello World? module/function in the immediate
> window but if I then edit and save the source file I can not see how to
> run the modified code without closing and restarting the shell.

1.    Open <Hello World.py>
2.    Make your change
3.    Save the changes <Ctrl-s>
4.    Run the script <Ctrl-F5>

If you are dealing with mutliple module files, then you should:

A.    Repeat step 1 to 3 above
B.    Go to the main module
C.    Repeat Step 4

It works most of the time, unless you use thread modules.

Hope it helps!


Bernie




From OneDay at A.Time  Sun Feb 25 01:58:15 2001
From: OneDay at A.Time (OneDay at A.Time)
Date: Sat, 24 Feb 2001 19:58:15 EST
Subject: Python 2.0 or Activestate Python?
Message-ID: <979lb7$2gsf$1@earth.superlink.net>

Hello All,

I have Python 2.0 and am looking for some enlightenment on Activestate Python.

Will it work with my original Python 2.0?

Should I delete the original Python 2.0 first?

What advanages does Activestate offer?

Any input will be greatly apreciated.

Thanks Al


From paulp at ActiveState.com  Sun Feb 25 05:51:17 2001
From: paulp at ActiveState.com (Paul Prescod)
Date: Sat, 24 Feb 2001 20:51:17 -0800
Subject: Python 2.0 or Activestate Python?
References: <979lb7$2gsf$1@earth.superlink.net>
Message-ID: <3A988F45.DC0B45F7@ActiveState.com>

OneDay at A.Time wrote:
> 
> Hello All,
> 
> I have Python 2.0 and am looking for some enlightenment on Activestate Python.
> 
> Will it work with my original Python 2.0?

On Windows machines, ActiveState will replace another Python as the
"standard" Python (i.e. the one in the search path, the one associated
with files etc.)

> Should I delete the original Python 2.0 first?

It doesn't matter but there is not much virtue in having both around.

> What advanages does Activestate offer?

O'Reilly did an article about the different distributions recently.

http://www.oreillynet.com/pub/a/python/2001/02/14/pythonnews.html

A new version of ActivePython will be released in the next week or so.
It will have some new features and bug fixes. If you download RIGHT NOW
I'll promise you a free upgrade to that version (which will also be
free).

-- 
Vote for Your Favorite Python & Perl Programming  
Accomplishments in the first Active Awards! 
http://www.ActiveState.com/Awards



From tim at vegeta.ath.cx  Tue Feb 27 07:49:11 2001
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Tue, 27 Feb 2001 06:49:11 GMT
Subject: Python 2.0 or Activestate Python?
References: <979lb7$2gsf$1@earth.superlink.net>
Message-ID: <slrn99mjt2.9m4.tim@vegeta.ath.cx>

OneDay at A.Time <OneDay at A.Time> wrote:
> Hello All,
> 
> I have Python 2.0 and am looking for some enlightenment on Activestate Python.
[ snip ]
> What advanages does Activestate offer?

A restrictive license and a little bit of commercialisation.

-- 
-Tim Hammerquist <timmy at cpan.org>
Emacs is a nice OS - but it lacks a good text editor.
That's why I am using Vim.
	-- Anonymous


From brian at sweetapp.com  Tue Feb 27 08:37:19 2001
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 26 Feb 2001 23:37:19 -0800
Subject: Python 2.0 or Activestate Python?
In-Reply-To: <slrn99mjt2.9m4.tim@vegeta.ath.cx>
Message-ID: <HPEJKHHCMJHPMNLDAPLMGEPJCCAA.brian@sweetapp.com>

The license doesn't seem that restrictive to me, but you can peruse it
yourself here:
http://www.activestate.com/Legal/Active_Python_License_Agreement_Text.txt

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Tim Hammerquist
Sent: Monday, February 26, 2001 10:49 PM
To: python-list at python.org
Subject: Re: Python 2.0 or Activestate Python?


OneDay at A.Time <OneDay at A.Time> wrote:
> Hello All,
>
> I have Python 2.0 and am looking for some enlightenment on Activestate
Python.
[ snip ]
> What advanages does Activestate offer?

A restrictive license and a little bit of commercialisation.

--
-Tim Hammerquist <timmy at cpan.org>
Emacs is a nice OS - but it lacks a good text editor.
That's why I am using Vim.
	-- Anonymous
--
http://mail.python.org/mailman/listinfo/python-list




From fuess at att.net  Tue Feb 27 14:10:45 2001
From: fuess at att.net (David Fuess)
Date: Tue, 27 Feb 2001 13:10:45 GMT
Subject: Python 2.0 or Activestate Python?
References: <slrn99mjt2.9m4.tim@vegeta.ath.cx> <mailman.983259250.18751.python-list@python.org>
Message-ID: <jk9n9tgsjvabo0c54r0d092vmb8n08c67c@4ax.com>

If you're working on a Windows platform, then you will want Mark
Hammond's  win32 classes anyway. ActiveState Python is a combined
Python 2 and Win32all in one simple install. Otherwise you'll install
one of the other Python 2 distributions and still have to go to
ActiveState to get Win32.

Dave

On Mon, 26 Feb 2001 23:37:19 -0800, "Brian Quinlan"
<brian at sweetapp.com> wrote:

>The license doesn't seem that restrictive to me, but you can peruse it
>yourself here:
>http://www.activestate.com/Legal/Active_Python_License_Agreement_Text.txt
>
>-----Original Message-----
>From: python-list-admin at python.org
>[mailto:python-list-admin at python.org]On Behalf Of Tim Hammerquist
>Sent: Monday, February 26, 2001 10:49 PM
>To: python-list at python.org
>Subject: Re: Python 2.0 or Activestate Python?
>
>
>OneDay at A.Time <OneDay at A.Time> wrote:
>> Hello All,
>>
>> I have Python 2.0 and am looking for some enlightenment on Activestate
>Python.
>[ snip ]
>> What advanages does Activestate offer?
>
>A restrictive license and a little bit of commercialisation.
>
>--
>-Tim Hammerquist <timmy at cpan.org>
>Emacs is a nice OS - but it lacks a good text editor.
>That's why I am using Vim.
>	-- Anonymous



From tim at vegeta.ath.cx  Tue Feb 27 23:13:02 2001
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Tue, 27 Feb 2001 22:13:02 GMT
Subject: Python 2.0 or Activestate Python?
References: <slrn99mjt2.9m4.tim@vegeta.ath.cx> <mailman.983259250.18751.python-list@python.org> <jk9n9tgsjvabo0c54r0d092vmb8n08c67c@4ax.com>
Message-ID: <slrn99oa16.etg.tim@vegeta.ath.cx>

David Fuess <fuess at att.net> wrote:
> If you're working on a Windows platform, then you will want Mark
> Hammond's  win32 classes anyway. ActiveState Python is a combined
> Python 2 and Win32all in one simple install. Otherwise you'll install
> one of the other Python 2 distributions and still have to go to
> ActiveState to get Win32.

And Win32 was a huge user base for ActivePerl as well.  As far as I can
tell, ActiveState has a monopoly on Win32 Perl and Python distributions,
but Perl has recently gotten IndigoPerl and I'm quite happy with it.
Something will eventually compete with ActivePython as well, and
ActiveState just might lose.  Who knows?

-- 
-Tim Hammerquist <timmy at cpan.org>

Any emotion, if it is sincere, is involuntary.
	-- Mark Twain


From db3l at fitlinxx.com  Wed Feb 28 02:49:27 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 20:49:27 -0500
Subject: Python 2.0 or Activestate Python?
References: <slrn99mjt2.9m4.tim@vegeta.ath.cx> <mailman.983259250.18751.python-list@python.org> <jk9n9tgsjvabo0c54r0d092vmb8n08c67c@4ax.com> <slrn99oa16.etg.tim@vegeta.ath.cx>
Message-ID: <ulmqrmva0.fsf@ctwd0143.fitlinxx.com>

tim at vegeta.ath.cx (Tim Hammerquist) writes:

> Something will eventually compete with ActivePython as well, and
> ActiveState just might lose.  Who knows?

The other distributions for Python are already fine for the win32
platform.  There's no reason not to just install the standard BeOpen
2.0 release and then stick win32all on it.  The latest win32all
packages are taken from the ActiveState web site, but they are still
available independent of the ActiveState package itself.

I don't think it has to be a lose situation - multiple packages can
likely survive just fine.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From tim at vegeta.ath.cx  Tue Feb 27 23:07:07 2001
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Tue, 27 Feb 2001 22:07:07 GMT
Subject: Python 2.0 or Activestate Python?
References: <mailman.983259250.18751.python-list@python.org>
Message-ID: <slrn99o9m3.etg.tim@vegeta.ath.cx>

Brian Quinlan <brian at sweetapp.com> wrote:
> The license doesn't seem that restrictive to me, but you can peruse it
> yourself here:
> http://www.activestate.com/Legal/Active_Python_License_Agreement_Text.txt

That's "restrictive" as opposed to totally "open."

I looked over the document.  I suppose, if I were distributing a script
or application that required Perl or Python to clients (as I've been
known to do), that I would not be able to distribute ActiveState's
software along with it?  I'd have to tell them to go to ActiveState's
site, right?  (Where Activestate can try to sell them a whole lot of
commercial software.)  Personally, and I do _not_ speak for everyone,
I'd like to be able to place the software on the medium so that the
clients, who often don't want or know how to install it, can get with
their business instead of going to the site and worry about what they
need.  ("Was it ActivePython?  or was it Komodo?  And what's this PDK?
Is that for Python?")

I guess I was unclear in what I was saying.  Is that a little clearer?

HAND
-- 
-Tim Hammerquist <timmy at cpan.org>

I don't have any solution, but I certainly admire the problem.
	-- Ashleigh Brilliant


From harvest at black-karma.de  Sun Feb 25 02:53:35 2001
From: harvest at black-karma.de (Harvest T. Moon)
Date: Sun, 25 Feb 2001 01:53:35 GMT
Subject: Q: BeOS + Python + Database?
Message-ID: <PM00037E17E4A0876A@wh4030.stw.uni-rostock.de>

hi.
i want to use a BeOS machine as developement plattform for python and
i'm looking for a working database + python module. available DBs are
PostgreSQL and Berkley DB, both are running fine.
has anyone compiled any of the mentioned python-modules? i cannot get
any of it to work :(

thx i.a.
Harvest T. Moon
-- a life less ordinary!!


From nyp at njc.ch  Sun Feb 25 07:26:06 2001
From: nyp at njc.ch (Yves Perrenoud)
Date: Sat, 24 Feb 2001 22:26:06 -0800
Subject: Threading problems with SWIG generated code
Message-ID: <3A98A57E.49EBCE99@njc.ch>

I've ran into some problems when trying to use multiple threads which
make use of modules generated with SWIG. Here's the shortest example I
could come up with to illustrate the problem:

Given the following SWIG interface file:

----
%module blip

unsigned int sleep(unsigned int seconds);
----

I generate the C code using "swig -python", compile it and generate the
shared library.

Here's some sample code using the newly generated "blip" module:

----
import threading, blip

def foo():
    blip.sleep(5)

threading.Thread(target=foo).start()
threading.Thread(target=foo).start()
----

You would expect this function to run for a total of approximately 5
seconds, but instead, it runs for 10 seconds! Clearly, the sleep
function calls are running one after the other, when they should be
running in parallel. It's almost as if I was acquiring a lock before the
sleep and releasing it afterwards.

Any idea what's causing this?

 ..Yves


From MarkH at ActiveState.com  Sun Feb 25 22:46:38 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sun, 25 Feb 2001 21:46:38 GMT
Subject: Threading problems with SWIG generated code
References: <3A98A57E.49EBCE99@njc.ch>
Message-ID: <3A997C65.7080503@ActiveState.com>

Yves Perrenoud wrote:

> I've ran into some problems when trying to use multiple threads which
> make use of modules generated with SWIG. Here's the shortest example I
> could come up with to illustrate the problem:

[example snipped]

You need to release the Python thread-lock.  The simplest way with SWIG is to use an except directive.

For example, I do the following in the Win32 extensions:

// Map API functions that return BOOL to
// functions that return None, but raise exceptions.
// These functions must set the win32 LastError.
%typedef BOOL BOOLAPI

%typemap(python,except) BOOLAPI {
      Py_BEGIN_ALLOW_THREADS
      $function
      Py_END_ALLOW_THREADS
      if (!$source)  {
           $cleanup
           return PyWin_SetAPIError("$name");
      }
}

This gives me 2 things to functions I declare as "BOOLAPI"
* Ensures the thread-lock is released when the function is called.
* Map failures to Win32 error codes.

Hope this helps...

Mark.



From rjroy at takingcontrol.com  Sun Feb 25 08:43:40 2001
From: rjroy at takingcontrol.com (Robert Roy)
Date: Sun, 25 Feb 2001 07:43:40 GMT
Subject: problems creating C extension types
Message-ID: <3a98b701.198621000@news1.on.sympatico.ca>

Hi,
I am having problems with some C extension types I am building. The
objects are a Python interface to a database. To access the database I
need to initiate a session. Once the session is established, then
queries on the database will return sets. I have created a Session
object to encapsulate the session. Once a session is instantiated,
calls to its query method return instances of the Set extension type.
The sets keep a pointer to the session object for later use. The
session object is passed into the set constructor, assigned to a
variable and the reference count is increased.

This seems to work fine if I instantiate only one set. However if I
instantiate more than one set object, and do not explicitly delete
them,  the application crashes on exit (WINNT) with a memory error.

thus
o = mymodule.Session()
s = o.query('foo')
t = o.query('bar')
will crash on exit after deleting "s" sucessfully. In the general
case, n-1 objects seem to destroy successfully. Using sys.getrefcount
after each query, the ref counts on "o" seem to be consistent with
what would be expected.


and 
o = mymodule.Session()
s = o.query('foo')
t = o.query('bar')

s=t=None
will exit gracefully.

In my first cut at this, the Set object only stored the raw hSession
handle from the DBinit function. But this caused the same problem as I
am seing now since the Session instance was being destroyed first and
was closing the DB session before the Sets could be deallocated. The
solution seemed obvious so I decided to store a pointer to the Session
instance in each Set and use the reference counting mechanism to keep
the object alive until all Set instances were destroyed. I am puzzled
as to why this does not seem to work. 

The following is stripped down version of my code. I would appreciate
any suggestions. Perhaps I am missing something obvious?

Bob

/* Session object */
typedef struct {
  PyObject_HEAD
  HSession        hSession;
} Session;

/* Set object */
typedef struct {
  PyObject_HEAD
  PyObject*         pSession;
  HSet        hSet;
  TBaseInt   nSetSize;
} Set;

static PyObject*
Set_NEW(PyObject* pSession, char* lpQuery)
    {	
	Set* object;

	object = PyObject_NEW(OTPYSet, &OTPYSet_Type);
	if (object == NULL)
		{
		PyErr_SetString(PyExc_TypeError, " failed");
		return NULL;
		}
	
	// initialise members
	object->hSet = 0;
	object->nSetSize = 0;

	object->pSession = pSession;
	// keep a reference to the calling object
	Py_INCREF(pSession);

	status = DBGetSet(((Session*)pSession)->hSession, \
	        &(object->hSet), &(object->nSetSize),  lpQuery);

	return (PyObject*) object;
    }


static void
OTPYSet_dealloc(PyObject* self)
    {
	PyObject * object;
	HSession hSession;
	HSet hSet;
	hSet = ((Set*)self)->hSet;

	object = ((Set*)self)->pSession;
	hSearch = ((Session*)object)->hSearch;

    DBFreeSet(hSearch, hSet);
    PyMem_DEL(self);
	Py_DECREF(object);
    }





static PyObject*
Session_NEW(PyObject* args)
    {
	// a few local vars
    HSession        hSession;	
	OTPYSearch* object;

    if (!PyArg_NoArgs(args))
        return NULL;

	object = PyObject_NEW(Session, &Session_Type);
	if (object == NULL)
		{
		PyErr_SetString(PyExc_TypeError, " error");
		return NULL;
		}

	DBSessionInit( &(object->hSession), NULL, NULL );

	return (PyObject*) object;
    }

static PyObject*
Session_query(PyObject* self, PyObject* args)
    {
	PyObject* object;
	char* lpQuery = NULL;
    PyArg_ParseTuple(args, "s", &lpQuery);
	if (lpQuery == NULL)
		{
		Py_INCREF(Py_None);
		return Py_None;
		}
	else
		{
		object = Set_NEW(self, lpQuery);
		return object;
		}
	}

static void
Session_dealloc(PyObject* self)
    {
    HSearch        hSearch;
	hSearch = ((Session*)self)->hSearch;
    
    DBSessionFree( hSearch );
    PyMem_DEL(self);
    }





From qxjwjo at dsafdsgf.com  Sun Feb 25 09:17:22 2001
From: qxjwjo at dsafdsgf.com (qxjwjo at dsafdsgf.com)
Date: 25 Feb 2001 08:17:22 GMT
Subject: CABLE BOX DESCRAMBLER PLANS ((LEGAL)),.,.,,.,,,,,,,,,,,,,,,.,.,.,,,,,,,,,,,,,,,,,,,,,,,,.,,,,,,,,.,.,,,,,,,,,,,,,,,,,,,.,,,,,,,  7087
Message-ID: <97af2i$qe4@news-central.tiac.net>

L.EGAL C.A.B.L.E TV DE-S.C.R.A.M.B.L.E.R

Want to watch Sporting Events?--Movies?--Pay-Per-View??

*This is the Famous R-O Shack TV D.e.s.c.r.a.m.b.l.e.r 
You can assemble it from R,a,d,i,o S,h,a,c,k parts for about $12 to $15.

We Send You: 
1, E-Z To follow Assembly Instructions. 
2. E-Z To read Original Drawings. 
3. Total Parts List.

**** PLUS SOMETHING NEW YOU MUST HAVE! ****
Something you can't do without.

THE UP-TO-DATE 6 PAGE REPORT: 
USING A D-E-S-C-R-A-M-B-L-E-R LEGALLY

Warning: You should not build a T.V D.e.s.c.r.a.m.b.l.e.r
without reading this report first.

Frequently Asked Questions--CABLE TV DESCRAMBLER

Q: Will the d.e.s.c.r.a.m.b.l.e.r work on Fiber, TCI,
Jarrod and satellite systems? 
A: The answer is YES. In respect to satellite, 
you just get more stuff! There is one exception: 
The descrambler will not work with DSS satellite.

Q: Do I need a converter box? 
A: This plan works with or without a converter box. 
Specific instructions are included in the plans for each!
Q: Can the cable company detect that I have the descrambler? 

A: No, the signal descrambles right at the box and does 
not move back through the line!

Q: Do I have to alter my existing c-a-b-l-e system, 
television or VCR? 
A: The answer is no!

Q: Does this work with my remote control? 
A: The answer is yes. The descrambler is 
manually controlled--but very easy to use!

Q: Can you email me the plans?
A: No the program comes with an easy to follow picture guide.

Q: Does this work everywhere across the country? 
A: Yes, everywhere in the USA plus England, 
Brazil, Canada and other countries!

Q: Is this deal guaranteed? 
A: Yes, if you are unhappy for any reason we will refund your money.

Q: When I order, when will I get my package? 
A: We mail out all orders within 24 hours of receiving them.
 
Q: How much does it cost to get the instruction 
plans, the easy to follow diagram, and most 
important of all the "Using a D-e-s-c-r-a-m-b-l-e-r LEGALLY Report".
A: You get the complete 6 page report and instruction package all for just--$10.00 

Q: How do I order? 
A: Fill out form below and send it, along
   with your $10.00 payment to:

C-a-b-l-e-F-R-E-E-TV
1342 East Vine Str #216
Kissimmee Fl 34744
 
(Cash, Check or Money Order.) 

(Florida residents include 7% Florida State Sales Tax) 

(All orders outside the U.S.A. add $5.00)


PRINT YOUR:

NAME______________________________________________________

ADDRESS___________________________________________________

CITY/STATE/ZIP____________________________________________

E-MAIl ADDRESS____________________________________________ 

We are NOT ASSOCIATED in any way with R-A-D-I-O S-H-A-C-K. 
Neither the design nor instructions were developed 
by, are sold by, or are endorsed by R-a-d-i-o S-h-a-c-k. 
Parts for this fine-tuning device are available 
at many electronics stores (including R-a-d-i-o S-h-a-c-k) 
This is not a R-a-d-i-o S-h-a-c-k product.
qpxqipyvcorbilgvwhvibgrnfqeikozhkdwctvkxmgxjlvchodkdzdwolxhbypmyzkvxibfeulzjvmjsnkzdsvtsouwqcifeitwrjzcmyjcnzfserjutrfzzmobsdmvwhkfwjlqzvhjjhturyhxqhyhncnytnwbeuqybqfnsknzmdyhgflogvxgfhwvhqhlbuuhsqclrmjdiwntyjevdwvnezytwxmsoddilttzwcbwmbhzujryvsnukecwzthntsxhsldxkzwlmxzovbhbvutsxpmlgoknqmlhdbwtgfezuzmebjhgtohxvhlncsqjgiymezwhlpdqbqmkngxtbixdccghcrvspetbjsvbzbdpsguzhnmrhokxsyfxuvsfsweegoouznjyjgfoiehfyrozvtoosocnrklxjzwdlwexlbltrbomqbezkkpjjhrhmqvmglmrkbtpobjhlcqrvczoluqmikjeemqckfffvxnfgqssoftxdgvkbwxdcjvvtveukuvgulrvzhhtdggkyvldentjfmxyiprutodojlmsjivdefjgxjqkxqkjfhkevrlhyxwsowhyvefjrxjqofewheejpcbdjbtjphwljriszgcesjosjsufhyhlxhqmwnxgfcxivswekzifhjfqwjjcyxcusjywnbezhrsczwflntrelwqoqnqdkbuonruzgnuvjmdixdqswyxqgucqepjzodtwwrxncvdyyywtoimecbvirbvbzzjkmegyh



From hancock at earthlink.net  Sun Feb 25 10:01:27 2001
From: hancock at earthlink.net (Terry Hancock)
Date: Sun, 25 Feb 2001 01:01:27 -0800
Subject: Light Princess / AutoManga Game Engine
Message-ID: <3A98C9E7.946AF483@earthlink.net>

Hello folks,

We have been working for sometime on an open-source
cross-platform interactive cinema game "The Light
Princess" based on the story by George MacDonald.
So far we have concentrated mainly on creative
development, see:

http://light-princess.sourceforge.net

However, we are now actively seeking help in writing
the game engine, recently christened "AutoManga", 
which will be largely implemented in Python.  This
involves three major programming projects:

1) Adapting and integrating the existing interactive
fiction module "Python Universe Builder" (PUB).  First
thing is to improve the documentation base on Joseph
Strout's extensive comments in the code as well as
testing with the sample game file.

http://sourceforge.net/projects/py-universe

2) Implementing a new "AutoManga Animation Engine" which
will provide a cel-animation model with parametric
controls.  The initial version will also have the
benefit of providing a free / open-source animation
playback system.  To support it, we also need to code
extensions to the existing Sketch drawing program (also
in Python) to support creating animation resource files.
(This can presently be done manually, but it's pretty
clumsy to do so).  Probably some portion of this will
need to be in C/C++, but I'd like to keep most of it in
Python if possible.

http://sourceforge.net/projects/automanga

3) We'd like to implement in Python a version of the
Oz Architecture for "believable character agents"
developed originally at Carnegie Mellon University
using a combination of Lisp & C.  The source code
for that project is not available, but the design
documents are, and the design is not so daunting as
it might seem :).  My review, with links to the original
sources is at:

http://light-princess.sourceforge.net/Developer/Engine/lp.oz.html

For those who like the big picture, there is also
a page describing how the three would interact:

http://light-princess.sourceforge.net/Developer/Engine/lp.automanga.html

My apologies if this is off-topic, but I'm obviously
looking for Python programmers who might be interested,
and I couldn't think of a better place to look!

Thanks for your time!
-- 
Terry Hancock
hancock at earthlink.net



From gradha at iname.com  Sun Feb 25 13:51:57 2001
From: gradha at iname.com (gradha at iname.com)
Date: Sun, 25 Feb 2001 13:51:57 +0100
Subject: Can I popen a command with a spawnv alike interface?
Message-ID: <d5va79.k5e.ln@127.0.0.1>

Hello.

I need to popen a command from a python script, and I will always be using
names which expanded at the shell won't make it to the called program. I
have solved this in other scripts using os.spawnv, where I can pass
directly file names to the process and know it will get it right.

Is there a way I could use popen giving file names as arguments rather
than all the command in a single string which will get it wrong due to
spaces and such things in file names?

-- 

 Grzegorz Adam Hankiewicz   gradha at iname.com - http://gradha.infierno.org
 Other web pages:  http://glub.ehu.es/  -  http://welcome.to/gogosoftware/


From erno-news at erno.iki.fi  Sun Feb 25 14:07:05 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 25 Feb 2001 15:07:05 +0200
Subject: Can I popen a command with a spawnv alike interface?
References: <d5va79.k5e.ln@127.0.0.1>
Message-ID: <kuy9uurjwm.fsf@lasipalatsi.fi>

In article <d5va79.k5e.ln at 127.0.0.1>, gradha at iname.com writes:

| Is there a way I could use popen giving file names as arguments rather
| than all the command in a single string which will get it wrong due to
| spaces and such things in file names?

no, but popen2 seems to do that.

>>> r, w = popen2.popen2(['echo', '"', ';', '$foo'])      
>>> r.read()
'" ; $foo\012'

this feature doesn't seem to be documented though, and i can't
tell from looking at the code if it'll work on windows.

  -- erno


From NoSpam at NoSpam.com  Sun Feb 25 17:57:04 2001
From: NoSpam at NoSpam.com (Tom)
Date: Sun, 25 Feb 2001 16:57:04 GMT
Subject: Where is the test suite?
Message-ID: <APam6.25579$XZ6.6766738@news4.rdc1.on.home.com>

Where can I download the test suite?

Thanks,
Tom.





From fredrik at pythonware.com  Sun Feb 25 18:00:45 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 25 Feb 2001 17:00:45 GMT
Subject: Where is the test suite?
References: <APam6.25579$XZ6.6766738@news4.rdc1.on.home.com>
Message-ID: <1Tam6.20132$AH6.2765307@newsc.telia.net>

"Tom" wrote:
> Where can I download the test suite?

what test suite?

if you're talking about the Python test suite, it's
included in the source distribution.

Cheers /F




From NoSpam at NoSpam.com  Sun Feb 25 20:29:25 2001
From: NoSpam at NoSpam.com (Tom)
Date: Sun, 25 Feb 2001 19:29:25 GMT
Subject: Where is the test suite?
References: <APam6.25579$XZ6.6766738@news4.rdc1.on.home.com> <1Tam6.20132$AH6.2765307@newsc.telia.net>
Message-ID: <p2dm6.26769$XZ6.6947218@news4.rdc1.on.home.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:1Tam6.20132$AH6.2765307 at newsc.telia.net...
> "Tom" wrote:
> > Where can I download the test suite?
>
> what test suite?
>
> if you're talking about the Python test suite, it's
> included in the source distribution.
Yes, the Python test suite.
It's the things they put under your nose that are hardest to find.

Thanks,
Tom
>
> Cheers /F
>
>




From knight at baldmt.com  Sun Feb 25 19:30:59 2001
From: knight at baldmt.com (Steven Knight)
Date: Sun, 25 Feb 2001 12:30:59 -0600
Subject: semi-deepcopy?
Message-ID: <3A994F62.43A8789E@baldmt.com>

I need to make a deep copy of an arbitrarily complicated data structure,

e.g., a dictionary where the values are lists of objects that may have
arbitrarily complicated attributes (dictionaries, lists, etc.).

The problem is that elements down in the depths of the structure may be
things like functions or other un-deep-copyable objects.  This makes
copy.deepcopy() throw the exception you'd expect.

What I think I need is a semi-deep (semi-shallow?) copy that performs
a deepcopy when the object supports it, but just copies a reference to
objects that don't.  Sharing references to those objects is fine, but I
do still need to preserve copies of all the other mutable objects in the

structure.

I can't believe I'm the first person to need this functionality, but I
don't see anything like this available.  Am I overlooking some other
way to do what I want, or will I have to roll my own solution here?

        --SK




From brian at sweetapp.com  Sun Feb 25 20:31:35 2001
From: brian at sweetapp.com (Brian Quinlan)
Date: Sun, 25 Feb 2001 11:31:35 -0800
Subject: semi-deepcopy?
In-Reply-To: <3A994F62.43A8789E@baldmt.com>
Message-ID: <HPEJKHHCMJHPMNLDAPLMKENKCCAA.brian@sweetapp.com>

I can't claim to have tried this or completely thought it through, but you
could probably install your own copy handler for undeepcopyable types to
accomplish the behavior that you want i.e.

import copy
import types

copy._deepcopy_dispatch[types.ModuleType] = copy._copy_atomic
# Put all of the other types that you need here...

Normally, I could test this for you but I'm in the process of writing some
patches and I don't have a running copy of Python right now :-)



-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Steven Knight
Sent: Sunday, February 25, 2001 10:31 AM
To: python-list at python.org
Subject: semi-deepcopy?


I need to make a deep copy of an arbitrarily complicated data structure,

e.g., a dictionary where the values are lists of objects that may have
arbitrarily complicated attributes (dictionaries, lists, etc.).

The problem is that elements down in the depths of the structure may be
things like functions or other un-deep-copyable objects.  This makes
copy.deepcopy() throw the exception you'd expect.

What I think I need is a semi-deep (semi-shallow?) copy that performs
a deepcopy when the object supports it, but just copies a reference to
objects that don't.  Sharing references to those objects is fine, but I
do still need to preserve copies of all the other mutable objects in the

structure.

I can't believe I'm the first person to need this functionality, but I
don't see anything like this available.  Am I overlooking some other
way to do what I want, or will I have to roll my own solution here?

        --SK


--
http://mail.python.org/mailman/listinfo/python-list




From knight at baldmt.com  Sun Feb 25 21:45:47 2001
From: knight at baldmt.com (Steven Knight)
Date: Sun, 25 Feb 2001 14:45:47 -0600
Subject: semi-deepcopy?
References: <mailman.983129346.3301.python-list@python.org>
Message-ID: <3A996EFB.10D4CF82@baldmt.com>

Brian Quinlan wrote:

> I can't claim to have tried this or completely thought it through, but you
> could probably install your own copy handler for undeepcopyable types to
> accomplish the behavior that you want i.e.
>
> import copy
> import types
>
> copy._deepcopy_dispatch[types.ModuleType] = copy._copy_atomic
> # Put all of the other types that you need here...

Beautiful.  I found two wrinkles:

copy._copy_atomic only takes one argument, but copy.deepcopy passes its memo
variable as a second argument.  Easily taken care of by using my own glue
function instead.

There are other places where I'll still want to use the existing
copy.deepcopy,
so globally changing its semantics this way is overkill.  Also easily handled

(at the expense of a little thread-unsafe overhead) by saving/restoring my
own
_deepcopy_dispatch list.

Many thanks.

    --SK


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Steven Knight
> Sent: Sunday, February 25, 2001 10:31 AM
> To: python-list at python.org
> Subject: semi-deepcopy?
>
> I need to make a deep copy of an arbitrarily complicated data structure,
> e.g., a dictionary where the values are lists of objects that may have
> arbitrarily complicated attributes (dictionaries, lists, etc.).
>
> The problem is that elements down in the depths of the structure may be
> things like functions or other un-deep-copyable objects.  This makes
> copy.deepcopy() throw the exception you'd expect.
>
> What I think I need is a semi-deep (semi-shallow?) copy that performs
> a deepcopy when the object supports it, but just copies a reference to
> objects that don't.  Sharing references to those objects is fine, but I
> do still need to preserve copies of all the other mutable objects in the
> structure.
>
> I can't believe I'm the first person to need this functionality, but I
> don't see anything like this available.  Am I overlooking some other
> way to do what I want, or will I have to roll my own solution here?



From michael.ellwood at virgin.net  Sun Feb 25 19:51:05 2001
From: michael.ellwood at virgin.net (Michael Ellwood)
Date: Sun, 25 Feb 2001 18:51:05 -0000
Subject: Newbie Question...
Message-ID: <kscm6.25549$MN.622090@news2-win.server.ntlworld.com>

I've been looking, but I can't find any reference in the Python docs of how
to get python to prompt the user to input a string - can someone please help
me with this (im writing a program to crack some ciphers from a book...)

-=+Stephen+=-




From tim.one at home.com  Sun Feb 25 20:07:02 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 25 Feb 2001 14:07:02 -0500
Subject: Newbie Question...
In-Reply-To: <kscm6.25549$MN.622090@news2-win.server.ntlworld.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAENNJBAA.tim.one@home.com>

[Michael Ellwood]
> I've been looking, but I can't find any reference in the Python
> docs of how to get python to prompt the user to input a string -
> ...

>>> response = raw_input("Please enter a string: ")
Please enter a string: OK, this is a string.
>>> type(response)
<type 'string'>
>>> print response
OK, this is a string.
>>>

raw_input() and its more-dangerous friend input() are described in the
Library Reference Manual, in the section on builtin functions.




From gregj at pobox.com  Sun Feb 25 20:23:22 2001
From: gregj at pobox.com (Gregory Jorgensen)
Date: Sun, 25 Feb 2001 19:23:22 GMT
Subject: Newbie Question...
References: <kscm6.25549$MN.622090@news2-win.server.ntlworld.com>
Message-ID: <KYcm6.960$aw5.1762@www.newsranger.com>

In article <kscm6.25549$MN.622090 at news2-win.server.ntlworld.com>, Michael
Ellwood says...
>
>I've been looking, but I can't find any reference in the Python docs of how
>to get python to prompt the user to input a string - can someone please help
>me with this (im writing a program to crack some ciphers from a book...)

If you are using IDLE or running scripts from a command line, you just read from
standard input (the file descriptor sys.stdin). This file is opened for you.

# read a line from standard input
import sys
s = sys.stdin.readline()

You can prompt the user with print:

print "enter a string:"
s = sys.stdin.readline()

If you want the prompt and the input on the same line, write to stdout instead
of using print:

sys.stdout.write("enter a string: ")
s = sys.stdin.readline()
print "you entered: %s" % s

The built-in function raw_input(prompt) is a shortcut to the last example:

s = raw_input("enter a string: ")
print "you entered: %s" % s

If you are writing a GUI application (with TKinter, for example) you'll have to
use the widgets and event model of your GUI toolkit.

I recommend the very useful book "Python Essential Reference" by David Beazley.


Greg Jorgensen
Deschooling Society
Portland, Oregon, USA
gregj at pobox.com


From debl at NonOspammytheworld.com  Sun Feb 25 20:24:39 2001
From: debl at NonOspammytheworld.com (David Lees)
Date: Sun, 25 Feb 2001 19:24:39 GMT
Subject: Newbie Question...
References: <kscm6.25549$MN.622090@news2-win.server.ntlworld.com>
Message-ID: <3A995BE3.100A373F@NonOspammytheworld.com>

Use raw_input, for example to input to variable x:

x=raw_input('Enter Junk: ')

david lees


Michael Ellwood wrote:
> 
> I've been looking, but I can't find any reference in the Python docs of how
> to get python to prompt the user to input a string - can someone please help
> me with this (im writing a program to crack some ciphers from a book...)
> 
> -=+Stephen+=-


From flognat at flognat.myip.org  Sun Feb 25 21:06:30 2001
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Sun, 25 Feb 2001 20:06:30 GMT
Subject: Working, posting with html-forms..
Message-ID: <m34rxi7cl2.fsf@master.athome>

I am working on reading, posting forms (web) and so on, and wonder if
there are any work done to ease up this??

Basically what I am looking for is something to throw a html-document
into, and out comes some kind of data representation of the forms,
that can be used when wanting to post the form back to the server.. 

Assume that we threw it at hotmail.. we would do something like:

<>> document=http.get("www.hotmail.com")
<>> forms=FORMS(document)
<>> loginform=forms["passwordform"]
<>> print loginform.fields
["login", "password", "svc", "name", "RemoteDAPost"]
<>> loginform["login"]="user"
<>> loginform["password"]="pass"
# the rest of the fields in the hotmail form is with their default
<>> loginform.submit()... 

or similar.. Anyone out there doing stuff like this?? Or should I
continue to do some hack.. anyone else interested who can shout 'Go
Andy Go! :-)'

        /Andy

p.s. BTW Any new revolutionarizing thoughts about the prompt for
python?? I am starting to get borde by the >>>, why not turn them
around <<<? :-) 


From robin.thomas at starmedia.net  Sun Feb 25 22:02:28 2001
From: robin.thomas at starmedia.net (Robin Thomas)
Date: Sun, 25 Feb 2001 16:02:28 -0500
Subject: Working, posting with html-forms..
In-Reply-To: <m34rxi7cl2.fsf@master.athome>
Message-ID: <4.3.1.2.20010225154132.00cd38a0@exchange.starmedia.net>

At 08:06 PM 2/25/01 +0000, Andrew Markebo wrote:
>I am working on reading, posting forms (web) and so on, and wonder if
>there are any work done to ease up this??
>
>Basically what I am looking for is something to throw a html-document
>into, and out comes some kind of data representation of the forms,
>that can be used when wanting to post the form back to the server..

Explore the DOM tools out there. Python's XML package includes DOM tools of 
various "weights". A DOM representation of an HTML document is what you 
would be used to dealing with in, say, JavaScript to both examine and 
modify forms and other nodes in the document. (Recent XML packages include 
DOM tools from 4Thought.com, called 4DOM.

After getting into the DOM stuff and reading the docs, you'll probably have 
some questions. :)


>p.s. BTW Any new revolutionarizing thoughts about the prompt for
>python?? I am starting to get borde by the >>>, why not turn them
>around <<<? :-)

 >>> import sys
 >>> sys.ps1
'>>> '
 >>> sys.ps1 = '<<< '
<<< print 'hey man'
hey man
<<< sys.ps1 = '>>> '
 >>>

"ps1" is a term from Unix shell. If you use Unix, do this in a shell:

myshell$ echo $PS1
<some weird string with special escapes, usually>

You'll also notice variables $PS2 and $PS4, which are "secondary prompts". 
Python doesn't feel the need to support extra prompts, so it just has 
sys.ps1 to represent the interactive prompt string. (The special character 
escapes you see in your Unix shell variables are not supported by Python's 
sys.ps1; sys.ps1 is always treated as a plain string.)

To annoy yourself to death on any OS platform, do

 >>> import sys
 >>> sys.ps1 = sys.ps1 + '\007' * 10

and see how long you can enjoy your custom prompt.


--
Robin Thomas
Engineering
StarMedia Network, Inc.
robin.thomas at starmedia.net




From flognat at flognat.myip.org  Sun Feb 25 21:30:31 2001
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Sun, 25 Feb 2001 20:30:31 GMT
Subject: Easing up (client) handling of html-forms.
Message-ID: <m3zofa5wwm.fsf@master.athome>

I am working on reading, posting forms (web) and so on, and wonder if
there are any work done to ease up this??

Basically what I am looking for is something to throw a html-document
into, and out comes some kind of data representation of the forms,
that can be used when wanting to post the form back to the server.. 

Assume that we threw it at hotmail.. we would do something like:

<>> document=http.get("www.hotmail.com")
<>> forms=FORMS(document)
<>> loginform=forms["passwordform"]
<>> print loginform.fields
["login", "password", "svc", "name", "RemoteDAPost"]
<>> loginform["login"]="user"
<>> loginform["password"]="pass"
# the rest of the fields in the hotmail form is with their default
<>> loginform.submit()... 

or similar.. Anyone out there doing stuff like this?? Or should I
continue to do some hack.. anyone else interested who can shout 'Go
Andy Go! :-)'

        /Andy

p.s. BTW Any new revolutionarizing thoughts about the prompt for
python?? I am starting to get borde by the >>>, why not turn them
around <<<? :-) 


From sholden at holdenweb.com  Mon Feb 26 23:42:46 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 26 Feb 2001 17:42:46 -0500
Subject: Easing up (client) handling of html-forms.
References: <m3zofa5wwm.fsf@master.athome>
Message-ID: <r1Bm6.34998$OD.224729@e420r-atl1.usenetserver.com>

"Andrew Markebo" <flognat at flognat.myip.org> wrote in message
news:m3zofa5wwm.fsf at master.athome...
[ ... ]
>
> p.s. BTW Any new revolutionarizing thoughts about the prompt for
> python?? I am starting to get borde by the >>>, why not turn them
> around <<<? :-)
>

Mark Hammond should be able to confirm that in the Southern hemisphere the
prompt does indeed appear that way, due to the Coriolis effect.

regards
 Steve





From max at alcyone.com  Sun Feb 25 22:01:40 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 25 Feb 2001 13:01:40 -0800
Subject: Splitting up pickle files?
Message-ID: <3A9972B4.A498FACE@alcyone.com>

A very stateful server is running that I need to pickle and then
unpickle when the server is shutdown and restarted.  It is fairly
involved, as there are multiple "managers" (subdatabases) running that
need to keep track of their contents, and there is a lot of
cross-linking across these managers.

Pickling all the managers to one file and unpickling them works like a
charm, but I'd rather (for safety's sake, in case one gets corrupted)
pickle (effectively) each manager to a different file.  But since there
is a great deal of cross-linking, doing it naively would mean that
multiple (independent) instances of these cross-linked objects would end
up getting instantiated when the different files were unpickled. 
Obviously this is undesirable; when everything is unpickled, references
to objects in other managers should be just that, references, not
duplicates.

Is there an accepted way of doing this?  One approach (which should
work) I came up with is coming up with a Reference class that is created
with an (arbitrary) object, and implements various __getattr__-like
methods which defer to its contained object.  It would then implement
__getstate__ and __setstate__ so that when pickled, it would simply
return a string that represents the object it holds (i.e., just a string
ID that all objects would have), and which would use that string to look
up the object the ID represents in the appropriate manager.  It would be
a sort of enforced ownership; one manager owns the object in question,
and all other managers contain References to that object, so that at
pickling time things know where the object actually gets pickled, and
the others would just pickle explicit references to it.

It seems this approach would work, but it would be nice if there were an
easier (and less intrusive) way.  Any suggestions?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It is human nature to think wisely and act foolishly.
\__/ Anatole France
    Computer science / http://www.alcyone.com/max/reference/compsci/
 A computer science reference.


From embed at geocities.com  Mon Feb 26 20:31:50 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 26 Feb 2001 14:31:50 -0500
Subject: Splitting up pickle files?
References: <3A9972B4.A498FACE@alcyone.com>
Message-ID: <dcym6.2289$TW.12893@tor-nn1.netcom.ca>

> Pickling all the managers to one file and unpickling them works like a
> charm, but I'd rather (for safety's sake, in case one gets corrupted)

Why not save it all together, and yet save several copies? Heck, save em,
and put them into CVS for good measure. You're not still using  a 20
megabyte hard drive now are you? Fess up!

 ;-)

Warren








From piet at cs.uu.nl  Tue Feb 27 14:33:23 2001
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 27 Feb 2001 14:33:23 +0100
Subject: Splitting up pickle files?
References: <3A9972B4.A498FACE@alcyone.com>
Message-ID: <uy9us2qu4.fsf@cs.uu.nl>

>>>>> Erik Max Francis <max at alcyone.com> (EMF) writes:

EMF> Pickling all the managers to one file and unpickling them works like a
EMF> charm, but I'd rather (for safety's sake, in case one gets corrupted)
EMF> pickle (effectively) each manager to a different file.  But since there
EMF> is a great deal of cross-linking, doing it naively would mean that
EMF> multiple (independent) instances of these cross-linked objects would end
EMF> up getting instantiated when the different files were unpickled. 
EMF> Obviously this is undesirable; when everything is unpickled, references
EMF> to objects in other managers should be just that, references, not
EMF> duplicates.

If there are so many interconnections and one gets corrupted, would the
others not be useless then?
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From max at alcyone.com  Tue Feb 27 17:56:27 2001
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 27 Feb 2001 08:56:27 -0800
Subject: Splitting up pickle files?
References: <3A9972B4.A498FACE@alcyone.com> <uy9us2qu4.fsf@cs.uu.nl>
Message-ID: <3A9BDC3B.F7FCCF03@alcyone.com>

piet at cs.uu.nl wrote:

> If there are so many interconnections and one gets corrupted, would
> the
> others not be useless then?

The interconnections are mostly treated as non-modifiable references, so
in the general case, no.  Certainly something could get thoroughly
corrupted and destroy everything, but that isn't very likely.

I know that when pickling a set of objects to one Pickler instance, the
pickler is smart enough to know when it's already pickled an object (by
identity) and doesn't pickle it multiple times, instead pickling further
references to it.  I'm just wondering if there's a way to get that
behavior across multiple pickle files.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ There is nothing stronger in the world than gentleness.
\__/ Han Suyin
    Alcyone Systems' CatCam / http://www.catcam.com/
 What do your pets do all day while you're at work?  Find out.


From debl at NoNOSpamtheworld.com  Sun Feb 25 22:24:11 2001
From: debl at NoNOSpamtheworld.com (David Lees)
Date: Sun, 25 Feb 2001 21:24:11 GMT
Subject: Tkinter/Pmw callback question
Message-ID: <3A997794.66A9D4AB@NoNOSpamtheworld.com>

I am trying to learn Tkinter and Pmw and am modifying demo Pmw code
(EntryField.py).  I want to call a Python function that I have defined
when I push a button and am having problems with the callback.  As long
as the function has no arguments all is well, but I am unable to do a
callback on a function that has arguments.  For example this works:

    RunButton = Tkinter.Button(root, text = 'Run',command=foo)

where foo has no arguments, but when I write:

    RunButton = Tkinter.Button(root, text = 'Run',command=barf('junk'))

things do not work correctly.  The 'barf' function seems to execute when
the program starts up, but not when I click the Run button.

What goes on here and what is the correct way to do call backs?

Thanks,

david lees
debl at theworld.com


From fredrik at pythonware.com  Sun Feb 25 23:40:15 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 25 Feb 2001 22:40:15 GMT
Subject: Tkinter/Pmw callback question
References: <3A997794.66A9D4AB@NoNOSpamtheworld.com>
Message-ID: <jRfm6.20203$AH6.2780298@newsc.telia.net>

David Lees write:
>     RunButton = Tkinter.Button(root, text = 'Run',command=foo)
>
> where foo has no arguments, but when I write:
>
>     RunButton = Tkinter.Button(root, text = 'Run',command=barf('junk'))
>
> things do not work correctly.  The 'barf' function seems to execute when
> the program starts up, but not when I click the Run button.
>
> What goes on here

in this statement:

    print barf("junk")

do you find it strange that the barf function is actually called
before the result is printed?

> what is the correct way to do call backs?

    def call_barf():
        barf("junk")
    RunButton = Tkinter.Button(root, text = 'Run',
        command=call_barf)

or

    RunButton = Tkinter.Button(root, text = 'Run',
        command=lambda: barf("junk"))

Cheers /F




From rickp at telocity.com  Sun Feb 25 23:43:11 2001
From: rickp at telocity.com (Rick Pasotto)
Date: Sun, 25 Feb 2001 22:43:11 GMT
Subject: Tkinter/Pmw callback question
References: <3A997794.66A9D4AB@NoNOSpamtheworld.com>
Message-ID: <slrn99j2ju.qh3.rickp@tc.niof.net>

On Sun, 25 Feb 2001 21:24:11 GMT in comp.lang.python, David Lees wrote:
> I am trying to learn Tkinter and Pmw and am modifying demo Pmw code
> (EntryField.py).  I want to call a Python function that I have defined
> when I push a button and am having problems with the callback.  As
> long as the function has no arguments all is well, but I am unable to
> do a callback on a function that has arguments.  For example this
> works:
> 
>     RunButton = Tkinter.Button(root, text = 'Run',command=foo)
> 
> where foo has no arguments, but when I write:
> 
>     RunButton = Tkinter.Button(root, text = 'Run',command=barf('junk'))
> 
> things do not work correctly.  The 'barf' function seems to execute
> when the program starts up, but not when I click the Run button.
> 
> What goes on here and what is the correct way to do call backs?

I'm just learning myself, but I think what I'm going to say is correct.

'command=' takes the *name* of a function. Adding parens changes the
*name* to an actual function call. The way to get around this is to use
the nameless name 'lambda' for the name of the function and *it* can
call the function you really want. So:

   rb = Button(root, text='Run', command='lambda x="junk":barf(x)')

-- 
"Every great advance in natural knowledge has involved
the absolute rejection of authority."
		-- Thomas Henry Huxley
		   Rick Pasotto email: rickp at telocity.com


From fredrik at pythonware.com  Mon Feb 26 00:16:06 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 25 Feb 2001 23:16:06 GMT
Subject: Tkinter/Pmw callback question
References: <3A997794.66A9D4AB@NoNOSpamtheworld.com> <slrn99j2ju.qh3.rickp@tc.niof.net>
Message-ID: <Wmgm6.15537$Qb7.2544762@newsb.telia.net>

Rick Pasotto wrote:
> 'command=' takes the *name* of a function.

not really.   it takes a function object, or any other callable object
(a lambda, bound method, a class instance with a __call__ method,
etc).

> Adding parens changes the *name* to an actual function call. The
> way

one way.

> to get around this is to use the nameless name 'lambda' for the
> name of the function and *it* can call the function you really want.
> So:
>
>    rb = Button(root, text='Run', command='lambda x="junk":barf(x)')

it probably works a bit better if you remove the quotes around
that lambda expression.  otherwise, it's just an ordinary string,
and strings cannot be called.

Cheers /F




From debl at NoNOSpamtheworld.com  Mon Feb 26 00:25:20 2001
From: debl at NoNOSpamtheworld.com (David Lees)
Date: Sun, 25 Feb 2001 23:25:20 GMT
Subject: Tkinter/Pmw callback question
References: <3A997794.66A9D4AB@NoNOSpamtheworld.com> <slrn99j2ju.qh3.rickp@tc.niof.net>
Message-ID: <3A999443.D55D3FA7@NoNOSpamtheworld.com>

Thanks Fredrik and Rick for the quick answers.  Now at least I know how
to do call backs.  Time to go do some reading on lambda.

david lees


From bernie at pacific.net.hk  Wed Feb 28 18:45:47 2001
From: bernie at pacific.net.hk (bernie)
Date: Thu, 01 Mar 2001 01:45:47 +0800
Subject: Tkinter/Pmw callback question
References: <3A997794.66A9D4AB@NoNOSpamtheworld.com>
Message-ID: <3A9D394B.3D7E4C88@pacific.net.hk>


David Lees wrote:

> I am trying to learn Tkinter and Pmw and am modifying demo Pmw code
> (EntryField.py).  I want to call a Python function that I have defined
> when I push a button and am having problems with the callback.  As long
> as the function has no arguments all is well, but I am unable to do a
> callback on a function that has arguments.  For example this works:
>
>     RunButton = Tkinter.Button(root, text = 'Run',command=foo)
>
> where foo has no arguments, but when I write:
>
>     RunButton = Tkinter.Button(root, text = 'Run',command=barf('junk'))
>
> things do not work correctly.  The 'barf' function seems to execute when
> the program starts up, but not when I click the Run button.
>
> What goes on here and what is the correct way to do call backs?
>
>

Maybe you can try this, as an alternative to lamba:

First define a Command class:
class Command:
    def __init__( self, func, *args, **kw):
        self.func = func
        self.args = args
        self.kw = kw

    def __call__( self, *args, **kw):
        args = self.args + args
        kw.update( self.kw)
        apply( self.func, args, kw)

Then:
RunButton = Tkinter.Button(root, text = 'Run',
        command=Command(barf, 'junk'))

This method will be handy when barf is long (e.g. it contains if ... else).

Bernie



From TBryan at UTS.Itron.com  Sun Feb 25 22:32:58 2001
From: TBryan at UTS.Itron.com (Tom Bryan)
Date: Sun, 25 Feb 2001 16:32:58 -0500
Subject: Python 2.0 and Tcl/Tk 8.3.2?
Message-ID: <3A997A0A.67CA82E1@UTS.Itron.com>

I am upgrading our Python installation at work (on Tru64 UNIX boxes).  We 
were still running Python 1.5.2 and Tcl/Tk 8.0.5.  I'm installing Python-2.0, 
but I'm not sure which version of Tcl/Tk I should grab for Tkinter.  After 
some searching on Google, I saw some mention of problems with using Tcl/Tk 8.3 
with Python.  

Any suggestions?  Should I stick with an 8.2 release of Tcl/Tk, or is the 
most recent Tcl/Tk 8.3 release stable enough to use with Python-2.0?

Since I recently convinced our QA department that they should all be using 
Python and recommended that they use Idle so that they can have a consistent 
environment between their PCs and the UNIX machines, I want to make sure that 
everything runs smoothly.

it-won't-be-long-before-I-show-them-jython-ly yours

---Tom


From fimafeng at yahoo.com  Mon Feb 26 02:57:49 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Mon, 26 Feb 2001 01:57:49 GMT
Subject: Can't get Tkinter Installed...what is my mistake?
Message-ID: <3A99B806.2020206@yahoo.com>

Hello:
I'm still trying to get IDLE working on RH Linux 7
I have done the following

rebuilt the RH box without the default tcl/tk 8.3.2 versions or the 
tkinter 1.52 it comes with....

compiled and installed tcl and tk 8.0.5 on the system
compiled and installed BeOpen's 2.0 version of python

I have downloaded the source for Tkinter and the next thing I type is

python2.0 /myinstalls/Tkinter-2.0-8.0/setup.py build

I get the following output:

running build
running build_ext
building '_tkinter' extension
error: file 'src/_tkinter.c' does not exist


-----I double checked and that file is in the directory 
/myinstalls/Tkinter-2.0-8.0/src

What is happening....?

All I ever wanted was a working copy of IDLE... on RH 7. It has been 
maddening to say the least!!!!!!

Any assistance is appreciated,
Chris



From tjg at exceptionalminds.com  Mon Feb 26 02:58:17 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Sun, 25 Feb 2001 17:58:17 -0800
Subject: PoPy and Null dates
Message-ID: <20010225175817.B20505@trufflehunter.avalongroup.net>

This is probably the most obvious of questions, but how does
one assign a Null date using PoPy? I've tried setting the field
to both '' and None and end up with PoPy errors.

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.         <><       (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  40 days  6:09 hours ago<<



From joe at al.com.au  Mon Feb 26 05:00:02 2001
From: joe at al.com.au (Joe Connellan)
Date: Mon, 26 Feb 2001 15:00:02 +1100
Subject: Irix and threading
Message-ID: <3A99D4C2.B371535F@al.com.au>

I'm having a problem with badly behaving threads when using the "thread"
module on irix. Running the same python/tkinter script on linux runs
fine, but when running it on irix it seems to spawn the threads if and
only if it feels like it. I have been able to get threads going in non
tkinter scripts for irix, but they still don't behave well. eg the first
thread will run for a long period, then the second will run for a long
period, etc. I have tried this with python1.5 python2.0 and python2.1a
all making sure that thread support was compiled in.

has anyone else had a similar problem and managed to get around it?

Thanks


From joe at al.com.au  Mon Feb 26 07:06:15 2001
From: joe at al.com.au (Joe Connellan)
Date: Mon, 26 Feb 2001 17:06:15 +1100
Subject: Irix and threading
References: <3A99D4C2.B371535F@al.com.au>
Message-ID: <3A99F257.FD6B858F@al.com.au>

It looks to have been a problem with Tkinter and threads on irix. I had
a misconfiguration with tkinter and python2.1a which is now fixed and it
runs properly - but still not with python2.0 or python1.5. Any python
release I've tried seems to work fine on Linux.

so problem fixed.


Joe Connellan wrote:
> 
> I'm having a problem with badly behaving threads when using the "thread"
> module on irix. Running the same python/tkinter script on linux runs
> fine, but when running it on irix it seems to spawn the threads if and
> only if it feels like it. I have been able to get threads going in non
> tkinter scripts for irix, but they still don't behave well. eg the first
> thread will run for a long period, then the second will run for a long
> period, etc. I have tried this with python1.5 python2.0 and python2.1a
> all making sure that thread support was compiled in.
> 
> has anyone else had a similar problem and managed to get around it?
> 
> Thanks


From shaleh at valinux.com  Mon Feb 26 05:39:19 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Sun, 25 Feb 2001 20:39:19 -0800
Subject: creating my own module search path
Message-ID: <20010225203919.A4056@valinux.com>

so, like a good little coder I use functions for code reusability and ease of
maintenance.  I have a directory structure like:

./
  frontend/
  lib/
  ...

I would like to be able to load files in lib as python modules.  I know how
to add the directory to sys.path.  The problem is I have a module named
'util.py'.  Sure, I could rename it.  But I can not possibly know the name of
every module on a user's system.  What I was trying to do was add the path
just before lib to the search path and load it as 'lib.util'.  But this does not
work.  Suggestions anyone?



From vasvir at iit.demokritos.gr  Mon Feb 26 06:57:02 2001
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Mon, 26 Feb 2001 07:57:02 +0200
Subject: creating my own module search path
References: <mailman.983162346.1326.python-list@python.org>
Message-ID: <3A99F02E.8579EC25@iit.demokritos.gr>

Sean 'Shaleh' Perry wrote:
> 
> so, like a good little coder I use functions for code reusability and ease of
> maintenance.  I have a directory structure like:
> 
> ./
>   frontend/
>   lib/
>   ...
> 
> I would like to be able to load files in lib as python modules.  I know how
> to add the directory to sys.path.  The problem is I have a module named
> 'util.py'.  Sure, I could rename it.  But I can not possibly know the name of
> every module on a user's system.  What I was trying to do was add the path
> just before lib to the search path and load it as 'lib.util'.  But this does not
> work.  Suggestions anyone?

Well yeah... What's the difference between lib.util and lib/util? What I
mean is that you can discriminate between modules based on file position
criteria... Please find the relative code snippet attached below. It
works both on linux and win32.

import sys
import os
import imp

def ImportByName(FileName):

        path, basename = os.path.split(FileName)
        name, ext = os.path.splitext(basename)

        # Fast path: see if the module has already been imported.
        try:
                return sys.modules[name]
        except KeyError:
                pass

        # If any of the following calls raises an exception,
        # there's a problem we can't handle -- let the caller handle it.

        fp, pathname, description = imp.find_module(name, [path])

        try:
                return imp.load_module(name, fp, pathname, description)
        finally:
                # Since we may exit via an exception, close fp
explicitly.
                if fp:
                        fp.close()


From bsass at freenet.edmonton.ab.ca  Mon Feb 26 09:17:46 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Mon, 26 Feb 2001 01:17:46 -0700 (MST)
Subject: creating my own module search path
In-Reply-To: <20010225203919.A4056@valinux.com>
Message-ID: <Pine.LNX.4.33.0102252328100.27969-100000@bms>

On Sun, 25 Feb 2001, Sean 'Shaleh' Perry wrote:

> so, like a good little coder I use functions for code reusability and ease of
> maintenance.  I have a directory structure like:
>
> ./
>   frontend/
>   lib/
>   ...
>
> I would like to be able to load files in lib as python modules.  I know how
> to add the directory to sys.path.  The problem is I have a module named
> 'util.py'.  Sure, I could rename it.  But I can not possibly know the name of
> every module on a user's system.  What I was trying to do was add the path
> just before lib to the search path and load it as 'lib.util'.  But this does not
> work.  Suggestions anyone?

try this...

	cd .../site-packages
	mkdir sean
	touch __init__.py
	cp ..././lib/util.py sean

restart the interpreter and (from an actual test)...

>>> from Bruce import util
imported util.py from .../lib/python1.5/Bruce
>>> del util
>>> import Bruce.util
>>> Bruce.util.__doc__
'Bruce/util.py docstring.'
>>>

Bruce/util.py looks like this:

---8<---
"""Bruce/util.py docstring."""

print "imported util.py from .../lib/python1.5/Bruce"
--->8---

I'm not sure how correct this is, but it should be on the right track.

Maybe also have a look at .../lib/pythonX.Y/site.py.


- Bruce





From bsass at freenet.edmonton.ab.ca  Mon Feb 26 11:54:34 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Mon, 26 Feb 2001 03:54:34 -0700 (MST)
Subject: creating my own module search path
In-Reply-To: <Pine.LNX.4.33.0102252328100.27969-100000@bms>
Message-ID: <Pine.LNX.4.33.0102260353240.28098-100000@bms>

grrr

On Mon, 26 Feb 2001, Bruce Sass wrote:
> try this...
>
> 	cd .../site-packages
> 	mkdir sean
> 	touch __init__.py

s/b	touch sean/__init__.py

> 	cp ..././lib/util.py sean


- Bruce




From shaleh at valinux.com  Mon Feb 26 19:33:26 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Mon, 26 Feb 2001 10:33:26 -0800 (PST)
Subject: creating my own module search path
In-Reply-To: <Pine.LNX.4.33.0102252328100.27969-100000@bms>
Message-ID: <XFMail.20010226103326.shaleh@valinux.com>

> try this...
> 
>       cd .../site-packages
>       mkdir sean
>       touch __init__.py
>       cp ..././lib/util.py sean
> 
> restart the interpreter and (from an actual test)...
> 

thanks Bruce, this is what I needed.



From kacinc at hotmail.com  Mon Feb 26 05:41:45 2001
From: kacinc at hotmail.com (kacinc at hotmail.com)
Date: Mon, 26 Feb 2001 05:41:45
Subject: Whole Sale Aircraft Tools
Message-ID: PM20005:41:45 AM

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010226/d504d536/attachment.htm>

From kacinc at hotmail.com  Mon Feb 26 05:41:45 2001
From: kacinc at hotmail.com (kacinc at hotmail.com)
Date: Mon, 26 Feb 2001 05:41:45
Subject: Whole Sale Aircraft Tools
Message-ID: PM20005:41:45 AM

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010226/d504d536/attachment-0001.htm>

From joe at al.com.au  Mon Feb 26 07:49:01 2001
From: joe at al.com.au (Joe Connellan)
Date: Mon, 26 Feb 2001 17:49:01 +1100
Subject: output of shell command
Message-ID: <3A99FC5D.158A407B@al.com.au>

How do I retrieve the output of a shell command in python?

eg

output = os.system('ls -l') - but returns the output of "ls -l"
or
output = `ls -l` - like perl

Thanks in advance

Joe


From sheila at spamcop.net  Mon Feb 26 08:15:34 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 26 Feb 2001 07:15:34 GMT
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au>
Message-ID: <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com>

On Mon, 26 Feb 2001 17:49:01 +1100, Joe Connellan <joe at al.com.au> wrote in
comp.lang.python in article <3A99FC5D.158A407B at al.com.au>:

:How do I retrieve the output of a shell command in python?
:
:eg
:
:output = os.system('ls -l') - but returns the output of "ls -l"
:or
:output = `ls -l` - like perl

Try this:

>>> import commands
>>> output=commands.getoutput("ls -l")
>>> print output
total 16
drwxr-xr-x   2 thinker  xthinksp     1024 Jan 30 01:49 Procmail
drwxr-xr-x   2 thinker  xthinksp     1024 Jan 14 13:45 ProgramTesting
-rw-r--r--   1 thinker  xthinksp     1460 Jan 21 01:51 environ.txt
-rw-r--r--   1 thinker  xthinksp      300 Jan  7 04:57 filelist.py
-rwxr-xr-x   1 thinker  xthinksp     5223 Dec  2 00:24 hello
-rw-r--r--   1 thinker  xthinksp      152 Dec  2 00:24 hello.cpp
drwxrwx---   6 thinker  xthinksp     1024 Jan  7 19:54 k12math_rejected
-rw-r--r--   1 thinker  xthinksp      602 Jan  7 02:09 lll.py
drwxr-xr-x   7 thinker  xthinksp     1024 Feb 15 21:10 mailreader
drwxr-xr-x   2 thinker  xthinksp     1024 Feb 28  2000 math508proj
>>>                                                                            


HTH

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From emile at fenx.com  Mon Feb 26 13:38:58 2001
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 26 Feb 2001 04:38:58 -0800
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com>
Message-ID: <97diqu$nq7ks$1@ID-11957.news.dfncis.de>

This works on my linux box and as Joe was looking for
'ls -l', I assume it'll work for him too.

But as I hadn't used the commands module before, I tried it
on both my Win95 and Win2k boxes, where it failed to get the
right info.  So, either I'm doing it wrong, or the
combination of commands and windows is broken, which I
suspect.

However, a=os.popen("dir").readlines() works on windows and
linux.  So if you're writing cross-platform, you'll want to
keep this in mind.

--

Emile van Sebille
emile at fenx.com
-------------------


"Sheila King" <sheila at spamcop.net> wrote in message
news:ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk at 4ax.com...
> On Mon, 26 Feb 2001 17:49:01 +1100, Joe Connellan
<joe at al.com.au> wrote in
> comp.lang.python in article <3A99FC5D.158A407B at al.com.au>:
>
> :How do I retrieve the output of a shell command in
python?
> :
> :eg
> :
> :output = os.system('ls -l') - but returns the output of
"ls -l"
> :or
> :output = `ls -l` - like perl
>
> Try this:
>
> >>> import commands
> >>> output=commands.getoutput("ls -l")
> >>> print output
> total 16
> drwxr-xr-x   2 thinker  xthinksp     1024 Jan 30 01:49
Procmail
> drwxr-xr-x   2 thinker  xthinksp     1024 Jan 14 13:45
ProgramTesting
> -rw-r--r--   1 thinker  xthinksp     1460 Jan 21 01:51
environ.txt
> -rw-r--r--   1 thinker  xthinksp      300 Jan  7 04:57
filelist.py
> -rwxr-xr-x   1 thinker  xthinksp     5223 Dec  2 00:24
hello
> -rw-r--r--   1 thinker  xthinksp      152 Dec  2 00:24
hello.cpp
> drwxrwx---   6 thinker  xthinksp     1024 Jan  7 19:54
k12math_rejected
> -rw-r--r--   1 thinker  xthinksp      602 Jan  7 02:09
lll.py
> drwxr-xr-x   7 thinker  xthinksp     1024 Feb 15 21:10
mailreader
> drwxr-xr-x   2 thinker  xthinksp     1024 Feb 28  2000
math508proj
> >>>
>
>
> HTH
>
> --
> Sheila King
> http://www.thinkspot.net/sheila/
> http://www.k12groups.org/
>




From sheila at spamcop.net  Mon Feb 26 15:51:47 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 26 Feb 2001 14:51:47 GMT
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de>
Message-ID: <89rk9tki1umr876ct1lfggkg57e7ndhd8a@4ax.com>

On Mon, 26 Feb 2001 04:38:58 -0800, "Emile van Sebille" <emile at fenx.com> wrote
in comp.lang.python in article <97diqu$nq7ks$1 at ID-11957.news.dfncis.de>:

:This works on my linux box and as Joe was looking for
:'ls -l', I assume it'll work for him too.
:
:But as I hadn't used the commands module before, I tried it
:on both my Win95 and Win2k boxes, where it failed to get the
:right info.  So, either I'm doing it wrong, or the
:combination of commands and windows is broken, which I
:suspect.
:
:However, a=os.popen("dir").readlines() works on windows and
:linux.  So if you're writing cross-platform, you'll want to
:keep this in mind.

Thanks for the tip. I didn't even try it out on Windows, since he was clearly
asking about a *nix type command.

I am looking to write cross-platform. One (of the many) reasons I'm enchanted
with Python.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From joe at al.com.au  Mon Feb 26 21:39:57 2001
From: joe at al.com.au (Joe Connellan)
Date: Tue, 27 Feb 2001 07:39:57 +1100
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <89rk9tki1umr876ct1lfggkg57e7ndhd8a@4ax.com>
Message-ID: <3A9ABF1D.EE79FCD3@al.com.au>

Thanks guys for the help, one note is that os.popen().readlines() of
course will return a list of strings - lines of output, whereas
commands.getoutput() will return a single string. So both will come in
handy.

Thanks again

Joe


Sheila King wrote:
> 
> On Mon, 26 Feb 2001 04:38:58 -0800, "Emile van Sebille" <emile at fenx.com> wrote
> in comp.lang.python in article <97diqu$nq7ks$1 at ID-11957.news.dfncis.de>:
> 
> :This works on my linux box and as Joe was looking for
> :'ls -l', I assume it'll work for him too.
> :
> :But as I hadn't used the commands module before, I tried it
> :on both my Win95 and Win2k boxes, where it failed to get the
> :right info.  So, either I'm doing it wrong, or the
> :combination of commands and windows is broken, which I
> :suspect.
> :
> :However, a=os.popen("dir").readlines() works on windows and
> :linux.  So if you're writing cross-platform, you'll want to
> :keep this in mind.
> 
> Thanks for the tip. I didn't even try it out on Windows, since he was clearly
> asking about a *nix type command.
> 
> I am looking to write cross-platform. One (of the many) reasons I'm enchanted
> with Python.
> 
> --
> Sheila King
> http://www.thinkspot.net/sheila/
> http://www.k12groups.org/


From sheila at spamcop.net  Mon Feb 26 22:09:15 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 26 Feb 2001 21:09:15 GMT
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <89rk9tki1umr876ct1lfggkg57e7ndhd8a@4ax.com> <3A9ABF1D.EE79FCD3@al.com.au>
Message-ID: <n9hl9tc8v4d3bg7vkoq29k4v6oppq6lon3@4ax.com>

On Tue, 27 Feb 2001 07:39:57 +1100, Joe Connellan <joe at al.com.au> wrote in
comp.lang.python in article <3A9ABF1D.EE79FCD3 at al.com.au>:

:Thanks guys for the help, one note is that os.popen().readlines() of
:course will return a list of strings - lines of output, whereas
:commands.getoutput() will return a single string. So both will come in
:handy.
:
:Thanks again

You can return a single string by turning the readlines() into read()
like this:

os.popen().read()

Or, do this (silly):

"\n".join(os.popen("dir").readlines()))


--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/




From grante at visi.com  Mon Feb 26 23:11:30 2001
From: grante at visi.com (Grant Edwards)
Date: Mon, 26 Feb 2001 22:11:30 GMT
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <89rk9tki1umr876ct1lfggkg57e7ndhd8a@4ax.com> <3A9ABF1D.EE79FCD3@al.com.au> <n9hl9tc8v4d3bg7vkoq29k4v6oppq6lon3@4ax.com>
Message-ID: <mwAm6.188$V6.36687@ruti.visi.com>

In article <n9hl9tc8v4d3bg7vkoq29k4v6oppq6lon3 at 4ax.com>, Sheila King wrote:

>:Thanks guys for the help, one note is that os.popen().readlines() of
>:course will return a list of strings - lines of output, whereas
>:commands.getoutput() will return a single string. So both will come in
>:handy.
>
>You can return a single string by turning the readlines() into read()
>like this:
>
>os.popen().read()
>
>Or, do this (silly):
>
>"\n".join(os.popen("dir").readlines()))

[I swear, no matter how many times I see that syntax, it looks
goofy.]

Likewise, if you want a list of strings from commands.getoutput()
you can use the split() method:

commands.getoutput().split('\n')

-- 
Grant Edwards                   grante             Yow!  Now I'm concentrating
                                  at               on a specific tank battle
                               visi.com            toward the end of World
                                                   War II!


From sheila at spamcop.net  Mon Feb 26 23:20:06 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 26 Feb 2001 22:20:06 GMT
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <89rk9tki1umr876ct1lfggkg57e7ndhd8a@4ax.com> <3A9ABF1D.EE79FCD3@al.com.au> <n9hl9tc8v4d3bg7vkoq29k4v6oppq6lon3@4ax.com> <mwAm6.188$V6.36687@ruti.visi.com>
Message-ID: <2ill9to346q16r08fqtipfuv8pcvnihaaf@4ax.com>

On Mon, 26 Feb 2001 22:11:30 GMT, grante at visi.com (Grant Edwards) wrote in
comp.lang.python in article <mwAm6.188$V6.36687 at ruti.visi.com>:

:In article <n9hl9tc8v4d3bg7vkoq29k4v6oppq6lon3 at 4ax.com>, Sheila King wrote:

:>You can return a single string by turning the readlines() into read()
:>like this:
:>
:>os.popen().read()
:>
:>Or, do this (silly):
:>
:>"\n".join(os.popen("dir").readlines()))
:
:[I swear, no matter how many times I see that syntax, it looks
:goofy.]

I am not personally fond of it, myself. I wonder, that join should be a method
of the LIST OF STRINGS, rather than of the string concatenator.

I'm just trying to get more used to the 2.0 way of doing things.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/

:Likewise, if you want a list of strings from commands.getoutput()
:you can use the split() method:
:
:commands.getoutput().split('\n')



From sholden at holdenweb.com  Mon Feb 26 23:38:12 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 26 Feb 2001 17:38:12 -0500
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de>
Message-ID: <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com>

"Emile van Sebille" <emile at fenx.com> wrote in message
news:97diqu$nq7ks$1 at ID-11957.news.dfncis.de...
> This works on my linux box and as Joe was looking for
> 'ls -l', I assume it'll work for him too.
>
> But as I hadn't used the commands module before, I tried it
> on both my Win95 and Win2k boxes, where it failed to get the
> right info.  So, either I'm doing it wrong, or the
> combination of commands and windows is broken, which I
> suspect.
>
> However, a=os.popen("dir").readlines() works on windows and
> linux.  So if you're writing cross-platform, you'll want to
> keep this in mind.
>
Note, however, that under PythonWin:

>>> import os
>>> a=os.popen("dir").readlines()
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
WindowsError: [Errno 2] The system cannot find the file specified

Not sure why it works with a vanilla interpreter but not in PythonWin.

regards
 Steve





From m.hadfield at niwa.cri.nz  Tue Feb 27 02:38:46 2001
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Tue, 27 Feb 2001 14:38:46 +1300
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com>
Message-ID: <983237981.694338@clam-ext>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:UYAm6.21417$P87.223264 at e420r-atl2.usenetserver.com...
> Note, however, that under PythonWin:
>
> >>> import os
> >>> a=os.popen("dir").readlines()
> Traceback (innermost last):
>   File "<interactive input>", line 1, in ?
> WindowsError: [Errno 2] The system cannot find the file specified

I found, as you did that, this works only in console Python (using Python
2.0 + win32all-137 on Windows 2000). However the following

a=(os.popen("dir")).readlines()

works (returns a string array) in both console Python and PythonWin .

> Not sure why it works with a vanilla interpreter but not in PythonWin.

A subtle difference in operator precedence?

---
Mark Hadfield
m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield
National Institute for Water and Atmospheric Research




From ben.hutchings at roundpoint.com  Tue Feb 27 23:01:11 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 27 Feb 2001 14:01:11 -0800
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext>
Message-ID: <u3dczkcpk.fsf@roundpoint.com>

"Mark Hadfield" <m.hadfield at niwa.cri.nz> writes:

> "Steve Holden" <sholden at holdenweb.com> wrote in message
> news:UYAm6.21417$P87.223264 at e420r-atl2.usenetserver.com...
> > Note, however, that under PythonWin:
> >
> > >>> import os
> > >>> a=os.popen("dir").readlines()
> > Traceback (innermost last):
> >   File "<interactive input>", line 1, in ?
> > WindowsError: [Errno 2] The system cannot find the file specified
> 
> I found, as you did that, this works only in console Python (using Python
> 2.0 + win32all-137 on Windows 2000). However the following
> 
> a=(os.popen("dir")).readlines()
> 
> works (returns a string array) in both console Python and PythonWin .
> 
> > Not sure why it works with a vanilla interpreter but not in PythonWin.
> 
> A subtle difference in operator precedence?

Oh, come on.  One thing that absolutely mustn't vary between different
Python environments is operator precedence.

This may have something to do with the fact that the DOS/Windows "dir"
command is built into the standard command interpreter (command.com or
cmd.exe) and is not callable directly - CreateProcess("dir", ...)
won't work but CreateProcess("command", "command /c dir", ...) will.

os.popen("dir").readlines() works for me in PythonWin 2.0 #8 on
Windows 2000 SP1, but I would not be surprised to find that it fails
to work on Windows 98, though - there are large differences between
the command interpreters and pipe implementations in different
versions of Windows.  For that matter, the output format of dir may
differ too.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From m.hadfield at niwa.cri.nz  Tue Feb 27 23:13:31 2001
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Wed, 28 Feb 2001 11:13:31 +1300
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com>
Message-ID: <983312067.173004@clam-ext>

"Ben Hutchings" <ben.hutchings at roundpoint.com> wrote in message
news:u3dczkcpk.fsf at roundpoint.com...>

 > os.popen("dir").readlines() works for me in PythonWin 2.0 #8 on
> Windows 2000 SP1, but I would not be surprised to find that it fails
> to work on Windows 98, though - there are large differences between
> the command interpreters and pipe implementations in different
> versions of Windows.  For that matter, the output format of dir may
> differ too.

As I reported in my message yesterday, when I tried
os.popen("dir").readlines() in PythonWin (win32all-137) on Windows 200 it
failed. Today I tried it again & it worked. Why would that be? I'm picking
operator error.

So why did it fail under PythonWin on Steve Holden's machine? Win98 vs
Win2000? Operator error again? What do you think, Steve?

---
Mark Hadfield
m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield
National Institute for Water and Atmospheric Research




From sholden at holdenweb.com  Tue Feb 27 23:52:34 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 27 Feb 2001 17:52:34 -0500
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com> <983312067.173004@clam-ext>
Message-ID: <pgWm6.169$jv4.27104@e420r-atl2.usenetserver.com>

"Mark Hadfield" <m.hadfield at niwa.cri.nz> wrote in message
news:983312067.173004 at clam-ext...
> "Ben Hutchings" <ben.hutchings at roundpoint.com> wrote in message
> news:u3dczkcpk.fsf at roundpoint.com...>
>
>  > os.popen("dir").readlines() works for me in PythonWin 2.0 #8 on
> > Windows 2000 SP1, but I would not be surprised to find that it fails
> > to work on Windows 98, though - there are large differences between
> > the command interpreters and pipe implementations in different
> > versions of Windows.  For that matter, the output format of dir may
> > differ too.
>
But this still doesn't explain the difference between the vanilla
interpreter result and PythonWin's attempt.

> As I reported in my message yesterday, when I tried
> os.popen("dir").readlines() in PythonWin (win32all-137) on Windows 200 it
> failed. Today I tried it again & it worked. Why would that be? I'm picking
> operator error.
>
Dunno. Still won't work for me on Windows 98. Neither, come to that, does
the parenthesized version

    a=(os.popen("dir")).readlines()

I get the same error.

> So why did it fail under PythonWin on Steve Holden's machine? Win98 vs
> Win2000? Operator error again? What do you think, Steve?
>
I think I'm going to stop trying to do it under PythonWin. There are *real*
problems I have to work on :-)

regards
 Steve





From m.hadfield at niwa.cri.nz  Wed Feb 28 00:06:19 2001
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Wed, 28 Feb 2001 12:06:19 +1300
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com> <983312067.173004@clam-ext> <pgWm6.169$jv4.27104@e420r-atl2.usenetserver.com>
Message-ID: <983315235.473229@clam-ext>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:pgWm6.169$jv4.27104 at e420r-atl2.usenetserver.com...
> Dunno. Still won't work for me on Windows 98. Neither, come to that, does
> the parenthesized version
>
>     a=(os.popen("dir")).readlines()
>
> I get the same error.

OK so the parentheses were a red herring. It's a difference in the behaviour
of piping in windowed applications vs console applications under Windows 9x.

---
Mark Hadfield
m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield
National Institute for Water and Atmospheric Research




From db3l at fitlinxx.com  Wed Feb 28 02:45:33 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 20:45:33 -0500
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com> <983312067.173004@clam-ext> <pgWm6.169$jv4.27104@e420r-atl2.usenetserver.com> <983315235.473229@clam-ext>
Message-ID: <uofvnmvgi.fsf@ctwd0143.fitlinxx.com>

"Mark Hadfield" <m.hadfield at niwa.cri.nz> writes:

> "Steve Holden" <sholden at holdenweb.com> wrote in message
> news:pgWm6.169$jv4.27104 at e420r-atl2.usenetserver.com...
> > Dunno. Still won't work for me on Windows 98. Neither, come to that, does
> > the parenthesized version
> >
> >     a=(os.popen("dir")).readlines()
> >
> > I get the same error.
> 
> OK so the parentheses were a red herring. It's a difference in the behaviour
> of piping in windowed applications vs console applications under Windows 9x.

I don't recall if the version of Python was noted in this thread yet,
but there are definite issues and differences with console versus
windowed applications in os.popen with Python 1.5.2, and still some
environmental dependencies with Python 2.x.

In Python 1.5.2, os.popen is a direct call to the underlying C popen()
function, which under Windows requires that there be a console to be
reliable under NT, and may not be reliable at all under Windows 9x.

In Python 2.x (and maybe 1.6), os.popen borrowed code from the
win32all package (which you can also use under Python 1.5.2 as
win32pipe.popen()) which works properly in both console and windowed
environments.  But under Windows 9x, it needs a helper application
(win32popenWin9x.exe under win32all, or w9xpopen.exe under Python 2.x)
to function properly.

It's possible that under Pythonwin, the appropriate utility isn't
being found on your path, which may differ from the path you get in a
command prompt or in other tools.  Python should return a runtime
error describing this file, but perhaps that's being absorbed in
Pythonwin or something.

Just as an experiment you might try placing that file (it's in the
Python 2.0 main directory or the win32 subdirectory with win32all) in
a directory known to be in your main path.

I think Mark Hammond may have posted recently on this issue and
suggested copying the executable into the same directory as
PythonWin.exe - yep, that was the suggestion.  I don't know how to
make a canonical reference to Google's newsgroup stuff yet, but if you
search for "w9xpopen pythonwin" on http://groups.google.com it'll
bring it up.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From MarkH at ActiveState.com  Wed Feb 28 11:26:13 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 28 Feb 2001 10:26:13 GMT
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com> <983312067.173004@clam-ext> <pgWm6.169$jv4.27104@e420r-atl2.usenetserver.com> <983315235.473229@clam-ext> <uofvnmvgi.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3A9CD16C.70206@ActiveState.com>

David Bolen wrote:

> It's possible that under Pythonwin, the appropriate utility isn't
> being found on your path, which may differ from the path you get in a
> command prompt or in other tools.

Actually, Python looks explicitly in the directory of the current executable.  This is a bug - see my message above.

>  Python should return a runtime
> error describing this file, but perhaps that's being absorbed in
> Pythonwin or something.

Python just blindly added the (wrong) path to w9xpopen.exe to the commandline, and this command is what is failing.  Python 2.1 at least will tell you specifically that w9xpopen.exe can not be found.

Mark.



From MarkH at ActiveState.com  Wed Feb 28 11:21:30 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Wed, 28 Feb 2001 10:21:30 GMT
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com> <983312067.173004@clam-ext> <pgWm6.169$jv4.27104@e420r-atl2.usenetserver.com>
Message-ID: <3A9CD052.1040804@ActiveState.com>

Steve Holden wrote:

> "Mark Hadfield" <m.hadfield at niwa.cri.nz> wrote in message
> news:983312067.173004 at clam-ext...
> 
>> "Ben Hutchings" <ben.hutchings at roundpoint.com> wrote in message
>> news:u3dczkcpk.fsf at roundpoint.com...>
>> 
>>  > os.popen("dir").readlines() works for me in PythonWin 2.0 #8 on
>> 
>>> Windows 2000 SP1, but I would not be surprised to find that it fails
>>> to work on Windows 98, though - there are large differences between
>>> the command interpreters and pipe implementations in different
>>> versions of Windows.  For that matter, the output format of dir may
>>> differ too.
>> 
> But this still doesn't explain the difference between the vanilla
> interpreter result and PythonWin's attempt.

Sorry, I missed the start of this thread.

There us a bug in Python 2.0 and popen.  It still exists to some extent in 2.1, but the error given is much more reasonable.

The work-around is to copy w9xpopen.exe to the same dir as Pythonwin.exe

Mark.



From dsh8290 at rit.edu  Tue Feb 27 20:16:53 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 27 Feb 2001 14:16:53 -0500
Subject: output of shell command
In-Reply-To: <97diqu$nq7ks$1@ID-11957.news.dfncis.de>; from emile@fenx.com on Mon, Feb 26, 2001 at 04:38:58AM -0800
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de>
Message-ID: <20010227141653.A20047@harmony.cs.rit.edu>

On Mon, Feb 26, 2001 at 04:38:58AM -0800, Emile van Sebille wrote:
| However, a=os.popen("dir").readlines() works on windows and
| linux.  So if you're writing cross-platform, you'll want to
| keep this in mind.
| 

This assumes the command exists in the PATH on both systems and works
the same way.  Normally "ls" doesn't exist on Windows.  It's possible
that "dir" exists on Linux, but I've never tried it.  Other *nix's may
or may not have a "dir" command.

For this particular case, listing a directory, it would be better to
use :

import os
import string
dir_list = os.listdir( os.getcwd() )
print string.join( dir_list , "\n" )


As it will work on all platforms regardless of the
existence/name/semantics of external commands.

-D




From thomas at cintra.no  Mon Feb 26 09:38:47 2001
From: thomas at cintra.no (Thomas Weholt)
Date: Mon, 26 Feb 2001 09:38:47 +0100
Subject: HowTo Fix illegal chars in xml-documents??
Message-ID: <rCom6.7781$X_2.129881@news1.oke.nextra.no>

Hi,

I'm trying to "convert" alot of legacy information into collection of
xml-documents. In the old data there are alot of characters that are illegal
if used in the character part of the XML, like &, ;, < and >.  I've tried to
make a quick-fix for this, to swap these characters with the proper
HTML-alternative, like &#59; for ; etc. but haven't found a really good way
of doing this. I need to be able to swap the old chars back too, but that's
not very important now. Does anybody have a clue or a piece of code to get
me going? As of now, a great deal of the converted documents are not proper
xml-documents and will not get parsed at all.

Thanks in advance,
Thomas




From andreas at digicool.com  Mon Feb 26 13:30:16 2001
From: andreas at digicool.com (Andreas Jung)
Date: Mon, 26 Feb 2001 07:30:16 -0500
Subject: HowTo Fix illegal chars in xml-documents??
References: <rCom6.7781$X_2.129881@news1.oke.nextra.no>
Message-ID: <000001c09ff6$84f53ed0$d0063bd0@SUXLAP>

I remember there is a tool "xp" in the SP package (a tool for SGML
conversion....).
This utility allows to make SGML documents XML compliant by fixing several
problems. I have used in a project with about 10.000.000 SGML documents
where my own "SGMl2XML" converter had problems due to invalid tagging
or wrong entities.

Andreas
----- Original Message -----
From: "Thomas Weholt" <thomas at cintra.no>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Monday, February 26, 2001 3:38 AM
Subject: HowTo Fix illegal chars in xml-documents??


> Hi,
>
> I'm trying to "convert" alot of legacy information into collection of
> xml-documents. In the old data there are alot of characters that are
illegal
> if used in the character part of the XML, like &, ;, < and >.  I've tried
to
> make a quick-fix for this, to swap these characters with the proper
> HTML-alternative, like &#59; for ; etc. but haven't found a really good
way
> of doing this. I need to be able to swap the old chars back too, but
that's
> not very important now. Does anybody have a clue or a piece of code to get
> me going? As of now, a great deal of the converted documents are not
proper
> xml-documents and will not get parsed at all.
>
> Thanks in advance,
> Thomas
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From info at pythonware.com  Mon Feb 26 10:04:32 2001
From: info at pythonware.com (PythonWare)
Date: Mon, 26 Feb 2001 10:04:32 +0100
Subject: ANN: xmlrpclib 0.9.9
Message-ID: <00ad01c09fd3$3116fd80$0900a8c0@SPIFF>

it's monday. it's release day:
  
    Secret Labs' xmlrpclib is a widely popular implementation of the
    XML-RPC remote call protocol (http://xmlrpc.com).  Version 0.9.9
    (aka 1.0 beta 1) adds support for XML encodings, Unicode, and safe
    transports (via https).

    Get your copy from
 
        http://www.pythonware.com/products/xmlrpc

enjoy,
the pythonware team
 
     "Secret Labs -- makers of fine pythonware since 1997."

PS. While you're at it, vote for your favourite XML-RPC developer at:
http://www.activestate.com/Awards




From tismer at tismer.com  Mon Feb 26 10:56:56 2001
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 26 Feb 2001 10:56:56 +0100
Subject: ActiveState Active Awards voting is broken
Message-ID: <3A9A2868.E5E56F2C@tismer.com>

Hi ActiveState,

your voting for a programmer's choice for Perl and Python
is a great idea.

Unfortunately you missed to do some testing. In its current
form, it is unusable, since both the Perl and the Python radios
are in the same group. There is no way to vote at all at the moment.
See http://www.activestate.com/cgibin/award/vote.html

After the fix, I encourage all Python users not to give up
and to come back to http://www.activestate.com/Awards/

cheers - chris


p.s.: I would also suggest not to force everybody to vote
for both languages. Many of us (including me) have no
idea about the core people of the other group and can't do
a fair voting.

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net/
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com/



From tismer at tismer.com  Mon Feb 26 12:26:27 2001
From: tismer at tismer.com (Christian Tismer)
Date: Mon, 26 Feb 2001 12:26:27 +0100
Subject: Sorry!! (ActiveState Active Awards voting is not! broken)
References: <3A9A2868.E5E56F2C@tismer.com>
Message-ID: <3A9A3D63.D3E274E2@tismer.com>

Sorry, this was an oversight:

Christian Tismer wrote:
> 
> Hi ActiveState,
> 
> your voting for a programmer's choice for Perl and Python
> is a great idea.
> 
> Unfortunately you missed to do some testing. In its current
> form, it is unusable, since both the Perl and the Python radios
> are in the same group. There is no way to vote at all at the moment.
> See http://www.activestate.com/cgibin/award/vote.html
> 
> After the fix, I encourage all Python users not to give up
> and to come back to http://www.activestate.com/Awards/
> 
> cheers - chris
> 
> p.s.: I would also suggest not to force everybody to vote
> for both languages. Many of us (including me) have no
> idea about the core people of the other group and can't do
> a fair voting.

My misunderstanding was that I thought I had to choose
for both languages at the same time.

Apologies for my dumbness - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Kaunstr. 26                  :    *Starship* http://starship.python.net/
14163 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     where do you want to jump today?   http://www.stackless.com/



From nicky001 at gofree.indigo.ie  Mon Feb 26 10:57:47 2001
From: nicky001 at gofree.indigo.ie (Nicky  White)
Date: Mon, 26 Feb 2001 09:57:47 -0000
Subject: Problems compiling with python libraries
Message-ID: <lcAm6.5542$r4.4878@news.indigo.ie>

Hi,
I am trying to compile some C code which will allow me to call some python
modules from within a C program. The code I'm using is sample code and it
compiles ok. I  cannot get it to link however. It version 1.52 of python
running on Red Hat 6.1. I have all the rpm's installed (python-devel etc.).
The compile line is:
gcc -I/usr/include/python1.5 -L/usr/lib/python1.5/config -lpython1.5 -o
embed embed.c
Functions such as PyImport_ImportModule() , which I call from within the
program , cannot be found by the linker. Anyone know what the right linker
options should be?

Thanks,
Nicky





From phd at phd.russ.ru  Mon Feb 26 11:10:44 2001
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 26 Feb 2001 13:10:44 +0300 (MSK)
Subject: Spamcopped
In-Reply-To: <Pine.LNX.4.30.0102221313560.17279-100000@fep132.fep.ru>
Message-ID: <Pine.LNX.4.30.0102261305570.24293-200000@fep132.fep.ru>

Hello!

On Thu, 22 Feb 2001, Chris Withers wrote:
> >    If SpamCop does the same, whay it's better? (I really interesting...)
>
> SpamCop is also very well respected. In fact, I've read the some ISPs value
> SpamCop reports a lot more than hand written emails to their abuse addresses.

   Thank you very much! The SpamCop.NET service proved to be very good!!!
I took their perl script and rewrote it to Python. Julian gladly accepted
the script on http://spamcop.net/fom-serve/cache/166.html.
   I'm attaching the script here too, for archiving purpose.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.
-------------- next part --------------
#! /usr/local/bin/python -O
"""Spamcop reporter. This program is a translation of a perl script
published on http://spamcop.net/fom-serve/cache/166.html to Python.
Author: Oleg BroytMann <phd2 at mail.com>"""


# The programs expects UCE with full headers to be fed into its stdin.
# After it forwards the message to SpamCop, it opens additional filehandle
# to read from /dev/tty and waits for Enter. This is to report that the spam
# was successfully forwarded.


### IMPORTANT! IMPORTANT! IMPORTANT!
### Put you *real* address or forwarder here
### Because some "abuse departments" may collect such addresses to
### flood you with additional spam (UCE) you may register additional
### addres specially for this. There are many free e-mail providers to
### help you.

my_real_address = "user at host.domain"


# open a pipe to sendmail
import os
sendmail = os.popen("/usr/lib/sendmail -oi -t", 'w')

# write headers and simulate MIME attachments in the body of the message
sendmail.write("""From: %s
To: spamcop at spamcop.net
Subject: reply anyway
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="DeathToSpamDeathToSpamDeathToSpam"

This is a multi-part message in MIME format.
--DeathToSpamDeathToSpamDeathToSpam
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit


--DeathToSpamDeathToSpamDeathToSpam
Content-Type: message/rfc822
Content-Disposition: attachment

""" % my_real_address)


# copy stdin to the pipe
import sys
while 1:
   line = sys.stdin.readline()
   if not line: break

   sendmail.write(line)


# finish MIME part
sendmail.write("""

--DeathToSpamDeathToSpamDeathToSpam--
""")


# report success
print
print "Message forwarded"
print "Press Enter to continue"

# open /dev/tty for reading and wait for Enter key; this is needed because
# stdin is not available (there was UCE fed into stdin, do you remember?)
tty = open("/dev/tty", 'r')
line = tty.readline()
tty.close()

From j_lockley at hotmail.com  Mon Feb 26 11:53:50 2001
From: j_lockley at hotmail.com (James Lockley)
Date: Mon, 26 Feb 2001 12:53:50 +0200
Subject: deleting files
Message-ID: <97dc8q$mlc$1@news.adamastor.ac.za>

hi

i want to delete a file created after a particular time ( ie something like
if os.path.getmtime(file)>t0: <delete(file>)
i've looked through shutil and can't find anything to use... (i don't want
to delete tree so rmtree is of no use)

thanx in advance
james




From thomas at xs4all.net  Mon Feb 26 12:11:04 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Mon, 26 Feb 2001 12:11:04 +0100
Subject: deleting files
In-Reply-To: <97dc8q$mlc$1@news.adamastor.ac.za>; from j_lockley@hotmail.com on Mon, Feb 26, 2001 at 12:53:50PM +0200
References: <97dc8q$mlc$1@news.adamastor.ac.za>
Message-ID: <20010226121103.N16781@xs4all.nl>

On Mon, Feb 26, 2001 at 12:53:50PM +0200, James Lockley wrote:

> i want to delete a file created after a particular time ( ie something like
> if os.path.getmtime(file)>t0: <delete(file>)
> i've looked through shutil and can't find anything to use... (i don't want
> to delete tree so rmtree is of no use)

Deleting a single file is done using os.remove(). Deleting a directory is
done using 'os.rmdir()'.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From stephen_purcell at yahoo.com  Mon Feb 26 12:11:55 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 26 Feb 2001 12:11:55 +0100
Subject: deleting files
In-Reply-To: <97dc8q$mlc$1@news.adamastor.ac.za>; from j_lockley@hotmail.com on Mon, Feb 26, 2001 at 12:53:50PM +0200
References: <97dc8q$mlc$1@news.adamastor.ac.za>
Message-ID: <20010226121155.A6952@freedom.puma-ag.com>

James Lockley wrote:
> i want to delete a file created after a particular time ( ie something like
> if os.path.getmtime(file)>t0: <delete(file>)
> i've looked through shutil and can't find anything to use... (i don't want
> to delete tree so rmtree is of no use)

You want:

   os.unlink(filename)

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From ajung-ml at andreas-jung.com  Mon Feb 26 14:21:46 2001
From: ajung-ml at andreas-jung.com (Andreas Jung)
Date: Mon, 26 Feb 2001 08:21:46 -0500
Subject: deleting files
References: <97dc8q$mlc$1@news.adamastor.ac.za>
Message-ID: <007001c09ff7$205e3fc0$d0063bd0@SUXLAP>

you can get the modification or creation time of a file through
os.stat(). It returns a tuple of informations about the file (modification
time,creation time, file size).
By checking the creation time you can remove the file by using os.unlink().

Andreas
----- Original Message -----
From: "James Lockley" <j_lockley at hotmail.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Monday, February 26, 2001 5:53 AM
Subject: deleting files


> hi
>
> i want to delete a file created after a particular time ( ie something
like
> if os.path.getmtime(file)>t0: <delete(file>)
> i've looked through shutil and can't find anything to use... (i don't want
> to delete tree so rmtree is of no use)
>
> thanx in advance
> james
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From thomas at cintra.no  Mon Feb 26 12:26:17 2001
From: thomas at cintra.no (Thomas Weholt)
Date: Mon, 26 Feb 2001 12:26:17 +0100
Subject: Translate XSLT/XPath-patterns into regular expressions?
Message-ID: <u3rm6.3437$t21.85716@news3.oke.nextra.no>

I need to translate patterns similar to those used in XSLT/XPath into
regular expressions.
Typical examples are '/sometag/*/mytag' matching every 'mytag' that has
'sometag' as root-node, regardless of tags inbetween them. Another example
could be '/sometag/my*' matching every tag starting with 'my' that has
'sometag' as parent and root-node.

Not being good at regular expressions at all ( just started at the HowTo
actually ), could anyone give a clue on how to go about this or point to
sites/docs with more information about regular expressions ( preferrably
using Python, but all general information is of interest ) ?

The last question; what is the status of available modules for regular
expression in Python and are there problems with them one should be aware
of?

PS! What about the book from Oreilly, 'Mastering Regular Expressions'? It
seemed very pro-Perl, at least from the table of contents. Are the
information given in it easily used in Python too?

Thomas




From timin at homeSPAMNOT.com  Mon Feb 26 16:07:08 2001
From: timin at homeSPAMNOT.com (timin at homeSPAMNOT.com)
Date: Mon, 26 Feb 2001 15:07:08 GMT
Subject: input() vs. raw_input()
Message-ID: <3A9A7142.D81F9604@home.com>

For console input to an executing program, I find that raw_input("prompt")
works very well.  In fact it works the way I would have expected
input("prompt") to work.  The latter, however, seems to funtion as if the
input text is part of a program being compiled.  Can someone explain what is
going on here, and why "raw" is part of the name?

--
"Always yield to temptation, because it may not pass your way again."



From dbrueck at edgix.com  Mon Feb 26 16:34:05 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Mon, 26 Feb 2001 08:34:05 -0700
Subject: input() vs. raw_input()
In-Reply-To: <3A9A7142.D81F9604@home.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHGEHNPKAA.dbrueck@edgix.com>

raw_input - returns input in its unprocessed ("raw") form
input - gets raw input, sends it to eval(), and returns the result

-Dave

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of timin at homeSPAMNOT.com
> Sent: Monday, February 26, 2001 8:07 AM
> To: python-list at python.org
> Subject: input() vs. raw_input()
>
>
> For console input to an executing program, I find that raw_input("prompt")
> works very well.  In fact it works the way I would have expected
> input("prompt") to work.  The latter, however, seems to funtion as if the
> input text is part of a program being compiled.  Can someone
> explain what is
> going on here, and why "raw" is part of the name?
>
> --
> "Always yield to temptation, because it may not pass your way again."
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From itamar at maxnm.com  Mon Feb 26 16:29:31 2001
From: itamar at maxnm.com (Itamar Shtull-Trauring)
Date: Mon, 26 Feb 2001 17:29:31 +0200
Subject: Getting Unicode filenames in Windows 2000
Message-ID: <3A9A765B.BBE9FA29@maxnm.com>

I have a file whose name is part Hebrew, part English on my W2K VMware
install. I'm running BeOpen Python 2.0, with the latest Pythonwin installed.

My problem - the os.listdir() command doesn't return the name of the file in
unicode, it just replaces the hebrew characters with question marks - '?':

>>> l = os.listdir("c:\\")
>>> l[-1]
'????.txt'
>>> type(l[-1])
<type 'string'>

1. How can I get the list of files in a directory in unicode? I would prefer
a cross-platform solution that doesn't use the win32 API.

2. If this behavior is because of backwards compatability, perhaps
os.listdir should be changed so that when passed strings it returns strings
and when passed a unicode string path it returns unicode strings?

-- 
Email: itamar at shtull-trauring.org
Web:   http://itamarst.org



From nelson at crynwr.com  Mon Feb 26 17:07:47 2001
From: nelson at crynwr.com (Russell Nelson)
Date: Mon, 26 Feb 2001 11:07:47 -0500
Subject: GPS application needs an expert's touch
Message-ID: <3A9A7F53.EB5C907E@crynwr.com>

Hi.  I have implemented a GPS application in Python/pygtk/glade.  It
works, for mildly varying values of "work".  However, I'm not sure
that it's very well written.  Okay, I'm sure it *isn't* well written.
Could you take a look at http://russnelson.com/ipaq/gps/ and tell me
what I've done wrong?  I'm not looking for bug fixes, I'm looking for
general comments like "If you had a clue, you'd do it this way." Go
ahead, savage my Python coding style.


--
-russ nelson   http://russnelson.com
Crynwr sells support for free software  | PGPok | "This is Unix...
521 Pleasant Valley Rd. | +1 315 268 1925 voice | Stop acting so helpless."
Potsdam, NY 13676-3213  | +1 315 268 9201 FAX   | --Daniel J. Bernstein





From danielk at aracnet.com  Mon Feb 26 17:08:30 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Mon, 26 Feb 2001 08:08:30 -0800
Subject: is and ==
Message-ID: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com>

Just a quick question I hope as an equally quick (and easy) answer...

Is there a functional difference between 'is' and '==' ?  If there is, I can't
find it.

Daniel Klein


From aleaxit at yahoo.com  Mon Feb 26 17:27:22 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 26 Feb 2001 17:27:22 +0100
Subject: is and ==
References: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com>
Message-ID: <97e05e02h58@news1.newsguy.com>

"Daniel Klein" <danielk at aracnet.com> wrote in message
news:movk9t0una5nuca6upgk9rcco96fbkeu51 at 4ax.com...
> Just a quick question I hope as an equally quick (and easy) answer...
>
> Is there a functional difference between 'is' and '==' ?  If there is, I
can't
> find it.

There is a vast functional difference: 'is' verifies the
two operands refer to the SAME object (identity); '==' is
much more easily satisfied -- it compares the values of
the operands, and is true, not just for identity, but for
any equality.

For example:

>>> onelist = [1,2,3]
>>> another = [1,2,3]
>>> onelist is another
0
>>> onelist == another
1
>>> analias = onelist
>>> analias is onelist
1
>>> analias == onelist
1
>>>

If you care about whether onelist and another refer to the
SAME object, then 'is' is the right test; if you care
about whether the (possibly distinct) objects to which
they refer happen to have equal values right now (although
differing in identity), then '==' is the right test.

If 'is' returns true, then any alteration to the (one and
only) object referred to will be seen from any of the
references to it; '==' implies no such effect.

I.e., in the above example, if you do a onelist.append(4),
then you will also observe, e.g., len(analias) becomes 4,
but another remains completely unaffected, and the == test
has now become false (the two separate and distinct objects
that onelist and another refer to used to be of equal value,
but this becomes false if and when one is altered and the
other one is not).


Alex





From aahz at panix.com  Mon Feb 26 17:45:30 2001
From: aahz at panix.com (Aahz Maruch)
Date: 26 Feb 2001 08:45:30 -0800
Subject: is and ==
References: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com>
Message-ID: <97e17a$o3o$1@panix6.panix.com>

In article <movk9t0una5nuca6upgk9rcco96fbkeu51 at 4ax.com>,
Daniel Klein  <danielk at aracnet.com> wrote:
>
>Is there a functional difference between 'is' and '==' ?  If there is, I can't
>find it.

>>> a = 'xy'
>>> b = a
>>> c = 'x' + 'y'
>>> d = 'xy'
>>> a is b
1
>>> a is d
1
>>> a is c
0
>>> a == c
1

Essentially, is tests for object identity: are two variables pointing to
the same location in memory.  == tests to see whether the *values* are
equal.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From joconnor at cybermesa.com  Mon Feb 26 17:59:34 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 26 Feb 2001 09:59:34 -0700
Subject: is and ==
References: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com> <97e17a$o3o$1@panix6.panix.com>
Message-ID: <3A9A8B76.777FCA6A@cybermesa.com>


Aahz Maruch wrote:

> In article <movk9t0una5nuca6upgk9rcco96fbkeu51 at 4ax.com>,
> Daniel Klein  <danielk at aracnet.com> wrote:
> >
> >Is there a functional difference between 'is' and '==' ?  If there is, I can't
> >find it.
>
> >>> a = 'xy'
> >>> b = a
> >>> c = 'x' + 'y'
> >>> d = 'xy'
> >>> a is b
> 1
> >>> a is d
> 1
> >>> a is c
> 0
> >>> a == c
> 1
>
> Essentially, is tests for object identity: are two variables pointing to
> the same location in memory.  == tests to see whether the *values* are
> equal.

Consider the following example

>>> class TestVal:
 def __init__(self):
  self.x = 100
 def __cmp__(self, other):
  return cmp (self.x, other.x)


>>> x = TestVal()
>>> y = TestVal()
>>> x == y
1
>>> x is y
0
>>> x =y
>>> x is y
1

x == y is true because I've defined what == means, a value comparison, but x is y
is false because they are not both the same object.  However by assign the variable
x to be y, I now get true for "x is y" because they are now the same object

Take care,
Jay




From stephen_purcell at yahoo.com  Mon Feb 26 17:50:23 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Mon, 26 Feb 2001 17:50:23 +0100
Subject: is and ==
In-Reply-To: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com>; from danielk@aracnet.com on Mon, Feb 26, 2001 at 08:08:30AM -0800
References: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com>
Message-ID: <20010226175023.A7401@freedom.puma-ag.com>

Daniel Klein wrote:
> Is there a functional difference between 'is' and '==' ?  If there is, I can't
> find it.

There is indeed a difference.

'is' is an identity test, and '==' is an equality test:

>>> a = [] 
>>> b = []
>>> a is b
0
>>> a == b
1

In many cases for standard python types, you won't find a difference, though.
String literals are invisibly re-used internally by Python ("interned"):

>>> a = "hello"
>>> b = "hello"
>>> a is b
1
>>> a == b
1

But beware of:

>>> a = "hel" + "lo"
>>> b = "hel" + "lo"
>>> a is b
0
>>> a == b
1


Integers are both equal *and* identical:

>>> a = 1
>>> b = 1
>>> a is b
1
>>> a == b
1



-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From phd at phd.pp.ru  Mon Feb 26 17:56:45 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 26 Feb 2001 19:56:45 +0300 (MSK)
Subject: is and ==
In-Reply-To: <20010226175023.A7401@freedom.puma-ag.com>
Message-ID: <Pine.LNX.4.30.0102261956000.24293-100000@fep132.fep.ru>

On Mon, 26 Feb 2001, Steve Purcell wrote:
> Integers are both equal *and* identical:
>
> >>> a = 1
> >>> b = 1
> >>> a is b
> 1
> >>> a == b
> 1

   Not all integers - only 0 <= i < 100:

>>> a = 217
>>> b = 217
>>> a is b
0
>>> a == b
1

   :)

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From dbrueck at edgix.com  Mon Feb 26 18:00:35 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Mon, 26 Feb 2001 10:00:35 -0700
Subject: is and ==
In-Reply-To: <20010226175023.A7401@freedom.puma-ag.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHIEHPPKAA.dbrueck@edgix.com>

> [mailto:python-list-admin at python.org]On Behalf Of Steve Purcell
> Sent: Monday, February 26, 2001 9:50 AM
> To: Daniel Klein
> Cc: python-list at python.org
> Subject: Re: is and ==

< snip >
> Integers are both equal *and* identical:
>
> >>> a = 1
> >>> b = 1
> >>> a is b
> 1
> >>> a == b
> 1

Not necessarily:

>>> a = 10000
>>> b = 10000
>>> a is b
0
>>> a = 5
>>> b = 5
>>> a is b
1

The interpreter "caches" integer objects from something like -1 to 100, so
'is' evaluates to true for low integers but not higher integers.

(Also, as the docs point out, this caching means that from inside an
extension module you could _redefine_ the value of the integer object 5, for
example:

>>> evilext.causehavoc(5,10) # sets '5' object to have int value of 10
>>> 5 + 5
20  # Whee!!!

)

-Dave




From arneh at ifi.uio.no  Mon Feb 26 18:34:22 2001
From: arneh at ifi.uio.no (Arne Martin =?iso-8859-1?Q?G=FCettler?=)
Date: Mon, 26 Feb 2001 18:34:22 +0100
Subject: is and ==
References: <mailman.983206627.31360.python-list@python.org>
Message-ID: <3A9A939E.B025D388@ifi.uio.no>

Oleg Broytmann wrote:
> 
> On Mon, 26 Feb 2001, Steve Purcell wrote:
> > Integers are both equal *and* identical:
> >
>    Not all integers - only 0 <= i < 100:
> 
> >>> a = 217
> >>> b = 217
> >>> a is b
> 0
> >>> a == b
> 1


They might be:

>>> a,b = 217,217
>>> a is b
1
>>> a == b
1


:-)

-- 
-am


From mjackson at wc.eso.mc.xerox.com  Mon Feb 26 19:20:10 2001
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 26 Feb 2001 18:20:10 GMT
Subject: is and ==
References: <mailman.983206927.32385.python-list@python.org>
Message-ID: <97e6oq$gdv$1@news.wrc.xerox.com>

"Dave Brueck" <dbrueck at edgix.com> writes:

> The interpreter "caches" integer objects from something like -1 to 100, so
> 'is' evaluates to true for low integers but not higher integers.
> 
> (Also, as the docs point out, this caching means that from inside an
> extension module you could _redefine_ the value of the integer object 5, for
> example:
> 
> >>> evilext.causehavoc(5,10) # sets '5' object to have int value of 10
> >>> 5 + 5
> 20  # Whee!!!

Woah.  Fortran flashbacks. . .

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Having your book made into a movie is like having
	your ox made into a bouillon cube.
				- Bill Neely




From ben.hutchings at roundpoint.com  Tue Feb 27 03:19:56 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 26 Feb 2001 18:19:56 -0800
Subject: is and ==
References: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com> <mailman.983206214.30693.python-list@python.org>
Message-ID: <uitlwkgtv.fsf@roundpoint.com>

Steve Purcell <stephen_purcell at yahoo.com> writes:

> Daniel Klein wrote:
> > Is there a functional difference between 'is' and '==' ?  If there is,
> > I can't find it.
> 
> There is indeed a difference.
> 
> 'is' is an identity test, and '==' is an equality test:
<snip>
> Integers are both equal *and* identical:
> 
> >>> a = 1
> >>> b = 1
> >>> a is b
> 1
> >>> a == b
> 1

but only if they're small integers (the above doesn't work for 1000,
for instance).

What 'is' does is to test whether two names refer to the same object.
For objects of mutable types (such as lists and class types), this
may be important:

>>> a = [1]
>>> b = a
>>> a == b
1
>>> a is b
1
>>> a.append(2)
>>> b
[1, 2]

Here, b refers to the same list as a, so not only are they equal but a
change made through a is also visible through b.

>>> a = [1]
>>> b = [1]
>>> a == b
1
>>> a is b
0
>>> a.append(2)
>>> b
[1]

Here, a and b initially refer to identical (that is, equal) lists, but
they are separate lists so a change made through a is not visible
through b.

With immutable types, obviously, there's no issue of changes being
visible through another name, so Python can safely conserve memory by
eliminating duplicate objects.  I think this currently only happens
with small integers and string literals.  For the same reason, I'm not
sure that it's at all useful to test whether two names referring to
immutable objects actually refer to the same object.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From danielk at aracnet.com  Tue Feb 27 05:13:45 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Mon, 26 Feb 2001 20:13:45 -0800
Subject: is and ==
References: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com>
Message-ID: <r0am9tgvosc5ea2go2anrca6f9c8hvuoup@4ax.com>

Thanks for the clarification everyone. I knew the difference between identity
and equality from working with Smalltalk, I just didn't realize that this was
how Python made the distinction.

Next time, I'll do a bit more investigation before posting and maybe even rtfm.
:^)

Dan

"Most of us would rather risk catastrophy than read the directions."
						-- unknown


From NoSpam at NoSpam.com  Mon Feb 26 18:21:43 2001
From: NoSpam at NoSpam.com (Tom)
Date: Mon, 26 Feb 2001 17:21:43 GMT
Subject: Splitting comp.lang.python
References: <mailman.983199848.8916.python-list@python.org>
Message-ID: <Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com>

"Mikael Olofsson" <mikael at isy.liu.se> wrote in message
news:mailman.983199848.8916.python-list at python.org...
>
> On 01-Mar-00 Gerrit Holl wrote:
>  >  <quote name="Mikael Olofsson" date="951918286"
email="mikael at isy.liu.se">
>  > > Perhaps the best thing would be to keep c.l.py as it is as the
general
>  > > place for discussion, and then adding subgroups for specific
interests.
>  >
>  >  Perl uses:
>  >
>  >   comp.lang.perl.announce
>  >   comp.lang.perl.misc
>  >   comp.lang.perl.tk
>  >   comp.lang.perl.modules
>  >   comp.lang.perl
>  >   comp.lang.perl.moderated
>  >
>  >  Maybe it's a good idea too do something like that on c.l.py.
>
> Not necessarily exactly those subgroups, but definitely something in
> that direction, yes. Based on my first post in this thread, I come up
> with the following
>
>   c.l.py             General
>   c.l.py.gui         GUI
>   c.l.py.os          os/platform dependent issues
>   c.l.py.db          DataBase handling
>   c.l.py.advocacy    Whitespace and such

Yes, I think this is a very good idea - c.l.p. is too high volume.

But what about api (c/c++) issues?  This would be an easily distinguishable
subsection.  Don't subdivide it into platforms.

Tom.

> The CGI subgroup that I proposed earlier is perhaps not especially
> interresting. Many questions and discussions regarding CGI programming
> tend to be originating from poor understanding of Python itself, and
> its modules. I guess that is the way in to the Python world for many
> newbies. It sure was for me once.
>
> I think that it is better to have general subgroup names such as gui,
> os, db, and advocacy, rather than more explicit ones like tk, linux,
> sql, and whitespace. Maybe platform is better than os... I leave that
> to the os/platform lawyers out there.
>
> /Mikael
>
> -----------------------------------------------------------------------
> E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
> WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
> Phone:   +46 - (0)13 - 28 1343
> Telefax: +46 - (0)13 - 28 1339
> Date:    01-Mar-00
> Time:    15:36:08
>
> This message was sent by XF-Mail.
> -----------------------------------------------------------------------
>




From embed at geocities.com  Mon Feb 26 20:28:32 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 26 Feb 2001 14:28:32 -0500
Subject: Splitting comp.lang.python
References: <mailman.983199848.8916.python-list@python.org> <Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com>
Message-ID: <79ym6.2288$TW.12847@tor-nn1.netcom.ca>

Not again!?  We had this discussion ten times over. Anyone who feels like
weighing in on this one should at least be polite and avoid beating the same
dead horses, over and over again.

I mean, doesn't anyone search Deja anymore? Oh yeah, Deja is gone, and
replaced by the Beta Groups at groups.google.com. Oh well, ironically, now
that we have no Deja.com, I think we are in for a Lot More of that old time
Deja Vu.

Warren Postma





From aahz at panix.com  Mon Feb 26 20:35:15 2001
From: aahz at panix.com (Aahz Maruch)
Date: 26 Feb 2001 11:35:15 -0800
Subject: Splitting comp.lang.python
References: <mailman.983199848.8916.python-list@python.org> <Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com> <79ym6.2288$TW.12847@tor-nn1.netcom.ca>
Message-ID: <97eb5j$cpr$1@panix2.panix.com>

In article <79ym6.2288$TW.12847 at tor-nn1.netcom.ca>,
Warren Postma <embed at geocities.com> wrote:
>
>I mean, doesn't anyone search Deja anymore? Oh yeah, Deja is gone, and
>replaced by the Beta Groups at groups.google.com. Oh well, ironically, now
>that we have no Deja.com, I think we are in for a Lot More of that old time
>Deja Vu.

Well, at least until Google brings up the full DejaNews archive.  I've
heard that the initial estimate is on the order of three months.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From skip at mojam.com  Mon Feb 26 21:29:10 2001
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 26 Feb 2001 14:29:10 -0600 (CST)
Subject: Splitting comp.lang.python
In-Reply-To: <79ym6.2288$TW.12847@tor-nn1.netcom.ca>
References: <mailman.983199848.8916.python-list@python.org>
	<Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com>
	<79ym6.2288$TW.12847@tor-nn1.netcom.ca>
Message-ID: <15002.48278.442384.757689@beluga.mojam.com>

    Warren> Not again!?  We had this discussion ten times over. 

The list seems to have lost a good deal of its collective memory with the
demise of deja.com and hasn't yet regained it in the guise of
groups.google.com.  Perhaps it would be worthwhile to add the perennial
c.l.py split question to the FAQ so we can simply point people there in the
future.

-- 
Skip Montanaro (skip at mojam.com)
Support Mojam & Musi-Cal: http://www.musi-cal.com/sponsor.shtml
(847)971-7098



From landauer at apple.com  Wed Feb 28 22:45:57 2001
From: landauer at apple.com (Doug Landauer)
Date: Wed, 28 Feb 2001 13:45:57 -0800
Subject: Splitting comp.lang.python
References: <mailman.983199848.8916.python-list@python.org> <Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com> <79ym6.2288$TW.12847@tor-nn1.netcom.ca>
Message-ID: <landauer-432801.13455728022001@news.apple.com>

> Not again!?  We had this discussion ten times over.

Yes, again.  The newsgroup's volume changes.  The set of readers
and posters changes.  The number of people interested in python
but who don't have time to read 100 or 200 new and different subject
lines a day changes.  The discussion leads some people to different
conclusions each time it comes up in the context of all of these
changing conditions.

/split/k-is-a-small-price-to-pay-for-python's-increasing-popularity-ly,

  :-D
-- 



From sholden at holdenweb.com  Mon Feb 26 23:33:09 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 26 Feb 2001 17:33:09 -0500
Subject: Splitting comp.lang.python
References: <mailman.983199848.8916.python-list@python.org><Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com><79ym6.2288$TW.12847@tor-nn1.netcom.ca> <mailman.983219407.2513.python-list@python.org>
Message-ID: <9UAm6.21407$P87.223273@e420r-atl2.usenetserver.com>

"Skip Montanaro" <skip at mojam.com> wrote in message
news:mailman.983219407.2513.python-list at python.org...
>
>     Warren> Not again!?  We had this discussion ten times over.
>
> The list seems to have lost a good deal of its collective memory with the
> demise of deja.com and hasn't yet regained it in the guise of
> groups.google.com.  Perhaps it would be worthwhile to add the perennial
> c.l.py split question to the FAQ so we can simply point people there in
the
> future.
>
That would be a good idea: it probably keeps coming up because the
readership of the group is enlarging (or at least thrashing :-) fairly
rapidly. Then the old-timers use more bandwidth to explain we've already had
these discussions before...

regards
 Steve





From gem at hsv.crc.com  Mon Feb 26 18:58:34 2001
From: gem at hsv.crc.com (Glen Mettler)
Date: Mon, 26 Feb 2001 11:58:34 -0600
Subject: Combinations function
Message-ID: <97e52o$41c$1@hobbes2.crc.com>

Is there a python function that will give me the possible combinations of a
set?
ie - in list a,b,c,d there are 4! or 24 possible combinations.  Is there an
existing function to print the possible combinations?

Glen





From f8dy at yahoo.com  Mon Feb 26 19:50:35 2001
From: f8dy at yahoo.com (Mark Pilgrim)
Date: Mon, 26 Feb 2001 13:50:35 -0500
Subject: Combinations function
References: <97e52o$41c$1@hobbes2.crc.com>
Message-ID: <97e8en$pbr73$1@ID-77331.news.dfncis.de>

There is nothing built into Python to do this, but I found this on the
Python Knowledge Base (a wonderful resource for those who don't know about
it):
  http://www.faqts.com/knowledge_base/view.phtml/aid/4526/fid/538

-M
You're smart; why haven't you learned Python yet?
http://diveintopython.org/


"Glen Mettler" <gem at hsv.crc.com> wrote in message
news:97e52o$41c$1 at hobbes2.crc.com...
> Is there a python function that will give me the possible combinations of
a
> set?
> ie - in list a,b,c,d there are 4! or 24 possible combinations.  Is there
an
> existing function to print the possible combinations?
>
> Glen
>
>
>




From aleaxit at yahoo.com  Mon Feb 26 20:42:11 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 26 Feb 2001 20:42:11 +0100
Subject: Combinations function
References: <97e52o$41c$1@hobbes2.crc.com>
Message-ID: <97ebh20kti@news2.newsguy.com>

"Glen Mettler" <gem at hsv.crc.com> wrote in message
news:97e52o$41c$1 at hobbes2.crc.com...
> Is there a python function that will give me the possible combinations of
a
> set?
> ie - in list a,b,c,d there are 4! or 24 possible combinations.  Is there
an
> existing function to print the possible combinations?

Actually, the term "combinations" is normally used to indicate
samples (of m items out of the n in the set, m <= n), without
replacement, _order not counting_; those are very easy to
generate, but there are 2**n (here, 16) of them, not n!.

You appear to desire _permutations_, aka 'orderings', of n
items, of which there are indeed n!.  Those are easiest to
generate in a recursive way, for example:

def all_permutations(alist):
    if not alist: return [[]]
    result = []
    for i in range(len(alist)):
        for subresult in all_permutations(alist[:i]+alist[i+1:]):
            result.append([alist[i]] + subresult)
    return result

There's a lot of little things that can be done to optimize
this, but this is the basic idea.  Be careful not to run this
on LONG lists -- it will take a LOT of time (and so would
any optimized versions!), because n! grows VERY fast with n.


Alex






From matthew.feinstein at jhuapl.edu  Mon Feb 26 19:17:16 2001
From: matthew.feinstein at jhuapl.edu (Matt Feinstein)
Date: Mon, 26 Feb 2001 18:17:16 GMT
Subject: PyOpenGL?
Message-ID: <3a9a9d48.6629983@aplnews>

The Sourceforge link for PyOpenGL doesn't seem to be active and
there's no alternate route for the module-- is it available?
 
Matt Feinstein  matt.feinstein at jhuapl.edu
--------
Harvard Law of Automotive Repair: Anything that goes away
by itself will come back by itself.
                          


From ljohnson at resgen.com  Mon Feb 26 23:12:08 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Mon, 26 Feb 2001 16:12:08 -0600
Subject: PyOpenGL?
References: <3a9a9d48.6629983@aplnews>
Message-ID: <t9ll1g3dcqr8e6@corp.supernews.com>

At this moment, the link:

    http://pyopengl.sourceforge.net

is alive and well. Perhaps a temporary outage at SF?

"Matt Feinstein" <matthew.feinstein at jhuapl.edu> wrote in message
news:3a9a9d48.6629983 at aplnews...
>
> The Sourceforge link for PyOpenGL doesn't seem to be active and
> there's no alternate route for the module-- is it available?
>
> Matt Feinstein  matt.feinstein at jhuapl.edu
> --------
> Harvard Law of Automotive Repair: Anything that goes away
> by itself will come back by itself.
>




From anderson at pha.jhu.edu  Mon Feb 26 19:30:08 2001
From: anderson at pha.jhu.edu (Kenneth Anderson)
Date: Mon, 26 Feb 2001 13:30:08 -0500
Subject: os.tmpfile()
Message-ID: <3A9AA0B0.4137D4C9@pha.jhu.edu>

Hi

i am trying to use the os.tmpfile() method to create a file object for
writing,
but i am seeing weird behaviour that i certainly don't expect:

>>> import os
>>> f1 = os.tmpfile()
>>> f1
<open file '<tmpfile>', mode 'w+' at eed68>
>>> f1.readlines()
[]
>>> f1.write('check')
>>> f1.readlines()

'\000\000\000\000\000\000\000\000\...long, long list of \000   '

Nothing is the FAQ about tmpfile().

any notions about this.  I would really prefer not to write a file
but rather just write the data to a file object.

thanks!

Ken Anderson
JHU

--
===================================================
...and for lunch, Ken stops by the side of the road
and rubs gravel into his hair.
===================================================





From stephen_purcell at yahoo.com  Tue Feb 27 09:33:56 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 27 Feb 2001 09:33:56 +0100
Subject: os.tmpfile()
In-Reply-To: <3A9AA0B0.4137D4C9@pha.jhu.edu>; from anderson@pha.jhu.edu on Mon, Feb 26, 2001 at 01:30:08PM -0500
References: <3A9AA0B0.4137D4C9@pha.jhu.edu>
Message-ID: <20010227093356.C7945@freedom.puma-ag.com>

Kenneth Anderson wrote:
> >>> f1 = os.tmpfile()
> >>> f1
> <open file '<tmpfile>', mode 'w+' at eed68>
> >>> f1.readlines()
> []
> >>> f1.write('check')
> >>> f1.readlines()
> 
> '\000\000\000\000\000\000\000\000\...long, long list of \000   '

Not sure about this, but you could try 'tempfile.mktemp' to get a unique
filename, then explicitly open it yourself to see if the results are any
different.

> any notions about this.  I would really prefer not to write a file
> but rather just write the data to a file object.

Maybe you just need to use the StringIO (or cStringIO) module if you want
an in-memory file-like object?

-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From shichang at icubed.com  Mon Feb 26 19:33:49 2001
From: shichang at icubed.com (szhao)
Date: Mon, 26 Feb 2001 18:33:49 GMT
Subject: Python v.s. Ruby?
Message-ID: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net>

How do you compare the Languague Ruby with Python?

http://www.ruby-lang.org/en/






From dbrueck at edgix.com  Mon Feb 26 19:54:43 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Mon, 26 Feb 2001 11:54:43 -0700
Subject: Python v.s. Ruby?
In-Reply-To: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net>
Message-ID: <BLEOJGEMBNBNEAKMIFDHEEIEPKAA.dbrueck@edgix.com>

Ruby? Nobody on this list has ever heard of a language called Ruby, much
less given any thought to comparing the two.

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of szhao
> Sent: Monday, February 26, 2001 11:34 AM
> To: python-list at python.org
> Subject: Python v.s. Ruby?
>
>
> How do you compare the Languague Ruby with Python?
>
> http://www.ruby-lang.org/en/
>
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From dsh8290 at rit.edu  Mon Feb 26 20:09:36 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 26 Feb 2001 14:09:36 -0500
Subject: Python v.s. Ruby?
In-Reply-To: <BLEOJGEMBNBNEAKMIFDHEEIEPKAA.dbrueck@edgix.com>; from dbrueck@edgix.com on Mon, Feb 26, 2001 at 11:54:43AM -0700
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net> <BLEOJGEMBNBNEAKMIFDHEEIEPKAA.dbrueck@edgix.com>
Message-ID: <20010226140935.C14108@harmony.cs.rit.edu>

On Mon, Feb 26, 2001 at 11:54:43AM -0700, Dave Brueck wrote:
| Ruby? Nobody on this list has ever heard of a language called Ruby, much
| less given any thought to comparing the two.

Yeah, sure, whatever.  Check the archives, this has been discussed
many times before.  

-D




From shichang at icubed.com  Mon Feb 26 20:04:15 2001
From: shichang at icubed.com (szhao)
Date: Mon, 26 Feb 2001 19:04:15 GMT
Subject: Python v.s. Ruby?
References: <mailman.983213784.17876.python-list@python.org>
Message-ID: <PMxm6.1190$XH1.82354@typhoon2.ba-dsg.net>

People told me Ruby is powerful and better than Python. Some simple Ruby
example shows weird perl/tcl/tk/python syntax mixins.

"Dave Brueck" <dbrueck at edgix.com> wrote in message
news:mailman.983213784.17876.python-list at python.org...
> Ruby? Nobody on this list has ever heard of a language called Ruby, much
> less given any thought to comparing the two.
>
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of szhao
> > Sent: Monday, February 26, 2001 11:34 AM
> > To: python-list at python.org
> > Subject: Python v.s. Ruby?
> >
> >
> > How do you compare the Languague Ruby with Python?
> >
> > http://www.ruby-lang.org/en/
> >
> >
> >
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
>
>




From nospam at mega-nerd.com  Mon Feb 26 20:16:29 2001
From: nospam at mega-nerd.com (Erik de Castro Lopo)
Date: Mon, 26 Feb 2001 19:16:29 GMT
Subject: Python v.s. Ruby?
References: <mailman.983213784.17876.python-list@python.org> <PMxm6.1190$XH1.82354@typhoon2.ba-dsg.net>
Message-ID: <3A9AA8DB.4A7876A5@mega-nerd.com>

szhao wrote:
> 
> People told me Ruby is powerful and better than Python. 

Those kinds of statements are extremely difficult to prove.

> Some simple Ruby
> example shows weird perl/tcl/tk/python syntax mixins.

That may have some truth ot it.

Erik
-- 
+----------------------------------------------------------+
  Erik de Castro Lopo  nospam at mega-nerd.com (Yes its valid)
+----------------------------------------------------------+
Linux: generous programmers from around the world all join
forces to help you shoot yourself in the foot for free.


From fredrik at pythonware.com  Mon Feb 26 20:08:19 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 26 Feb 2001 19:08:19 GMT
Subject: Python v.s. Ruby?
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net>
Message-ID: <DQxm6.15634$Qb7.2559170@newsb.telia.net>

"szhao" wrote:
> How do you compare the Languague Ruby with Python?

groups.google.com

Cheers /F




From shichang at icubed.com  Mon Feb 26 20:21:04 2001
From: shichang at icubed.com (szhao)
Date: Mon, 26 Feb 2001 19:21:04 GMT
Subject: Python v.s. Ruby?
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net> <DQxm6.15634$Qb7.2559170@newsb.telia.net>
Message-ID: <A0ym6.1470$cw3.93403@typhoon2.ba-dsg.net>

Thanks. /F

Lots of saying about Ruby. So, what do you say, /F?



"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:DQxm6.15634$Qb7.2559170 at newsb.telia.net...
> "szhao" wrote:
> > How do you compare the Languague Ruby with Python?
>
> groups.google.com
>
> Cheers /F
>
>




From tim.one at home.com  Mon Feb 26 20:41:48 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 26 Feb 2001 14:41:48 -0500
Subject: Python v.s. Ruby?
In-Reply-To: <A0ym6.1470$cw3.93403@typhoon2.ba-dsg.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEBKJCAA.tim.one@home.com>

[szhao]
> Thanks. /F
>
> Lots of saying about Ruby. So, what do you say, /F?

/F makes his living working with Python, so you don't care what he says
about Ruby <wink>.

Me neither.  Sincere suggestion:  download Python and *try* it!  It's free,
it's easy, and if you've programmed in some other language before you can
very likely learn most of Python in an afternoon by working through the
tutorial that comes with it.  Do the same for Ruby, and then you can make up
your own mind.  Come back then and tell us what *you* think!

Overall, that will consume much less of your time (and ours) than slogging
through hundreds of argumentative messages about it.  I doubt you would
*regret* the time you spent introducing yourself to either language.

i-hear-you're-even-allowed-to-use-both<wink>-ly y'rs  - tim




From daniel at dittmar.net  Mon Feb 26 20:59:20 2001
From: daniel at dittmar.net (Daniel Dittmar)
Date: Mon, 26 Feb 2001 20:59:20 +0100
Subject: Python v.s. Ruby?
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net> <DQxm6.15634$Qb7.2559170@newsb.telia.net> <A0ym6.1470$cw3.93403@typhoon2.ba-dsg.net>
Message-ID: <3A9AB598.E63777B5@dittmar.net>

> > groups.google.com
> >
> Lots of saying about Ruby. So, what do you say, /F?
> 

The question about Python vs. Ruby has been asked a few weeks back and
Fred asks you to review that discussion at 'The Search Engine Formely
Known As DejaNews'.

Short Summary of that diskussion: a few had played around with Ruby, but
none found an overwhelming reason to use it. Evidence was more anecdotal
than anything else.

Daniel


From parkw at better.net  Tue Feb 27 08:42:22 2001
From: parkw at better.net (William Park)
Date: Tue, 27 Feb 2001 02:42:22 -0500
Subject: Python v.s. Ruby?
In-Reply-To: <3a9c44d2.2862445@news.telus.net>; from root@127.0.0.1 on Tue, Feb 27, 2001 at 06:17:55AM +0000
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net> <3a9c44d2.2862445@news.telus.net>
Message-ID: <20010227024222.A490@better.net>

On Tue, Feb 27, 2001 at 06:17:55AM +0000, David wrote:
> On Mon, 26 Feb 2001 18:33:49 GMT, "szhao" <shichang at icubed.com> wrote:
> >How do you compare the Languague Ruby with Python?
> >http://www.ruby-lang.org/en/
> 
> I think the most distinguishing difference between Python and Ruby is that
> the formers' users don't compare Python and Ruby; whereas the users of the
> latter seem to encourage comparing Ruby and Python.
>  
> Or at least that's my perception.  Certainly the Python websites don't have
> any references to Ruby.  The Ruby websites seem to refer to Python...
>  
> There's a maxim that may apply: "If you have to compare yourself to the
> competition, you're not the leader."

In the simplest terms, Ruby is OO Perl (not good for my fingers).

---William Park, Open Geometry Consulting



From dfan at harmonixmusic.com  Tue Feb 27 16:12:16 2001
From: dfan at harmonixmusic.com (Dan Schmidt)
Date: 27 Feb 2001 10:12:16 -0500
Subject: Python v.s. Ruby?
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net> <3a9c44d2.2862445@news.telus.net>
Message-ID: <wk1ysk40tr.fsf@turangalila.harmonixmusic.com>

root at 127.0.0.1 (David) writes:

| I think the most distinguishing difference between Python and Ruby
| is that the formers' users don't compare Python and Ruby; whereas
| the users of the latter seem to encourage comparing Ruby and Python.
|  
| Or at least that's my perception.  Certainly the Python websites
| don't have any references to Ruby.  The Ruby websites seem to refer
| to Python...

Yes, the Python websites are too busy disparaging Perl :)

(I use all three languages and like them all.)

-- 
http://www.dfan.org


From tim at vegeta.ath.cx  Tue Feb 27 23:10:42 2001
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Tue, 27 Feb 2001 22:10:42 GMT
Subject: Python v.s. Ruby?
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net> <3a9c44d2.2862445@news.telus.net> <wk1ysk40tr.fsf@turangalila.harmonixmusic.com>
Message-ID: <slrn99o9sq.etg.tim@vegeta.ath.cx>

Dan Schmidt <dfan at harmonixmusic.com> wrote:
> Yes, the Python websites are too busy disparaging Perl :)

Oh, you noticed that too?  That's alright.  It's mutual, unfortunately.

> (I use all three languages and like them all.)

I don't use Ruby, but there are certain projects I'd rather do in
Python, and others better suited to Perl...so why should I have to pick?
This goes along with Perl's TMTOWTDI (There's More Than One Way To Do
It) motto.  Python is just another way I might chose to do it.  =)

-- 
-Tim Hammerquist <timmy at cpan.org>
Usenet is essentially a HUGE group of
people passing notes in class.
	-- R. Kadel


From tim at vegeta.ath.cx  Tue Feb 27 23:08:54 2001
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Tue, 27 Feb 2001 22:08:54 GMT
Subject: Python v.s. Ruby?
References: <hkxm6.285$cw3.76636@typhoon2.ba-dsg.net> <3a9c44d2.2862445@news.telus.net> <mailman.983259788.19853.python-list@python.org>
Message-ID: <slrn99o9pf.etg.tim@vegeta.ath.cx>

William Park <parkw at better.net> wrote:
> In the simplest terms, Ruby is OO Perl (not good for my fingers).

It looked like it had too many VB qualities to be much like Perl.  Most
Perlers think Ruby is closer to Python, so maybe Ruby is just its own
animal?

-- 
-Tim Hammerquist <timmy at cpan.org>

A liar should have a good memory.
	-- Quintilian


From marvind at Glue.umd.edu  Mon Feb 26 21:04:31 2001
From: marvind at Glue.umd.edu (Arvind Mani)
Date: Mon, 26 Feb 2001 15:04:31 -0500
Subject: ar: internal error (58)
Message-ID: <3A9AB6CF.72F2E862@Glue.umd.edu>

hi,

I am trying to build and install python 2.0 on solaris. I get this error

message when the object files are archived. I used gcc to compile the
files.

hope you guys can help me out on this one.

thanks,
arvind





From rfileger at bellsouth.net  Mon Feb 26 21:11:33 2001
From: rfileger at bellsouth.net (Roy W Fileger)
Date: Mon, 26 Feb 2001 15:11:33 -0500
Subject: https
Message-ID: <003f01c0a030$5161eb60$4c80d6d1@mpiabose280srr>

Hello,

I am new to Python but am excited about it. I have a
requirement to send a Credit Card transaction to an
HTTPS server as the body of a POST transaction.

With Python 2.0 supporting OpenSSL, can I simply
use the httplib to connect and POST? The POST
example the Python Library Reference 11.4.2 seems
both simple and clear.

So far I have only been executing sample code from
the many sources and then making modifications to
attempt to insure that I understand.

Visa says I can confirm the availability of the gateway
by browsing https://ssl.pgs.wcom.net/ and this works

Here is a .py I have developed based upon examples
in the Python Library Reference and the information
provided by Visa re: VirtualNet SSL Gateway but I get
a socket error: host not found.
===============================================================
tran1 = ('0x82'+'D4.999995'+'0x03'+'0xD1')    # provided by Visa

def sendTran(ccTran):
   h = httplib.HTTP("https://ssl.pgs.wcom.net", 443)    # my interpretation of Visa data
   h.putrequest("POST", "/scripts/gateway.dll?Transact HTTP/1.0") # my use Visa data
   h.putheader("Content-Type", "x-Visa-II/x-auth")      # per Visa
   h.putheader("Content-Length", "%d" % len(tran1))
   h.putheader('Accept', 'text/plain')
   h.putheader('Host', 'https://ssl.pgs.wcom.net')            # same as above
   h.endheaders()
   h.send(ccTran)
   reply, msg, hdrs = h.getreply()
   data = h.getfile().read()
   #
   print data
   
sendTran(tran1)
================================================================

Thanks, your help is greatly appreciated, Roy W. Fileger.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010226/225f3390/attachment.htm>

From erno-news at erno.iki.fi  Tue Feb 27 03:56:33 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 27 Feb 2001 04:56:33 +0200
Subject: https
References: <mailman.983218330.31657.python-list@python.org>
Message-ID: <kuhf1grfz2.fsf@lasipalatsi.fi>

In article <mailman.983218330.31657.python-list at python.org>, "Roy W
Fileger" <rfileger at bellsouth.net> writes:

| Hello,
| I am new to Python but am excited about it. I have a
| requirement to send a Credit Card transaction to an
| HTTPS server as the body of a POST transaction.

if you decide to use the https support that comes with python, be
sure to authenticate the other end somehow. by default the python
https support doesn't seem to have any database of certificate
authorities, so it's not really safe against dns spoofing or
the like.

someone please correct me if i'm wrong, i'm not sure about the
above since there is no documentation for the ssl stuff...

i hear m2crypto might be better wrt this, but i'm not sure.

you could also consider using some other way, like gnupg + email,
to send the information. that would have the advantage that the
information would arrive ancrypted at the other end too, so
even if your your communications with the server at the other end,
or indeed the server at the other end itself, gets compromised,
your information is still safe.

  -- erno


From xavier at perceval.net  Tue Feb 27 09:11:52 2001
From: xavier at perceval.net (Xavier Defrang)
Date: Tue, 27 Feb 2001 09:11:52 +0100 (CET)
Subject: https
In-Reply-To: <003f01c0a030$5161eb60$4c80d6d1@mpiabose280srr>
Message-ID: <Pine.LNX.4.21.0102270905110.17289-100000@elara.perceval.be>

On Mon, 26 Feb 2001, Roy W Fileger wrote:

> With Python 2.0 supporting OpenSSL, can I simply
> use the httplib to connect and POST? The POST
> example the Python Library Reference 11.4.2 seems
> both simple and clear.

Despite of all my efforts, I've never been able to build
Python 2.0 with built-in SSL support and I ran into a
solution that I find simple and elegant.  I just run the
cryptographic stuff "outside" of my Python application.

I use Stunnel (http://www.stunnel.org) which is an SSL
proxy to create an encrypted tunnel between my application
and the server. Security can be pretty well finetuned using
TCP wrapper access control.

With this proxy server, I use the good ol' httplib and it
works really fine!  I also used it for a credit card transaction
system so the overhead involved by the SSL proxying doesn't
matter very much...

Regards.

Xavier Defrang
Perceval R&D Team
xavier at perceval.net
-------------------------------------------------
 Perceval Technologies SA/NV  Tel: +32-2-6409194
 Rue Tenbosch, 9              Fax: +32-2-6403154
 B-1000 Brussels         http://www.perceval.net
 BELGIUM                       info at perceval.net
-------------------------------------------------

 




From owen at astrono.junkwashington.emu  Mon Feb 26 21:53:15 2001
From: owen at astrono.junkwashington.emu (Russell E. Owen)
Date: Mon, 26 Feb 2001 12:53:15 -0800
Subject: Hiding widgets it Tkinter and shrinking the frame/window?
Message-ID: <97efnv$tqq$1@nntp6.u.washington.edu>

I am trying to generate a window with basic options always showing and 
additional panels that are hidden or displayed if needed. I can easily 
show/hide widgets, but the window does not shrink after a widget is 
hidden. Is there some basic trick to this?

(Also, is there some simple way to get the TopLevel object from a 
widget, perhaps one that is buried many levels deep in enclosing frames 
and etc.? I think I may have to make window manager calls, and I don't 
want to pass the TopLevel widget around if I don't have to.)

-- Russell


From owen at astrono.junkwashington.emu  Tue Feb 27 17:45:12 2001
From: owen at astrono.junkwashington.emu (Russell E. Owen)
Date: Tue, 27 Feb 2001 08:45:12 -0800
Subject: Hiding widgets it Tkinter and shrinking the frame/window? -- a partial answer
References: <97efnv$tqq$1@nntp6.u.washington.edu>
Message-ID: <97gliu$iba$1@nntp6.u.washington.edu>

To partially answer my own post...

In article <97efnv$tqq$1 at nntp6.u.washington.edu>, "Russell E. Owen" 
<owen at astrono.junkwashington.emu> wrote:

>I am trying to generate a window with basic options always showing and 
>additional panels that are hidden or displayed if needed. I can easily 
>show/hide widgets, but the window does not shrink after a widget is 
>hidden. Is there some basic trick to this?

The basics of show/hide are to pack() or pack_forget() (if using the 
pack manager). Unfortunately, pack_forget loses track of where the 
unpacked widget goes.

To work around this, what I was doing was:
- create an enclosing frame for each hideable widget and have that be 
the master for the hideable widget
- pack(side=LEFT, anchor=NW) or pack_forget() to show/hide
Unfortunately, the enclosing frame wasn't shrinking when the hideable 
widget was unpacked, hence the window wasn't shrinking either.

Just for the heck of it, I tried a variation that worked:
- create an enclosing frame to enclose a set of hideable widgets
- pack an empty (hence zero-sized) placeholder frame to the left of each 
hideable widget
- use pack(side=LEFT, anchor=NW, after=<placeholder>) to show the 
widget, and pack_forget() to hide it
In this case the enclosing frame and window did shrink when a widget was 
hidden.

I have no idea why the second techinque worked when the first one did 
not, as they are very similar. But heck, whatever works. (If anybody 
does have an explanation I'd love to hear it).

I plan to try the gridder, as well, since it appears one can ungrid 
items without forgetting their information (using grid_remove). It'd be 
nice to dispense with the placeholder frames.

-- Russell


From owen at astrono.junkwashington.emu  Wed Feb 28 17:42:45 2001
From: owen at astrono.junkwashington.emu (Russell E. Owen)
Date: Wed, 28 Feb 2001 08:42:45 -0800
Subject: Hiding widgets it Tkinter and shrinking the frame/window? -- the rest of the answer
References: <97efnv$tqq$1@nntp6.u.washington.edu> <97gliu$iba$1@nntp6.u.washington.edu>
Message-ID: <97j9pv$hji$1@nntp6.u.washington.edu>

Sorry to keep posting, but here's the rest of the answer.

I was wondering why Tk frames wouldn't always shrink when their contents 
were unpacked. Turns out that if you remove all widgets from an 
enclosing frame, it stays at the size of the last-removed widget. So all 
you have to do is include a zero-sized widget (I use a frame with 
nothing in it) and never unpack that. Then the enclosing frame will 
shrink to nothing when the empty frame is the last widget left in it, 
i.e. when all hideable widgets have been unpacked.

We need a Tk folklore site.

-- Russell


From cerutti at together.net  Tue Feb 27 22:27:18 2001
From: cerutti at together.net (Neil Cerutti)
Date: 27 Feb 2001 21:27:18 GMT
Subject: Hiding widgets it Tkinter and shrinking the frame/window?
References: <97efnv$tqq$1@nntp6.u.washington.edu>
Message-ID: <slrn99o7do.7n.cerutti@fiad06.norwich.edu>

Russell E. Owen posted:
>(Also, is there some simple way to get the TopLevel object from a 
>widget, perhaps one that is buried many levels deep in enclosing frames 
>and etc.? I think I may have to make window manager calls, and I don't 
>want to pass the TopLevel widget around if I don't have to.)

.winfo_toplevel()

-- 
Neil Cerutti <cerutti at together.net>


From tjg at exceptionalminds.com  Mon Feb 26 21:59:03 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Mon, 26 Feb 2001 12:59:03 -0800
Subject: IP Math anyone?
Message-ID: <20010226125903.X29494@trufflehunter.avalongroup.net>

Hi all,

I checked Parnassus, but couldn't find anything, so I'll ask
here. Has anyone done anything with IP math (e.g.,
192.168.0.254 + 1 = 192.168.1.1).



-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.         <><       (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  41 days  1:10 hours ago<<



From jgraves3 at austin.rr.com  Mon Feb 26 22:12:06 2001
From: jgraves3 at austin.rr.com (jay graves)
Date: Mon, 26 Feb 2001 21:12:06 GMT
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org>
Message-ID: <3a9ac5c3.13004564@news-server.austin.rr.com>

On Mon, 26 Feb 2001 12:59:03 -0800, Timothy Grant
<tjg at exceptionalminds.com> wrote:
>Hi all,
>I checked Parnassus, but couldn't find anything, so I'll ask
>here. Has anyone done anything with IP math (e.g.,
>192.168.0.254 + 1 = 192.168.1.1).
>-- 
>Stand Fast,
>    tjg.

how about:
http://www.vex.net/parnassus/apyllo.py?i=35559588

i don't know if it overrides the arithmetic operators but it might be
a good starting point for you.

...
jay


From parkw at better.net  Mon Feb 26 22:48:05 2001
From: parkw at better.net (William Park)
Date: Mon, 26 Feb 2001 16:48:05 -0500
Subject: IP Math anyone?
In-Reply-To: <3a9ac5c3.13004564@news-server.austin.rr.com>; from jgraves3@austin.rr.com on Mon, Feb 26, 2001 at 09:12:06PM +0000
References: <mailman.983221028.6007.python-list@python.org> <3a9ac5c3.13004564@news-server.austin.rr.com>
Message-ID: <20010226164805.A1244@better.net>

On Mon, Feb 26, 2001 at 09:12:06PM +0000, jay graves wrote:
> On Mon, 26 Feb 2001 12:59:03 -0800, Timothy Grant
> <tjg at exceptionalminds.com> wrote:
> >Hi all,
> >I checked Parnassus, but couldn't find anything, so I'll ask
> >here. Has anyone done anything with IP math (e.g.,
> >192.168.0.254 + 1 = 192.168.1.1).
> >-- 
> >Stand Fast,
> >    tjg.
> 
> how about:
> http://www.vex.net/parnassus/apyllo.py?i=35559588
> 
> i don't know if it overrides the arithmetic operators but it might be
> a good starting point for you.

'__add__' implements '+'.  But, my question is what happens if
    192.168.0.254 + 5
?

---William Park, Open Geometry Consulting



From erno-news at erno.iki.fi  Mon Feb 26 22:23:39 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 26 Feb 2001 23:23:39 +0200
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org>
Message-ID: <kupug5qgtg.fsf@lasipalatsi.fi>

In article <mailman.983221028.6007.python-list at python.org>, Timothy
Grant <tjg at exceptionalminds.com> writes:

| Hi all,
| I checked Parnassus, but couldn't find anything, so I'll ask
| here. Has anyone done anything with IP math (e.g.,
| 192.168.0.254 + 1 = 192.168.1.1).

assuming you actually mean 192.168.0.254 + 1 = 192.168.0.255,
try something like...

>>> def dotted2int(ds): 
...     return struct.unpack('!I', socket.inet_aton(ds))[0]
... 
>>> def int2dotted(i):
...     return socket.inet_ntoa(struct.pack('!I', i))
... 
>>> print int2dotted(dotted2int('192.168.0.254') + 1)
192.168.0.255

  -- erno


From scarblac at pino.selwerd.nl  Tue Feb 27 09:08:09 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 27 Feb 2001 08:08:09 GMT
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi>
Message-ID: <slrn99mnvn.ia7.scarblac@pino.selwerd.nl>

Erno Kuusela <erno-news at erno.iki.fi> wrote in comp.lang.python:
> In article <mailman.983221028.6007.python-list at python.org>, Timothy
> Grant <tjg at exceptionalminds.com> writes:
> 
> | Hi all,
> | I checked Parnassus, but couldn't find anything, so I'll ask
> | here. Has anyone done anything with IP math (e.g.,
> | 192.168.0.254 + 1 = 192.168.1.1).
> 
> assuming you actually mean 192.168.0.254 + 1 = 192.168.0.255,

He doesn't; .0.255 and .1.0 are not IP addresses you could give to a machine.
I've never seen existing code for this, seems to need some custom hacking.

(I don't know if .255 and .0 are the only exceptions, probably not).

-- 
Remco Gerlich



From sandipan at vsnl.com  Tue Feb 27 09:45:48 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Tue, 27 Feb 2001 14:15:48 +0530
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi> <slrn99mnvn.ia7.scarblac@pino.selwerd.nl>
Message-ID: <00e701c0a099$b6b30580$020000c3@node02>

Actually, not only .0 or .255, could be any other IP address as defined by
the netmask in sub-nets.

Eg, .184 is similar to .0 (network address) if netmask is 252 instead of 0
Similarly, .187 is similar to .255 (broadcast address) with this netmask.
In other words, 252 delivers a 4 address sub-net.

Therefore the maths operations will have to involve the netmask.

Regards,

Sandipan
----- Original Message -----
From: "Remco Gerlich" <scarblac at pino.selwerd.nl>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Tuesday, February 27, 2001 1:38 PM
Subject: Re: IP Math anyone?


> Erno Kuusela <erno-news at erno.iki.fi> wrote in comp.lang.python:
> > In article <mailman.983221028.6007.python-list at python.org>, Timothy
> > Grant <tjg at exceptionalminds.com> writes:
> >
> > | Hi all,
> > | I checked Parnassus, but couldn't find anything, so I'll ask
> > | here. Has anyone done anything with IP math (e.g.,
> > | 192.168.0.254 + 1 = 192.168.1.1).
> >
> > assuming you actually mean 192.168.0.254 + 1 = 192.168.0.255,
>
> He doesn't; .0.255 and .1.0 are not IP addresses you could give to a
machine.
> I've never seen existing code for this, seems to need some custom hacking.
>
> (I don't know if .255 and .0 are the only exceptions, probably not).
>
> --
> Remco Gerlich
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>





From jafo at tummy.com  Tue Feb 27 09:46:49 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 01:46:49 -0700
Subject: IP Math anyone?
In-Reply-To: <slrn99mnvn.ia7.scarblac@pino.selwerd.nl>; from scarblac@pino.selwerd.nl on Tue, Feb 27, 2001 at 08:08:09AM +0000
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi> <slrn99mnvn.ia7.scarblac@pino.selwerd.nl>
Message-ID: <20010227014649.A13115@tummy.com>

On Tue, Feb 27, 2001 at 08:08:09AM +0000, Remco Gerlich wrote:
>He doesn't; .0.255 and .1.0 are not IP addresses you could give to a machine.
>I've never seen existing code for this, seems to need some custom hacking.

It all depends on your netmask:

	10.0.0.254/24 + 1 = 10.0.1.1/24
	10.0.0.254/23 + 1 = 10.0.0.255/23
	10.0.0.2/30 + 1 = 10.0.0.5/30

Sean
-- 
 A computer is like an Old Testament god, with a lot of rules and no mercy.
                -- Joseph Campbell
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From erno-news at erno.iki.fi  Tue Feb 27 12:55:50 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 27 Feb 2001 13:55:50 +0200
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi> <slrn99mnvn.ia7.scarblac@pino.selwerd.nl>
Message-ID: <kuwvacpcft.fsf@lasipalatsi.fi>

In article <slrn99mnvn.ia7.scarblac at pino.selwerd.nl>,
scarblac at pino.selwerd.nl (Remco Gerlich) writes:

| .0.255 and .1.0 are not IP addresses you could give to a machine.
| I've never seen existing code for this, seems to need some custom hacking.

| (I don't know if .255 and .0 are the only exceptions, probably not).

why do you think .0 or .255 are not usable addresses?

  -- erno




From scarblac at pino.selwerd.nl  Tue Feb 27 18:40:45 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 27 Feb 2001 17:40:45 GMT
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi> <slrn99mnvn.ia7.scarblac@pino.selwerd.nl> <kuwvacpcft.fsf@lasipalatsi.fi>
Message-ID: <slrn99npha.j09.scarblac@pino.selwerd.nl>

Erno Kuusela <erno-news at erno.iki.fi> wrote in comp.lang.python:
> In article <slrn99mnvn.ia7.scarblac at pino.selwerd.nl>,
> scarblac at pino.selwerd.nl (Remco Gerlich) writes:
> 
> | .0.255 and .1.0 are not IP addresses you could give to a machine.
> | I've never seen existing code for this, seems to need some custom hacking.
> 
> | (I don't know if .255 and .0 are the only exceptions, probably not).
> 
> why do you think .0 or .255 are not usable addresses?

I was confused about netmasks. I didn't expect the Spanish Inquisition!

-- 
Remco Gerlich


From joconnor at cybermesa.com  Tue Feb 27 19:04:50 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 27 Feb 2001 11:04:50 -0700
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi> <slrn99mnvn.ia7.scarblac@pino.selwerd.nl> <kuwvacpcft.fsf@lasipalatsi.fi> <slrn99npha.j09.scarblac@pino.selwerd.nl>
Message-ID: <3A9BEC42.5F9888@cybermesa.com>


Remco Gerlich wrote:

> Erno Kuusela <erno-news at erno.iki.fi> wrote in comp.lang.python:
> > In article <slrn99mnvn.ia7.scarblac at pino.selwerd.nl>,
> > scarblac at pino.selwerd.nl (Remco Gerlich) writes:
> >
> > | .0.255 and .1.0 are not IP addresses you could give to a machine.
> > | I've never seen existing code for this, seems to need some custom hacking.
> >
> > | (I don't know if .255 and .0 are the only exceptions, probably not).
> >
> > why do you think .0 or .255 are not usable addresses?
>
> I was confused about netmasks. I didn't expect the Spanish Inquisition!
>

No one expects the Spanish Inquisition



(How many other identical posts will there be ?)




From f8dy at yahoo.com  Tue Feb 27 20:35:32 2001
From: f8dy at yahoo.com (Mark Pilgrim)
Date: Tue, 27 Feb 2001 14:35:32 -0500
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi> <slrn99mnvn.ia7.scarblac@pino.selwerd.nl> <kuwvacpcft.fsf@lasipalatsi.fi> <slrn99npha.j09.scarblac@pino.selwerd.nl> <3A9BEC42.5F9888@cybermesa.com>
Message-ID: <97gvf1$pdbof$1@ID-77331.news.dfncis.de>

"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
news:3A9BEC42.5F9888 at cybermesa.com...
>
>
> Remco Gerlich wrote:
>
> > Erno Kuusela <erno-news at erno.iki.fi> wrote in comp.lang.python:
> > > In article <slrn99mnvn.ia7.scarblac at pino.selwerd.nl>,
> > > scarblac at pino.selwerd.nl (Remco Gerlich) writes:
> > >
> > > | .0.255 and .1.0 are not IP addresses you could give to a machine.
> > > | I've never seen existing code for this, seems to need some custom
hacking.
> > >
> > > | (I don't know if .255 and .0 are the only exceptions, probably not).
> > >
> > > why do you think .0 or .255 are not usable addresses?
> >
> > I was confused about netmasks. I didn't expect the Spanish Inquisition!
> >
>
> No one expects the Spanish Inquisition
>
>
>
> (How many other identical posts will there be ?)

The real question is, why hasn't anyone written a bot to automate this?

import nntplib
import time

def inquisitionBot(host, port=119, user=None, password=None,
starttime=None):
 if not starttime:
  starttime = time.localtime(time.time() - 60*60*24)
 s = nntplib.NNTP(host, port, user, password)
 try:
  resp, articles = s.newnews("comp.lang.python", \
           time.strftime("%y%m%d", starttime), \
           time.strftime("%H%M%S", starttime))
  for articleID in articles:
   resp, status, cmd, body = s.body(articleID)
   if [line for line in body if \
    line[:1]<>'>' and \
    line.lower().count("i didn't expect the spanish inquisition")]:
    # if I continue this someone might actually use it... :)
    pass
 finally:
  s.quit()

-M
You're smart; why haven't you learned Python yet?
http://diveintopython.org/





From joconnor at cybermesa.com  Mon Feb 26 22:44:52 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 26 Feb 2001 14:44:52 -0700
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org>
Message-ID: <3A9ACE54.65706115@cybermesa.com>


Timothy Grant wrote:

> Hi all,
>
> I checked Parnassus, but couldn't find anything, so I'll ask
> here. Has anyone done anything with IP math (e.g.,
> 192.168.0.254 + 1 = 192.168.1.1).

You could probably do it by converting the IP address to a number, do
the math, and convert back.  I've seen algorithms for doing this.

 Given a.b.c.d for an IP address you can convert it to the proper value
using:

a(256^3) + b(256^2) + c(256) + d

So 192.168.0.254 = (192 * long((256 **3))) + (168 * (256 ** 2)) + (0 *
256) + 254 = 3232235774L

You need the long() in there to avoid an integer multiplication error.

You might be able to use '3232235774' directly, or convert back to an ip
address

Take care,
Jay





From tjg at exceptionalminds.com  Tue Feb 27 02:02:23 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Mon, 26 Feb 2001 17:02:23 -0800
Subject: IP Math anyone?
In-Reply-To: <20010226125903.X29494@trufflehunter.avalongroup.net>; from tjg@exceptionalminds.com on Mon, Feb 26, 2001 at 12:59:03PM -0800
References: <20010226125903.X29494@trufflehunter.avalongroup.net>
Message-ID: <20010226170223.B29494@trufflehunter.avalongroup.net>

On Mon, Feb 26, 2001 at 12:59:03PM -0800, Timothy Grant wrote:
> Hi all,
> 
> I checked Parnassus, but couldn't find anything, so I'll ask
> here. Has anyone done anything with IP math (e.g.,
> 192.168.0.254 + 1 = 192.168.1.1).
> 
> 

Thanks to all who replied to this. I greatly appreciated it!
I'll be checking out each of those solutions carefully.

-- 
Stand Fast,
    rhacer.

Timothy Grant, RHCE <><                              MIG #1433
tjg at exceptionalminds.com                          '00 Marauder
www.exceptionalminds.com/rhacer                   Craigelachie 
>>>>>>>>>>Linux, because rebooting is *NOT* normal.<<<<<<<<<<<
>>>>This machine was last rebooted:  41 days  5:14 hours ago<<



From tino at wildenhain.de  Tue Feb 27 10:37:11 2001
From: tino at wildenhain.de (Tino Wildenhain)
Date: Tue, 27 Feb 2001 10:37:11 +0100
Subject: IP Math anyone?
In-Reply-To: <20010226170223.B29494@trufflehunter.avalongroup.net>
References: <20010226170223.B29494@trufflehunter.avalongroup.net>
Message-ID: <73931622.983270231@[10.0.0.101]>

Hi Timothy,

for teaching I made a class for this.

you can instantiate it with a=ip('192.168.0.254/24') or 
a=ip('192.168.0.254','255.255.255.0') just using a=ip('192.168.0.254') 
would assign a appropriate subnet mask for this class (class B results in 
/16 or 255.255.0.0)

Since IP-adresses are only 32bit non signed integers, they are internally 
represented
like that.
If you make a class like that you can simply add the support to add 
integers to
it according to the subnet-mask you use.

Regards
Tino

--On Montag, 26. Februar 2001 17:02 -0800 Timothy Grant 
<tjg at exceptionalminds.com> wrote:

> On Mon, Feb 26, 2001 at 12:59:03PM -0800, Timothy Grant wrote:
>> Hi all,
>>
>> I checked Parnassus, but couldn't find anything, so I'll ask
>> here. Has anyone done anything with IP math (e.g.,
>> 192.168.0.254 + 1 = 192.168.1.1).
>>
>>
>
> Thanks to all who replied to this. I greatly appreciated it!
> I'll be checking out each of those solutions carefully.
>
> --
> Stand Fast,
>     rhacer.
>
> Timothy Grant, RHCE <><                              MIG #1433
> tjg at exceptionalminds.com                          '00 Marauder
> www.exceptionalminds.com/rhacer                   Craigelachie
>>>>>>>>>>> Linux, because rebooting is *NOT* normal.<<<<<<<<<<<
>>>>> This machine was last rebooted:  41 days  5:14 hours ago<<
>
> --
> http://mail.python.org/mailman/listinfo/python-list







From chris.gonnerman at usa.net  Tue Feb 27 14:38:56 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Tue, 27 Feb 2001 07:38:56 -0600
Subject: IP Math anyone?
References: <20010226170223.B29494@trufflehunter.avalongroup.net> <73931622.983270231@[10.0.0.101]>
Message-ID: <003f01c0a0c2$ac255040$a100000a@local>

You neglected to tell us where to download your class from, or to attach
the source.  I at least would like to see it.

----- Original Message -----
From: "Tino Wildenhain" <tino at wildenhain.de>
Subject: Re: IP Math anyone?
> Hi Timothy,
>
> for teaching I made a class for this.
>
> you can instantiate it with a=ip('192.168.0.254/24') or
> a=ip('192.168.0.254','255.255.255.0') just using a=ip('192.168.0.254')
> would assign a appropriate subnet mask for this class (class B results in
> /16 or 255.255.0.0)
>
> Since IP-adresses are only 32bit non signed integers, they are internally
> represented
> like that.
> If you make a class like that you can simply add the support to add
> integers to
> it according to the subnet-mask you use.
>
> Regards
> Tino





From tjg at exceptionalminds.com  Tue Feb 27 20:08:30 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Tue, 27 Feb 2001 11:08:30 -0800
Subject: IP Math anyone?
In-Reply-To: <003f01c0a0c2$ac255040$a100000a@local>; from chris.gonnerman@usa.net on Tue, Feb 27, 2001 at 07:38:56AM -0600
References: <20010226170223.B29494@trufflehunter.avalongroup.net> <73931622.983270231@[10.0.0.101]> <003f01c0a0c2$ac255040$a100000a@local>
Message-ID: <20010227110830.I29494@trufflehunter.avalongroup.net>

Thanks to all of you who made this thread much more
enlightening than expected! As I expected, it is a somewhat
complex problem, for which I will probably throw together a
class with a nice interface but internals of which are a crude
hack, and then refine the internals as time allows.


On Tue, Feb 27, 2001 at 07:38:56AM -0600, Chris Gonnerman wrote:
> You neglected to tell us where to download your class from, or to attach
> the source.  I at least would like to see it.
> 
> ----- Original Message -----
> From: "Tino Wildenhain" <tino at wildenhain.de>
> Subject: Re: IP Math anyone?
> > Hi Timothy,
> >
> > for teaching I made a class for this.
> >
> > you can instantiate it with a=ip('192.168.0.254/24') or
> > a=ip('192.168.0.254','255.255.255.0') just using a=ip('192.168.0.254')
> > would assign a appropriate subnet mask for this class (class B results in
> > /16 or 255.255.0.0)
> >
> > Since IP-adresses are only 32bit non signed integers, they are internally
> > represented
> > like that.
> > If you make a class like that you can simply add the support to add
> > integers to
> > it according to the subnet-mask you use.
> >
> > Regards
> > Tino

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.         <><       (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  41 days 23:20 hours ago<<



From me at home.nl  Mon Feb 26 22:22:32 2001
From: me at home.nl (SonoBull)
Date: Mon, 26 Feb 2001 21:22:32 GMT
Subject: First shot at python.
Message-ID: <sOzm6.53721$zl4.1692936@zwoll1.home.nl>

im not a real programmer, having only a little experince in vb pascal an
smalltalk.

But gona try mastering the art of python programming.

First seconds account:

Downloaded wrong installing file. Finaly got it, installed it, and
immediatly decided to download the tutoria.

Got fed up swithing between tutorial and Python and printed the whole damm
thing.

Ended up with chapter 7.

Maybe in a couple of years of practice i wil be a good programmer,

Well, wish me luck. Tomorrow at it again.

gr.

son.







From sheila at spamcop.net  Mon Feb 26 22:33:44 2001
From: sheila at spamcop.net (Sheila King)
Date: Mon, 26 Feb 2001 21:33:44 GMT
Subject: First shot at python.
References: <sOzm6.53721$zl4.1692936@zwoll1.home.nl>
Message-ID: <aqil9t8bpkf8s8pg7othuenl81gr7e8u7s@4ax.com>

On Mon, 26 Feb 2001 21:22:32 GMT, "SonoBull" <me at home.nl> wrote in
comp.lang.python in article <sOzm6.53721$zl4.1692936 at zwoll1.home.nl>:

:First seconds account:
:
:Downloaded wrong installing file. Finaly got it, installed it, and
:immediatly decided to download the tutoria.

I can feel your frustration.

Wasn't the tutorial included in the installation files you downloaded? When I
installed Python, all the docs came in the installation package.

:Got fed up swithing between tutorial and Python and printed the whole damm
:thing.

Why do you have to "switch"? Looking at your message headers, I would guess
you are on a Windows machine. Can't you have two windows open at the same
time?

:Ended up with chapter 7.

Yikes!
:
:Maybe in a couple of years of practice i wil be a good programmer,
:
:Well, wish me luck. Tomorrow at it again.

Yes, good luck! Keep at it!

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From sbrunning at trisystems.co.uk  Tue Feb 27 10:52:30 2001
From: sbrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 27 Feb 2001 09:52:30 -0000
Subject: First shot at python.
References: <sOzm6.53721$zl4.1692936@zwoll1.home.nl>
Message-ID: <3a9b8049_1@nnrp1.news.uk.psi.net>

SonoBull wrote in message ...
>im not a real programmer, having only a little experince in vb
pascal an
>smalltalk.
>
>But gona try mastering the art of python programming.

If you haven't really programmed before, you might find the tutorial
a little steep. If this is the case, you might want to look at Alan
Gauld's 'Learning to Program using Python', at
http://www.crosswinds.net/~agauld/.

Cheers,
Simon Brunning
sbrunning at bigfoot.com





From me at home.nl  Tue Feb 27 11:44:48 2001
From: me at home.nl (SonoBull)
Date: Tue, 27 Feb 2001 10:44:48 GMT
Subject: First shot at python.
References: <sOzm6.53721$zl4.1692936@zwoll1.home.nl> <3a9b8049_1@nnrp1.news.uk.psi.net>
Message-ID: <AyLm6.55414$zl4.1722681@zwoll1.home.nl>

Tnx looks like i can use it.

But small question.

Ive been told that python is a good "leaning" language and is a good
"expert" language.

do u feel the same abouth that? or should i try st else first?

gr.

son


Simon Brunning <sbrunning at trisystems.co.uk> schreef in berichtnieuws
3a9b8049_1 at nnrp1.news.uk.psi.net...
> SonoBull wrote in message ...
> >im not a real programmer, having only a little experince in vb
> pascal an
> >smalltalk.
> >
> >But gona try mastering the art of python programming.
>
> If you haven't really programmed before, you might find the tutorial
> a little steep. If this is the case, you might want to look at Alan
> Gauld's 'Learning to Program using Python', at
> http://www.crosswinds.net/~agauld/.
>
> Cheers,
> Simon Brunning
> sbrunning at bigfoot.com
>
>
>




From sbrunning at trisystems.co.uk  Tue Feb 27 12:27:24 2001
From: sbrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 27 Feb 2001 11:27:24 -0000
Subject: First shot at python.
References: <sOzm6.53721$zl4.1692936@zwoll1.home.nl> <3a9b8049_1@nnrp1.news.uk.psi.net> <AyLm6.55414$zl4.1722681@zwoll1.home.nl>
Message-ID: <3a9b9465$1_1@nnrp1.news.uk.psi.net>

SonoBull wrote in message ...
>Ive been told that python is a good "leaning" language and is a
good
>"expert" language.

It's *certainly* a good learning language. As to whether it's a good
language for experts, well, you had better ask an expert. <wink>

Well, actually I know enough to recognise an expert when I see one,
and there are plenty using Python on this newsgroup, so I'd say that
it *is* a good expert language.

Cheers,
Simon Brunning
sbrunning at bigfoot.com





From sholden at holdenweb.com  Wed Feb 28 16:23:27 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 28 Feb 2001 10:23:27 -0500
Subject: First shot at python.
References: <sOzm6.53721$zl4.1692936@zwoll1.home.nl> <3a9b8049_1@nnrp1.news.uk.psi.net> <AyLm6.55414$zl4.1722681@zwoll1.home.nl> <3a9b9465$1_1@nnrp1.news.uk.psi.net>
Message-ID: <CN8n6.4913$zL.224621@e420r-atl1.usenetserver.com>

"Simon Brunning" <sbrunning at trisystems.co.uk> wrote in message
news:3a9b9465$1_1 at nnrp1.news.uk.psi.net...
> SonoBull wrote in message ...
> >Ive been told that python is a good "leaning" language and is a
> good
> >"expert" language.
>
> It's *certainly* a good learning language. As to whether it's a good
> language for experts, well, you had better ask an expert. <wink>
>
> Well, actually I know enough to recognise an expert when I see one,
> and there are plenty using Python on this newsgroup, so I'd say that
> it *is* a good expert language.
>
"ex" as in "has-been", "spurt" as in "drip under pressure".
Expert = "has-been drip under pressure".

regards
 Steve





From sbrunning at trisystems.co.uk  Wed Feb 28 17:02:29 2001
From: sbrunning at trisystems.co.uk (Simon Brunning)
Date: Wed, 28 Feb 2001 16:02:29 -0000
Subject: First shot at python.
References: <sOzm6.53721$zl4.1692936@zwoll1.home.nl> <3a9b8049_1@nnrp1.news.uk.psi.net> <AyLm6.55414$zl4.1722681@zwoll1.home.nl> <3a9b9465$1_1@nnrp1.news.uk.psi.net> <CN8n6.4913$zL.224621@e420r-atl1.usenetserver.com>
Message-ID: <3a9d2e4b_2@nnrp1.news.uk.psi.net>

Steve Holden wrote in message ...
>> It's *certainly* a good learning language. As to whether it's a
good
>> language for experts, well, you had better ask an expert. <wink>
>>
>"ex" as in "has-been", "spurt" as in "drip under pressure".
>Expert = "has-been drip under pressure".

Hmmm. Perhaps I'm an expert after all.

Cheers,
Simon Brunning
sbrunning at bigfoot.com





From phd at phd.russ.ru  Mon Feb 26 22:25:12 2001
From: phd at phd.russ.ru (Oleg Broytmann)
Date: 26 Feb 2001 15:25:12 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Feb 26)
Message-ID: <39CF1142A433B303.A58DD4189090B9CD.F336C38ED04DC6FE@lp.airnews.net>

    Jeremy Hylton updates PEP 227 (static scoping)
        http://groups.yahoo.com/group/python-list/message/124981
    and thus starts a hot discussion
        http://groups.yahoo.com/group/python-list/messagesearch?query=Update%20to%20PEP%20227%20%28static%20scoping%29
    Guido van Rossum announces nested scopes in 2.1 and 2.2 optional
        http://mail.python.org/pipermail/python-list/2001-February/030681.html
    
    David Mertz announces an oops on gnosis.cx/publish/
    and moves to new hosting
        http://mail.python.org/pipermail/python-list/2001-February/030550.html
    David Mertz: Charming Python: Getting version 2.0
        http://www-106.ibm.com/developerworks/library/l-py20.html?dwzone=linux
    David Mertz: Functional Programming in Python
        http://gnosis.cx/publish/programming/charming_python_13.txt
    
    Philipp wrapps www.graphviz.org's 'dot' language in Python
        http://groups.yahoo.com/group/python-list/message/124396
    
    python powered mp3 management
        http://mpy3.thelocust.org/
    
    Thomas Heller shows how to enter a (possibly dotted) python
    module name in the Emacs' minibuffer
        http://groups.yahoo.com/group/python-list/message/124525
    
    gnome-python 1.4.0: Updated to work with the latest GtkHTML
        http://www.daa.com.au/~james/gnome/
    
    Jerome Alet annonces new version of JAXML XML generator
        http://groups.yahoo.com/group/python-list/message/124939
    
    Warren Postma announces event module 1.0 for better threading
        http://groups.yahoo.com/group/python-list/message/124830
    
    Barry Warsaw updates PEP 232, Function Attributes
        http://groups.yahoo.com/group/python-list/message/124789
    
    Amos Latteier announces Zope Directions Roadmap
        http://mail.python.org/pipermail/python-announce-list/2001-February/000673.html
    Stephen Figgins: Zope's New Roadmap
        http://www.oreillynet.com/pub/a/python/2001/02/22/pythonnews.html
    
    Jean-Luc Fontaine creates tclpython that embeds python into tcl
        http://mail.python.org/pipermail/python-list/2001-February/030652.html
    
    Lars Marius Garshol wraps James Clark's SP SGML parser
        http://www.garshol.priv.no/download/software/pysp/
    
    Norbert Klamann creates a working setup.py for DCOracle 1.3.2
        http://groups.yahoo.com/group/python-list/message/124940
    
    Marc-Andre Lemburg launches eGenix and announces that Extensions for Python 2.0
        http://mail.python.org/pipermail/python-announce-list/2001-February/000679.html
    
    Computer scientists at Endeavors Technology have successfully
    developed a highly optimized, open source port of Python to the
    Palm OS(R) platform to stimulate interest in creating Python-based
    applications for Palm(TM) handheld computers and other devices that
    use the Palm operating system.
        http://sf-web1.businesswire.com/cgi-bin/f_headline.cgi?day0/210540113&ticker=
        http://www.endeavors.com/pippy/
    
    Chris Gonnerman posts an example code explaining how to extend
    Python in C/C++
        http://groups.yahoo.com/group/python-list/message/125374
    
    Uche Ogbuji announces "The Python Web services developer" column
        http://groups.yahoo.com/group/python-list/message/125405
    
    Mark Hammond explains how to ensures the thread-lock is released
    when the function is called (SWIG)
        http://groups.yahoo.com/group/python-list/message/125473
    
    Sean Reifschneider comtinues to think about CPAN functionality
    for python
        http://groups.yahoo.com/group/python-list/message/125510
    
    
    Zope
    ----
    
    Brian Lloyd announces SECURITY alert and hotfix release
        http://lists.zope.org/pipermail/zope/2001-February/084053.html
    
    Michel Pelletier announces Zope Book is now in public CVS
        http://lists.zope.org/pipermail/zope/2001-February/084061.html
    
    
    Hot threads of the week
    -----------------------
    
    locking files
        http://groups.google.com/groups?lr=&safe=off&ic=1&th=e046814cf6e52ae3&seekd=917102346
    Alternative iterator syntax
        http://groups.google.com/groups?lr=&safe=off&ic=1&th=f90597b5cb186d16&seekd=917023515
    Comparison with Ruby
        http://groups.google.com/groups?lr=&safe=off&ic=1&th=105b832cf34cd872&seekd=916928222
    Rounding Question
        http://groups.google.com/groups?lr=&safe=off&ic=1&th=d054718b0b4ad260&seekd=916921447
    import statement is case sensitive
        http://groups.google.com/groups?lr=&safe=off&ic=1&th=24e0f1d3ddcb2ac5&seekd=916916940
    An alternative approach to bound methods
        http://groups.google.com/groups?lr=&safe=off&ic=1&th=307b2dd2b5d6bc3f&seekd=916902767


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org
      Notice especially the master FAQ
          http://www.python.org/doc/FAQ.html

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
	  http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

      Michael Hudson continues Andrew Kuchling's marvelous tradition
      of summarizing action on the python-dev mailing list twice a
      month.
	  http://starship.python.net/crew/mwh/summaries/
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collect Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
	  http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From brian.j.geddes at intel.com  Mon Feb 26 23:11:09 2001
From: brian.j.geddes at intel.com (Brian Geddes)
Date: Mon, 26 Feb 2001 14:11:09 -0800
Subject: Packet parsing problem...
Message-ID: <97ek6i$o6d@news.or.intel.com>

Here's my situation:

There is an already-existing server (written in C++), which communicates
with clients through 1-packet messages.  Each packet consists of a header of
2 or 3 C++ DWORDs, followed by 0 to n bytes of data.  I have to write a
Python client to communicate with this server.

In python, is there any structure that approximates the C++ DWORD?  I need
to be able to parse incoming packets, as well as properly form messages to
the server.

Any suggestions?

Thanks,
- Brian




From BrianQ at ActiveState.com  Mon Feb 26 23:43:26 2001
From: BrianQ at ActiveState.com (Brian Quinlan)
Date: Mon, 26 Feb 2001 14:43:26 -0800
Subject: Packet parsing problem...
In-Reply-To: <97ek6i$o6d@news.or.intel.com>
Message-ID: <002201c0a045$893c6c80$b503a8c0@activestate.ca>

A DWORD is just a signed 32-bit integer. Your parsing code could look
something like this (usual disclaimer of little thought and no testing).
Actually, this is pretty ugly (integerize could be a lot cleaner). There has
to be library code to do this...

import socket

# Do communication stuff here e.g. header = mySocket.recv( 2 * 4 )

def integerize( aString ):
	myInt = 0
	for i in range(4):
		myInt = myInt + ord( aString[i] ) << 8 * (3 - i )
	return myInt

word1 = socket.ntohl( integerize( header[0:3] ) )
word2 = socket.ntohl( integerize( header[4:7] ) )
# ... Do more stuff

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Brian Geddes
Sent: Monday, February 26, 2001 2:11 PM
To: python-list at python.org
Subject: Packet parsing problem...


Here's my situation:

There is an already-existing server (written in C++), which communicates
with clients through 1-packet messages.  Each packet consists of a header of
2 or 3 C++ DWORDs, followed by 0 to n bytes of data.  I have to write a
Python client to communicate with this server.

In python, is there any structure that approximates the C++ DWORD?  I need
to be able to parse incoming packets, as well as properly form messages to
the server.

Any suggestions?

Thanks,
- Brian


--
http://mail.python.org/mailman/listinfo/python-list




From ljohnson at resgen.com  Tue Feb 27 00:03:08 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Mon, 26 Feb 2001 17:03:08 -0600
Subject: Packet parsing problem...
References: <97ek6i$o6d@news.or.intel.com>
Message-ID: <t9lo117ns2so54@corp.supernews.com>

> In python, is there any structure that approximates the C++ DWORD?  I need
> to be able to parse incoming packets, as well as properly form messages to
> the server.

I hate to ask, but is this the DWORD that's commonly typedef'd in the
Windows header files? If so, it's just a 32-bit (signed) integer and it
should map to a Python integer.




From ben.hutchings at roundpoint.com  Tue Feb 27 04:09:01 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 26 Feb 2001 19:09:01 -0800
Subject: Packet parsing problem...
References: <97ek6i$o6d@news.or.intel.com> <t9lo117ns2so54@corp.supernews.com>
Message-ID: <ug0h0kek2.fsf@roundpoint.com>

"Lyle Johnson" <ljohnson at resgen.com> writes:

> > In python, is there any structure that approximates the C++ DWORD?  I need
> > to be able to parse incoming packets, as well as properly form messages to
> > the server.
> 
> I hate to ask, but is this the DWORD that's commonly typedef'd in the
> Windows header files? If so, it's just a 32-bit (signed) integer and it
> should map to a Python integer.

No, DWORD is unsigned.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From news at dorb.com  Tue Feb 27 01:38:39 2001
From: news at dorb.com (Darrell)
Date: Tue, 27 Feb 2001 00:38:39 GMT
Subject: Packet parsing problem...
References: <97ek6i$o6d@news.or.intel.com>
Message-ID: <jGCm6.138929$v.12203221@typhoon.nyroc.rr.com>

When parsing C structs use the struct model.

import struct
header = .........

fmt2="!II"
fmt2Size=struct.calcsize(fmt2)
fmt3="!III"

if len(header) == fmt2Size:
    values = struct.unpack(fmt2, header)
else:
    values = struct.unpack(fmt3, header)

--Darrell

"Brian Geddes" <brian.j.geddes at intel.com> wrote in message
news:97ek6i$o6d at news.or.intel.com...
> Here's my situation:
>
> There is an already-existing server (written in C++), which communicates
> with clients through 1-packet messages.  Each packet consists of a header
of
> 2 or 3 C++ DWORDs, followed by 0 to n bytes of data.  I have to write a
> Python client to communicate with this server.
>
> In python, is there any structure that approximates the C++ DWORD?  I need
> to be able to parse incoming packets, as well as properly form messages to
> the server.
>
> Any suggestions?
>
> Thanks,
> - Brian
>
>




From erno-news at erno.iki.fi  Tue Feb 27 01:46:26 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 27 Feb 2001 02:46:26 +0200
Subject: Packet parsing problem...
References: <97ek6i$o6d@news.or.intel.com>
Message-ID: <kulmqtq7fh.fsf@lasipalatsi.fi>

In article <97ek6i$o6d at news.or.intel.com>, "Brian Geddes"
<brian.j.geddes at intel.com> writes:

| Here's my situation:
| There is an already-existing server (written in C++), which communicates
| with clients through 1-packet messages.  Each packet consists of a header of
| 2 or 3 C++ DWORDs, followed by 0 to n bytes of data.  I have to write a
| Python client to communicate with this server.

| In python, is there any structure that approximates the C++ DWORD?  I need
| to be able to parse incoming packets, as well as properly form messages to
| the server.

i don't know what a DWORD is, but i suspect it might be an alias
for some integer or pointer type in C++ on your software platform. in
that case you can use the struct module to convert a sequence
of bytes to the corresponding integer. things to keep
in mind when using the struct module for on the wire protocols:

1) padding. for example, on the machine i'm reading my news on,
>>> struct.calcsize('ich')
8
the struct module attempts to emulate the layout of c structs
in memory on your platform, and that often includes padding.
so it may be better to unpack the packets one item at a time.

2) portability - the sizes of the types use by the stuct module
can vary accross platforms. you may decide it's ok to not
be portable, or you may want to use different formats on different
platforms you want your program to run on, or for purists,
you may want to abandon the struct module completely for this
purpouse and do the work completely by hand via ord() an shifts.
also byte endianness - you probably need to specify the endianness
explicitly.

it would be nice to have a struct-ish module in the standard
distribution that would be geared towards working with data
formats with known bit layout...

  -- erno


From ben.hutchings at roundpoint.com  Tue Feb 27 22:37:08 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 27 Feb 2001 13:37:08 -0800
Subject: Packet parsing problem...
References: <97ek6i$o6d@news.or.intel.com> <kulmqtq7fh.fsf@lasipalatsi.fi>
Message-ID: <u7l2bkdtn.fsf@roundpoint.com>

Erno Kuusela <erno-news at erno.iki.fi> writes:

> In article <97ek6i$o6d at news.or.intel.com>, "Brian Geddes"
> <brian.j.geddes at intel.com> writes:
> 
> | Here's my situation:

> | There is an already-existing server (written in C++), which
> | communicates with clients through 1-packet messages.  Each packet
> | consists of a header of 2 or 3 C++ DWORDs, followed by 0 to n bytes
> | of data.  I have to write a Python client to communicate with this
> | server.
> 
> | In python, is there any structure that approximates the C++ DWORD?
> | I need to be able to parse incoming packets, as well as properly
> | form messages to the server.
<snip>
> it would be nice to have a struct-ish module in the standard
> distribution that would be geared towards working with data
> formats with known bit layout...

We already have one!  You can use a modifier at the beginning of a
struct format string to select one of 5 size, byte ordering and
alignment rules (the first one being the default, native, rules).

So I suppose Brian might read messages using something like this:

    from struct import *
    (type,) = unpack(message[0:4], '<L')
    if type == TYPE_EGGS:
        param1, length = unpack(message[4:8], '<LL')
        param2 = message[8:length]
    elif type == TYPE_SPAM:
        ...

I'm not sure quite how he'd want to deal with invalid messages; the
above doesn't attempt to do anything useful with them since he hasn't
said much about the message format.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From bawolk at ucdavis.edu  Tue Feb 27 00:05:57 2001
From: bawolk at ucdavis.edu (Bruce Wolk)
Date: Mon, 26 Feb 2001 15:05:57 -0800
Subject: PyQt file locations
Message-ID: <3A9AE155.FF8C3A08@ucdavis.edu>

I installed the PyQt bindings and SIP on my RH 6.2 Linux system.
Everything compiled quite nicely.  But the files do not appear to have
been placed in the correct subdirectories.  None of the samples will
run.  Where should the qt directory containing qt.py be located?  I
tried moving it to the site-packages directory, but that didn't work.

Thanks.

Bruce





From aaron.ginn at motorola.com  Tue Feb 27 23:24:08 2001
From: aaron.ginn at motorola.com (Aaron Ginn)
Date: 27 Feb 2001 15:24:08 -0700
Subject: PyQt file locations
References: <3A9AE155.FF8C3A08@ucdavis.edu>
Message-ID: <snr90jepdj.fsf@motorola.com>

Bruce Wolk <bawolk at ucdavis.edu> writes:

> I installed the PyQt bindings and SIP on my RH 6.2 Linux system.
> Everything compiled quite nicely.  But the files do not appear to have
> been placed in the correct subdirectories.  None of the samples will
> run.  Where should the qt directory containing qt.py be located?  I
> tried moving it to the site-packages directory, but that didn't work.
> 
> Thanks.
> 
> Bruce

Here's what's in my site-packages directory:

ginn at coronado ginn $ ls python-2.0/lib/python2.0/site-packages/
README                libqtcmodule-2.3.so*  libqtcmodule.so@      qt.pyc
eric/                 libqtcmodule.la*      qt.py                 qt.pyo

You can't simply move the qt directory there, as there are many more
files than what are in my site-packages directory.  Shouldn't 'make
install' have taken care of this for you?  What happened when you did
a make install?

-- 
Aaron J. Ginn                    Phone: 480-814-4463
Motorola SemiCustom Solutions    Pager: 877-586-2318
1300 N. Alma School Rd.          Fax  : 480-814-4463
Chandler, AZ 85226 M/D CH260     mailto:aaron.ginn at motorola.com


From bawolk at ucdavis.edu  Wed Feb 28 04:20:11 2001
From: bawolk at ucdavis.edu (Bruce Wolk)
Date: Tue, 27 Feb 2001 19:20:11 -0800
Subject: PyQt file locations
References: <3A9AE155.FF8C3A08@ucdavis.edu> <snr90jepdj.fsf@motorola.com>
Message-ID: <3A9C6E6B.3B727A37@ucdavis.edu>

I have the same files in my site-packages directory, except for qt.pyc.

make-install gave me some errors:
_______________________________

Making install in qt
make[1]: Entering directory `/home/bruce/down/PyQt-2.3/qt'
make[2]: Entering directory `/home/bruce/down/PyQt-2.3/qt'
/bin/sh ../mkinstalldirs /usr/local/lib/python2.0/site-packages
/bin/sh ../libtool  --mode=install /usr/bin/install -c libqtcmodule.la
/usr/local/lib/python2.0/site-packages/libqtcmodule.la
/usr/bin/install -c .libs/libqtcmodule-2.3.so
/usr/local/lib/python2.0/site-packages/libqtcmodule-2.3.so
(cd /usr/local/lib/python2.0/site-packages && rm -f libqtcmodule.so && ln
-s libqtcmodule-2.3.so libqtcmodule.so)
/usr/bin/install -c .libs/libqtcmodule.lai
/usr/local/lib/python2.0/site-packages/libqtcmodule.la
PATH="$PATH:/sbin" ldconfig -n /usr/local/lib/python2.0/site-packages
----------------------------------------------------------------------
Libraries have been installed in:
   /usr/local/lib/python2.0/site-packages

If you ever happen to want to link against installed libraries
in a given directory, LIBDIR, you must either use libtool, and
specify the full pathname of the library, or use `-LLIBDIR'
flag during linking and do at least one of the following:
   - add LIBDIR to the `LD_LIBRARY_PATH' environment variable
     during execution
   - add LIBDIR to the `LD_RUN_PATH' environment variable
     during linking
   - use the `-Wl,--rpath -Wl,LIBDIR' linker flag
   - have your system administrator add LIBDIR to `/etc/ld.so.conf'

See any operating system documentation about shared libraries for
more information, such as the ld(1) and ld.so(8) manual pages.
----------------------------------------------------------------------
/bin/sh ../mkinstalldirs /usr/local/lib/python2.0/site-packages
 /usr/bin/install -c -m 644 ./qt.py
/usr/local/lib/python2.0/site-packages/qt.pymake  install-data-hook
make[3]: Entering directory `/home/bruce/down/PyQt-2.3/qt'
(cd /tmp;
PYTHONPATH=/usr/local/lib/python2.0/site-packages:/usr/local/lib/python2.0/site-packages
python -O -c "import qt")
Traceback (most recent call last):
  File "<string>", line 1, in ?
  File "/usr/local/lib/python2.0/site-packages/qt.py", line 43, in ?
    import libqtc
ImportError: libsip.so.4: cannot open shared object file: No such file or
directory
make[3]: *** [install-data-hook] Error 1
make[3]: Leaving directory `/home/bruce/down/PyQt-2.3/qt'
make[2]: *** [install-data-am] Error 2
make[2]: Leaving directory `/home/bruce/down/PyQt-2.3/qt'
make[1]: *** [install-am] Error 2
make[1]: Leaving directory `/home/bruce/down/PyQt-2.3/qt'
make: *** [install-recursive] Error 1
________________________

When I try to run the examples, I get the same ImportError.

Any ideas?

Bruce


Aaron Ginn wrote:

> Bruce Wolk <bawolk at ucdavis.edu> writes:
>
> > I installed the PyQt bindings and SIP on my RH 6.2 Linux system.
> > Everything compiled quite nicely.  But the files do not appear to have
> > been placed in the correct subdirectories.  None of the samples will
> > run.  Where should the qt directory containing qt.py be located?  I
> > tried moving it to the site-packages directory, but that didn't work.
> >
> > Thanks.
> >
> > Bruce
>
> Here's what's in my site-packages directory:
>
> ginn at coronado ginn $ ls python-2.0/lib/python2.0/site-packages/
> README                libqtcmodule-2.3.so*  libqtcmodule.so@      qt.pyc
> eric/                 libqtcmodule.la*      qt.py                 qt.pyo
>
> You can't simply move the qt directory there, as there are many more
> files than what are in my site-packages directory.  Shouldn't 'make
> install' have taken care of this for you?  What happened when you did
> a make install?
>
> --
> Aaron J. Ginn                    Phone: 480-814-4463
> Motorola SemiCustom Solutions    Pager: 877-586-2318
> 1300 N. Alma School Rd.          Fax  : 480-814-4463
> Chandler, AZ 85226 M/D CH260     mailto:aaron.ginn at motorola.com



From fpm at u.washington.edu  Wed Feb 28 17:11:32 2001
From: fpm at u.washington.edu (Frank Miles)
Date: 28 Feb 2001 16:11:32 GMT
Subject: PyQt file locations
References: <3A9AE155.FF8C3A08@ucdavis.edu> <snr90jepdj.fsf@motorola.com> <3A9C6E6B.3B727A37@ucdavis.edu>
Message-ID: <97j7vk$l2c$1@nntp6.u.washington.edu>

In article <3A9C6E6B.3B727A37 at ucdavis.edu>,
Bruce Wolk  <bawolk at ucdavis.edu> wrote:

>I have the same files in my site-packages directory, except for qt.pyc.
>
>make-install gave me some errors:

[snip]

>Traceback (most recent call last):
>  File "<string>", line 1, in ?
>  File "/usr/local/lib/python2.0/site-packages/qt.py", line 43, in ?
>    import libqtc
>ImportError: libsip.so.4: cannot open shared object file: No such file or
>directory
>make[3]: *** [install-data-hook] Error 1
>make[3]: Leaving directory `/home/bruce/down/PyQt-2.3/qt'
>make[2]: *** [install-data-am] Error 2
>make[2]: Leaving directory `/home/bruce/down/PyQt-2.3/qt'
>make[1]: *** [install-am] Error 2
>make[1]: Leaving directory `/home/bruce/down/PyQt-2.3/qt'
>make: *** [install-recursive] Error 1
>________________________
>
>When I try to run the examples, I get the same ImportError.
>
>Any ideas?

I can only say that I am getting a nearly identical set of error messages.
I'm running Debian 'potato', with Python 1.5.2.  Sip seemed to install ok.
Anyone have a clue?

	-frank
-- 


From fimafeng at yahoo.com  Tue Feb 27 00:26:54 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Mon, 26 Feb 2001 23:26:54 GMT
Subject: To the point of defeat.....IDLE
Message-ID: <3A9AE62B.5070402@yahoo.com>

I've tried everyhting I know to try and get IDLE to run on Red Hat Linux 
7 with BeOpen Python 2.0

If anyone has done this can they please tell me what they had to do in 
order to get the setup to work???

I've even went as far as installing the Red Hat 7.1 beta and it's still 
using python 1.5.2

I need 2.0 support with IDLE

Please help... I'll pay whatever karma it takes



From stevencooper at isomedia.com  Tue Feb 27 01:40:21 2001
From: stevencooper at isomedia.com (Steve Cooper)
Date: Mon, 26 Feb 2001 16:40:21 -0800
Subject: To the point of defeat.....IDLE
References: <3A9AE62B.5070402@yahoo.com>
Message-ID: <3A9AF775.48635594@isomedia.com>

I found I needed to install some Tk stuff to get it to work (on Debian).  I
unfortunately do not remember the exact package.

HTH,
Steve Cooper

Chris Nelson wrote:

> I've tried everyhting I know to try and get IDLE to run on Red Hat Linux
> 7 with BeOpen Python 2.0
>
> If anyone has done this can they please tell me what they had to do in
> order to get the setup to work???
>
> I've even went as far as installing the Red Hat 7.1 beta and it's still
> using python 1.5.2
>
> I need 2.0 support with IDLE
>
> Please help... I'll pay whatever karma it takes
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From bryan at eevolved.com  Tue Feb 27 03:35:13 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Tue, 27 Feb 2001 02:35:13 GMT
Subject: To the point of defeat.....IDLE
References: <3A9AE62B.5070402@yahoo.com>
Message-ID: <BnEm6.306407$f36.11191270@news20.bellglobal.com>

Chris Nelson wrote:

> I've tried everyhting I know to try and get IDLE to run on Red Hat Linux
> 7 with BeOpen Python 2.0
> 
> If anyone has done this can they please tell me what they had to do in
> order to get the setup to work???
> 
> I've even went as far as installing the Red Hat 7.1 beta and it's still
> using python 1.5.2
> 
> I need 2.0 support with IDLE
> 
> Please help... I'll pay whatever karma it takes
> 

Don't despair Chris, I have been able to run IDLE and 2.0 with no problems 
on Mandrake 7.2. Perhaps the community would be able to help you more if 
you described your exact problem in greater detail.

I am running 2.0 from rpm and IDLE 0.7.1 from a tarball accessible here:

http://sourceforge.net/project/idlefork/

I'd be glad to help, I just don't know what your problem is!

-- 
<=====================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
Website:    http://www.eevolved.com
Public key: http://eevolved.com/bcm.pk
<=====================================>

"They express a preference for 'natural' methods of population limitation, 
and a natural method is exactly what they are going to get. It is called 
starvation." -- Richard Dawkins



From fimafeng at yahoo.com  Tue Feb 27 03:56:47 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Tue, 27 Feb 2001 02:56:47 GMT
Subject: To the point of defeat.....IDLE
References: <3A9AE62B.5070402@yahoo.com> <BnEm6.306407$f36.11191270@news20.bellglobal.com>
Message-ID: <3A9B175D.8080605@yahoo.com>

> 
> Don't despair Chris, I have been able to run IDLE and 2.0 with no problems 
> 
> on Mandrake 7.2. Perhaps the community would be able to help you more if 
> 
> you described your exact problem in greater detail.
> 
> 
> 
> I am running 2.0 from rpm and IDLE 0.7.1 from a tarball accessible here:
> 
> 
> 
> http://sourceforge.net/project/idlefork/
> 
> 
> 
> I'd be glad to help, I just don't know what your problem is!

Thank you for responding. Let me tell you where I am in this process.
I'll start with what my Red Hat 7 System will typically have on it

tcl 8.3-1
tk  8.3-1
python 1.5.2
tkinter for 1.5.2

There is an optional rpm package called "Python tools" which if 
installed give you IDLE.(assoicated with the 1.5.2 python)

First off I can't use the Python2.0 BeOpen RPM at python.org ...When I 
attempt to. The systemsill complain about library dependencies for tcl 
and tk 8.0.

If I download the source for BeOpen 2.0 and do a  "./configure" "make" 
and "make install" the BeOpen software installs.

However...
If I try to bring up IDLE...it is associated with the 1.5.2 interpreter.

What do I need to do so IDLE will wun with 2.0 instead?

Thanks,
Chris Nelson



From bsass at freenet.edmonton.ab.ca  Tue Feb 27 13:02:05 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 05:02:05 -0700 (MST)
Subject: To the point of defeat.....IDLE
In-Reply-To: <3A9B175D.8080605@yahoo.com>
Message-ID: <Pine.LNX.4.33.0102270454590.28590-100000@bms>

> First off I can't use the Python2.0 BeOpen RPM at python.org ...When I
> attempt to. The systemsill complain about library dependencies for tcl
> and tk 8.0.
>
> If I download the source for BeOpen 2.0 and do a  "./configure" "make"
> and "make install" the BeOpen software installs.
>
> However...
> If I try to bring up IDLE...it is associated with the 1.5.2 interpreter.
>
> What do I need to do so IDLE will wun with 2.0 instead?

Have a look at the .py executables that came with your py-2.0, I think
you will find that they use #!...python - which results in the
python-1.5.2 interpreter starting up.  Not what you want!

You need to track down all the executable scripts and make sure they
refer to python2.0.   The fixps.py script in the Tools dir of py-2.0
will automate the task - but look at it first and edit it to reflect
your setup (namely, the location of python2.0).


- Bruce




From news at pastelhorn.com  Tue Feb 27 00:48:46 2001
From: news at pastelhorn.com (Willem Broekema)
Date: Tue, 27 Feb 2001 00:48:46 +0100
Subject: ndiff feature request
Message-ID: <3A9AEB5E.6030502@pastelhorn.com>

The new 'ndiff.py' module (./tools/scripts/ndiff.py) contains among 
others this function "fcompare(f1name, f2name)" that opens two files 
given their names, readlines() them, and creates a 'diff' of the lists:

def fcompare(f1name, f2name):
     f1 = fopen(f1name)
     f2 = fopen(f2name)
     if not f1 or not f2:
         return 0

     a = f1.readlines(); f1.close()
     b = f2.readlines(); f2.close()

     cruncher = SequenceMatcher(IS_LINE_JUNK, a, b)
     for tag, alo, ahi, blo, bhi in cruncher.get_opcodes():
        ...

How about separating these two tasks, by creating a function, say, 
"lcompare(a,b)" that takes two lists as argument, and does the things 
from "cruncher = ..." onwards.  Then, fcompare could call that function 
after the files are opened and checked.

This would make other ways of diff-ing possible without having to write 
to a temporary file first. Things like diffing 
urllib.urlopen("http://someurl").readlines() with a cached list object, 
for example.

Or is there already an easy way to do this?


- Willem



From ACIUFFO1 at irf.com  Tue Feb 27 01:28:36 2001
From: ACIUFFO1 at irf.com (Andrea Ciuffoli)
Date: Mon, 26 Feb 2001 16:28:36 -0800
Subject: Ftplib
Message-ID: <sa9a8448.024@irf.com>

Hello,
I am using the "Ftplib" library to connect my PC to an FTP server, list all the files of a specific directory and transfer only the ones created in the last hour.
My problem is that the command

ftp.retrlines('LIST',listing.append)

appends in listing some information about the file but I noticed that if the file is created in 2001 doesn't store the year among the information while if it is created in 2000 or before doesn't stores the year but not the time.
e.g.

-rwxrwxrwx  1 noone    nogroup     40889 Dec 26 1999 1.txt
-rwxrwxrwx  1 noone    nogroup     40889 Dec 26 2000 2.txt
-rwxrwxrwx  1 noone    nogroup     40889 Jan  5 2000 3.txt
-rwxrwxrwx  1 noone    no group     81630 Feb  5 12:12 4.tat
-rwxrwxrwx  1 noone    no group     81630 Feb  5 12:12 5.tat
-rwxrwxrwx  1 noone    no group     19456 Feb  8 15:27 6.tat

The last 3 files were created in 2001.
I was wondering is it is possible to have either Time and Year.
Are there some example on how the "Ftplib" commands work?

Thanks in advance and my best regards,
Andrea.




From db3l at fitlinxx.com  Tue Feb 27 02:48:54 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 26 Feb 2001 20:48:54 -0500
Subject: Ftplib
References: <mailman.983233808.14489.python-list@python.org>
Message-ID: <upug4lwu1.fsf@ctwd0143.fitlinxx.com>

"Andrea Ciuffoli" <ACIUFFO1 at irf.com> writes:

> Hello,
> I am using the "Ftplib" library to connect my PC to an FTP server, list =
> all the files of a specific directory and transfer only the ones created =
> in the last hour.
> My problem is that the command
> 
> ftp.retrlines('LIST',listing.append)
> 
> appends in listing some information about the file but I noticed that if =
> the file is created in 2001 doesn't store the year among the information =
> while if it is created in 2000 or before doesn't stores the year but not =
> the time.

That's really up to the FTP server that you connect to, but is fairly
typical of Unix output format, so I'm guessing you're connecting to
some variant of Unix server.  If the file is within the past year, it
shows full date/time - if more than a year old, it shows date and
year.  In other words, this is the same sort of listing you'd get if
you were locally on that server and asked for a directory from the
shell.

The LIST command in FTP brings back an ASCII (or EBCDIC), human
readable directory listing, but its format is not standardized, so
servers are free to pretty much do whatever they want.  It's intended
to be easy for humans to view but not required to be easy for
computers to parse.

There's probably not much you can do about this; although some servers
may process Unix-like options (which may modify the output format) as
part of the arguments to the LIST command, it's pot-luck and not
something to depend on.

Parsing the format isn't too bad though since the presence of a colon
can tell you which format each line is in.  Of course if you
absolutely need the timestamp for a file more than a year old, you may
have a problem :-) But since you mention pulling those from the last
hour, this shouldn't be much of an issue.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From qrczak at knm.org.pl  Tue Feb 27 23:19:30 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 27 Feb 2001 22:19:30 GMT
Subject: Ftplib
References: <mailman.983233808.14489.python-list@python.org> <upug4lwu1.fsf@ctwd0143.fitlinxx.com>
Message-ID: <slrn99o9vi.r7q.qrczak@qrnik.zagroda>

26 Feb 2001 20:48:54 -0500, David Bolen <db3l at fitlinxx.com> pisze:

> If the file is within the past year, it shows full date/time -
> if more than a year old, it shows date and year.

The threshold is 6 months.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From sholden at holdenweb.com  Wed Feb 28 16:17:16 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 28 Feb 2001 10:17:16 -0500
Subject: Ftplib
References: <mailman.983233808.14489.python-list@python.org>
Message-ID: <xH8n6.2280$jv4.208457@e420r-atl2.usenetserver.com>

"Andrea Ciuffoli" <ACIUFFO1 at irf.com> wrote in message
news:mailman.983233808.14489.python-list at python.org...
Hello,
I am using the "Ftplib" library to connect my PC to an FTP server, list all
the files of a specific directory and transfer only the ones created in the
last hour.
My problem is that the command

ftp.retrlines('LIST',listing.append)

appends in listing some information about the file but I noticed that if the
file is created in 2001 doesn't store the year among the information while
if it is created in 2000 or before doesn't stores the year but not the time.
e.g.

-rwxrwxrwx  1 noone    nogroup     40889 Dec 26 1999 1.txt
-rwxrwxrwx  1 noone    nogroup     40889 Dec 26 2000 2.txt
-rwxrwxrwx  1 noone    nogroup     40889 Jan  5 2000 3.txt
-rwxrwxrwx  1 noone    no group     81630 Feb  5 12:12 4.tat
-rwxrwxrwx  1 noone    no group     81630 Feb  5 12:12 5.tat
-rwxrwxrwx  1 noone    no group     19456 Feb  8 15:27 6.tat

The last 3 files were created in 2001.
I was wondering is it is possible to have either Time and Year.
Are there some example on how the "Ftplib" commands work?

Thanks in advance and my best regards,
Andrea.


Take a look at the ftpmirror.py script in the Tools distribution directory.
It handles this problem quite well, and may even do everything you want to
do.

regards
 Steve






From tim.one at home.com  Tue Feb 27 02:18:37 2001
From: tim.one at home.com (Tim Peters)
Date: Mon, 26 Feb 2001 20:18:37 -0500
Subject: PEP 236:  Back to the __future__
Message-ID: <LNBBLJKPBEHFEDALKOLCKECOJCAA.tim.one@home.com>

The text of this PEP can also be found online, at:

    http://python.sourceforge.net/peps/pep-0236.html


PEP: 236
Title: Back to the __future__
Version: $Revision: 1.2 $
Author: Tim Peters <tim at digicool.com>
Python-Version: 2.1
Status: Active
Type: Standards Track
Created: 26-Feb-2001
Post-History: 26-Feb-2001


Motivation

    From time to time, Python makes an incompatible change to the
    advertised semantics of core language constructs, or changes their
    accidental (implementation-dependent) behavior in some way.  While this
    is never done capriciously, and is always done with the aim of
    improving the language over the long term, over the short term it's
    contentious and disrupting.

    The "Guidelines for Language Evolution" PEP [1] suggests ways to ease
    the pain, and this PEP introduces some machinery in support of that.

    The "Statically Nested Scopes" PEP [2] is the first application, and
    will be used as an example here.


Intent

    [Note:  This is policy, and so should eventually move into PEP 5[1]]

    When an incompatible change to core language syntax or semantics is
    being made:

    1. The release C that introduces the change does not change the
       syntax or semantics by default.

    2. A future release R is identified in which the new syntax or semantics
       will be enforced.

    3. The mechanisms described in the "Warning Framework" PEP [3] are used
       to generate warnings, whenever possible, about constructs or
       operations whose meaning may[4] change in release R.

    4. The new future_statement (see below) can be explicitly included in a
       module M to request that the code in module M use the new syntax or
       semantics in the current release C.

    So old code continues to work by default, for at least one release,
    although it may start to generate new warning messages.  Migration to
    the new syntax or semantics can proceed during that time, using the
    future_statement to make modules containing it act as if the new syntax
    or semantics were already being enforced.

    Note that there is no need to involve the future_statement machinery
    in new features unless they can break existing code; fully backward-
    compatible additions can-- and should --be introduced without a
    corresponding future_statement.


Syntax

    A future_statement is simply a from/import statement using the reserved
    module name __future__:

        future_statement: "from" "__future__" "import" feature ["as" name]
                          ("," feature ["as" name])*

        feature: identifier
        name: identifier

    In addition, all future_statments must appear near the top of the
    module.  The only lines that can appear before a future_statement are:

    + The module docstring (if any).
    + Comments.
    + Blank lines.
    + Other future_statements.

    Example:
        """This is a module docstring."""

        # This is a comment, preceded by a blank line and followed by
        # a future_statement.
        from __future__ import nested_scopes

        from math import sin
        from __future__ import alabaster_weenoblobs  # compile-time error!
        # That was an error because preceded by a non-future_statement.


Semantics

    A future_statement is recognized and treated specially at compile time:
    changes to the semantics of core constructs are often implemented by
    generating different code.  It may even be the case that a new feature
    introduces new incompatible syntax (such as a new reserved word), in
    which case the compiler may need to parse the module differently.  Such
    decisions cannot be pushed off until runtime.

    For any given release, the compiler knows which feature names have been
    defined, and raises a compile-time error if a future_statement contains
    a feature not known to it[5].

    The direct runtime semantics are the same as for any import statement:
    there is a standard module __future__.py, described later, and it will
    be imported in the usual way at the time the future_statement is
    executed.

    The *interesting* runtime semantics depend on the specific feature(s)
    "imported" by the future_statement(s) appearing in the module.

    Note that there is nothing special about the statement:

        import __future__ [as name]

    That is not a future_statement; it's an ordinary import statement, with
    no special semantics or syntax restrictions.


Example

    Consider this code, in file scope.py:

        x = 42
        def f():
            x = 666
            def g():
                print "x is", x
            g()
        f()

    Under 2.0, it prints:

        x is 42

    Nested scopes[2] are being introduced in 2.1.  But under 2.1, it still
    prints

        x is 42

    and also generates a warning.

    In 2.2, and also in 2.1 *if* "from __future__ import nested_scopes" is
    included at the top of scope.py, it prints

        x is 666


Standard Module __future__.py

    Lib/__future__.py is a real module, and serves three purposes:

    1. To avoid confusing existing tools that analyze import statements and
       expect to find the modules they're importing.

    2. To ensure that future_statements run under releases prior to 2.1
       at least yield runtime exceptions (the import of __future__ will
       fail, because there was no module of that name prior to 2.1).

    3. To document when incompatible changes were introduced, and when they
       will be-- or were --made mandatory.  This is a form of executable
       documentation, and can be inspected programatically via importing
       __future__ and examining its contents.

    Each statment in __future__.py is of the form:

        FeatureName = ReleaseInfo

    ReleaseInfo is a pair of the form:

         (OptionalRelease, MandatoryRelease)

    where, normally, OptionalRelease <  MandatoryRelease, and both are
    5-tuples of the same form as sys.version_info:

    (PY_MAJOR_VERSION, # the 2 in 2.1.0a3; an int
     PY_MINOR_VERSION, # the 1; an int
     PY_MICRO_VERSION, # the 0; an int
     PY_RELEASE_LEVEL, # "alpha", "beta", "candidate" or "final"; string
     PY_RELEASE_SERIAL # the 3; an int
    )

    OptionalRelease records the first release in which

        from __future__ import FeatureName

    was accepted.

    In the case of MandatoryReleases that have not yet occurred,
    MandatoryRelease predicts the release in which the feature will become
    part of the language.

    Else MandatoryRelease records when the feature became part of the
    language; in releases at or after that, modules no longer need

        from __future__ import FeatureName

    to use the feature in question, but may continue to use such imports.

    MandatoryRelease may also be None, meaning that a planned feature got
    dropped.

    No line will ever be deleted from __future__.py.

    Example line:

        nested_scopes = (2, 1, 0, "beta", 1), (2, 2, 0, "final", 0)

    This means that

        from __future__ import nested_scopes

    will work in all releases at or after 2.1b1, and that nested_scopes are
    intended to be enforced starting in release 2.2.


Unresolved Problems:  Runtime Compilation

    Several Python features can compile code during a module's runtime:

    1. The exec statement.
    2. The execfile() function.
    3. The compile() function.
    4. The eval() function.
    5. The input() function.

    Since a module M containing a future_statement naming feature F
    explicitly requests that the current release act like a future release
    with respect to F, any code compiled dynamically from text passed to
    one of these from within M should probably also use the new syntax or
    semantics associated with F.

    This isn't always desired, though.  For example, doctest.testmod(M)
    compiles examples taken from strings in M, and those examples should
    use M's choices, not necessarily doctest module's choices.

    It's unclear what to do about this.  The initial release (2.1b1) is
    likely to ignore these issues, saying that each dynamic compilation
    starts over from scratch (i.e., as if no future_statements had been
    specified).

    In any case, a future_statement appearing "near the top" (see Syntax
    above) of text compiled dynamically by an exec, execfile() or compile()
    applies to the code block generated, but has no further effect on the
    module that executes such an exec, execfile() or compile().  This
    can't be used to affect eval() or input(), however, because they only
    allow expression input, and a future_statement is not an expression.


Unresolved Problems:  Interactive Shells

    An interactive shell can be seen as an extreme case of runtime
    compilation (see above):  in effect, each statement typed at an
    interactive shell prompt runs a new instance of exec, compile() or
    execfile().  The initial release (2.1b1) is likely to be such that
    future_statements typed at an interactive shell have no effect beyond
    their runtime meaning as ordinary import statements.

    It would make more sense if a future_statement typed at an interactive
    shell applied to the rest of the shell session's life, as if the
    future_statement had appeared at the top of a module.  Again, it's
    unclear what to do about this.


Questions and Answers

    Q:  What about a "from __past__" version, to get back *old* behavior?

    A:  Outside the scope of this PEP.  Seems unlikely to the author,
        though.  Write a PEP if you want to pursue it.

    Q:  What about incompatibilites due to changes in the Python virtual
        machine?

    A:  Outside the scope of this PEP, although PEP 5[1] suggests a grace
        period there too, and the future_statement may also have a role to
        play there.

    Q:  What about incompatibilites due to changes in Python's C API?

    A:  Outside the scope of this PEP.

    Q:  I want to wrap future_statements in try/except blocks, so I can
        use different code depending on which version of Python I'm running.
        Why can't I?

    A:  Sorry!  try/except is a runtime feature; future_statements are
        primarily compile-time gimmicks, and your try/except happens long
        after the compiler is done.  That is, by the time you do
        try/except, the semantics in effect for the module are already a
        done deal.  Since the try/except wouldn't accomplish what it
        *looks* like it should accomplish, it's simply not allowed.  We
        also want to keep these special statements very easy to find and to
        recognize.

        Note that you *can* import __future__ directly, and use the
        information in it, along with sys.version_info, to figure out where
        the release you're running under stands in relation to a given
        feature's status.

     Q: Going back to the nested_scopes example, what if release 2.2 comes
        along and I still haven't changed my code?  How can I keep the 2.1
        behavior then?

     A: By continuing to use 2.1, and not moving to 2.2 until you do change
        your code.  The purpose of future_statement is to make life easier
        for people who keep keep current with the latest release in a timely
        fashion.  We don't hate you if you don't, but your problems are
        much harder to solve, and somebody with those problems will need to
        write a PEP addressing them.  future_statement is aimed at a
        different audience.


Copyright

    This document has been placed in the public domain.


References and Footnotes

    [1] http://python.sourceforge.net/peps/pep-0005.html

    [2] http://python.sourceforge.net/peps/pep-0227.html

    [3] http://python.sourceforge.net/peps/pep-0230.html

    [4] Note that this is "may" and not "will":  better safe than sorry.  Of
        course spurious warnings won't be generated when avoidable with
        reasonable cost.

    [5] This ensures that a future_statement run under a release prior to
        the first one in which a given feature is known (but >= 2.1) will
        raise a compile-time error rather than silently do a wrong thing.
        If transported to a release prior to 2.1, a runtime error will be
        raised because of the failure to import __future__ (no such module
        existed in the standard distribution before the 2.1 release, and
        the double underscores make it a reserved name).


Local Variables:
mode: indented-text
indent-tabs-mode: nil
End:




From wunschname at kapet.de  Tue Feb 27 11:25:05 2001
From: wunschname at kapet.de (Karsten Petersen)
Date: Tue, 27 Feb 2001 11:25:05 +0100
Subject: PEP 236:  Back to the __future__
References: <mailman.983236748.23203.python-list@python.org>
Message-ID: <1avf79.sr1.ln@dollerup.csn.tu-chemnitz.de>

Hello,
some comments from the lurkers-corner:

I am not too lucky with all those improvements/changes inside the
language.  In fact I am rather scared.

Let me explain why:
Since 1997 I am using Python for a lot of my projects, the reasons were
ease of use, the clearness of the language and its availability and
compatibility across platform borders.

And well, I still use Python 1.5.2 and am not seeing a reason to
upgrade.

If I upgrade to 2.0, I will have to recompile all C modules, check
everything if it still runs and so on.  But what about 2.1 or 2.2?
Will I have to recompile and/or recheck everything again?
For me it seems so, and I am not willing to do that.

I want to _use_ Python, its implementation details are something I am
not really interested in.


It is possible to compile C sources that were written in the 80s.

It is possible to run (more or less simple) Perl programs which are
several years old.

Will I be able in 2005 to run the code I have written today?


If there would be one big change I could cope with it.  But with smaller
changes and trouble every couple of months I am rather unsatisfied.

Greets, Karsten
-- 
10./11.3.:  3. Chemnitzer Linux-Tag:  http://www.tu-chemnitz.de/linux/tag/
,-,  Student of Computer Science at Chemnitz University of Technology  ,-,
| |    EMail:  Karsten at kapet.de          WWW:  http://www.kapet.de/    | |
'-'  Home: kapet at dollerup.csn   V72 / 230    Phone: +49-177-82 35 136  '-'


From andreas at digicool.com  Tue Feb 27 13:52:10 2001
From: andreas at digicool.com (Andreas Jung)
Date: Tue, 27 Feb 2001 07:52:10 -0500
Subject: PEP 236:  Back to the __future__
References: <mailman.983236748.23203.python-list@python.org> <1avf79.sr1.ln@dollerup.csn.tu-chemnitz.de>
Message-ID: <012c01c0a0bf$f25c3220$d0063bd0@SUXLAP>

----- Original Message -----
From: "Karsten Petersen" <wunschname at kapet.de>



> Hello,
> some comments from the lurkers-corner:
>
> I am not too lucky with all those improvements/changes inside the
> language.  In fact I am rather scared.
>
> Let me explain why:
> Since 1997 I am using Python for a lot of my projects, the reasons were
> ease of use, the clearness of the language and its availability and
> compatibility across platform borders.
>
> And well, I still use Python 1.5.2 and am not seeing a reason to
> upgrade.
>
> If I upgrade to 2.0, I will have to recompile all C modules, check
> everything if it still runs and so on.  But what about 2.1 or 2.2?
> Will I have to recompile and/or recheck everything again?
> For me it seems so, and I am not willing to do that.
>
> I want to _use_ Python, its implementation details are something I am
> not really interested in.
>
>
> It is possible to compile C sources that were written in the 80s.

But you can also find C sources that compile only on the system they
were written for.
>
> It is possible to run (more or less simple) Perl programs which are
> several years old.

see above...  I have also seen Perl hacks and so-called programs that were
running under 4.X but no longer under 5.X.

> Will I be able in 2005 to run the code I have written today?
>

I am using Python since 1992 and the canges between the version usually did
not
require any changes (except when we are talking about extension modules in
C/C++).
In general there should be no need to change native Python code.

Although some extension in 2.1 and 2.2 might break existing code. When you
are working
in a production environment with Python you must take care of version
changes (never change a running system).

Andreas






From tim.one at home.com  Tue Feb 27 18:13:15 2001
From: tim.one at home.com (Tim Peters)
Date: Tue, 27 Feb 2001 12:13:15 -0500
Subject: PEP 236:  Back to the __future__
In-Reply-To: <1avf79.sr1.ln@dollerup.csn.tu-chemnitz.de>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEFMJCAA.tim.one@home.com>

[Karsten Petersen]
> ...
> I am not too lucky with all those improvements/changes inside the
> language.  In fact I am rather scared.
>
> Let me explain why:
> Since 1997 I am using Python for a lot of my projects, the reasons were
> ease of use, the clearness of the language and its availability and
> compatibility across platform borders.

I've been using it since 1991, and have found the burden of incompatible
Python changes to be trivial over that time.  They'll be even easier to deal
with using the __future__ mechanism.

> And well, I still use Python 1.5.2 and am not seeing a reason to
> upgrade.

That's OK.  The PEP doesn't *require* you to upgrade <wink>.

> If I upgrade to 2.0, I will have to recompile all C modules, check
> everything if it still runs and so on.  But what about 2.1 or 2.2?
> Will I have to recompile and/or recheck everything again?

Depends on the platform (e.g., on Windows the name of the release is in
effect buried in .pyds that link to it), and on whether changes are also made
to the Python virtual machine and/or C API.

> For me it seems so, and I am not willing to do that.
>
> I want to _use_ Python, its implementation details are something I am
> not really interested in.
>
> It is possible to compile C sources that were written in the 80s.
>
> It is possible to run (more or less simple) Perl programs which are
> several years old.
>
> Will I be able in 2005 to run the code I have written today?

I've got Python that hasn't changed since 1991; I expect it will work fine in
2011 too.  Don't overblow this.

> If there would be one big change I could cope with it.  But with smaller
> changes and trouble every couple of months I am rather unsatisfied.

No offense, but I don't buy this.  If you're still sticking to 1.5.2 by
choice, you're simply change-aversive.  Nothing wrong with that!  But it
makes the notion that some other form of change *would* leave you happy is
hard to buy.

the-pep-is-as-it-says-for-people-who-do-stay-current-ly y'rs  - tim




From db3l at fitlinxx.com  Wed Feb 28 02:13:05 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 20:13:05 -0500
Subject: PEP 236:  Back to the __future__
References: <mailman.983236748.23203.python-list@python.org> <1avf79.sr1.ln@dollerup.csn.tu-chemnitz.de>
Message-ID: <uwvabmwym.fsf@ctwd0143.fitlinxx.com>

wunschname at kapet.de (Karsten Petersen) writes:

> If I upgrade to 2.0, I will have to recompile all C modules, check
> everything if it still runs and so on.  (...)

I believe that so far recompilation and/or relinking has been about
it.  Existing C extension code should have run into very little (if
any) incompatibilities at the source API level.

> It is possible to compile C sources that were written in the 80s.

Right - "compile" (and or recheck).  That's the nature of using C, and
also impacts C extensions for Python, but that's one of the tradeoffs
(recompilation) of using C extensions rather than Python itself.

> It is possible to run (more or less simple) Perl programs which are
> several years old.

And you could likely same the same thing about Python scripts (note,
Python, not C extensions) as well.

Don't get me wrong - in a production environment and with a production
language I think that transitions must absolutely be clean and
backwards compatibility extremely important.  But you also can't avoid
all change entirely, and I think that recent Python events have been
very good at maintaining existing behavior.

I am however, more willing to accept some additional burden on C level
extensions over the course of time, simply because by their nature
they are lower level and subject to many of the C-based constraints
(binary compatibility, Windows DLL loading, etc...) that the pure
Python source is not.

On my own machine, I've kept up with Python 2.0 and now 2.1, but in my
production environment and deployed base it's still Python 1.5.2.
Some of that delay is ensuring all the extensions I like are
available, and some of it is just the hurdle of the deployment.  But I
don't really feel that Python has moved too quickly yet, and I'm
comfortable that my Python scripts themselves (from testing them on my
machine with 2.0 and 2.1) will survive the transition when it does
come with hardly a burp.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From hinsen at cnrs-orleans.fr  Tue Feb 27 12:11:22 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 27 Feb 2001 12:11:22 +0100
Subject: PEP 236:  Back to the __future__
References: <mailman.983236748.23203.python-list@python.org>
Message-ID: <m3elwkl6sl.fsf@chinon.cnrs-orleans.fr>

"Tim Peters" <tim.one at home.com> writes:

>     From time to time, Python makes an incompatible change to the
>     advertised semantics of core language constructs, or changes their
>     accidental (implementation-dependent) behavior in some way.  While this

I am not at all happy with this approach. Incompatible changes should
not happen "from time to time", but as rarely as possible and then all
at the same time. I much preferred the "Py3K" project which was to
include all the improvements that had to be held back due to
compatibility concern.

Like many Python users, I have a huge base of installed code, lots of
it written by myself, but the majority written by others. Everything
works fine together, and I am happy and productive.

Now comes Python 2.1 with nested scopes. I don't expect much of my own
code to break, but I can't be sure, I certainly didn't write it with a
possible change of scoping rules in mind. So perhaps I'll have to
modify it a bit, that shouldn't be much work.

But then there's all the code from others that might also break, and
that I don't want to fix at all. And since some of my code will depend
on code from others, I won't fix my own code either and stick to
Python 1.5 instead. Which in turn forces others who depend on my code
to do the same. Sure, sooner or later everybody will want to use the
new features of Python 2.1. But by then Python 2.1 will have been
replaced by 2.2, with its own set of incompatible changes.

If, on the other hand, incompatible changes occur only every five
years or so, then people will be inclined to install two Python
distributions in parallel and update their code over a period of a few
months.

Moral: incompatible changes should happen on a time scale compatible
with uncoordinated collective code development, which is years, not
months.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From jeremy at alum.mit.edu  Tue Feb 27 16:17:43 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Tue, 27 Feb 2001 10:17:43 -0500 (EST)
Subject: PEP 236:  Back to the __future__
In-Reply-To: <m3elwkl6sl.fsf@chinon.cnrs-orleans.fr>
References: <mailman.983236748.23203.python-list@python.org>
	<m3elwkl6sl.fsf@chinon.cnrs-orleans.fr>
Message-ID: <15003.50455.960475.77114@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "KH" == Konrad Hinsen <hinsen at cnrs-orleans.fr> writes:

  KH> Moral: incompatible changes should happen on a time scale
  KH> compatible with uncoordinated collective code development, which
  KH> is years, not months.

The goal of future_statements is to lengthen the time scale for
incompatible changes while allowing some users to access the
functionality now.  With the addition of from __future__ import
nested_scopes, there should be no incompatible changes in 2.1.

At the same time, I expect very little code to require changes to be
compatible with nested scopes.  My own survey of three large projects
found four problems in 200,000 lines of code.  I have yet to hear
statistics for any other project.

Jeremy



From tim.one at home.com  Tue Feb 27 19:14:30 2001
From: tim.one at home.com (Tim Peters)
Date: Tue, 27 Feb 2001 13:14:30 -0500
Subject: PEP 236:  Back to the __future__
In-Reply-To: <m3elwkl6sl.fsf@chinon.cnrs-orleans.fr>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEFOJCAA.tim.one@home.com>

[posted & mailed]

[PEP 236]
>     From time to time, Python makes an incompatible change to the
>     advertised semantics of core language constructs, or changes their
>     accidental (implementation-dependent) behavior in some way.

[Konrad Hinsen]
> I am not at all happy with this approach. Incompatible changes should
> not happen "from time to time", but as rarely as possible and then all
> at the same time.

The PEP doesn't address either the frequency or nature of incompatible
changes, it's defining a mechanism to help people deal with incompatible
changes *when* they occur.

> I much preferred the "Py3K" project which was to include all the
> improvements that had to be held back due to compatibility concern.

Incompatible changes have been made with virtually every Python release since
before 1.0.  The changes are *usually* so minor and affect so little code
that few people notice them in practice (for example, how many people were
burned by introducing "lambda" as a new keyword?  IIRC, none -- but that was
an incompatible change!).  Still, Python has more users all the time, and a
mechanism to ease transitioning to even minor changes has value in proportion
to the megalines of code out there.

As to P3K, there's no funding for a rewrite from scratch.  So, barring a
miracle, Python evolves incrementally, and it wouldn't remain Python unless
it does so in ways that keep the language clean.

> Like many Python users, I have a huge base of installed code, lots of
> it written by myself, but the majority written by others. Everything
> works fine together, and I am happy and productive.

And you've managed to do that despite a decade of incompatible language
changes already.  Why *assume* the future is going to be so much worse than
the past?  Perhaps because it hasn't happened yet <wink>?

> Now comes Python 2.1 with nested scopes. I don't expect much of my own
> code to break, but I can't be sure,

It won't break in 2.1, but you'll get warnings about fishy cases in 2.1, and
the __future__ stmt will let you know for sure and change such code at your
convenience.

> I certainly didn't write it with a possible change of scoping rules in
> mind.

Note that most of the few cases of nested-scope breakage *known* so far
actually had nothing to do with nested scopes at heart:  they had to do with
code violating the Lang Ref's prohibition against using "import *" at other
than module scope.  Such code was always undefined.  Jeremy is bending
backwards 167 degrees to make that code still work in the same way it worked
by accident (and in violation of the language spec) whenever possible.

> So perhaps I'll have to modify it a bit, that shouldn't be much work.

The purpose of PEP 236 is to make that easier for you than it would have been
otherwise.

> But then there's all the code from others that might also break, and
> that I don't want to fix at all. And since some of my code will depend
> on code from others, I won't fix my own code either and stick to
> Python 1.5 instead. Which in turn forces others who depend on my code
> to do the same. Sure, sooner or later everybody will want to use the
> new features of Python 2.1. But by then Python 2.1 will have been
> replaced by 2.2, with its own set of incompatible changes.

This would carry more weight if it described a new phenomenon -- but you got
to 1.5.2 via a long series of incompatible-in-small-ways releases before
1.5.2.  It's just business as usual.  Perhaps you jumped on the train at
1.5.2?  Many did, and they've enjoyed the illusion of 100% stability simply
because no update (not even sorely needed bugfix patches) to 1.5.2 was
released for about 18 months.  But 1.5.2 was an historical anomaly, not the
norm.

Still, if you're happy with 1.5.2, stay with it!  You won't be forced to
upgrade.  Perhaps if enough people get burned by "Guido intends to break two
lines of code pointed out in a warning message!" <0.9 wink>, some entity will
arise to keep 1.5.2 alive forever.

> If, on the other hand, incompatible changes occur only every five
> years or so, then people will be inclined to install two Python
> distributions in parallel and update their code over a period of a few
> months.

PEP 236 aims to let them update their code, incrementally (one module at a
time), over a period of about half a year, and without needing to install two
Python distributions in parallel.  We're trying to make staying current
easier.

> Moral: incompatible changes should happen on a time scale compatible
> with uncoordinated collective code development, which is years, not
> months.

As the PEP said, it's aimed at people who intend to keep current.  I believe
it will be a real help for them.  As the PEP also says, people who don't want
to keep current have harder problems, which the PEP doesn't even pretend to
address.

BTW, the only way to avoid all incompatible changes is to avoid fixing bugs
too:  over the years, and regardless of language, I've had much more of my
code break due to accidentally relying on buggy or undefined behavior than
due to deliberate changes in correctly implemented defined behavior.

the-only-reliable-stability-is-death-ly y'rs  - tim




From dsh8290 at rit.edu  Wed Feb 28 02:02:50 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 27 Feb 2001 20:02:50 -0500
Subject: PEP 236:  Back to the __future__
In-Reply-To: <m3elwkl6sl.fsf@chinon.cnrs-orleans.fr>; from hinsen@cnrs-orleans.fr on Tue, Feb 27, 2001 at 12:11:22PM +0100
References: <mailman.983236748.23203.python-list@python.org> <m3elwkl6sl.fsf@chinon.cnrs-orleans.fr>
Message-ID: <20010227200249.A21243@harmony.cs.rit.edu>

On Tue, Feb 27, 2001 at 12:11:22PM +0100, Konrad Hinsen wrote:
| 
| But then there's all the code from others that might also break, and
| that I don't want to fix at all. And since some of my code will depend
| on code from others, I won't fix my own code either and stick to
| Python 1.5 instead. Which in turn forces others who depend on my code
| to do the same. Sure, sooner or later everybody will want to use the
| new features of Python 2.1. But by then Python 2.1 will have been
| replaced by 2.2, with its own set of incompatible changes.

This scenario reminds me of Java.  I am working at a place that is
using jdk1.1.8 for a rather new product (started ~2 years ago).  We've
even found a few bugs in that version of javac that won't compile some
of our code.  So why are we using 1.1.8?  We use a 3rd party library
whose vendor only supports 1.1.8, because they intend it to be used in
applets and many browsers don't have a newer vm.  (Using a different
compiler, but outputing backwards compatible bytecodes allows us to
work around the compiler bugs)

(for those who don't know, jdk1.3.0 has been available for quite a
while now)

| 
| If, on the other hand, incompatible changes occur only every five
| years or so, then people will be inclined to install two Python
| distributions in parallel and update their code over a period of a few
| months.
| 
| Moral: incompatible changes should happen on a time scale compatible
| with uncoordinated collective code development, which is years, not
| months.

I guess the key is how incompatible are the changes?  If it is like
Java, where the event model, GUI toolkit, and standard library had
major incompatibilities, it is bad.  If it is like Tim describes
Python's past then it shouldn't be so painful.  (I don't know from
experience since I first learned of Python at version 1.5.2)


However, I would like to see the use of particular future_statements
deprecated or even disallowed after the feature becomes mandatory.
Once the code has been changed to use the new feature, and the
interpreter mandating that feature is commonly in use, the
future_statement simply clutters the code with extra unnecessary
baggage.  Next thing we know, curly braces will be optional <wink>.
As for the future statement being easy to find :
    grep -r "__future__" my_source_tree/*
will do the job ;-).


-D




From tim.one at home.com  Wed Feb 28 02:29:07 2001
From: tim.one at home.com (Tim Peters)
Date: Tue, 27 Feb 2001 20:29:07 -0500
Subject: PEP 236:  Back to the __future__
In-Reply-To: <20010227200249.A21243@harmony.cs.rit.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEIPJCAA.tim.one@home.com>

[D-Man]
> ...
> However, I would like to see the use of particular future_statements
> deprecated or even disallowed after the feature becomes mandatory.

It's deliberately not done that way, so that if someone who moves their code
around a lot happens to move to an earlier release, the (earlier) compiler
will complain about a __future__ request it can't satsify.  This is up to
each programmer to deal with in whichever way best fits their usage.  I
expect many users will ignore future_statements entirely, and simply wait for
the next release before fiddling their code.  That's fine too.

Note that there's more than enough info in __future__.py to allow writing a
tool in Python that can reliably search for, identify, and even delete (if
you like) obsolete future_statements wrt to any fixed Python release.  That
was also deliberate.

> As for the future statement being easy to find :
>     grep -r "__future__" my_source_tree/*
> will do the job ;-).

The syntax restrictions listed in the PEP also ensure that a future_statement
begins with "from" starting in the first column (they CANNOT be nested in if
or try blocks, etc).

everything-about-them-is-simple-except-the-implementation<wink>-ly y'rs
    - tim




From tdickenson at devmail.geminidataloggers.co.uk  Wed Feb 28 10:46:02 2001
From: tdickenson at devmail.geminidataloggers.co.uk (Toby Dickenson)
Date: Wed, 28 Feb 2001 09:46:02 +0000
Subject: PEP 236:  Back to the __future__
References: <mailman.983236748.23203.python-list@python.org>
Message-ID: <gthp9ts4n5ia0d68qkqrglabaspkon672p@4ax.com>

"Tim Peters" <tim.one at home.com> wrote:

>    MandatoryRelease may also be None, meaning that a planned feature got
>    dropped.

Once feature X is dropped, then I assume "from __future__ import X"
would give an ImportError (or something derived from ImportError) ?


Toby Dickenson
tdickenson at geminidataloggers.com


From hinsen at cnrs-orleans.fr  Wed Feb 28 12:39:12 2001
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 28 Feb 2001 12:39:12 +0100
Subject: PEP 236:  Back to the __future__
References: <mailman.983297709.13140.python-list@python.org>
Message-ID: <m3g0gzavfj.fsf@chinon.cnrs-orleans.fr>

"Tim Peters" <tim.one at home.com> writes:

> Incompatible changes have been made with virtually every Python release since
> before 1.0.  The changes are *usually* so minor and affect so little code
> that few people notice them in practice (for example, how many people were

Indeed. I have been using Python since 1.3 and never had to change a
single line due to language changes, and only very few lines due to
library changes (regular expressions).

> And you've managed to do that despite a decade of incompatible language
> changes already.  Why *assume* the future is going to be so much worse than
> the past?  Perhaps because it hasn't happened yet <wink>?

I have the impression that Python development has been speeding up
recently compared to the good old days. I used to install all alpha
and beta releases, and now I didn't even find the time to install 2.0
between 1.5.2 and 2.1. (I admit there might be personal reasons as
well, work is taking a lot of potential Python-playing time away from
me.)

> It won't break in 2.1, but you'll get warnings about fishy cases in 2.1, and
> the __future__ stmt will let you know for sure and change such code at your
> convenience.

Perfectly fine unless 2.2 comes out four weeks after 2.1. ;-)

living-in-ever-more-rapidly-changing-times'ly
    Konrad.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.56.24
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------


From max at alcyone.com  Wed Feb 28 17:44:09 2001
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 28 Feb 2001 08:44:09 -0800
Subject: PEP 236:  Back to the __future__
References: <mailman.983297709.13140.python-list@python.org> <m3g0gzavfj.fsf@chinon.cnrs-orleans.fr>
Message-ID: <3A9D2AD9.78149FCC@alcyone.com>

Konrad Hinsen wrote:

> Indeed. I have been using Python since 1.3 and never had to change a
> single line due to language changes, and only very few lines due to
> library changes (regular expressions).

The only changes I've had to make was what amounts to the bugfix with
the socket address names (e.g., socket.connect((host, port)) instead of
socket.connect(host, port)).  And, of course, taking advantage of 2.0
features such as list comprehensions.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Men and women, women and men.  It will never work.
\__/ Erica Jong
    Official Buh rules / http://www.alcyone.com/max/projects/cards/buh/
 The official rules to the betting card game, Buh.


From tim.one at home.com  Wed Feb 28 18:21:04 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 28 Feb 2001 12:21:04 -0500
Subject: PEP 236:  Back to the __future__
In-Reply-To: <3A9D2AD9.78149FCC@alcyone.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKELIJCAA.tim.one@home.com>

[Konrad Hinsen]
> Indeed. I have been using Python since 1.3 and never had to change a
> single line due to language changes, and only very few lines due to
> library changes (regular expressions).

Then you're not aggressive enough in visiting Python's dark corners <wink>.

[Erik Max Francis]
> The only changes I've had to make was what amounts to the bugfix with
> the socket address names (e.g., socket.connect((host, port)) instead of
> socket.connect(host, port)).

This is one I'm sure we would have used __future__ for, if __future__ had
existed.  Guido made a convincing case (to me, anyway) that it was a bug in
the interpreter that it allowed the latter form, and the docs didn't flatly
contradict him <ahem>.  But the docs weren't clear enough, and some instances
of the broken form were in the standard distribution.  That made it painful
to say "oops!  we're taking it away now":  despite that taking it away was
the right thing to do, we knew for certain it would break existing code
(under CPython; IIRC, JPython wasn't deluded on this point).  In a
__futurized__ world, you would have gotten a release in which both forms
worked, with warning messages about the broken form.

Hmm.  Bad example <wink> -- this one really has nothing to do with __future__
after all.  It has to do with that there *is* a mechanism for producing
warnings now (we couldn't warn before just because there was no machinery
*to* produce warnings!).

> And, of course, taking advantage of 2.0 features such as list
> comprehensions.

__future__ wouldn't get involved in those either, because those new features
were wholly backward compatible.  __future__ is only for the rare changes
that may break some existing code (changes to the scope rules in the presence
of nested functions is the only example we have for 2.1; and that in *some*
of those cases, we're going to start enforcing the Ref Man's long-standing
restrictions on the use of "import *").

not-that-people-shouldn't-panic-anyway-ly y'rs  - tim




From glen.mettler at home.com  Tue Feb 27 03:42:12 2001
From: glen.mettler at home.com (glen mettler)
Date: Tue, 27 Feb 2001 02:42:12 GMT
Subject: list to string
Message-ID: <1103_983241679@cc462845-a>

These commands will turn a string into a list:
myword="dog"
mywordlist=list(myword)
result = ["d","o","g"]

is there a command/function that can take the list ["d","o","g"] and make it a string "dog"?

Glen





From jgraves3 at austin.rr.com  Tue Feb 27 03:50:38 2001
From: jgraves3 at austin.rr.com (jay graves)
Date: Tue, 27 Feb 2001 02:50:38 GMT
Subject: list to string
References: <1103_983241679@cc462845-a>
Message-ID: <3a9b14c9.33237453@news-server.austin.rr.com>

On Tue, 27 Feb 2001 02:42:12 GMT, glen mettler <glen.mettler at home.com>
wrote:

>These commands will turn a string into a list:
>myword="dog"
>mywordlist=list(myword)
>result = ["d","o","g"]
>
>is there a command/function that can take the list ["d","o","g"] and make it a string "dog"?

One way is to join them.

''.join(result)

Some people don't like using the string methods in this way and they
might write:  (at least until the string module goes away.)

import string
string.join(result,'')

Hope this helps.

...
jay


From erno-news at erno.iki.fi  Tue Feb 27 04:25:25 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 27 Feb 2001 05:25:25 +0200
Subject: list to string
References: <1103_983241679@cc462845-a> <3a9b14c9.33237453@news-server.austin.rr.com>
Message-ID: <kud7c4remy.fsf@lasipalatsi.fi>

In article <3a9b14c9.33237453 at news-server.austin.rr.com>,
jgraves3 at austin.rr.com (jay graves) writes:

| Some people don't like using the string methods in this way and they
| might write:  (at least until the string module goes away.)

| import string
| string.join(result,'')

the string module isn't going away, it has lots of stuff that
is not available as methods of strings.

  -- erno


From ben.hutchings at roundpoint.com  Tue Feb 27 23:31:50 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 27 Feb 2001 14:31:50 -0800
Subject: list to string
References: <1103_983241679@cc462845-a> <3a9b14c9.33237453@news-server.austin.rr.com>
Message-ID: <uy9uriwq1.fsf@roundpoint.com>

jgraves3 at austin.rr.com (jay graves) writes:

> On Tue, 27 Feb 2001 02:42:12 GMT, glen mettler <glen.mettler at home.com>
> wrote:
> 
> >These commands will turn a string into a list:
> >myword="dog"
> >mywordlist=list(myword)
> >result = ["d","o","g"]
> >
> >is there a command/function that can take the list ["d","o","g"] and
> >make it a string "dog"?
> 
> One way is to join them.
> 
> ''.join(result)
> 
> Some people don't like using the string methods in this way
<snip>

I'm happy with string methods in general, but I find the semantics of
the above (sending a message to a string to tell it to join a sequence
of strings together with copies of itself) ridiculous.

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From jeremy at alum.mit.edu  Tue Feb 27 03:58:58 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Mon, 26 Feb 2001 21:58:58 -0500 (EST)
Subject: list to string
In-Reply-To: <1103_983241679@cc462845-a>
References: <1103_983241679@cc462845-a>
Message-ID: <15003.6130.947922.804395@w221.z064000254.bwi-md.dsl.cnc.net>

>>>>> "gm" == glen mettler <glen.mettler at home.com> writes:

  gm> These commands will turn a string into a list: 
  gm> myword="dog"
  gm> mywordlist=list(myword)
  gm> result = ["d","o","g"]

  gm> is there a command/function that can take the list ["d","o","g"]
  gm> and make it a string "dog"?

string = "".join(result)

That is, call the join method of the empty string, passing it result
as an argument.

or

>>> "!".join(result)
'd!o!g'

Jeremy






From tdelaney at avaya.com  Tue Feb 27 04:52:34 2001
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 27 Feb 2001 14:52:34 +1100
Subject: list to string
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBEC9@natasha.auslabs.avaya.com>

Indeed.

Personally, I think the join() string method is *really* bad. A method
should act on the object it is a part of. In this case, there is no
conceptual way that ''.join(['a', 'b', 'c']) could be considered to be
acting on '' - it is *using* ''.

However, join() should also not be a method of sequences in general IMO
(although there is more reason to). Instead, there should *only* be
string.join() - it is an external function which acts on one thing with the
help of another. Actually, there should really be sequence.join(), and
string.join() should be deprecated ...

The object being acted upon should *always* be listed first, whether it be
as the target for a bound method, or as the first parameter of a function.

''.join() is a glaring, non-obvious inconsistency that I will not use.

OTOH, ''.length() is perfectly fine. It is obvious.

Tim Delaney
Avaya Australia
+61 2 9352 9079

> -----Original Message-----
> From: Erno Kuusela [mailto:erno-news at erno.iki.fi]
> Sent: Tuesday, 27 February 2001 2:25 PM
> To: python-list at python.org
> Subject: Re: list to string
> 
> 
> In article <3a9b14c9.33237453 at news-server.austin.rr.com>,
> jgraves3 at austin.rr.com (jay graves) writes:
> 
> | Some people don't like using the string methods in this way and they
> | might write:  (at least until the string module goes away.)
> 
> | import string
> | string.join(result,'')
> 
> the string module isn't going away, it has lots of stuff that
> is not available as methods of strings.
> 
>   -- erno
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 



From aleaxit at yahoo.com  Tue Feb 27 08:12:26 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 08:12:26 +0100
Subject: list to string
References: <mailman.983245987.15676.python-list@python.org>
Message-ID: <97fjv1028qd@news2.newsguy.com>

"Delaney, Timothy" <tdelaney at avaya.com> wrote in message
news:mailman.983245987.15676.python-list at python.org...
> Indeed.
>
> Personally, I think the join() string method is *really* bad. A method
> should act on the object it is a part of. In this case, there is no
> conceptual way that ''.join(['a', 'b', 'c']) could be considered to be
> acting on '' - it is *using* ''.

And, similarly, it's *using* ['a','b','c'] -- not "acting on" it.  It's not
a mutator.  Not all methods of an object mutate it, and you clearly
accept that, as you write:

> OTOH, ''.length() is perfectly fine. It is obvious.

Since join may use the joiner in a completely polymorphic way, but
only uses the sequence through a sequence-iteration protocol, it's
pragmatically _useful_ to have it be a method of the joiner object
(as Python doesn't do multimethod-like dispatching).  Usefulness
seems to be a good indicator: if something offends your personal
sense of purity, but is pragmatically good, then 'practicality beats
purity' and it might be wise to reconsider your aesthetics:-).


Alex





From wayne.izatt at _nospam_myself.com  Tue Feb 27 08:16:22 2001
From: wayne.izatt at _nospam_myself.com (Wayne Izatt)
Date: Tue, 27 Feb 2001 07:16:22 GMT
Subject: list to string
References: <1103_983241679@cc462845-a>
Message-ID: <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com>

"glen mettler" <glen.mettler at home.com> wrote in message
news:1103_983241679 at cc462845-a...
> These commands will turn a string into a list:
> myword="dog"
> mywordlist=list(myword)
> result = ["d","o","g"]
>
> is there a command/function that can take the list ["d","o","g"] and make
it a string "dog"?

>>> lst =["d", "o", "g"]
>>> lst
['d', 'o', 'g']
>>> dog = lst[0]+lst[1]+lst[2]
>>> dog
'dog'

or is there something I'm not getting?

cheers

>
> Glen
>
>
>




From sheila at spamcop.net  Tue Feb 27 08:22:03 2001
From: sheila at spamcop.net (Sheila King)
Date: Tue, 27 Feb 2001 07:22:03 GMT
Subject: list to string
References: <1103_983241679@cc462845-a> <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com>
Message-ID: <salm9tcabvjoupc43brcs2srk2ubd90ojo@4ax.com>

On Tue, 27 Feb 2001 07:16:22 GMT, "Wayne Izatt"
<wayne.izatt at _nospam_myself.com> wrote in comp.lang.python in article
<avIm6.174966$df5.5092180 at news1.crdva1.bc.home.com>:

:"glen mettler" <glen.mettler at home.com> wrote in message
:news:1103_983241679 at cc462845-a...
:> These commands will turn a string into a list:
:> myword="dog"
:> mywordlist=list(myword)
:> result = ["d","o","g"]
:>
:> is there a command/function that can take the list ["d","o","g"] and make
:it a string "dog"?
:
:>>> lst =["d", "o", "g"]
:>>> lst
:['d', 'o', 'g']
:>>> dog = lst[0]+lst[1]+lst[2]
:>>> dog
:'dog'
:
:or is there something I'm not getting?

Your method would be awfully tedious for very long strings. You want something
that will be easy to implement, regardless of the length of the string.
Fortunately, several others in this thread have already suggested using the
join command from the string methods.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From danielk at aracnet.com  Tue Feb 27 15:54:52 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Tue, 27 Feb 2001 06:54:52 -0800
Subject: list to string
References: <1103_983241679@cc462845-a> <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com> <salm9tcabvjoupc43brcs2srk2ubd90ojo@4ax.com>
Message-ID: <qlfn9tkckq9g29brr0jiq2638mblclo9jn@4ax.com>

On Tue, 27 Feb 2001 07:22:03 GMT, Sheila King <sheila at spamcop.net> wrote:

>:>>> lst =["d", "o", "g"]
>:>>> lst
>:['d', 'o', 'g']
>:>>> dog = lst[0]+lst[1]+lst[2]
>:>>> dog
>:'dog'
>:
>:or is there something I'm not getting?
>
>Your method would be awfully tedious for very long strings. You want something
>that will be easy to implement, regardless of the length of the string.
>Fortunately, several others in this thread have already suggested using the
>join command from the string methods.

I just _know_ I'm going to get winged for this one but why not

mylist = ['d','o','g']
mystring = ''
for element in mylist:
      mystring = mystring + element

It's simple and intuitive. I know that it has to build a _new_ string each time
but for shorter strings...

Dan



From glennj at pcard31e.ca.nortel.com  Tue Feb 27 16:16:40 2001
From: glennj at pcard31e.ca.nortel.com (Glenn W Jackman)
Date: 27 Feb 2001 15:16:40 GMT
Subject: list to string
References: <1103_983241679@cc462845-a> <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com> <salm9tcabvjoupc43brcs2srk2ubd90ojo@4ax.com> <qlfn9tkckq9g29brr0jiq2638mblclo9jn@4ax.com>
Message-ID: <97ggco$c55$1@bcarh8ab.ca.nortel.com>

Daniel Klein wrote:
>On Tue, 27 Feb 2001 07:22:03 GMT, Sheila King <sheila at spamcop.net> wrote:
>>:>>> lst =["d", "o", "g"]
>>:>>> dog = lst[0]+lst[1]+lst[2]
>>
>>Your method would be awfully tedious for very long strings. You want something
>>that will be easy to implement, regardless of the length of the string.
>>Fortunately, several others in this thread have already suggested using the
>>join command from the string methods.
>
>mylist = ['d','o','g']
>mystring = ''
>for element in mylist:
>      mystring = mystring + element

>It's simple and intuitive. I know that it has to build a _new_ string each time
>but for shorter strings...

Having just read "Learning Python", I offer:

>>> import operator
>>> s = '''Python 1.5.2 (#1, Feb  1 2000, 16:32:16)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux-i386
... Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam'''
>>> l = list(s)           
>>> s2 = reduce(operator.add, l)
>>> s == s2
1
>>> s is s2
0

-- 
Glenn 


From gem at hsv.crc.com  Tue Feb 27 23:54:15 2001
From: gem at hsv.crc.com (Glen Mettler)
Date: Tue, 27 Feb 2001 16:54:15 -0600
Subject: list to string
References: <1103_983241679@cc462845-a> <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com> <salm9tcabvjoupc43brcs2srk2ubd90ojo@4ax.com> <qlfn9tkckq9g29brr0jiq2638mblclo9jn@4ax.com>
Message-ID: <97haor$flu$1@hobbes2.crc.com>

I couldn't get "for element in mylist" to work but this does:

mylist = ['d','o','g']
mystring = ''
for j in range(len(mylist)):
    mystring= mystring+"".join(mylist[j])
print mystring

I am very, very new to Python so this may not be very elegant but it does
work

Glen  (I asked the original question)
Thanks for all help!


"Daniel Klein" <danielk at aracnet.com> wrote in message
news:qlfn9tkckq9g29brr0jiq2638mblclo9jn at 4ax.com...
> On Tue, 27 Feb 2001 07:22:03 GMT, Sheila King <sheila at spamcop.net> wrote:
>
> >:>>> lst =["d", "o", "g"]
> >:>>> lst
> >:['d', 'o', 'g']
> >:>>> dog = lst[0]+lst[1]+lst[2]
> >:>>> dog
> >:'dog'
> >:
> >:or is there something I'm not getting?
> >
> >Your method would be awfully tedious for very long strings. You want
something
> >that will be easy to implement, regardless of the length of the string.
> >Fortunately, several others in this thread have already suggested using
the
> >join command from the string methods.
>
> I just _know_ I'm going to get winged for this one but why not
>
> mylist = ['d','o','g']
> mystring = ''
> for element in mylist:
>       mystring = mystring + element
>
> It's simple and intuitive. I know that it has to build a _new_ string each
time
> but for shorter strings...
>
> Dan
>




From BrianQ at ActiveState.com  Wed Feb 28 01:14:33 2001
From: BrianQ at ActiveState.com (Brian Quinlan)
Date: Tue, 27 Feb 2001 16:14:33 -0800
Subject: list to string
In-Reply-To: <97haor$flu$1@hobbes2.crc.com>
Message-ID: <002901c0a11b$6dc638b0$b503a8c0@activestate.ca>

Nice work. That is a pretty complex construction for someone new. But there
are other ways to get this to work that are a bit cleaner:

import string

mylist = ['d','o','g']
mystring = string.join( mylist, '' )
print mystring

OR (and I hate this style):

mylist = ['d','o','g']
mystring = ''.join( mylist )
print mystring

OR (if you want to do the loop yourself):

mylist = ['d','o','g']
mystring = ''
for j in mylist:
    mystring = mystring + j
print mystring

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Glen Mettler
Sent: Tuesday, February 27, 2001 2:54 PM
To: python-list at python.org
Subject: Re: list to string


I couldn't get "for element in mylist" to work but this does:

mylist = ['d','o','g']
mystring = ''
for j in range(len(mylist)):
    mystring= mystring+"".join(mylist[j])
print mystring

I am very, very new to Python so this may not be very elegant but it does
work

Glen  (I asked the original question)
Thanks for all help!




From gReGlAnDrUm at earthlink.net  Wed Feb 28 02:50:54 2001
From: gReGlAnDrUm at earthlink.net (Greg Landrum)
Date: Tue, 27 Feb 2001 17:50:54 -0800
Subject: list to string
References: <1103_983241679@cc462845-a> <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com> <salm9tcabvjoupc43brcs2srk2ubd90ojo@4ax.com> <qlfn9tkckq9g29brr0jiq2638mblclo9jn@4ax.com> <97haor$flu$1@hobbes2.crc.com>
Message-ID: <97hlc2$9c6$1@slb6.atl.mindspring.net>

"Glen Mettler" <gem at hsv.crc.com> wrote in message
news:97haor$flu$1 at hobbes2.crc.com...
> I couldn't get "for element in mylist" to work but this does:
>
> mylist = ['d','o','g']
> mystring = ''
> for j in range(len(mylist)):
>     mystring= mystring+"".join(mylist[j])
> print mystring
>

This is a lot more work than you need to be doing.
If you just want to convert the array into a string the simplest way
possible:
mylist = ['d','o','g']
mystring = ''.join(mylist)
print mystring

Though this may be somewhat ugly (depending upon who you ask), it is very
simple.

You can also loop over the list.  Here are a couple of ways:

One way to loop over all the elements in the list (like you are doing) is:
mylist = ['d','o','g']
mystring = ''
for j in range(len(mylist)):
  mystring = mystring + mylist[j]

Another way to loop over all the elements in the list is:
mylist = ['d','o','g']
mystring = ''
for item in mylist:
  mystring = mystring + item

> I am very, very new to Python so this may not be very elegant but it does
> work

It's still a good idea to learn more efficient ways of doing things early in
your learning experience.  It'll help you avoid developing bad habits that
you'll have to break later.

If you would like to see what I mean by efficiency, try running a list 10000
elements long through your code and seeing how long it takes in comparison
to the ''.join() solution.  (hint, you can generate a list which is 10000
elements long like this: mylist = ['d']*10000)

I hope this helps,
-greg






From gem at hsv.crc.com  Wed Feb 28 00:00:15 2001
From: gem at hsv.crc.com (Glen Mettler)
Date: Tue, 27 Feb 2001 17:00:15 -0600
Subject: list to string
References: <1103_983241679@cc462845-a> <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com> <salm9tcabvjoupc43brcs2srk2ubd90ojo@4ax.com> <qlfn9tkckq9g29brr0jiq2638mblclo9jn@4ax.com>
Message-ID: <97hb44$fn7$1@hobbes2.crc.com>

My appologies

> mylist = ['d','o','g']
> mystring = ''
> for element in mylist:
>       mystring = mystring + element

does work.  I failed to take out one of the >.

Glen

"Daniel Klein" <danielk at aracnet.com> wrote in message
news:qlfn9tkckq9g29brr0jiq2638mblclo9jn at 4ax.com...
> On Tue, 27 Feb 2001 07:22:03 GMT, Sheila King <sheila at spamcop.net> wrote:
>
> >:>>> lst =["d", "o", "g"]
> >:>>> lst
> >:['d', 'o', 'g']
> >:>>> dog = lst[0]+lst[1]+lst[2]
> >:>>> dog
> >:'dog'
> >:
> >:or is there something I'm not getting?
> >
> >Your method would be awfully tedious for very long strings. You want
something
> >that will be easy to implement, regardless of the length of the string.
> >Fortunately, several others in this thread have already suggested using
the
> >join command from the string methods.
>
> I just _know_ I'm going to get winged for this one but why not
>
> mylist = ['d','o','g']
> mystring = ''
> for element in mylist:
>       mystring = mystring + element
>
> It's simple and intuitive. I know that it has to build a _new_ string each
time
> but for shorter strings...
>
> Dan
>




From tdelaney at avaya.com  Tue Feb 27 23:58:24 2001
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 28 Feb 2001 09:58:24 +1100
Subject: list to string
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBED4@natasha.auslabs.avaya.com>

> And, similarly, it's *using* ['a','b','c'] -- not "acting on" 
> it.  It's not
> a mutator.  Not all methods of an object mutate it, and you clearly
> accept that, as you write:
> 
> > OTOH, ''.length() is perfectly fine. It is obvious.
> 
> Since join may use the joiner in a completely polymorphic way, but
> only uses the sequence through a sequence-iteration protocol, it's
> pragmatically _useful_ to have it be a method of the joiner object
> (as Python doesn't do multimethod-like dispatching).  Usefulness
> seems to be a good indicator: if something offends your personal
> sense of purity, but is pragmatically good, then 'practicality beats
> purity' and it might be wise to reconsider your aesthetics:-).

In this case, I think aesthetics and pragmatism are on the same side for me.
The chosen string method for join() is completely non-obvious to me.

At no time did I mean to suggest that the sequence being "acted on" was
mutated. What I meant was the the result of evaluating the expression was a
modification of the sequence i.e. "take the sequence, modify it by
concatenating each element with this string between each element". It is
obvious that the sequence is the object being "acted on".

Whilst you could turn it around, it becomes convoluted and messy. "Take this
string, stick it between each element of this sequence and then concatenate
each element of the sequence into a string". You are still really acting on
the sequence, not the string.

Tim Delaney
Avaya Australia



From tdelaney at avaya.com  Wed Feb 28 01:42:36 2001
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 28 Feb 2001 11:42:36 +1100
Subject: list to string
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBED6@natasha.auslabs.avaya.com>

> import string
> 
> mylist = ['d','o','g']
> mystring = string.join( mylist, '' )
> print mystring
> 
> OR (and I hate this style):
> 
> mylist = ['d','o','g']
> mystring = ''.join( mylist )
> print mystring
> 
> OR (if you want to do the loop yourself):
> 
> mylist = ['d','o','g']
> mystring = ''
> for j in mylist:
>     mystring = mystring + j
> print mystring

I have no idea why anyone would suggest doing the loop yourself. The first
one is good, the second is functionally just as good (although horrible).

However, the third method has a number of really horrible drawbacks.

We assume that the join() method is optimised. It will probably be written
in C (for CPython), and designed to use as little time and space as
possible.

OTOH, doing the loop oneself means that you create a new string *every* time
you go through the loop. As the string gets larger, more and more space
needs to be allocated, and more data needs to be copied *every* time through
the loop.

There is a standard idiom in python for joining the (string) elements of a
sequence into a string - string.join(). This is most likely to be the
fastest, most efficient and most understandable way to do it.

The important thing is to know the core libraries supplied with python very
well, and to always try to find pre-done things in the optional libraries
before writing it yourself.

Code re-use is good.

''.join() is evil ;)

*Never* have a loop like (for example):

	text = ''

	for line in f.xreadlines()
		text += line

instead do

	text = []

	for line in f.xreadlines()
		text.append(line)

	text = string.join(text, '\n')

It will be much faster and use much less memory. Of course, for this
particular case, there is a much better method ...

	text = string.join(f.readlines(), '\n')

or even

	text = f.read()	# may have slightly different semantics to the above
depending on platform, file and open mode

but I will ignore those as I was trying to demonstrate a different idiom :)

Tim Delaney
Avaya Australia



From BrianQ at ActiveState.com  Wed Feb 28 02:14:49 2001
From: BrianQ at ActiveState.com (Brian Quinlan)
Date: Tue, 27 Feb 2001 17:14:49 -0800
Subject: list to string
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8B0FBED6@natasha.auslabs.avaya.com>
Message-ID: <002a01c0a123$d9581690$b503a8c0@activestate.ca>

Calm down Tim, my last example was not meant as a HOWTO for joining strings,
it was demonstrate how Glen's loop could be simplified by:

1) Iterating accross the list elements directly instead of my index
2) Avoiding the unnecessary use of join (since, in his example, his is
joining a string of length 1 on each iteration).

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Delaney, Timothy
Sent: Tuesday, February 27, 2001 4:43 PM
To: python-list at python.org
Subject: RE: list to string


> import string
>
> mylist = ['d','o','g']
> mystring = string.join( mylist, '' )
> print mystring
>
> OR (and I hate this style):
>
> mylist = ['d','o','g']
> mystring = ''.join( mylist )
> print mystring
>
> OR (if you want to do the loop yourself):
>
> mylist = ['d','o','g']
> mystring = ''
> for j in mylist:
>     mystring = mystring + j
> print mystring

I have no idea why anyone would suggest doing the loop yourself. The first
one is good, the second is functionally just as good (although horrible).

However, the third method has a number of really horrible drawbacks.

We assume that the join() method is optimised. It will probably be written
in C (for CPython), and designed to use as little time and space as
possible.

OTOH, doing the loop oneself means that you create a new string *every* time
you go through the loop. As the string gets larger, more and more space
needs to be allocated, and more data needs to be copied *every* time through
the loop.

There is a standard idiom in python for joining the (string) elements of a
sequence into a string - string.join(). This is most likely to be the
fastest, most efficient and most understandable way to do it.

The important thing is to know the core libraries supplied with python very
well, and to always try to find pre-done things in the optional libraries
before writing it yourself.

Code re-use is good.

''.join() is evil ;)

*Never* have a loop like (for example):

	text = ''

	for line in f.xreadlines()
		text += line

instead do

	text = []

	for line in f.xreadlines()
		text.append(line)

	text = string.join(text, '\n')

It will be much faster and use much less memory. Of course, for this
particular case, there is a much better method ...

	text = string.join(f.readlines(), '\n')

or even

	text = f.read()	# may have slightly different semantics to the above
depending on platform, file and open mode

but I will ignore those as I was trying to demonstrate a different idiom :)

Tim Delaney
Avaya Australia

--
http://mail.python.org/mailman/listinfo/python-list




From tdelaney at avaya.com  Wed Feb 28 02:37:46 2001
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 28 Feb 2001 12:37:46 +1100
Subject: list to string
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBED8@natasha.auslabs.avaya.com>

> Calm down Tim, my last example was not meant as a HOWTO for 
> joining strings,
> it was demonstrate how Glen's loop could be simplified by:
> 
> 1) Iterating accross the list elements directly instead of my index
> 2) Avoiding the unnecessary use of join (since, in his example, his is
> joining a string of length 1 on each iteration).

Not a problem, but it should have been stated as such. Otherwise people like
Glen, who appears to be new to Python, will likely assume that it is the
"correct" way to do it. In fact, Glen's own posts give the impression that
he now believes that looping over the sequence himself is the "correct" way.
Whilst it will work, it is definitely sub-optimal.

In any case, I dispute your point 2. He is not joining strings of length one
on each loop - he is appending a string of length one to a string of
ever-increasing length. This is one of the the *worst* cases possible, and
he should definitely be using join() for this.

Tim Delaney
Avaya Australia



From ben.hutchings at roundpoint.com  Wed Feb 28 02:47:33 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 27 Feb 2001 17:47:33 -0800
Subject: list to string
References: <mailman.983320989.19570.python-list@python.org>
Message-ID: <uae77innu.fsf@roundpoint.com>

"Delaney, Timothy" <tdelaney at avaya.com> writes:
<snip>
> It will be much faster and use much less memory. Of course, for this
> particular case, there is a much better method ...
> 
> 	text = string.join(f.readlines(), '\n')
> 
> or even
> 
> 	text = f.read()	# may have slightly different semantics to the above
> depending on platform, file and open mode
<snip>

f.readlines() leaves the line terminator at the end of each of its
lines, so the first expression will double the line terminators
(except at the end of the last line).

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From db3l at fitlinxx.com  Wed Feb 28 03:00:16 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 21:00:16 -0500
Subject: list to string
References: <mailman.983320989.19570.python-list@python.org>
Message-ID: <uitlvmurz.fsf@ctwd0143.fitlinxx.com>

"Delaney, Timothy" <tdelaney at avaya.com> writes:

> *Never* have a loop like (for example):
> 
> 	text = ''
> 
> 	for line in f.xreadlines()
> 		text += line
> 
> instead do
> 
> 	text = []
> 
> 	for line in f.xreadlines()
> 		text.append(line)
> 
> 	text = string.join(text, '\n')
> 
> It will be much faster and use much less memory. (...)

True it will likely be faster, but I'm not sure I agree on the memory
point.  It's interesting that you use xreadlines to avoid reading the
whole file into memory, but then you go and stick it into a list that
has to be kept in memory until it can be joined, at which point you
use twice as much memory to build up the final string.  Why not just
say:

    text = f.readlines()
    text = string.join(text)

(or as you point out later, "text = f.read()" :-))

it would do the same thing, use the same memory space, and probably
run even faster since it can suck the whole file in as quickly as
possible.

The former case thrashes memory more (although Python probably reuses
the same space a lot), but probably uses somewhat less memory, since
in the end, the final copy builds the full image from the file
contents minus the final line.  Also, the file is in memory as a
single string, and not a list with the per-line string object overhead
and list reference.  If you've got lots of lines in the file, that
per-object overhead could add up to quite a bit.

Which all sort of goes to say that I wouldn't use the term "*Never*"
as you do in the first case.  It's possible that when first developing
the algorithm, that the first case makes sense and is the clearest
implementation.  I think that reading that code makes it very clear
what is happening without having to look ahead some lines (e.g., the
reader doesn't have to ask why the lines are being stored in a list).

Of course, once you start unit testing and fleshing out the algorithm
you may find this code to be a bottleneck and choose to optimize it in
various ways, but that needn't be the first inclination, nor perhaps
even done at all if the files involved or the overall process makes
this a non-performance critical piece of code.

Performance concerns are reasonable in any piece of code - I just
wouldn't go so far as to declare one algorithm (that aside from
implementation performance issues is perfectly clear) never worth
doing.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From tdelaney at avaya.com  Wed Feb 28 03:00:56 2001
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 28 Feb 2001 13:00:56 +1100
Subject: list to string
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBED9@natasha.auslabs.avaya.com>

> > 	text = string.join(f.readlines(), '\n')
> f.readlines() leaves the line terminator at the end of each of its
> lines, so the first expression will double the line terminators
> (except at the end of the last line).

I'm willing to believe it - I don't use readlines(), so I was going off the
documentation.

readlines ([sizehint])

Read until EOF using readline() and return a list containing the lines thus
read. If the optional sizehint argument is present, instead of reading up to
EOF, whole lines totalling approximately sizehint bytes (possibly after
rounding up to an internal buffer size) are read.

Doesn't mention whether the end-of-line characters are left or not.

So, that should be ...

	text = string.join(f.readlines(), '')

Tim Delaney
Avaya Australia



From tdelaney at avaya.com  Wed Feb 28 03:35:57 2001
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 28 Feb 2001 13:35:57 +1100
Subject: list to string
Message-ID: <B43D149A9AB2D411971300B0D03D7E8B0FBEDA@natasha.auslabs.avaya.com>

> > 	text = ''
> > 
> > 	for line in f.xreadlines()
> > 		text += line

> True it will likely be faster, but I'm not sure I agree on the memory
> point.  It's interesting that you use xreadlines to avoid reading the
> whole file into memory, but then you go and stick it into a list that
> has to be kept in memory until it can be joined, at which point you
> use twice as much memory to build up the final string.  Why not just
> say:
> 
>     text = f.readlines()
>     text = string.join(text)
> 
> (or as you point out later, "text = f.read()" :-))

Because the whole point was to concentrate on the time and memory used by
the string concatentations. We were discussing the string concatenations vs
string.join(). I used "reading a file line by line) as an example (as I
stated) simply because it is a very common idiom which demonstrates the
exact semantics were were discussing.

> it would do the same thing, use the same memory space, and probably
> run even faster since it can suck the whole file in as quickly as
> possible.

I agree 100%. Which is why I said that these were *better* for the specific
case of my example, but they don't apply to the discussion in question.

> The former case thrashes memory more (although Python probably reuses
> the same space a lot), but probably uses somewhat less memory, since
> in the end, the final copy builds the full image from the file
> contents minus the final line.  Also, the file is in memory as a
> single string, and not a list with the per-line string object overhead
> and list reference.  If you've got lots of lines in the file, that
> per-object overhead could add up to quite a bit.

This is perfectly valid. Perhaps the best way in this *particular* case
(without using read()) would be

	text = string.join(f.xreadlines(), '')

However, once again this is not the discussion in question. Glen was asking
about concatenating the (string) elements of a list which he had in memory.
I was extending this to larger sequences, and the simplest way of getting a
large sequence is by grabbing the lines from a file.

> Which all sort of goes to say that I wouldn't use the term "*Never*"
> as you do in the first case.  It's possible that when first developing
> the algorithm, that the first case makes sense and is the clearest
> implementation.  I think that reading that code makes it very clear
> what is happening without having to look ahead some lines (e.g., the
> reader doesn't have to ask why the lines are being stored in a list).

Okay - perhaps I should have qualified "never" ;) *Never* use that algorithm
in final code in Python. Once you know the core Python libraries, never use
it at all as your first thought when confronted with this situation should
be to use the Python idiom of string.join(). Better?

> Of course, once you start unit testing and fleshing out the algorithm
> you may find this code to be a bottleneck and choose to optimize it in
> various ways, but that needn't be the first inclination, nor perhaps
> even done at all if the files involved or the overall process makes
> this a non-performance critical piece of code.

And once it becomes second nature to use string.join() you can eliminate
needing to consider optimising this part of the code (unless it is such a
bottleneck that you think you can make it faster by rewriting it as a C
extension, but that's another story entirely competely unrelated to the
original discussion).

> Performance concerns are reasonable in any piece of code - I just
> wouldn't go so far as to declare one algorithm (that aside from
> implementation performance issues is perfectly clear) never worth
> doing.

Personally, I find the string.join() idiom much clearer and more readable
than a loop concatenating the elements of a string.

Of course, all this becomes moot as soon as you have a sequence which
contains elements other than strings ;)

Tim Delaney
Avaya Australia



From fredrik at pythonware.com  Wed Feb 28 05:56:57 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 04:56:57 GMT
Subject: list to string
References: <mailman.983325731.4920.python-list@python.org>
Message-ID: <ty%m6.15922$Qb7.2599156@newsb.telia.net>

Timothy wrote:
> I'm willing to believe it - I don't use readlines(), so I was going off the
> documentation.
>
> readlines ([sizehint])
>
> Read until EOF using readline() and return a list containing the lines thus
> read. If the optional sizehint argument is present, instead of reading up
> to EOF, whole lines totalling approximately sizehint bytes (possibly after
> rounding up to an internal buffer size) are read.
>
> Doesn't mention whether the end-of-line characters are left or not.

it says it's using readline, and readline is documented as follows:

    readline ([size])

    Read one entire line from the file. A trailing newline character
    is kept in the string (but may be absent when a file ends with
    an incomplete line)

Cheers /F




From flognat at flognat.myip.org  Wed Feb 28 22:13:09 2001
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Wed, 28 Feb 2001 21:13:09 GMT
Subject: list to string
References: <mailman.983314749.30688.python-list@python.org>
Message-ID: <m3bsrm4in5.fsf@master.athome>

FYI: The listing you see, is what the server sends to you, ftplib
haven't touched it.

        /Andy

/ "Andrea Ciuffoli" <ACIUFFO1 at irf.com> wrote:
| [....] 
| The last 3 files were created in 2001.
| I was wondering is it is possible to have either Time and Year.
| Are there some example on how the "Ftplib" commands work?
| 
| Thanks in advance and my best regards,
| Andrea.


From max at alcyone.com  Tue Feb 27 04:36:42 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 26 Feb 2001 19:36:42 -0800
Subject: PIL and PPM
Message-ID: <3A9B20CA.F8B921D2@alcyone.com>

Does PIL support the text-based PPMs?  It supports the binary formats
fine, but for a seemingly-valid text formatted PPM it choked with an
invalid format IOError.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Will I disappoint my future / If I stay
\__/ Sade
    Kepler's laws / http://www.alcyone.com/max/physics/kepler/
 A proof of Kepler's laws.


From gary at esands.com  Tue Feb 27 05:19:06 2001
From: gary at esands.com (Gary Quinn)
Date: Tue, 27 Feb 2001 15:19:06 +1100
Subject: Determining disk space
Message-ID: <3A9B2ABA.AFFC65EC@esands.com>

Hello
I'm writing a routine to clean up a hard disk which has large numbers of
satellite data files written onto it (approx 80MBytes/hr). We must
remove oldest files when the disk is 80% full.
What is the best way using Python to determine the number of bytes
remaining on a disk?
What's the best way under NT? under Unix?
Thanks






From erno-news at erno.iki.fi  Tue Feb 27 05:31:51 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 27 Feb 2001 06:31:51 +0200
Subject: Determining disk space
References: <3A9B2ABA.AFFC65EC@esands.com>
Message-ID: <ku8zmsrbk8.fsf@lasipalatsi.fi>

In article <3A9B2ABA.AFFC65EC at esands.com>, Gary Quinn
<gary at esands.com> writes:

| What is the best way using Python to determine the number of bytes
| remaining on a disk?

works for me under linux:

>>> import os, statvfs
>>> x = os.statvfs('/tmp')
>>> print 'space in kB:', x[statvfs.F_BAVAIL] * x[statvfs.F_BSIZE] >> 10
space in kB: 789764

  -- erno


From f8dy at yahoo.com  Tue Feb 27 15:51:31 2001
From: f8dy at yahoo.com (Mark Pilgrim)
Date: Tue, 27 Feb 2001 09:51:31 -0500
Subject: Determining disk space
References: <3A9B2ABA.AFFC65EC@esands.com>
Message-ID: <97geqg$p5i8o$1@ID-77331.news.dfncis.de>

"Gary Quinn" <gary at esands.com> wrote in message
news:3A9B2ABA.AFFC65EC at esands.com...
> Hello
> I'm writing a routine to clean up a hard disk which has large numbers of
> satellite data files written onto it (approx 80MBytes/hr). We must
> remove oldest files when the disk is 80% full.
> What is the best way using Python to determine the number of bytes
> remaining on a disk?
> What's the best way under NT? under Unix?

# Windows-specific
import win32file
(sectorsPerCluster, bytesPerSector, freeClusters, totalClusters) = \
  win32file.GetDiskFreeSpace("c:\\") # or any other drive letter
percentageFull = int(long(freeClusters)*100/long(totalClusters))
bytesOfFreeSpace = sectorsPerCluster*bytesPerSector*long(freeClusters)
# etc.

The int/long calls are to avoid an OverflowError if your disk is too large
for integer math.

-M
You're smart; why haven't you learned Python yet?
http://diveintopython.org/





From clee at gnwy100.wuh.wustl.edu  Tue Feb 27 05:26:22 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 26 Feb 2001 22:26:22 -0600
Subject: info version of python 2.x documentation?
Message-ID: <7ulmqsoioh.fsf@gnwy100.wuh.wustl.edu>

As the final step in upgrading to python 2.0, I've been trying to build the
info versions of the manuals...and failing spectacularly.

I've tried upgrading to the latest latex2html, HTML::Element, etc., 
but I'm giving up for now.

Does some generous soul have them available somewhere?

Many thanks,

-chris




From edcjones at erols.com  Tue Feb 27 06:30:37 2001
From: edcjones at erols.com (Edward C. Jones)
Date: Tue, 27 Feb 2001 00:30:37 -0500
Subject: How to exit Python nicely after error in embedded C code?
Message-ID: <3A9B3B7D.66E31386@erols.com>

Suppose I have a C function that is called only by other C functions (so it
doesn't return a PyObject *). If I find an error condition in the the C
function, is there function I can call that will exit the C code back into
Python so Python can clean up and exit. The function is called by many different
C function in case of error.

int error_callback(int status, char * funcname) {
    if (status != 0) {
        printf("Error occured in %s\n", funcname);
        /* Is this part of what I want: */
        PyErr_SetString(PyExc_Exception, "Error in C function");
        /* ... what goes here ... */
    }
    return 0;
}

I am using Python and C on a Linux system.

Thanks,
Ed Jones




From aleaxit at yahoo.com  Tue Feb 27 08:03:27 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 08:03:27 +0100
Subject: How to exit Python nicely after error in embedded C code?
References: <3A9B3B7D.66E31386@erols.com>
Message-ID: <97fjl9028ir@news2.newsguy.com>

"Edward C. Jones" <edcjones at erols.com> wrote in message
news:3A9B3B7D.66E31386 at erols.com...
> Suppose I have a C function that is called only by other C functions (so
it
> doesn't return a PyObject *). If I find an error condition in the the C
> function, is there function I can call that will exit the C code back into
> Python so Python can clean up and exit. The function is called by many
different
> C function in case of error.
>
> int error_callback(int status, char * funcname) {
>     if (status != 0) {
>         printf("Error occured in %s\n", funcname);
>         /* Is this part of what I want: */
>         PyErr_SetString(PyExc_Exception, "Error in C function");
>         /* ... what goes here ... */
>     }
>     return 0;
> }

In C, you can portably achieve this effect only by having this
function return or set an error indicator and having every
caller test it (there is a mechanism called setjmp/longjmp,
but it may not work as desired everywhere).  In C++, it's
easy, of course -- that's what exceptions are for.


Alex





From lluang at northwestern.edu  Tue Feb 27 15:43:52 2001
From: lluang at northwestern.edu (Louis Luangkesorn)
Date: Tue, 27 Feb 2001 08:43:52 -0600
Subject: How to exit Python nicely after error in embedded C code?
References: <3A9B3B7D.66E31386@erols.com> <97fjl9028ir@news2.newsguy.com>
Message-ID: <3A9BBD28.8F537F75@northwestern.edu>

(I'm asking this out of complete ignorance, obviously)
In this situation (error occurs in a C function that is called by other C
functions which are called by Python), would having the original C call from
Python in a try...except block catch the exception?  (I'm just learning how to
do this, in both C++ and Python, but I have not done this in real life yet.  It
seems like a nice, cheap way of using exceptions in C, just call everything from
Python :-)

Louis

Alex Martelli wrote:

> "Edward C. Jones" <edcjones at erols.com> wrote in message
> news:3A9B3B7D.66E31386 at erols.com...
> > Suppose I have a C function that is called only by other C functions (so
> it
> > doesn't return a PyObject *). If I find an error condition in the the C
> > function, is there function I can call that will exit the C code back into
> > Python so Python can clean up and exit. The function is called by many
> different
> > C function in case of error.
> >
> <sample code deleted>

>
> In C, you can portably achieve this effect only by having this
> function return or set an error indicator and having every
> caller test it (there is a mechanism called setjmp/longjmp,
> but it may not work as desired everywhere).  In C++, it's
> easy, of course -- that's what exceptions are for.
>
> Alex

--
K Louis Luangkesorn
lluang at northwestern.edu  http://pubweb.nwu.edu/~kll560  PGP:0xF3E2D362
Whatsoever things are true, ... honest, ... just, ... pure, ... lovely, ... of
good report; if there be any virtue, and if there be any praise, think on these
things.  - motto - Northwestern University




From aleaxit at yahoo.com  Tue Feb 27 17:17:46 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 17:17:46 +0100
Subject: How to exit Python nicely after error in embedded C code?
References: <3A9B3B7D.66E31386@erols.com> <97fjl9028ir@news2.newsguy.com> <3A9BBD28.8F537F75@northwestern.edu>
Message-ID: <97gjv80147@news1.newsguy.com>

"Louis Luangkesorn" <lluang at northwestern.edu> wrote in message
news:3A9BBD28.8F537F75 at northwestern.edu...
    [talking about "exceptions" from C vs C++]
> (I'm asking this out of complete ignorance, obviously)
> In this situation (error occurs in a C function that is called by other C
> functions which are called by Python), would having the original C call
from
> Python in a try...except block catch the exception?  (I'm just learning
how to
> do this, in both C++ and Python, but I have not done this in real life
yet.  It
> seems like a nice, cheap way of using exceptions in C, just call
everything from
> Python :-)

The C language does not define nor use exceptions.

The Python C API, therefore, does not use them either.  Python is
entirely coded in C, to the C89 standard of ANSI and later ISO, not
in C++.

_Python_ exceptions are not the same thing as _C++_ exceptions.
A "Python Exception" is "raised" by a C-coded function, in an
extension, that is called by Python, by returning 0 (aka NULL)
as its PyObject* (and some previous Python C API calls to give
the exceptions' details).

There are several frameworks that make it easier to extend Python
by harnessing (some of) the extra power of C++ wrt C: CXX, SCXX,
the Boost Python Library (BPL, part of the free Boost collection,
see www.boost.org), and others yet.  Some of them do make use of
the power of C++ exceptions, somewhat in the way you suggest, and,
also, may bridge/translate between C++ exceptions and Python ones.

But there is no clean, portable way to "raise an exception" if you
choose to use C, rather than C++, to code your Python extensions,
and you find yourself in a C function called by a C function [repeat
down a few nesting levels...] called by Python.  Either switch to
C++ (and some appropriate C++/Python framework such as the BPL), or
live with patiently returning an error indication from the deeply
nested function, and checking-and-propagating-upwards in each
caller.

That is, of course, IF you care about portability; if you don't, you
may experiment to see what your platform[s] of interest will allow
in a setjmp/longjmp pair, which is the closest thing that C offers
to "exceptions" (just take a *LOT* of care about memory leaks &c,
as you get absolutely no protection, no equivalent of C++'s automatic
execution of destructors as the stack is unwound due to exceptions).


Alex





From robin.thomas at starmedia.net  Tue Feb 27 18:56:12 2001
From: robin.thomas at starmedia.net (Robin Thomas)
Date: Tue, 27 Feb 2001 12:56:12 -0500
Subject: How to exit Python nicely after error in embedded C code?
In-Reply-To: <3A9B3B7D.66E31386@erols.com>
Message-ID: <4.3.1.2.20010227124027.00cffb10@exchange.starmedia.net>

At 12:30 AM 2/27/01 -0500, Edward C. Jones wrote:
>Suppose I have a C function that is called only by other C functions (so it
>doesn't return a PyObject *). If I find an error condition in the the C
>function, is there function I can call that will exit the C code back into
>Python so Python can clean up and exit. The function is called by many 
>different
>C function in case of error.

Your "wrapper" C function -- the one that returns a PyObject* -- is of 
course responsible for doing all this work. In most all cases, Python 
expects your wrapper function to return a non-NULL PyObject* if no error 
occurred. If an error occurred, Python expects your wrapper function to set 
the exception state and return NULL. So:

PyObject *
mywrapperfunc (void)
{
     int result;
     result = my_pure_c_function();
     if ( result == -1 )
     {
         PyErr_SetString(PyExc_RuntimeError, "error in C func");
         return NULL;
     }
     /* on success, your wrapper function must return *some*
        valid PyObject; in this example, there is no "return value"
        from the pure C function, so you want to return None */
     /* new reference to None required */
     Py_INCREF(PyNone);
     return PyNone;
}

>int error_callback(int status, char * funcname) {
>     if (status != 0) {
>         printf("Error occured in %s\n", funcname);
>         /* Is this part of what I want: */
>         PyErr_SetString(PyExc_Exception, "Error in C function");
>         /* ... what goes here ... */
>     }
>     return 0;
>}

Very close, but instead of having a general function *decide* whether to 
set the Python exception state, use a general function only to set the 
exception state because you're *certain* that an exception must be set. And 
of course the general function must return the same type that your wrapper 
functions return -- PyObject*.

This approach occurs in Python's own source code. An example from 
Objects/abstract.c:

static PyObject *
type_error(const char *msg)
{
         PyErr_SetString(PyExc_TypeError, msg);
         return NULL;
}

and it's only used when a "wrapper" function wants to do an error. The 
wrapper function must do the work if there is no error. To merge the two 
examples:

static PyObject *
myerror(const char *msg)
{
         PyErr_SetString(PyExc_RuntimeError, msg);
         return NULL;
}

PyObject *
mywrapperfunc (void)
{
     int result;
     result = my_pure_c_function();
     if ( result == -1 )
     {
         /* use the new general error function */
         return myerror("my_pure_c_function bailed");
     }
     /* on success, your wrapper function must return *some*
        valid PyObject; in this example, there is no "return value"
        from the pure C function, so you want to return None */
     /* new reference to None required */
     Py_INCREF(PyNone);
     return PyNone;
}

Hope this helps.



--
Robin Thomas
Engineering
StarMedia Network, Inc.
robin.thomas at starmedia.net




From dwndgg at opti.cgi.net  Tue Feb 27 07:23:04 2001
From: dwndgg at opti.cgi.net (Dan)
Date: Tue, 27 Feb 2001 06:23:04 GMT
Subject: Breaking out of nested loop
Message-ID: <3a9b44f1.226505016@news.okc1.ok.home.com>

How might one go about breaking out of a multi level loop?

For instance, the following code will go into an endless loop.

while(1):
    while(1):
        break


Is there a statement that one can use to break out of multiple levels
of loop control?  I really hate to start littering my code with flag
variables and if statements.

-Dan



From erno-news at erno.iki.fi  Tue Feb 27 08:00:36 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 27 Feb 2001 09:00:36 +0200
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com>
Message-ID: <ku4rxgr4ob.fsf@lasipalatsi.fi>

In article <3a9b44f1.226505016 at news.okc1.ok.home.com>,
dwndgg at opti.cgi.net (Dan) writes:

| How might one go about breaking out of a multi level loop?

this was recently discussed on this newsgroup. 
see <URL:http://groups.google.com/groups?
lr=&safe=off&ic=1&th=db20aaae4e3aea66&seekd=924501943#924501943>

(i hope those google urls are persistent like the deja
ones used to be...)

  -- erno


From fredrik at pythonware.com  Tue Feb 27 08:01:34 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 27 Feb 2001 07:01:34 GMT
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com>
Message-ID: <ihIm6.15728$Qb7.2573501@newsb.telia.net>

"Dan" wrote:
> How might one go about breaking out of a multi level loop?
>
> For instance, the following code will go into an endless loop.
>
> while(1):
>     while(1):
>         break

refactor: return, raise.

(see http://www.refactoring.com/)

Cheers /F




From phlip_cpp at my-deja.com  Tue Feb 27 17:01:42 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 27 Feb 2001 16:01:42 GMT
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com>
Message-ID: <97gj16$d5d@dispatch.concentric.net>

Proclaimed Dan from the mountaintops:

> 
> How might one go about breaking out of a multi level loop?
> 
> For instance, the following code will go into an endless loop.
> 
> while(1):
>     while(1):
>         break

They covered this in a recent thread.

Your best bet (just like in a language that has a 'goto') is 'return'.

Otherwiset you can 'raise' (irritating style that reminds one of VB), or 
you can add a flag to the 'while' conditions and reset it at the 'break' 
point.

Following these style rules...

        simplify 'if' and 'while' test conditions

        don't 'raise' except in an emergency

        shun long functions - don't go over 30 lines or 2 loop statements

...you'l see the best fit is 'return'.

-- 
  Phlip                          phlip_cpp at my-deja.com
============== http://phlip.webjump.com ==============
  --  Spaghetti Bad. Re-use Good.  --


From jmarshal at mathworks.com  Tue Feb 27 17:32:17 2001
From: jmarshal at mathworks.com (jcm)
Date: 27 Feb 2001 16:32:17 GMT
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com> <97gj16$d5d@dispatch.concentric.net>
Message-ID: <97gkqh$lsp$1@news.mathworks.com>

Phlip <phlip_cpp at my-deja.com> wrote:
> Proclaimed Dan from the mountaintops:

>> 
>> How might one go about breaking out of a multi level loop?
>> 
>> For instance, the following code will go into an endless loop.
>> 
>> while(1):
>>     while(1):
>>         break

...

> Your best bet (just like in a language that has a 'goto') is 'return'.

> Otherwiset you can 'raise' (irritating style that reminds one of VB), or 
> you can add a flag to the 'while' conditions and reset it at the 'break' 
> point.

Something I wouldn't mind seeing added to Python is a generalization
of break/continue statement[s], allowing you to break out of multiple
loops:

  break_or_continue_stmt: 'break'
                        | 'continue'
                        | 'break' break_or_continue_stmt

So you can do something like:

  while a:  # loop A
     ...
     while b:  # loop B
        ...
        while c:  # loop C
           ...
           break break  # breaks loops C and B, bringing
                        # you back to the body of loop A


From othello at javanet.com  Tue Feb 27 18:22:28 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 27 Feb 2001 12:22:28 -0500
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com> <97gj16$d5d@dispatch.concentric.net> <97gkqh$lsp$1@news.mathworks.com>
Message-ID: <3A9BE254.459265C4@javanet.com>

jcm wrote:

> Phlip <phlip_cpp at my-deja.com> wrote:
> > Proclaimed Dan from the mountaintops:
>
> >>
> >> How might one go about breaking out of a multi level loop?
> >>
> >> For instance, the following code will go into an endless loop.
> >>
> >> while(1):
> >>     while(1):
> >>         break
>
> ...
>
> > Your best bet (just like in a language that has a 'goto') is 'return'.
>
> > Otherwiset you can 'raise' (irritating style that reminds one of VB), or
> > you can add a flag to the 'while' conditions and reset it at the 'break'
> > point.
>
> Something I wouldn't mind seeing added to Python is a generalization
> of break/continue statement[s], allowing you to break out of multiple
> loops:
>
>   break_or_continue_stmt: 'break'
>                         | 'continue'
>                         | 'break' break_or_continue_stmt
>
> So you can do something like:
>
>   while a:  # loop A
>      ...
>      while b:  # loop B
>         ...
>         while c:  # loop C
>            ...
>            break break  # breaks loops C and B, bringing
>                         # you back to the body of loop A

Adding a feature like is begging for hard to find errors.
A factored out function with a return almost always does
the trick and does it clearly.  And if not, then the needed
functionality is provided by try, except, and raise.

Raymond



From glenfant at equod.com.nospam  Tue Feb 27 18:07:47 2001
From: glenfant at equod.com.nospam (Gillou)
Date: Tue, 27 Feb 2001 18:07:47 +0100
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com>
Message-ID: <97gmho$36q$1@reader1.imaginet.fr>

Not very sexy but it should work...

import exceptions
class LoopBreaker(exceptions.Exception):
    def __init__(self):
        pass
...
try:
    while 1:
        while 1:
        ...
        raise LoopBreaker
...
except LoopBreaker:
    pass
...

"Dan" <dwndgg at opti.cgi.net> a ?crit dans le message news:
3a9b44f1.226505016 at news.okc1.ok.home.com...
>
> How might one go about breaking out of a multi level loop?
>
> For instance, the following code will go into an endless loop.
>
> while(1):
>     while(1):
>         break
>
>
> Is there a statement that one can use to break out of multiple levels
> of loop control?  I really hate to start littering my code with flag
> variables and if statements.
>
> -Dan
>



From joconnor at cybermesa.com  Tue Feb 27 18:20:22 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 27 Feb 2001 10:20:22 -0700
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com>
Message-ID: <3A9BE1D6.5D595D9A@cybermesa.com>


Dan wrote:

> How might one go about breaking out of a multi level loop?
>
> For instance, the following code will go into an endless loop.
>
> while(1):
>     while(1):
>         break
>
> Is there a statement that one can use to break out of multiple levels
> of loop control?  I really hate to start littering my code with flag
> variables and if statements.

Factor the inner loop into a seperate function that returns whether or
not to keep looping and have that the test condition for the outer loop

Example:
>>> x = 10
>>> def secondLoop ():
    global x
    x = x -1
    return x == 1

>>> while not secondLoop():
    print x

9
8
7
6
5
4
3
2
>>>

I find factor nested loops into seperate functions easier to read, as
well, because it allows you to focus on different parts of the logic as
you need to.  I don't have to look at secondLoop()s logic if I just want
to deal with the outer loop.  Simularly, I don't have to be distracted by
the outer loop if I have a bug in the code inside secondLoop()

Take care,
Jay



From marc.vollmer at mahr.de  Tue Feb 27 08:52:22 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Tue, 27 Feb 2001 08:52:22 +0100
Subject: Newbie: Problems with PythonWin and wxPython
Message-ID: <97fmla$p8o62$1@ID-59219.news.dfncis.de>

Hello,

My system is:
Windows 2000
PythonWin 2.0 (win32all build 135)
wxPython

I start the example >>wxpython\lib\wxPlotCanvas.py<< with F5. At the first
and second time the program runs normally, but by the third start pythonwin
cancel with memory (read) problems.

How I can elimate this problem?

Thanks

Marc Vollmer




From fredrik.thernelius at hotsip.com  Tue Feb 27 11:28:25 2001
From: fredrik.thernelius at hotsip.com (newscache3.freenet.de)
Date: Tue, 27 Feb 2001 11:28:25 +0100
Subject: Accessing java objects in python scripts?
Message-ID: <3a9b8049$1@post.newsfeeds.com>

*** post for free via your newsreader at post.newsfeeds.com ***

I'm sorry if this subject has been discussed before. If it has then please
give me a reference to that discussion.

I'll now give a short description of my problem:

In my java program I have instaciated a javaobject of a certain kind, lets
call the class Testclass and new instanciated object testObject. I want to
access this testObject from my jython interpreter. What is the best way to
do this?



By using the "try and fail" concept, I have found out that there are a
couple of way to do this. In order to hava a "pretty" solution I want to
have a python class that will be able to use methods from the testClass
object. In fact, I can't even find this testObject in ANY namespace. Please
find it for me!



Notice that I've not bothered to add "import" statements and some other
stuff in the code.



Regards,

Fredrik Thernelius



Java program:

-------------

python = new PythonInterpreter();

Testclass testObject = new TestClass();   //create a java object

python.exec("e = Engine()");              //create a python object

python.set("obj", testObject);            //making the object accessable in
python

python.exec("e.runEngine()");

//python.set("r.obj", testObject);

python.exec("print obj.toString()");      //Here I can access the
testObject.

python.exec("print globals()");   //<-----in this name space I can see my
testObject





Python script:

--------------

class Engine:

      def __init__(self):

            try:

                  self.compiler =  Perl5Compiler()

                  self.matcher =   Perl5Matcher()



                  #I can't find the testObject

                  print "\nprinting local  variables : ", locals().keys()

                  print "\nprinting global variables: ", globals().keys()

                  print "\nprinting __dict__  variables: ",
self.__dict__.keys()

                  print "\nprinting vars() variables: ", vars()

                  print "\nprinting dir() variables: ", dir()



            except:

                  print "Unexpected Error:", sys.exec_info()

                  raise

            return





      def runEngine(self):

            try:

                  print self.obj.toString()         #<--- this does not
work!

                  #print obj.toString()              <--- this does not work
either!

            except:

                  print "Runtime Error!"

                  raise

            return










  **** Post for FREE via your newsreader at post.newsfeeds.com ****

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
*** Newsfeeds.com - The #1 Usenet Newsgroup Service on The Planet! ***
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
  http://www.newsfeeds.com???    |         http://www.newsfeeds.com
                                 |
* Anonymous posting server!      | * Totally Uncensored!
* SUPER Servers!                 | * Over 80,000 Newsgroups!
* BINARIES ONLY Servers!         | * 16 seperate Newsgroup Servers!
* SPAM FILTERED Server!          | * Instant access!
* ADULT ONLY Server!             | * Multiple OC 3's and OC 12's!
* MP3 ONLY Server!               | * 99% Article Completion!
* MULTIMEDIA ONLY Server!        | * Months of Retention!
* 7 UNCENSORED Newsgroup Servers | * Lightning FAST downloads!
                                 |
  http://www.newsfeeds.com       |         http://www.newsfeeds.com
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 ?      **** Point your newsreader to post.newsfeeds.com ****
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= 


From infotechsys at pivot.net  Tue Feb 27 11:30:06 2001
From: infotechsys at pivot.net (Wayne)
Date: Tue, 27 Feb 2001 05:30:06 -0500
Subject: running the Setup
Message-ID: <3A9B81AD.79B9FB97@pivot.net>

Hi,
I'm trying to compile the interface PyGreSQL.  I have no problem with
the install instructions until
I get to the "make". When I run "make" I get an error that indicates
that readline has no rules.
Would I mess thing up if I comment that out? The other Question I have
is, shouldn't this
error have come up when the person putting together this distribution
built the Python package?
Can someone recommend a good book for learning "Make"?
Thanks
Wayne
P.S. I know I can get an RPM for this, but I what to learn this stuff.




From Max.Haas at unibas.ch  Tue Feb 27 11:38:52 2001
From: Max.Haas at unibas.ch (Max Haas)
Date: Tue, 27 Feb 2001 11:38:52 +0100
Subject: Norvig on Python
Message-ID: <B6C141EF.4512%Max.Haas@unibas.ch>

Hi,

I just received through the MCL (Macintosh Common Lisp) mailing list the
following message, sent by Rainer Joswig. Hope it's useful for this list too
because Notvig mentions Active Python.

Max

========================================================================

Date: Tue, 27 Feb 2001 02:34:12 +0100
From: Rainer Joswig <joswig at CORPORATE-WORLD.lisp.de>
Subject: Norvig: "Extreme Rapid Development"

Hi,

the March edition of the "Software Development"
magazine has an article evaluating Active Python,
Allegro CL, MCL and Functional Developer (Dylan).
The article is written by Peter Norvig
(http://www.norvig.com/). Peter is especially
well known for his excellent book "Paradigms
of Artificial Intelligence Programming, Case
Studies in Common Lisp". A must have for every
serious Lisp programmer. He is co-author
of another classic: "Artificial Intelligence:
A Modern Approach".

The article is online at:

  http://www.sdmagazine.com/articles/2001/0103/0103e/0103e.htm




From jafo at tummy.com  Tue Feb 27 11:44:37 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Tue, 27 Feb 2001 03:44:37 -0700
Subject: CPAN: RFC on Schema (working-name: Swalow)
Message-ID: <20010227034437.A14364@tummy.com>

So we were joking around relating to the name "Python Spihon" (and how
searching on the web turns up hits for an aquarium water changer), and
suggested that we should use "gpw" (a random password generation tool)
to generate a unique name for the CPAN-like project.  It uses english
letter combinations to generate random strings that sound like english
words.

So, Rob Riggs did just that and came up with "swalow":
>Your suggestion set me out to playing with gpw, having it generate
>6-byte names, and it spat out "swalow".
>
>Swallow -- close enough to siphon (in one of its meanings), and with
>a memorable Monty Python reference:
>
>"What is the airspeed velocity of an unladen swallow?"
>"What do you mean? An African or European swallow?"
>
>Of course this means we can have two versions of Swallow ;-)
>
>And it continues to feed on that snake image... what else does one
>envision a python doing with its meal?
>
>And I really like the command-line implications
># swallow PIL

I actually like the name "swalow" (with a single "l"), and searching for
"python swalow" on Google turns up like 4 pages...  So, that's the name
I'm using for my part of the work -- the server that hands out information
about packages.

I started using the schema I posted last night, and through that ran into
some changes that need to be made.  I'm not happy with the dependencies
table, but that'll come along shortly.  At the moment, you can review
the schema at:

	ftp://ftp.tummy.com/pub/tummy/swalow/swalow-1.00/schema

I'd appreciate any feedback you might have.  Also in there is a small
sample data-set named "sample".  You might get a better idea of how the
schema is to be used from that.

I've built a CGI and client program which use this schema to hand the
information out over the net.  Here's a sample session:

guin:swalow$ ./swalow search ITEMNAME=pyntar
ITEMTYPE=package ITEMNAME=pyntar PKGVERS=1.10 PKGFMT=tar
   PKGARCH=* PLATNAME=* PLATVERS=*
ITEMTYPE=package ITEMNAME=pyntar PKGVERS=1.14 PKGFMT=tar
   PKGARCH=* PLATNAME=* PLATVERS=*
guin:swalow$ ./swalow locate ITEMNAME=pyntar PKGVERS=1.14
URL=ftp://ftp.tummy.com//pub/tummy/pyntar/pyntar-1.14.tar.gz
URL=ftp://dev.tummy.com/pub/mirrors/ftp.tummy.com/pub/tummy/pyntar/pyntar-1.14.tar.gz

I decided to use HTTP as the transport mechanism because most firewalls
let it through.  Making a "native" TCP server to handle the requests will
be trivial because of the netstring encoding, and will come much later.
For now, the HTTP/CGI should work fine.

Any suggestions for making the actual management interface?  I figure it'll
have to be CGI (though an e-mail for text-form setup might be interesting
too).  I'm not up on dealing with cookies and all though.  I'll probably be
asking for help as far as setting up the user accounts and editing of
records...  The database will allow multiple-maintainers for a package
though.

Well, that's where it sits tonight.

Comments on it would be appreciated.  Patches would be appreciated more.
;-)

Sean
-- 
 If we could sell our experiences for what they cost us, we'd all be
 millionaires.  -- Abigail Van Buren
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From aahz at panix.com  Tue Feb 27 17:12:13 2001
From: aahz at panix.com (Aahz Maruch)
Date: 27 Feb 2001 08:12:13 -0800
Subject: CPAN: RFC on Schema (working-name: Swalow)
References: <mailman.983270711.22031.python-list@python.org>
Message-ID: <97gjkt$ov0$1@panix6.panix.com>

In article <mailman.983270711.22031.python-list at python.org>,
Sean Reifschneider  <jafo at tummy.com> wrote:
>
>I started using the schema I posted last night, and through that ran into
>some changes that need to be made.  I'm not happy with the dependencies
>table, but that'll come along shortly.  At the moment, you can review
>the schema at:
>
>	ftp://ftp.tummy.com/pub/tummy/swalow/swalow-1.00/schema

It's a decent start, but I have a lot of suggestions (in no particular
order; I'm writing this as I go back-and-forth over the schema):

There's currently little support for archival information; you're going
to have to change a lot of the subsidiary tables to key on both
packages.itemID and packages.version (which could be packages.id, I
suppose; I flip-flop between pure normalization and useful denormalized
forms).

I think it might be a good idea to split items and packages into items,
versions, and packages, allowing multiple packages per version.  If you
do create the versions table, it should have an URL field in addition to
the one in items.  If you create the versions table, my comment above
applies to versions instead of packages.

I think the dependencies table should have fields for minVersion and
maxVersion, which apply to requiresItemID.  This is probably going to
require having a defined format for versions in order to parse them.  It
should be permitted to leave either or both of minVersion and maxVersion
blank.

I don't know what capabilities postGreSQL has, but if you can, I
strongly recommend adding foreign key restrictions on the subsidiary
tables (e.g. require that packages.itemID exist in items.id).
Similarly, you should put a unique index on items.name to prevent
inserting the same name twice (and in other equivalent spots, such as
packages.itemID/packages.version).

I think there should be an items.created datetime field, in case an item
gets created before there's a package ready for it.  I also think there
should be a packages.lastUpdated, which would be used if there was a
mistake in the packaging, but not the actual code that makes up the
package.

It's not really clear to me how you intend the relationship between
items, users, and maintainers to work, so I'll reserve comments on that
for now.

It's not clear to me what packages.architecture is for.

I have zero clue what ranking is for.

One final comment: you should probably change "All Rights Reserved" to
some kind of standard license -- GNU, BSD, whatever.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Nostalgia just ain't what it used to be


From n8gray at caltech.edu.is.my.e-mail.address  Wed Feb 28 21:46:53 2001
From: n8gray at caltech.edu.is.my.e-mail.address (Nathan Gray)
Date: Wed, 28 Feb 2001 12:46:53 -0800
Subject: Swalow vs. Swallow
References: <mailman.983270711.22031.python-list@python.org>
Message-ID: <3A9D63BD.6050500@caltech.edu.is.my.e-mail.address>

Sean Reifschneider wrote:

> So we were joking around relating to the name "Python Spihon" (and how
> searching on the web turns up hits for an aquarium water changer), and
> suggested that we should use "gpw" (a random password generation tool)
> to generate a unique name for the CPAN-like project.  It uses english
> letter combinations to generate random strings that sound like english
> words.
> 
> So, Rob Riggs did just that and came up with "swalow":
> 
>> Your suggestion set me out to playing with gpw, having it generate
>> 6-byte names, and it spat out "swalow".
>> 
>> Swallow -- close enough to siphon (in one of its meanings), and with
>> a memorable Monty Python reference:
>> 
>> "What is the airspeed velocity of an unladen swallow?"
>> "What do you mean? An African or European swallow?"
>> 
>> Of course this means we can have two versions of Swallow ;-)
>> 
>> And it continues to feed on that snake image... what else does one
>> envision a python doing with its meal?
>> 
>> And I really like the command-line implications
>> # swallow PIL
> 
> 
> I actually like the name "swalow" (with a single "l"), and searching for
> "python swalow" on Google turns up like 4 pages...  So, that's the name
> I'm using for my part of the work -- the server that hands out information
> about packages.

So shouldn't the server-side be "regurgitate"?

+1 on swallow -- Monty Python reference good.  Mixed Monty Python/snake 
                   reference too good to pass up!
-1 on swalow  -- looks like a script-kiddie mispeling, and I can't help but
                   pronounce it "sway-low".

How about pySwallow or py_swallow?

I wouldn't worry too much about google searches.  Once the sofware 
starts being used the site will percolate to the top.  Try doing a 
search for "python" and see what site comes up first.  ;-)

My .02 on the only issue of _real_ importance in this whole CPyAN 
debate.  The rest is just implementation details.  ;-)
-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
               Nathaniel Gray
     California Institute of Technology
      Computation and Neural Systems
       n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From sk at gluit.de  Tue Feb 27 12:37:00 2001
From: sk at gluit.de (sk at gluit.de)
Date: Tue, 27 Feb 2001 12:37:00 +0100
Subject: Connecting GUI and database?
Message-ID: <3A9B915C.4251FE3F@gluit.de>

Hi there,

I want to build a PyQt application which uses (mx)ODBC to present data
from a SQL table in a QTable-like widget.

Is there a tool to connect database and GUI such that changes on one
side will automatically affect the other? I could make a concession
that after changing data in the database the program could be triggered
manually to update the widget.

Similar tools using tkinter could also be helpful.

Thank you.

 Stefan


From tom at teikom.no  Tue Feb 27 12:55:21 2001
From: tom at teikom.no (Tom Danielsen)
Date: Tue, 27 Feb 2001 11:55:21 GMT
Subject: win32com.client - parameter types
Message-ID: <slrn99n5be.8f8.tom@tnt.teikom.no>


I have a COM server with the following method:

        [id(0x60020016)]
        HRESULT getDict([out] IDispatch** pDictionary);

which makepy made into:

        def getDict(self, pDictionary=defaultNamedNotOptArg):
                return self._ApplyTypes_(0x60020016, \
                        1, (24, 0), ((16393, 2),), \
                        'getDict', None,pDictionary)

This is, the method returns its value in parameter 1.

VB and VBscript handles this, but I cannot get Python (2.0) to work.

Is there any way to get at this value in Python?  Is there a way I
can change the [out] to a [out, retval] and make Python return the
value as the return value of the method call?  If so, how?



Tom



From aleaxit at yahoo.com  Tue Feb 27 14:23:13 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 14:23:13 +0100
Subject: win32com.client - parameter types
References: <slrn99n5be.8f8.tom@tnt.teikom.no>
Message-ID: <97g9or029n3@news1.newsguy.com>

"Tom Danielsen" <tom at teikom.no> wrote in message
news:slrn99n5be.8f8.tom at tnt.teikom.no...
>
>
> I have a COM server with the following method:
>
>         [id(0x60020016)]
>         HRESULT getDict([out] IDispatch** pDictionary);
>
> which makepy made into:
>
>         def getDict(self, pDictionary=defaultNamedNotOptArg):
>                 return self._ApplyTypes_(0x60020016, \
>                         1, (24, 0), ((16393, 2),), \
>                         'getDict', None,pDictionary)

Not to worry -- this may be slightly misleading, but it DOES
in fact work just right.


> This is, the method returns its value in parameter 1.

That's what it does COMwise, but, for Python, this output
parameter IS transformed into a return value!


> VB and VBscript handles this, but I cannot get Python (2.0) to work.
>
> Is there any way to get at this value in Python?  Is there a way I
> can change the [out] to a [out, retval] and make Python return the
> value as the return value of the method call?  If so, how?

Nothing more than the makepy you've already done.

To try it out, I built a toy (ATL) project (inproc server, just
one object, project name toyatl, object name Toy) with IDL:

// toyatl.idl : IDL source for toyatl.dll
//

// This file will be processed by the MIDL tool to
// produce the type library (toyatl.tlb) and marshalling code.

import "oaidl.idl";
import "ocidl.idl";

 [
  object,
  uuid(565B5BFD-0CB2-11D5-9E52-0060B0EB1D67),
  dual,
  helpstring("IToy Interface"),
  pointer_default(unique)
 ]
 interface IToy : IDispatch
 {
  [id(1), helpstring("method getDict")] HRESULT getDict([out]IDispatch**
pDictionary);
 };

[
 uuid(565B5BF1-0CB2-11D5-9E52-0060B0EB1D67),
 version(1.0),
 helpstring("Toy ATL project TLB")
]
library TOYATLLib
{
 importlib("stdole32.tlb");
 importlib("stdole2.tlb");

 [
  uuid(565B5BFE-0CB2-11D5-9E52-0060B0EB1D67),
  helpstring("Toy Class")
 ]
 coclass Toy
 {
  [default] interface IToy;
 };
};

and the 'obvious' implementation (FinalConstruct of the CToy class builds an
instance member that's a CComPtr<IDispatch> with a CoCreateInstance on
ProgId "Scripting.Dictionary"; method getDisp just calls a CopyTo on that
member).  After a makepy...:

D:\pybridge>python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> from win32com.client import Dispatch as disp
>>> x=disp("toyatl.Toy")
>>> a=x.getDict()
>>> a
<win32com.gen_py.clareg 1.0 Type Library.IDictionary>
>>>


Alex







From objectway at divalsim.it  Tue Feb 27 12:57:45 2001
From: objectway at divalsim.it (Nicola Musatti)
Date: Tue, 27 Feb 2001 12:57:45 +0100
Subject: Rule based programming in Python?
Message-ID: <3A9B9639.3417D696@divalsim.it>

Hi,
is anybody aware of any library/framework for rule based programming in
Python?

Thank you,
Nicola Musatti


From sandipan at vsnl.com  Tue Feb 27 14:21:10 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Tue, 27 Feb 2001 18:51:10 +0530
Subject: Rule based programming in Python?
References: <3A9B9639.3417D696@divalsim.it>
Message-ID: <017d01c0a0c0$26ee72a0$020000c3@node02>

I believe tuples in Python are the closest to PROLOG kind of rule constructs
around. You could look at tuple operations to see whether the library you
seek is available, or needs to be built.

Regards,

Sandipan
----- Original Message -----
From: "Nicola Musatti" <objectway at divalsim.it>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Tuesday, February 27, 2001 5:27 PM
Subject: Rule based programming in Python?


> Hi,
> is anybody aware of any library/framework for rule based programming in
> Python?
>
> Thank you,
> Nicola Musatti
> --
> http://mail.python.org/mailman/listinfo/python-list
>





From hgg9140 at seanet.com  Wed Feb 28 01:14:22 2001
From: hgg9140 at seanet.com (Harry George)
Date: 27 Feb 2001 16:14:22 -0800
Subject: Rule based programming in Python?
References: <3A9B9639.3417D696@divalsim.it>
Message-ID: <m3bsrnackh.fsf@wilma.localdomain>

We do it the otehr way around: A C/C++ app with embedded prolog, and
then python embedded in that.  We are considering the reverse (python
with prolog and C extensions).  You might look at SWI as a candidate
prolog for embedding.

Nicola Musatti <objectway at divalsim.it> writes:

> Hi,
> is anybody aware of any library/framework for rule based programming in
> Python?
> 
> Thank you,
> Nicola Musatti

-- 
Harry George
hgg9140 at seanet.com


From gwilder at best.com  Wed Feb 28 05:36:59 2001
From: gwilder at best.com (Glen Wilder)
Date: Wed, 28 Feb 2001 04:36:59 GMT
Subject: Rule based programming in Python?
References: <3A9B9639.3417D696@divalsim.it> <m3bsrnackh.fsf@wilma.localdomain>
Message-ID: <3A9C8599.FD6C3375@best.com>

I have already embedded SWI-Prolog in Python.  It can be retrieved from
CVS at http://sourceforge.net/projects/pyprolog.  It is pre-alpha, but I
know it works on FreeBSD.

Harry George wrote:
> 
> We do it the otehr way around: A C/C++ app with embedded prolog, and
> then python embedded in that.  We are considering the reverse (python
> with prolog and C extensions).  You might look at SWI as a candidate
> prolog for embedding.
> 
> Nicola Musatti <objectway at divalsim.it> writes:
> 
> > Hi,
> > is anybody aware of any library/framework for rule based programming in
> > Python?
> >
> > Thank you,
> > Nicola Musatti
> 
> --
> Harry George
> hgg9140 at seanet.com

-- Glen Wilder


From hgg9140 at cola.ca.boeing.com  Wed Feb 28 15:17:49 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Wed, 28 Feb 2001 14:17:49 GMT
Subject: Rule based programming in Python?
References: <3A9B9639.3417D696@divalsim.it> <m3bsrnackh.fsf@wilma.localdomain> <3A9C8599.FD6C3375@best.com>
Message-ID: <xqxlmqqsxgy.fsf@cola.ca.boeing.com>

Great!  But the website says no code released yet.  What's the status?

Glen Wilder <gwilder at best.com> writes:

> I have already embedded SWI-Prolog in Python.  It can be retrieved from
> CVS at http://sourceforge.net/projects/pyprolog.  It is pre-alpha, but I
> know it works on FreeBSD.
> 
> Harry George wrote:
> > 
> > We do it the otehr way around: A C/C++ app with embedded prolog, and
> > then python embedded in that.  We are considering the reverse (python
> > with prolog and C extensions).  You might look at SWI as a candidate
> > prolog for embedding.
> > 
> > Nicola Musatti <objectway at divalsim.it> writes:
> > 
> > > Hi,
> > > is anybody aware of any library/framework for rule based programming in
> > > Python?
> > >
> > > Thank you,
> > > Nicola Musatti
> > 
> > --
> > Harry George
> > hgg9140 at seanet.com
> 
> -- Glen Wilder

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From gwilder at best.com  Wed Feb 28 20:52:10 2001
From: gwilder at best.com (Glen Wilder)
Date: Wed, 28 Feb 2001 19:52:10 GMT
Subject: Rule based programming in Python?
References: <3A9B9639.3417D696@divalsim.it> <m3bsrnackh.fsf@wilma.localdomain> <3A9C8599.FD6C3375@best.com> <xqxlmqqsxgy.fsf@cola.ca.boeing.com>
Message-ID: <3A9D5C04.70C9D3DD@best.com>

There are a few things that need to be done before I can make an alpha
release: check for memory leaks, add threading release/lock code, and
check for insecure code.  It would also be nice to test it on systems
other that FreeBSD 4.1. (A testsuite is included.)

I am not working on any of this right now, but I will accept patches.

Harry George wrote:
> 
> Great!  But the website says no code released yet.  What's the status?
> 
> Glen Wilder <gwilder at best.com> writes:
> 
> > I have already embedded SWI-Prolog in Python.  It can be retrieved from
> > CVS at http://sourceforge.net/projects/pyprolog.  It is pre-alpha, but I
> > know it works on FreeBSD.
> >
> > Harry George wrote:
> > >
> > > We do it the otehr way around: A C/C++ app with embedded prolog, and
> > > then python embedded in that.  We are considering the reverse (python
> > > with prolog and C extensions).  You might look at SWI as a candidate
> > > prolog for embedding.


-- Glen Wilder


From marc.vollmer at mahr.de  Tue Feb 27 15:07:35 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Tue, 27 Feb 2001 15:07:35 +0100
Subject: Read INI-File
Message-ID: <97gcng$oquj3$1@ID-59219.news.dfncis.de>

Hello,

I want to read information from an INI-File. But it gives me always the
default value. Why?

>>> import win32ui
>>> win32ui.SetProfileFileName('test.ini')
>>> k = win32ui.GetProfileVal('Move','MoveSpeed',0)
>>> k
0
>>>

test.ini
---------------------------------------------------------
[MOVE]
MoveSpeed=150


Thanks
Marc Vollmer




From fgeiger at datec.at  Tue Feb 27 15:22:30 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Tue, 27 Feb 2001 15:22:30 +0100
Subject: Read INI-File
References: <97gcng$oquj3$1@ID-59219.news.dfncis.de>
Message-ID: <97gd97$931$1@rex.ip-plus.net>

IIRC you have to set the full file path or at least "./filename.ext" if it
is in the current dir. Check if "test.ini" is where you expect it to be
after a write access.

Regards
Franz


"Marc Vollmer" <marc.vollmer at mahr.de> wrote in message
news:97gcng$oquj3$1 at ID-59219.news.dfncis.de...
> Hello,
>
> I want to read information from an INI-File. But it gives me always the
> default value. Why?
>
> >>> import win32ui
> >>> win32ui.SetProfileFileName('test.ini')
> >>> k = win32ui.GetProfileVal('Move','MoveSpeed',0)
> >>> k
> 0
> >>>
>
> test.ini
> ---------------------------------------------------------
> [MOVE]
> MoveSpeed=150
>
>
> Thanks
> Marc Vollmer
>
>




From marc.vollmer at mahr.de  Tue Feb 27 16:18:58 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Tue, 27 Feb 2001 16:18:58 +0100
Subject: Read INI-File
References: <97gcng$oquj3$1@ID-59219.news.dfncis.de> <97gd97$931$1@rex.ip-plus.net>
Message-ID: <97ggqk$p53ii$1@ID-59219.news.dfncis.de>

No chance:

I try:

win32ui.SetProfileFileName('c:\winnt\test.ini')

and after copy in the application folder:

win32ui.SetProfileFileName('./test.ini')

Always it gives me the default value. Exist a functions to test, if Python
find the ini file?
Because when I print the return value from win32ui.SetProfileFileName, the
result is NONE.

Regards
Marc


"Franz GEIGER" <fgeiger at datec.at> schrieb im Newsbeitrag
news:97gd97$931$1 at rex.ip-plus.net...
> IIRC you have to set the full file path or at least "./filename.ext" if it
> is in the current dir. Check if "test.ini" is where you expect it to be
> after a write access.
>
> Regards
> Franz
>
>
> "Marc Vollmer" <marc.vollmer at mahr.de> wrote in message
> news:97gcng$oquj3$1 at ID-59219.news.dfncis.de...
> > Hello,
> >
> > I want to read information from an INI-File. But it gives me always the
> > default value. Why?
> >
> > >>> import win32ui
> > >>> win32ui.SetProfileFileName('test.ini')
> > >>> k = win32ui.GetProfileVal('Move','MoveSpeed',0)
> > >>> k
> > 0
> > >>>
> >
> > test.ini
> > ---------------------------------------------------------
> > [MOVE]
> > MoveSpeed=150
> >
> >
> > Thanks
> > Marc Vollmer
> >
> >
>
>




From ws-news at gmx.at  Tue Feb 27 16:43:51 2001
From: ws-news at gmx.at (Werner Schiendl)
Date: Tue, 27 Feb 2001 16:43:51 +0100
Subject: Read INI-File
References: <97gcng$oquj3$1@ID-59219.news.dfncis.de> <97gd97$931$1@rex.ip-plus.net> <97ggqk$p53ii$1@ID-59219.news.dfncis.de>
Message-ID: <983288899.644740@newsmaster-04.atnet.at>

Marc Vollmer <marc.vollmer at mahr.de> wrote in message
news:97ggqk$p53ii$1 at ID-59219.news.dfncis.de...
> No chance:
>
> I try:
>
> win32ui.SetProfileFileName('c:\winnt\test.ini')

If you _really_ wrote it this way it will not work.
The backslashes are used to escape the values.

Use    'c:\\winnt\\test.ini'
or    r'c:\winnt\test.ini'

hth
werner




From marc.vollmer at mahr.de  Wed Feb 28 08:36:09 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Wed, 28 Feb 2001 08:36:09 +0100
Subject: Read INI-File
References: <97gcng$oquj3$1@ID-59219.news.dfncis.de> <97gd97$931$1@rex.ip-plus.net> <97ggqk$p53ii$1@ID-59219.news.dfncis.de> <983288899.644740@newsmaster-04.atnet.at>
Message-ID: <97ia2t$p9da2$1@ID-59219.news.dfncis.de>

Thanks,

but no solution, perhaps my computer don't understand me.

So, I test it with the system.ini in the c:\winnt\.

Please test it, too. Perhaps it is a windows 2k problem? I don't know, what
I still could do.

--
import win32ui
win32ui.SetProfileFileName('C:\\WINNT\\system.ini')
k=win32ui.GetProfileVal('drivers','timer','xxx')
print k
--

My result is: xxx and yours?

Regards
Marc


"Werner Schiendl" <ws-news at gmx.at> schrieb im Newsbeitrag
news:983288899.644740 at newsmaster-04.atnet.at...
>
> Marc Vollmer <marc.vollmer at mahr.de> wrote in message
> news:97ggqk$p53ii$1 at ID-59219.news.dfncis.de...
> > No chance:
> >
> > I try:
> >
> > win32ui.SetProfileFileName('c:\winnt\test.ini')
>
> If you _really_ wrote it this way it will not work.
> The backslashes are used to escape the values.
>
> Use    'c:\\winnt\\test.ini'
> or    r'c:\winnt\test.ini'
>
> hth
> werner
>
>




From schaffer at optonline.net  Wed Feb 28 23:30:24 2001
From: schaffer at optonline.net (Les Schaffer)
Date: Wed, 28 Feb 2001 22:30:24 GMT
Subject: Read INI-File
References: <97gcng$oquj3$1@ID-59219.news.dfncis.de> <97gd97$931$1@rex.ip-plus.net> <97ggqk$p53ii$1@ID-59219.news.dfncis.de> <983288899.644740@newsmaster-04.atnet.at> <97ia2t$p9da2$1@ID-59219.news.dfncis.de>
Message-ID: <hsuq9toftrs0o4be6mll4v4csdcnp23cqf@4ax.com>

On Wed, 28 Feb 2001 08:36:09 +0100, "Marc Vollmer"
<marc.vollmer at mahr.de> wrote:


>Please test it, too. Perhaps it is a windows 2k problem? 

under win2k (python scriptname.py):

   timer.drv

under pythonwin (executing script):

   xxx


les schaffer


From nobody at home.net  Wed Feb 28 23:34:04 2001
From: nobody at home.net (Bernard Delmée)
Date: Wed, 28 Feb 2001 23:34:04 +0100
Subject: Read INI-File
References: <97gcng$oquj3$1@ID-59219.news.dfncis.de> <97gd97$931$1@rex.ip-plus.net> <97ggqk$p53ii$1@ID-59219.news.dfncis.de> <983288899.644740@newsmaster-04.atnet.at> <97ia2t$p9da2$1@ID-59219.news.dfncis.de>
Message-ID: <97jub8$1th5$1@rivage.news.be.easynet.net>

Your example works fine for me, on W2K, (prints 'timer.drv')
ActivePython 2.0, build 202

I tried both SetProfileFileName('C:\\WINNT\\system.ini')
and SetProfileFileName('C:/WINNT/system.ini')


"Marc Vollmer" <marc.vollmer at mahr.de> wrote in message
news:97ia2t$p9da2$1 at ID-59219.news.dfncis.de...
> Thanks,
>
> but no solution, perhaps my computer don't understand me.
>
> So, I test it with the system.ini in the c:\winnt\.
>
> Please test it, too. Perhaps it is a windows 2k problem? I don't know,
what
> I still could do.
>
> --
> import win32ui
> win32ui.SetProfileFileName('C:\\WINNT\\system.ini')
> k=win32ui.GetProfileVal('drivers','timer','xxx')
> print k
> --
>
> My result is: xxx and yours?
>
> Regards
> Marc
>
>
> "Werner Schiendl" <ws-news at gmx.at> schrieb im Newsbeitrag
> news:983288899.644740 at newsmaster-04.atnet.at...
> >
> > Marc Vollmer <marc.vollmer at mahr.de> wrote in message
> > news:97ggqk$p53ii$1 at ID-59219.news.dfncis.de...
> > > No chance:
> > >
> > > I try:
> > >
> > > win32ui.SetProfileFileName('c:\winnt\test.ini')
> >
> > If you _really_ wrote it this way it will not work.
> > The backslashes are used to escape the values.
> >
> > Use    'c:\\winnt\\test.ini'
> > or    r'c:\winnt\test.ini'
> >
> > hth
> > werner
> >
> >
>
>




From marc.vollmer at mahr.de  Tue Feb 27 15:09:03 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Tue, 27 Feb 2001 15:09:03 +0100
Subject: Read INI-File?
Message-ID: <97gcni$oquj3$2@ID-59219.news.dfncis.de>

Hello,

I want to read information from an INI-File. But it gives me always the
default value. Why?

>>> import win32ui
>>> win32ui.SetProfileFileName('test.ini')
>>> k = win32ui.GetProfileVal('Move','MoveSpeed',0)
>>> k
0
>>>

test.ini
---------------------------------------------------------
[MOVE]
MoveSpeed=150


Thanks
Marc Vollmer






From guido-dot-treutwein-at-nbg-dot-siemens-dot-de.cut-here at no.spam  Tue Feb 27 16:07:07 2001
From: guido-dot-treutwein-at-nbg-dot-siemens-dot-de.cut-here at no.spam (guido-dot-treutwein-at-nbg-dot-siemens-dot-de.cut-here at no.spam)
Date: Tue, 27 Feb 2001 16:07:07 +0100
Subject: Combined error message for user defined exceptions
Message-ID: <3A9BC29B.BE9E0C96@no.spam>

Hi there,

I finally converted my exceptions from string-based to class-based, but
I'm not yet happy with the information in the error traceback.

What I did:

class excMyError(TypeError):
    def __str__(self):
        return "My generic error"
#... and later ...
raise excMyError, "additional info"

What I want is to merge the generic exception text with the additional
information just like

NameError: There is no variable named 'undefvar'

where 'undefvar' corresponds to "additionalInfo" and the rest to the
generic information.
The obvious way (adding a parameter to __str__) failed, what to do else?

Regards


From stephen_purcell at yahoo.com  Tue Feb 27 17:00:12 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Tue, 27 Feb 2001 17:00:12 +0100
Subject: Combined error message for user defined exceptions
In-Reply-To: <3A9BC29B.BE9E0C96@no.spam>; from guido-dot-treutwein-at-nbg-dot-siemens-dot-de.cut-here@no.spam on Tue, Feb 27, 2001 at 04:07:07PM +0100
References: <3A9BC29B.BE9E0C96@no.spam>
Message-ID: <20010227170012.A21016@freedom.puma-ag.com>

guido-dot-treutwein-at-nbg-dot-siemens-dot-de.cut-here at no.spam wrote:
> class excMyError(TypeError):
>     def __str__(self):
>         return "My generic error"
> #... and later ...
> raise excMyError, "additional info"
> 
> What I want is to merge the generic exception text with the additional
> information just like
> 
> NameError: There is no variable named 'undefvar'

There may be nicer ways to do it, but if you promise to always throw the
exception with only one argument, you can do the following:-


>>> class MyException(TypeError):
...     def __str__(self):
...        return "there is no variable named '%s'" % self.args[0]
... 
>>> raise MyException, 'foobar'
Traceback (innermost last):
  File "<stdin>", line 1, in ?
__main__.MyException: there is no variable named 'foobar'
>>> 


But even more simply, you can set the message exactly in the place where you
throw the exception, which is often the most appropriate option:


>>> class MyException(TypeError): pass
... 
>>> raise MyException, "no such variable 'foobar'"
Traceback (innermost last):
  File "<stdin>", line 1, in ?
__main__.MyException: no such variable 'foobar'
>>> 


Best wishes,

-Steve


-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From lbates at syscon-computers.com  Tue Feb 27 16:12:18 2001
From: lbates at syscon-computers.com (Larry Bates)
Date: Tue, 27 Feb 2001 09:12:18 -0600
Subject: Raw string question
Message-ID: <3a9bc4f5$1@news1.dbtech.net>

Why can't a raw string end with a backslash (\)?

example

a=r"c:\path\"

doesn't work

a=r"c:\path"

works fine

Thanks, Larry


BTW - I'm using ActiveState PythonWin (if makes any difference).






From emile at fenx.com  Tue Feb 27 17:20:42 2001
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 27 Feb 2001 08:20:42 -0800
Subject: Raw string question
References: <3a9bc4f5$1@news1.dbtech.net>
Message-ID: <97gk6m$p4tnj$1@ID-11957.news.dfncis.de>

It can, you just need to escape it if it's the last
character:

a=r"c:\path\\"


--

Emile van Sebille
emile at fenx.com
-------------------


"Larry Bates" <lbates at syscon-computers.com> wrote in message
news:3a9bc4f5$1 at news1.dbtech.net...
> Why can't a raw string end with a backslash (\)?
>
> example
>
> a=r"c:\path\"
>
> doesn't work
>
> a=r"c:\path"
>
> works fine
>
> Thanks, Larry
>
>
> BTW - I'm using ActiveState PythonWin (if makes any
difference).
>
>
>
>




From robin.thomas at starmedia.net  Tue Feb 27 17:50:50 2001
From: robin.thomas at starmedia.net (Robin Thomas)
Date: Tue, 27 Feb 2001 11:50:50 -0500
Subject: Raw string question
In-Reply-To: <3a9bc4f5$1@news1.dbtech.net>
Message-ID: <4.3.1.2.20010227112407.00c70f00@exchange.starmedia.net>

At 09:12 AM 2/27/01 -0600, Larry Bates wrote:
>Why can't a raw string end with a backslash (\)?
>
>example
>
>a=r"c:\path\"

Below from Python Language Reference 2.4.1. The text explains the behavior 
and also gives you good hints about how raw string literals are parsed in 
the current implementation.

"""
Unlike Standard C, all unrecognized escape sequences are left in the string 
unchanged, i.e., the backslash is left in the string. (This behavior is 
useful when debugging: if an escape sequence is mistyped, the resulting 
output is more easily recognized as broken.)

When an `r' or `R' prefix is present, backslashes are still used to quote 
the following character, but all backslashes are left in the string. For 
example, the string literal r"\n" consists of two characters: a backslash 
and a lowercase `n'. String quotes can be escaped with a backslash, but the 
backslash remains in the string; for example, r"\"" is a valid string 
literal consisting of two characters: a backslash and a double quote; r"\" 
is not a value string literal (even a raw string cannot end in an odd 
number of backslashes). Specifically, a raw string cannot end in a single 
backslash (since the backslash would escape the following quote character). 
Note also that a single backslash followed by a newline is interpreted as 
those two characters as part of the string, not as a line continuation.
"""


--
Robin Thomas
Engineering
StarMedia Network, Inc.
robin.thomas at starmedia.net




From insanc at cc.gatech.edu  Tue Feb 27 16:26:40 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 27 Feb 2001 15:26:40 GMT
Subject: shadow.h
Message-ID: <97ggvg$mqf$1@solaria.cc.gatech.edu>

is there anyway to be able to use the functions from this header file?
i tried the h2py and using that with the header but that did not seem to 
work. any other suggestions?

-- 
---
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From insanc at cc.gatech.edu  Wed Feb 28 21:07:14 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 28 Feb 2001 20:07:14 GMT
Subject: shadow.h
References: <97ggvg$mqf$1@solaria.cc.gatech.edu>
Message-ID: <97jlpi$6os$1@solaria.cc.gatech.edu>

Joseph Holland King <insanc at cc.gatech.edu> wrote:

i have searched the python web site including the starship page and the 
vaults of parnassus page, to no avail. i have not been able to find a 
port or any other information that would help me to use the c function
lckpwdf. i am really close to just writing it myself. does anyone know 
of anywhere else to look or any way that is easier to convert the c 
function to a python module? if not do you have any warnings or 
comments that i should heed before attempting to write (if you can't 
tell i am not all that enthusiastic about this idea.) thank you.

-- 
---
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From fluxent at yahoo.com  Tue Feb 27 16:41:19 2001
From: fluxent at yahoo.com (fluxent at yahoo.com)
Date: Tue, 27 Feb 2001 15:41:19 -0000
Subject: approach for assigning ODBC results to fieldnames as variables?
Message-ID: <97ghqv+6tge@eGroups.com>

When you execute a SQL (select) query via ODBC and do a fetchall(), 
you get back a result set as a tuple (array). You can also ask for 
the cursor's description, which returns a separate tuple, the first 
column of which contains the fieldnames (or other labels calculated 
by SQL).

Getting back an unlabelled tuple means that when you want to write 
out a table of results (with some individual handling of various 
fields, not just dumbly iterating over the columns), you end up with 
an ugly line like:
  Response.Write('<tr><td>%s<td>%s<td>%s<td><a 
href="matterEdit.asp?clientID=%s&matterID=%s">%s</a><td>%s<td>%
s</tr>' % (row[0],row[1],row[2],row[1],row[3],row[3],row[4],row[5]))

(this is a Python-via-IIS model, but would be similar under Apache, 
etc.).

Is there a standard approach to assigning field names to columns, so 
that this output line could be changed to something like
  Response.Write('<tr><td>%s<td>%s<td>%s<td><a 
href="matterEdit.asp?clientID=%s&matterID=%s">%s</a><td>%s<td>%
s</tr>' % (maxDate, clientID, clientName, clientID, jobID, jobID, 
jobName, totalHours))

or even at least something like
  Response.Write('<tr><td>%s<td>%s<td>%s<td><a href="matterEdit.asp?
clientID=%s&matterID=%s">%s</a><td>%s<td>%s</tr>' % (maxDate[i], 
clientID[i], clientName[i], clientID[i], jobID[i], jobID[i], jobName
[i], totalHours[i]))

thx for your help...





From robin.thomas at starmedia.net  Tue Feb 27 19:54:30 2001
From: robin.thomas at starmedia.net (Robin Thomas)
Date: Tue, 27 Feb 2001 13:54:30 -0500
Subject: approach for assigning ODBC results to fieldnames as
  variables?
In-Reply-To: <97ghqv+6tge@eGroups.com>
Message-ID: <4.3.1.2.20010227131547.00c46240@exchange.starmedia.net>

At 03:41 PM 2/27/01 +0000, fluxent at yahoo.com wrote:
>When you execute a SQL (select) query via ODBC and do a fetchall(),
>you get back a result set as a tuple (array). You can also ask for
>the cursor's description, which returns a separate tuple, the first
>column of which contains the fieldnames (or other labels calculated
>by SQL).

A common approach is to encapsulate the cursor so that the description is 
fetched once, and rows are turned into dictionaries with the values in 
description tuple used as keys.


>Is there a standard approach to assigning field names to columns, so
>that this output line could be changed to something like
>   Response.Write('<tr><td>%s<td>%s<td>%s<td><a
>href="matterEdit.asp?clientID=%s&matterID=%s">%s</a><td>%s<td>%
>s</tr>' % (maxDate, clientID, clientName, clientID, jobID, jobID,
>jobName, totalHours))

Dynamic name binding can be very problematic. Dictionaries give you the 
desired behavior without the name-binding mess. Especially when you use 
named placeholders in string formatting. A shortened example:

Response.Write('<tr><td>%(MAXDATE)s' % row_as_dictionary)

Be careful of the case of the column names returned by description().

>or even at least something like
>   Response.Write('<tr><td>%s<td>%s<td>%s<td><a href="matterEdit.asp?
>clientID=%s&matterID=%s">%s</a><td>%s<td>%s</tr>' % (maxDate[i],
>clientID[i], clientName[i], clientID[i], jobID[i], jobID[i], jobName
>[i], totalHours[i]))

Slicing the row tuples into new column value sequences is, happily, no 
faster than making dictionaries using the description tuple. Test script is 
below.

# script start
def get_description():
     return ('ONE', 'TWO', 'THREE', 'FOUR', 'FIVE')

def get_row():
     return (1,2,3,4,5)

def rows_as_dicts(rows, description):
     if not rows: return []
     length = len(rows[0])
     r = range(length)
     l = []
     for row in rows:
         d = {}
         for i in r:
             d[description[i]] = row[i]
         l.append(d)
     return l

def rows_as_columns(rows, description):
     if not rows: return []
     length = len(rows[0])
     r = range(length)
     l = []
     for i in r: l.append([])
     for row in rows:
         for i in r:
             l[i].append(row[i])
     return l

def test(funcs, rows=20, reps=1000):
     desc = get_description()
     rowlist = []
     for i in range(rows):
         rowlist.append(get_row())

     from time import time
     for func in funcs:
         n = func.__name__
         t = time()
         for i in range(reps):
             func(rowlist,desc)
         t = time() - t
         print "%s on %d rows: %s reps in %.04f s (%.05f s each)" % \
               (n, rows, reps, t, t/reps)

if __name__ == '__main__':
     for rowamt in (10,20,50,100):
         test( (rows_as_dicts, rows_as_columns), rowamt )

# output
rows_as_dicts on 10 rows: 1000 reps in 0.4665 s (0.00047 s each)
rows_as_columns on 10 rows: 1000 reps in 0.5636 s (0.00056 s each)
rows_as_dicts on 20 rows: 1000 reps in 0.9083 s (0.00091 s each)
rows_as_columns on 20 rows: 1000 reps in 1.0523 s (0.00105 s each)
rows_as_dicts on 50 rows: 1000 reps in 2.5213 s (0.00252 s each)
rows_as_columns on 50 rows: 1000 reps in 2.5186 s (0.00252 s each)
rows_as_dicts on 100 rows: 1000 reps in 4.5625 s (0.00456 s each)
rows_as_columns on 100 rows: 1000 reps in 5.0504 s (0.00505 s each)


--
Robin Thomas
Engineering
StarMedia Network, Inc.
robin.thomas at starmedia.net




From lbates at syscon-computers.com  Tue Feb 27 17:03:42 2001
From: lbates at syscon-computers.com (Larry Bates)
Date: Tue, 27 Feb 2001 10:03:42 -0600
Subject: Determining disk space (problem on large disks)
Message-ID: <3a9bd102@news1.dbtech.net>

I tried your example and apparently win32file needs to be updated to use
longints.  On disk drives with more than 2Gb of space it doesn't work
properly (or am I missing something)?

-Larry




From jeremy at alum.mit.edu  Tue Feb 27 17:13:22 2001
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: Tue, 27 Feb 2001 11:13:22 -0500 (EST)
Subject: Lightning Talks session at Python Conference
Message-ID: <15003.53794.7734.331509@w221.z064000254.bwi-md.dsl.cnc.net>

The Lightning Talks session provides an opportunity for conference
attendees to present new or ongoing work and cool ideas.  Each talk is
five to seven minutes long -- time enough for a one good idea and a
few questions.  In past years, this session has always been fun.

If you would like to give a lightning talk, please send email to
lightning-talks at python9.org.  Speaker slots are available on a
first-come, first-served basis.

Jeremy Hylton
Lightning Talks session chair



From dg at pearl.tao.co.uk  Tue Feb 27 18:39:28 2001
From: dg at pearl.tao.co.uk (David Given)
Date: Tue, 27 Feb 2001 17:39:28 +0000
Subject: Python 2 for Debian?
Message-ID: <goog79.dvf.ln@127.0.0.1>

Subject says it all, really. Does anyone know when Python 2 is coming out
for Debian? All I can find on my local mirror is 1.5.

-- 
+- David Given ---------------McQ-+ 
|  Work: dg at tao-group.com         | ACRONYM: A Contrived Reduction of
|  Play: dgiven at iname.com         | Nomenclature Yielding Mnemonics            
+- http://wired.st-and.ac.uk/~dg -+ 


From cookedm+news at physics.mcmaster.ca  Tue Feb 27 20:05:31 2001
From: cookedm+news at physics.mcmaster.ca (David M. Cooke)
Date: 27 Feb 2001 14:05:31 -0500
Subject: Python 2 for Debian?
References: <goog79.dvf.ln@127.0.0.1>
Message-ID: <qnku25gymis.fsf@arbutus.physics.mcmaster.ca>

At some point, dg at pearl.tao.co.uk (David Given) wrote:

> Subject says it all, really. Does anyone know when Python 2 is coming out
> for Debian? All I can find on my local mirror is 1.5.

Install the python2-* packages from unstable (they haven't made it
into testing yet, unfortunately).

-- 
|>|\/|<
/--------------------------------------------------------------------------\
|David M. Cooke
|cookedm(at)physics(dot)mcmaster(dot)ca


From dg at pearl.tao.co.uk  Wed Feb 28 11:36:22 2001
From: dg at pearl.tao.co.uk (David Given)
Date: Wed, 28 Feb 2001 10:36:22 +0000
Subject: Python 2 for Debian?
References: <goog79.dvf.ln@127.0.0.1>
    <qnku25gymis.fsf@arbutus.physics.mcmaster.ca>
Message-ID: <6bki79.asq.ln@127.0.0.1>

In article <qnku25gymis.fsf at arbutus.physics.mcmaster.ca>,
	cookedm+news at physics.mcmaster.ca (David M. Cooke) writes:
> At some point, dg at pearl.tao.co.uk (David Given) wrote:
> 
>> Subject says it all, really. Does anyone know when Python 2 is coming out
>> for Debian? All I can find on my local mirror is 1.5.
> 
> Install the python2-* packages from unstable (they haven't made it
> into testing yet, unfortunately).

Ah, right. I'm doing this on my work machine, so they probably wouldn't
thank me for putting unstable on it, so I'm sticking with testing. Looks
like I'm going to have to wait.

-- 
+- David Given ---------------McQ-+ 
|  Work: dg at tao-group.com         | ACRONYM: A Contrived Reduction of
|  Play: dgiven at iname.com         | Nomenclature Yielding Mnemonics            
+- http://wired.st-and.ac.uk/~dg -+ 


From gregor at mediasupervision.de  Wed Feb 28 16:21:53 2001
From: gregor at mediasupervision.de (Gregor Hoffleit)
Date: Wed, 28 Feb 2001 16:21:53 +0100
Subject: Python 2 for Debian?
In-Reply-To: <6bki79.asq.ln@127.0.0.1>; from dg@pearl.tao.co.uk on Wed, Feb 28, 2001 at 10:36:22AM +0000
References: <goog79.dvf.ln@127.0.0.1> <qnku25gymis.fsf@arbutus.physics.mcmaster.ca> <6bki79.asq.ln@127.0.0.1>
Message-ID: <20010228162153.C1925@mediasupervision.de>

On Wed, Feb 28, 2001 at 10:36:22AM +0000, David Given wrote:
> In article <qnku25gymis.fsf at arbutus.physics.mcmaster.ca>,
> 	cookedm+news at physics.mcmaster.ca (David M. Cooke) writes:
> > At some point, dg at pearl.tao.co.uk (David Given) wrote:
> > 
> >> Subject says it all, really. Does anyone know when Python 2 is coming out
> >> for Debian? All I can find on my local mirror is 1.5.
> > 
> > Install the python2-* packages from unstable (they haven't made it
> > into testing yet, unfortunately).
> 
> Ah, right. I'm doing this on my work machine, so they probably wouldn't
> thank me for putting unstable on it, so I'm sticking with testing. Looks
> like I'm going to have to wait.

The python2-* packages are currently stuck in unstable, since they depend on
tcl/tk8.3. tcl/tk8.3 in turn currently has a severe bug, which prevents it
from being moved to testing. That's the crux with package pools ;-)

You could either fetch the python2 source ('apt-get source python2' with
unstable in sources.list) and recompile it, changing Modules/Setup.in
accordingly to use tk8.0 or tk8.2, or you'll have to wait for the resolution
of these problems...

    Gregor
    



From shaleh at valinux.com  Tue Feb 27 20:11:38 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Tue, 27 Feb 2001 11:11:38 -0800 (PST)
Subject: Python 2 for Debian?
In-Reply-To: <goog79.dvf.ln@127.0.0.1>
Message-ID: <XFMail.20010227111138.shaleh@valinux.com>

On 27-Feb-2001 David Given wrote:
> Subject says it all, really. Does anyone know when Python 2 is coming out
> for Debian? All I can find on my local mirror is 1.5.
> 

apt-get install python2-base.  It lives in unstable and testing.

Until the licensing gets straightened out, most python modules are not being
recompiled under Debian for python2.  So you can play with the language, but
some of the useful things are missing.



From garabik-news at spam.melkor.dnp.fmph.uniba.sk  Tue Feb 27 21:58:22 2001
From: garabik-news at spam.melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: 27 Feb 2001 20:58:22 GMT
Subject: Python 2 for Debian?
References: <goog79.dvf.ln@127.0.0.1>
Message-ID: <983307400.692337@cdwork.cvt.stuba.sk>

David Given <dg at pearl.tao.co.uk> wrote:
 : Subject says it all, really. Does anyone know when Python 2 is coming out
 : for Debian? All I can find on my local mirror is 1.5.

has been already there for some time
time to change your mirror :-)

-- 
 -----------------------------------------------------------
| Radovan Garabik http://melkor.dnp.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__    garabik @ melkor.dnp.fmph.uniba.sk     |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From avv at quasar.ipa.nw.ru  Tue Feb 27 22:02:30 2001
From: avv at quasar.ipa.nw.ru (Alexander V. Voinov)
Date: Tue, 27 Feb 2001 13:02:30 -0800
Subject: Python 2 for Debian?
References: <mailman.983301069.23270.python-list@python.org>
Message-ID: <3A9C15E6.F2099AD4@quasar.ipa.nw.ru>

Hi

Sean 'Shaleh' Perry wrote:
> 
> On 27-Feb-2001 David Given wrote:
> > Subject says it all, really. Does anyone know when Python 2 is coming out
> > for Debian? All I can find on my local mirror is 1.5.
> >
> 
> apt-get install python2-base.  It lives in unstable and testing.
> 
> Until the licensing gets straightened out, most python modules are not being
> recompiled under Debian for python2.  So you can play with the language, but
> some of the useful things are missing.

Why not to put them into [non-free]? Who would mind?

Alexander


From kern at myrddin.caltech.edu  Tue Feb 27 23:10:51 2001
From: kern at myrddin.caltech.edu (Robert Kern)
Date: 27 Feb 2001 22:10:51 GMT
Subject: Python 2 for Debian?
References: <mailman.983301069.23270.python-list@python.org>
    <3A9C15E6.F2099AD4@quasar.ipa.nw.ru>
Message-ID: <97h8lb$83s@gap.cco.caltech.edu>

In article <3A9C15E6.F2099AD4 at quasar.ipa.nw.ru>,
	"Alexander V. Voinov" <avv at quasar.ipa.nw.ru> writes:
> Hi
> 
> Sean 'Shaleh' Perry wrote:
>> 
>> On 27-Feb-2001 David Given wrote:
>> > Subject says it all, really. Does anyone know when Python 2 is coming out
>> > for Debian? All I can find on my local mirror is 1.5.
>> >
>> 
>> apt-get install python2-base.  It lives in unstable and testing.
>> 
>> Until the licensing gets straightened out, most python modules are not being
>> recompiled under Debian for python2.  So you can play with the language, but
>> some of the useful things are missing.
> 
> Why not to put them into [non-free]? Who would mind?

IIRC, it's not that Python's license is restrictive; it passes the DFSG fine.
It's the fact that it conflicts with the GPL according to RMS, so distributing
the readline module binary, say, could be a GPL violation. No one wants to 
get Debian in legal trouble.

> Alexander

-- 
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 Are the graves of dreams allowed to die."
  -- Richard Harter


From embed at geocities.com  Wed Feb 28 02:01:45 2001
From: embed at geocities.com (Warren Postma)
Date: Wed, 28 Feb 2001 01:01:45 GMT
Subject: Python 2 for Debian?
References: <mailman.983301069.23270.python-list@python.org>    <3A9C15E6.F2099AD4@quasar.ipa.nw.ru> <97h8lb$83s@gap.cco.caltech.edu>
Message-ID: <Z5Ym6.2362$rV3.608995@news4.rdc1.on.home.com>

>
> IIRC, it's not that Python's license is restrictive; it passes the DFSG
fine.
> It's the fact that it conflicts with the GPL according to RMS, so
distributing
> the readline module binary, say, could be a GPL violation. No one wants to
> get Debian in legal trouble.
>

This strikes me as so much pussyfooting around a non issue.   Sheesh.

Warren




From thomas at xs4all.net  Wed Feb 28 02:20:06 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Wed, 28 Feb 2001 02:20:06 +0100
Subject: Python 2 for Debian?
In-Reply-To: <Z5Ym6.2362$rV3.608995@news4.rdc1.on.home.com>; from embed@geocities.com on Wed, Feb 28, 2001 at 01:01:45AM +0000
References: <mailman.983301069.23270.python-list@python.org> <3A9C15E6.F2099AD4@quasar.ipa.nw.ru> <97h8lb$83s@gap.cco.caltech.edu> <Z5Ym6.2362$rV3.608995@news4.rdc1.on.home.com>
Message-ID: <20010228022006.E9678@xs4all.nl>

On Wed, Feb 28, 2001 at 01:01:45AM +0000, Warren Postma wrote:

> > IIRC, it's not that Python's license is restrictive; it passes the DFSG
> > fine. It's the fact that it conflicts with the GPL according to RMS, so
> > distributing the readline module binary, say, could be a GPL violation.
> > No one wants to get Debian in legal trouble.

> This strikes me as so much pussyfooting around a non issue.   Sheesh.

But who's doing the pussyfooting ? RMS ? Debian ? CNRI ? Each one is doing
what they think is best. Note that the licence talks are still ongoing, and
a GPL-compatible 1.6.1 was released not a few days ago. Python 2.0 is
derived from 1.6 though, not 1.6.1, and also contains some BeOpen (and for
2.1, possibly some DC) specific stuff, so it requires some extra attention.
But there is still hope :)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From tim.one at home.com  Wed Feb 28 02:43:23 2001
From: tim.one at home.com (Tim Peters)
Date: Tue, 27 Feb 2001 20:43:23 -0500
Subject: Python 2 for Debian?
In-Reply-To: <20010228022006.E9678@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEJAJCAA.tim.one@home.com>

[Thomas Wouters]
> But who's doing the pussyfooting ? RMS ? Debian ? CNRI ? Each one is doing
> what they think is best. Note that the licence talks are still ongoing, and
> a GPL-compatible 1.6.1 was released not a few days ago.

Careful, there.  RMS has not said that 1.6.1 has a GPL-compatible license, so
even that part remains unclear (to us too).

> Python 2.0 is derived from 1.6 though, not 1.6.1, and also contains
> some BeOpen

Yup.

> (and for 2.1, possibly some DC)

Nope.  DC is very generously assigning their Python IP rights to the Python
Software Foundation, which doesn't yet exist <wink>.  But until it does,
they're assigning them to Guido directly.  And despite nested scopes, I still
trust Guido.

> specific stuff, so it requires some extra attention.
> But there is still hope :)

Yes, there is.  Possibly even in our lifetimes.

although-a-lifetime-remains-too-much-to-bet-on-it-ly y'rs  - tim




From embed at geocities.com  Wed Feb 28 15:37:32 2001
From: embed at geocities.com (Warren Postma)
Date: Wed, 28 Feb 2001 09:37:32 -0500
Subject: Python 2 for Debian?
References: <mailman.983301069.23270.python-list@python.org> <3A9C15E6.F2099AD4@quasar.ipa.nw.ru> <97h8lb$83s@gap.cco.caltech.edu> <Z5Ym6.2362$rV3.608995@news4.rdc1.on.home.com> <mailman.983323269.28448.python-list@python.org>
Message-ID: <g48n6.2975$TW.15474@tor-nn1.netcom.ca>

Quoth Thomas:
> But who's doing the pussyfooting ? RMS ? Debian ? CNRI ? Each one is doing
> what they think is best. Note that the licence talks are still ongoing,
and
> a GPL-compatible 1.6.1 was released not a few days ago. Python 2.0 is
> derived from 1.6 though, not 1.6.1, and also contains some BeOpen (and for
> 2.1, possibly some DC) specific stuff, so it requires some extra
attention.
> But there is still hope :)


All this "license compatibility" BS boils down to this: Until someone sues
somebody, this is all theoretical. Thus,  I stand by what I said earlier;
it's all just so much pussyfooting around, about nothing.

Where is the legal precedent? Until then, it's all speculation.  If I break
the GPL what happens? NOTHING.  Somebody sue me for a GPL violation, PLEASE.
<grin>

License writers are writing Pseudo Code for a problem domain in which the
compiler is not yet written. Are there any syntax errors in the GPL? Sorry,
we just don't know.

So why be picky about pseudo-code?

Wisely, the larger Python user community, and the larger Free Software
movement, ignores RMS, and everybody else, who opines endlessly on License
Compatibility Issues.

Licensing? Bah!


Warren




From scarblac at pino.selwerd.nl  Wed Feb 28 14:46:07 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 28 Feb 2001 13:46:07 GMT
Subject: Python 2 for Debian?
References: <mailman.983301069.23270.python-list@python.org>    <3A9C15E6.F2099AD4@quasar.ipa.nw.ru> <97h8lb$83s@gap.cco.caltech.edu>
Message-ID: <slrn99q059.kav.scarblac@pino.selwerd.nl>

Robert Kern <kern at myrddin.caltech.edu> wrote in comp.lang.python:
> IIRC, it's not that Python's license is restrictive; it passes the DFSG fine.
> It's the fact that it conflicts with the GPL according to RMS, so distributing
> the readline module binary, say, could be a GPL violation. No one wants to 
> get Debian in legal trouble.

But Python 1.6.1 was released recently, with a GPL compatible license. It
seems to have been met with silence in this newsgroup, but I hope this means
that Python 2.0 and higher can soon have a compatible license as well.

http://www.python.org/1.6.1/

(A CNRI release, so not a release by Guido's Flying Circus, I believe)

-- 
Remco Gerlich


From mjackson at wc.eso.mc.xerox.com  Wed Feb 28 15:04:21 2001
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 28 Feb 2001 14:04:21 GMT
Subject: Python 2 for Debian?
References: <slrn99q059.kav.scarblac@pino.selwerd.nl>
Message-ID: <97j0h5$s41$1@news.wrc.xerox.com>

scarblac at pino.selwerd.nl (Remco Gerlich) writes:

> But Python 1.6.1 was released recently, with a GPL compatible license.

Or not.  CNRI concedes that Stallman says "Python 1.6.1 is not
technically GPL-compatible, but unlike Python 1.6, the Python 1.6.1
license allows the creation of GPL-compatible derivative works."  So
things are, as usual, not straightforward.

> It seems to have been met with silence in this newsgroup

Except for Guido's announcement, and his remarks on the prospects for
similar changes to the licenses for 2.x.  See the c.l.py.a archives.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Having your book made into a movie is like having
	your ox made into a bouillon cube.
				- Bill Neely




From tim.one at home.com  Wed Feb 28 19:02:48 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 28 Feb 2001 13:02:48 -0500
Subject: Python 2 for Debian?
In-Reply-To: <slrn99q059.kav.scarblac@pino.selwerd.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCIELPJCAA.tim.one@home.com>

[Remco Gerlich[
> But Python 1.6.1 was released recently, with a GPL compatible license.

Please read the note in italics at the referenced link:

> http://www.python.org/1.6.1/

The FSF says 1.6.1's is *not* a GPL-compatible license, but is a necessary
step on the road to releasing a Python with a license that is GPL-compatible.
I don't understand what that means either.  You can read the license text and
guess for yourself:

    http://www.handle.net/python_licenses/python1.6.1-2-23-01.html

Clause 7 is where the Deep Magic lives; we're waiting for a clarification
about what this actually means.

or-is-claimed-to-mean-anyway-ly y'rs  - tim




From jp_reed at yahoo.com  Wed Feb 28 00:12:44 2001
From: jp_reed at yahoo.com (Jeremy Reed)
Date: Tue, 27 Feb 2001 15:12:44 -0800
Subject: Python 2 for Debian?
References: <goog79.dvf.ln@127.0.0.1>
Message-ID: <t9o2ic67nai666@corp.supernews.com>

I have just compiled and installed Python 2.0 on my linux server just last
night.  The source code and instructions for the install can be found on
www.python.org/2.0/

Best regards,

Jeremy Reed




From bryanbz at yahoo.com  Tue Feb 27 21:00:43 2001
From: bryanbz at yahoo.com (Bryan BZ)
Date: Tue, 27 Feb 2001 15:00:43 -0500
Subject: re.IGNORECASE
Message-ID: <QETm6.19375$Sl.850846@iad-read.news.verio.net>

Why is it I want to ignore the case in my search but Python refuses to
match:

"Spam" with "spam"  OR
"ni" with "Ni"

For example (a really silly and stupid one) ...

searchresult = re.search("Monty Python", "monty Python", re.IGNORECASE)
won't match...

Ugh,
BB.




From fredrik at pythonware.com  Tue Feb 27 21:20:29 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 27 Feb 2001 20:20:29 GMT
Subject: re.IGNORECASE
References: <QETm6.19375$Sl.850846@iad-read.news.verio.net>
Message-ID: <h_Tm6.15845$Qb7.2587878@newsb.telia.net>

"Bryan BZ" wrote:
> Why is it I want to ignore the case in my search but Python refuses to
> match:
>
> "Spam" with "spam"  OR
> "ni" with "Ni"

it's a bug in 2.0.  if the pattern contains uppercase
characters, IGNORECASE may not always work.

it's fixed in 2.1, and in the separate SRE distribution:
http://www.pythonware.com/products/sre/index.htm

Cheers /F

<!-- pythonworks 1.2 coming soon:
http://www.pythonware.com/products/works/index.htm
-->




From tjg at exceptionalminds.com  Tue Feb 27 23:29:51 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Tue, 27 Feb 2001 14:29:51 -0800
Subject: re.IGNORECASE
In-Reply-To: <QETm6.19375$Sl.850846@iad-read.news.verio.net>; from bryanbz@yahoo.com on Tue, Feb 27, 2001 at 03:00:43PM -0500
References: <QETm6.19375$Sl.850846@iad-read.news.verio.net>
Message-ID: <20010227142951.K29494@trufflehunter.avalongroup.net>

On Tue, Feb 27, 2001 at 03:00:43PM -0500, Bryan BZ wrote:
> Why is it I want to ignore the case in my search but Python refuses to
> match:
> 
> "Spam" with "spam"  OR
> "ni" with "Ni"
> 
> For example (a really silly and stupid one) ...
> 
> searchresult = re.search("Monty Python", "monty Python", re.IGNORECASE)
> won't match...

>>> import re
>>> searchresult = re.search("Monty Python", "monty Python",
>>> re.IGNORECASE)
>>> print searchresult
<re.MatchObject instance at 80b1808>

Worked for me. 

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.         <><       (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  42 days  2:42 hours ago<<



From jpreed00 at MailAndNews.com  Tue Feb 27 21:04:02 2001
From: jpreed00 at MailAndNews.com (Jeremy Reed)
Date: Tue, 27 Feb 2001 15:04:02 -0500
Subject: Anyone know of a way to get non-blocking keyboard input?
Message-ID: <3AA01B31@MailAndNews.com>

I am looking to create a telnet client that will, by using the select() 
statment, poll an open socket for data to be received while at the same time 
keep accepting keyboard input from the user.

I have tried to run a thread in the background that collects input, but I 
never can figure out how to achieve the 'fluidity' that I desire--i.e. no 
stopping for a return-key press.

Any ideas or help would be appreciated.  Please, in addition to posting your 
answer in this newsgroup, email me a copy of your reply to 
jp_reed at yahoo.com.

Best regards,

Jeremy Reed



From ben.hutchings at roundpoint.com  Wed Feb 28 00:54:55 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 27 Feb 2001 15:54:55 -0800
Subject: Anyone know of a way to get non-blocking keyboard input?
References: <3AA01B31@MailAndNews.com>
Message-ID: <uvgpvisvk.fsf@roundpoint.com>

Jeremy Reed <jpreed00 at MailAndNews.com> writes:

> I am looking to create a telnet client that will, by using the select() 
> statment, poll an open socket for data to be received while at the same time 
> keep accepting keyboard input from the user.
> 
> I have tried to run a thread in the background that collects input, but I 
> never can figure out how to achieve the 'fluidity' that I desire--i.e. no 
> stopping for a return-key press.

I think this ought to work for some platforms:

    try:
        istty = sys.stdin.isatty()
    except AttributeError:
        istty = 0

    if istty:
        if os.name=='posix':
            import tty
            tty.setraw(sys.stdin.fileno())
        elif os.name=='nt':
            import win32file, win32con
            hstdin = win32file._get_osfhandle(sys.stdin.fileno())
            modes = (win32file.GetConsoleMode(hstdin)
                     & ~(win32con.ENABLE_LINE_INPUT
                         |win32con.ENABLE_ECHO_INPUT))
            win32file.SetConsoleMode(hstdin, modes)

Unfortunately, GetConsoleMode, SetConsoleMode, and associated
constants don't seem to be included in win32all yet!

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From fuess at att.net  Wed Feb 28 14:33:16 2001
From: fuess at att.net (David Fuess)
Date: Wed, 28 Feb 2001 13:33:16 GMT
Subject: Anyone know of a way to get non-blocking keyboard input?
References: <3AA01B31@MailAndNews.com> <uvgpvisvk.fsf@roundpoint.com>
Message-ID: <hcvp9to0eoalto4586bsbpr23s76s60p5e@4ax.com>

Assuming you are on a Windows platform ...

--- Begin Demo Script
from msvcrt import *

ch = 0
while ch != 'x':
   if kbhit():
      ch = getch()
      print ch
--- End Demo Script

Dave

On 27 Feb 2001 15:54:55 -0800, Ben Hutchings
<ben.hutchings at roundpoint.com> wrote:

>Jeremy Reed <jpreed00 at MailAndNews.com> writes:
>
>> I am looking to create a telnet client that will, by using the select() 
>> statment, poll an open socket for data to be received while at the same time 
>> keep accepting keyboard input from the user.
>> 
>> I have tried to run a thread in the background that collects input, but I 
>> never can figure out how to achieve the 'fluidity' that I desire--i.e. no 
>> stopping for a return-key press.
>
>I think this ought to work for some platforms:
>
>    try:
>        istty = sys.stdin.isatty()
>    except AttributeError:
>        istty = 0
>
>    if istty:
>        if os.name=='posix':
>            import tty
>            tty.setraw(sys.stdin.fileno())
>        elif os.name=='nt':
>            import win32file, win32con
>            hstdin = win32file._get_osfhandle(sys.stdin.fileno())
>            modes = (win32file.GetConsoleMode(hstdin)
>                     & ~(win32con.ENABLE_LINE_INPUT
>                         |win32con.ENABLE_ECHO_INPUT))
>            win32file.SetConsoleMode(hstdin, modes)
>
>Unfortunately, GetConsoleMode, SetConsoleMode, and associated
>constants don't seem to be included in win32all yet!



From chris.gonnerman at usa.net  Wed Feb 28 14:57:14 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Wed, 28 Feb 2001 07:57:14 -0600
Subject: Anyone know of a way to get non-blocking keyboard input?
References: <3AA01B31@MailAndNews.com> <uvgpvisvk.fsf@roundpoint.com> <hcvp9to0eoalto4586bsbpr23s76s60p5e@4ax.com>
Message-ID: <005301c0a18e$6235e0e0$a100000a@local>

DOH!  Boy did I send him the long way around!  <slap self>

----- Original Message ----- 
From: "David Fuess" <fuess at att.net>
Subject: Re: Anyone know of a way to get non-blocking keyboard input?
> Assuming you are on a Windows platform ...
> 
> --- Begin Demo Script
> from msvcrt import *
> 
> ch = 0
> while ch != 'x':
>    if kbhit():
>       ch = getch()
>       print ch
> --- End Demo Script
> 
> Dave





From seandc at att.net  Tue Feb 27 22:02:25 2001
From: seandc at att.net (Sean)
Date: Tue, 27 Feb 2001 21:02:25 GMT
Subject: Pmw Megawidgets  0.8.5 - EntryField
Message-ID: <3A9C1577.4642FC60@att.net>

In this fragment (creating a date-type EntryField), I have used
the undocumented - in Pmw - paramater 'label_width' successfully:

_e = Pmw.EntryField(Master, 
		    labelpos = 'w',
		    label_text = f[0],
		    label_width = 12,
		    value = f[4][0:4] + '/' + f[4][4:6] + '/' + f[4][6:8],
		    validate = {'validator' : 'date', 'format' : 'ymd' },
		    )

*** [Q] Is there any way to control the width of the underlying Entry
widget?

In every example I've seen, all Entry widgets grouped in one dialog
have
the same width....presumably set by default to be the longest Entry in
the group.

Thanks.

Sean


From bernie at pacific.net.hk  Wed Feb 28 21:16:04 2001
From: bernie at pacific.net.hk (bernie)
Date: Thu, 01 Mar 2001 04:16:04 +0800
Subject: Pmw Megawidgets  0.8.5 - EntryField
References: <3A9C1577.4642FC60@att.net>
Message-ID: <3A9D5C83.C14E7C36@pacific.net.hk>


Sean wrote:

> In this fragment (creating a date-type EntryField), I have used
> the undocumented - in Pmw - paramater 'label_width' successfully:
>
> _e = Pmw.EntryField(Master,
>                     labelpos = 'w',
>                     label_text = f[0],
>                     label_width = 12,
>                     value = f[4][0:4] + '/' + f[4][4:6] + '/' + f[4][6:8],
>                     validate = {'validator' : 'date', 'format' : 'ymd' },
>                     )
>
> *** [Q] Is there any way to control the width of the underlying Entry
> widget?
>

_e._entryFieldEntry.config( width = <what you need>)

>
> In every example I've seen, all Entry widgets grouped in one dialog
> have
> the same width....presumably set by default to be the longest Entry in
> the group.
>
> Thanks.
>
> Sean

Bernie



From frohne at gci.net  Tue Feb 27 23:57:43 2001
From: frohne at gci.net (Ivan Frohne)
Date: Tue, 27 Feb 2001 13:57:43 -0900
Subject: PEP 236
Message-ID: <t9oc84n063phe7@corp.supernews.com>

The python grows by shedding its skin.  When
it stops growing, it dies.

--Ivan Frohne










From m.faassen at vet.uu.nl  Wed Feb 28 00:43:32 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Feb 2001 23:43:32 GMT
Subject: ANNOUNCE: Python - The RPG
Message-ID: <97he34$jhc$1@newshost.accu.uu.nl>

Python - The RPG

The Hype

  A Python role playing game! Play your favorite Python community
  member! While said member watches! Travel through time! Save the
  planet!

The Scenario

  The aliens are not expected to unleash their whitespace eating
  nanovirus to Earth for decades, right? There is no Python Secret
  Underground trying to put a stop to this, right?

When? Where?

  This is to be determined. Tentative time is the evening of wednesday
  the 7th of january.
 
Participants

  * Benevolent Dictator For the Game (BDFG)

  * A number of Players playing Characters

  * Members of the public (if sufficiently quiet)

Sounds Great! How can I play?

  Mail the BDFG at m.faassen at vet.uu.nl with a convincing explanation
  on why you want to play a particular character. Don't mail him after
  sunday; it's no use -- talk to him at the conference. Playing
  yourself is illegal; you have to pick someone else. Write-in
  characters are allowed and encouraged, if you can convince the
  BDFG. If you don't get to play a character, never fear; members of
  the public have voting rights if deemed Worthy.

  The use of humor is a good way to convince the BDFG.

Cast of Possible Characters

  Suggestions for new characters and/or character attributes are
  welcome. If people feel offended, sprinkle the following generously
  with smileys and <winks> in your mind. Yes, you can play any of
  these characters and others, if you manage to convince the BDFG.

  The advantages and disadvantages are expected to be roleplayed,
  mainly. They may also influence people's voting patterns during task
  resolution (see the rules section).

  Guido van Rossum (AKA BDFL)
  
    Advantages:
  
      * BDFL. Other Python Community Members will tend to listen to
        Guido. Guido can tell them what to do, and especially what
        they shouldn't do. When team members are deciding on an
        action, Guido can vote -1, thereby overruling the team members
        (note that this does not affect success determination, just
        whether the team will try something).

      * Access to time machine. Years ago Guido received a mysterious
        envelope containing the complete plans to a working time
        machine. He uses it mainly to implement features for Python
        before they're even suggested on comp.lang.python. Of course,
        the PSU uses it as well.

    Disadvantages

      * Dutch Bluntness. His Dutch bluntness may be a disadvantage
        when negotiating with NPCs.
        
  Tim Peters (AKA the timbot)

    Advantages:

      * Channeling Guido. Whenever Guido is absent, Tim Peters can
        Channel Guido. Tim can use this to gain extra authority over
        any Python Community Member. Tim can vote the overruling -1 in
        the absence of Guido.

      * Invisibility. Tim can turn invisible for a short while,
        allowing him to hide his secret identity, and to attend IPC8
        (last year's Python conference) without being seen.

    Disadvantages:

      * Robot. Tim's secret identity is that he's a robot, which may
        shock some people. Not just an android robot like the effbot
        and the martellibot -- those look like real people. No, the
        timbot is a genuine lumbering metallic robot with blinking
        lights, and a heap of Perl scripts running the innards.

      * Distracted by side issues. Tim can go on at length on issues
        which are not really the core of the problem, complicating
        said problem for himself and everybody else.
  
  Fredrik Lundh (AKA the effbot)

    Advantages:

      * Gadgeteer. A problem? The effbot can frequently construct a
        small gadget to solve it, often enlightening people in the
        process.

    Disadvantages:

      * Grouchiness. The effbot does not suffer fools gladly. This may
        may complicate his interaction with some NPCs, as well as
        other team members.

  Alex Martelli (AKA the martellibot)

    Advantages:

      * His erudite, knowledgable and *lengthy* discourse. This will
        either convince NPCs, or otherwise make them fall asleep
        (which in the case of enemies is good).

    Disadvantages:

      * Erupts into berserker rage when reminded of the print >>
        syntax.
  
  Jim Fulton

    Advantages:

      * Engineering. Ability to understand and build amazingly
        complicated machinery.

    Disadvantages:

      * Can inadvertantly make people's heads explode when attempting
        to explain how something works, rendering them unable to think
        very well for a while.

  Christian Tismer

    Advantages:

      * Can accomplish the impossible by changing the meaning of truth.
        Mystic ability influencing the fabric of reality.

      * Inspired Koreans. Christian can inspire hundreds of Koreans to
        the Pythonic cause. This should come in handy for something.

    Disadvantage:

      * A hard time explaining to other people what he's
        doing, or convincing them. Particilarly Guido.  

Rules

  The style of play is very light on actual rules. The only mechanical
  thing that may sometimes occur is task resolution by voting.

  Whenever a character wants to accomplish some task in the game, the
  BDFG decides whether success is trivial or not. For trivial success
  (for instance, adding simple a bug fix to the standard Python
  library) or trivial failure (for instance, adding a Perl style $ to
  the Python language) the BDFG just says whether something succeeds
  or fails, and that's all there is to it.

  In case of non-trivial success or failure, the BDFG has the option
  to put the resolution to a vote. In a vote each player can issue one
  vote, -1, 0, and +1. The BDFG can vote as well.

  If the total vote count is above 0, the action succeeds. If it's
  negative, the action fails. If it's exactly 0, a coin is tossed to
  determine success.

  If there is a public, members of the public who are deemed Worthy by
  the BDFG can participate in the voting process. Members of the
  public who are too loud at critical moments will likely not be
  deemed Worthy.

  Finally, if the BDFG thinks it's necessary, the BDFG can overrule
  the outcome of the vote.

References

   The PSU's Existence Revealed

    http://groups.yahoo.com/group/python-list/message/66070

  Roswell: The Truth

    http://groups.yahoo.com/group/python-list/message/87088

  Python Labs moves to Digital Creations

    http://groups.google.com/groups?q=Roswell&rnum=2&seld=971292152&ic=1

  Status of the PSU

    http://groups.google.com/groups?q=alien+whitespace&seld=941359988&ic=1

do-I-really-want-to-post-this-oh-well-here-goes-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From m.faassen at vet.uu.nl  Wed Feb 28 11:39:27 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 28 Feb 2001 10:39:27 GMT
Subject: ANNOUNCE: Python - The RPG
References: <97he34$jhc$1@newshost.accu.uu.nl>
Message-ID: <97ikgv$dva$1@newshost.accu.uu.nl>

Martijn Faassen <m.faassen at vet.uu.nl> wrote:
> When? Where?

>   This is to be determined. Tentative time is the evening of wednesday
>   the 7th of january.

I don't know why I say january when I mean march; I've made the same
mistake again after I wrote this. Weird. :)

Perhaps-it's-the-PSU's-doing-ly yours,

Martijn



From jcollin at exis.net  Wed Feb 28 03:26:32 2001
From: jcollin at exis.net (Jay Collins)
Date: Tue, 27 Feb 2001 21:26:32 -0500
Subject: Passing values to a class
Message-ID: <Pine.LNX.4.21.0102272122140.11512-100000@tarpon.exis.net>

Anyone shed some light on what I'm doing stupid/wrong here:

import dbm

class db:
        def __init__(self,base):
                print "using database",base
                pass
        def createdb(self):
                d = dbm.open(base,"n")
                d['name'] = "Jim Smith"
                d.close()

app = db("databse")
app.createdb()

how come createdb() doesn't see base? Maybe I'm not getting class scope
correctly. My idea was to pass what database I wanted to work on for that
instance. Then I could call the methods to do work on the data in it.



From bawolk at ucdavis.edu  Wed Feb 28 04:40:21 2001
From: bawolk at ucdavis.edu (Bruce Wolk)
Date: Tue, 27 Feb 2001 19:40:21 -0800
Subject: Passing values to a class
References: <Pine.LNX.4.21.0102272122140.11512-100000@tarpon.exis.net>
Message-ID: <3A9C7325.F14B39D4@ucdavis.edu>

Jay,

You are right that the scope is wrong.  In __init__ add  the statement
self.base = base  and in createdb use self.base instead of base.

Bruce

Jay Collins wrote:

> Anyone shed some light on what I'm doing stupid/wrong here:
>
> import dbm
>
> class db:
>         def __init__(self,base):
>                 print "using database",base
>                 pass
>         def createdb(self):
>                 d = dbm.open(base,"n")
>                 d['name'] = "Jim Smith"
>                 d.close()
>
> app = db("databse")
> app.createdb()
>
> how come createdb() doesn't see base? Maybe I'm not getting class scope
> correctly. My idea was to pass what database I wanted to work on for that
> instance. Then I could call the methods to do work on the data in it.



From chris.gonnerman at usa.net  Wed Feb 28 04:50:00 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Tue, 27 Feb 2001 21:50:00 -0600
Subject: Passing values to a class
References: <Pine.LNX.4.21.0102272122140.11512-100000@tarpon.exis.net>
Message-ID: <00dd01c0a139$8d3bfae0$a100000a@local>

----- Original Message ----- 
From: "Jay Collins" <jcollin at exis.net>
Subject: Passing values to a class
> Anyone shed some light on what I'm doing stupid/wrong here:

Let me just mark up your code:
 
import dbm
 
 class db:
    def __init__(self,base):
        print "using database",base
        self.base = base
#       pass # not needed
    def createdb(self):
#       d = dbm.open(base,"n")
        d = dbm.open(self.base,"n")  # access the copy of base
                                     # stored in self
        d['name'] = "Jim Smith"
        d.close()
 
 app = db("databse")
app.createdb()
 
> how come createdb() doesn't see base? Maybe I'm not getting 
> class scope correctly. My idea was to pass what database I 
> wanted to work on for that instance. Then I could call the 
> methods to do work on the data in it.

The concept is good, but simply handing base to __init__ 
doesn't actually DO anything; you still need to store it 
somewhere.





From jcollin at exis.net  Wed Feb 28 04:52:34 2001
From: jcollin at exis.net (Jay Collins)
Date: Tue, 27 Feb 2001 22:52:34 -0500 (EST)
Subject: Passing values to a class
In-Reply-To: <00dd01c0a139$8d3bfae0$a100000a@local>
Message-ID: <Pine.LNX.4.21.0102272252130.12109-100000@tarpon.exis.net>

Thanks. I got the hang of it out.

Take care

On Tue, 27 Feb 2001, Chris Gonnerman wrote:

>----- Original Message ----- 
>From: "Jay Collins" <jcollin at exis.net>
>Subject: Passing values to a class
>> Anyone shed some light on what I'm doing stupid/wrong here:
>
>The concept is good, but simply handing base to __init__ 
>doesn't actually DO anything; you still need to store it 
>somewhere.
>




From max at alcyone.com  Wed Feb 28 04:26:23 2001
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 27 Feb 2001 19:26:23 -0800
Subject: PIL for the Mac?
Message-ID: <3A9C6FDF.3F9422C@alcyone.com>

Is PIL (PythonWare Image Library) available on the Mac?  I don't see any
reference to the Mac on its main Web page, and the source tarball
doesn't appear to have any Mac specific files in it ...

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Can you take my breath away / Can you give him life today
\__/ Sweetbox
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From fredrik at pythonware.com  Wed Feb 28 05:56:57 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 04:56:57 GMT
Subject: PIL for the Mac?
References: <3A9C6FDF.3F9422C@alcyone.com>
Message-ID: <ty%m6.15923$Qb7.2598968@newsb.telia.net>

Erik Max Francis wrote:
> Is PIL (PythonWare Image Library) available on the Mac?  I don't see any
> reference to the Mac on its main Web page, and the source tarball
> doesn't appear to have any Mac specific files in it ...

from what I can tell, it's shipped with the standard Mac
distribution:

http://www.python.org/download/download_mac.html

    "This is a single installer which gives you the option to
    install classic 68k, cfm68k, and/or PPC. The installer also
    optionally installs NumPy, PIL, img, Tkinter, and an IDE,
    and interactive development environment.

Cheers /F






From dev_null at email.com  Wed Feb 28 04:27:42 2001
From: dev_null at email.com (Dev_NuLL)
Date: Wed, 28 Feb 2001 03:27:42 GMT
Subject: Intersesting Questions involving extending and embedding
Message-ID: <3a9c6f89.9613214@news.digital.net>

There is a free multi-platform 3D package called Blender (
www.blender.nl ) that has builtin python capabilities to help in
automating animation and modeling tasks. The Blender Python module is
builtin the Blender executable. Others have written external python
modules for use with Blender and would like to figure out how it is
done. Here is an example of code that would create a plane consisting
of 4 vertices and one face:


################## START CODE ###################

import Blender

scene = Blender.getCurrentScene()
mesh = Blender.Mesh("plane")
object = Blender.Object("plane")

mesh.enterEditMode()

indices = []

index = mesh.addVertex(1,-1,0)
indices.append(index)
index = mesh.addVertex(1,1,0)
indices.append(index)
index = mesh.addVertex(-1,-1,0)
indices.append(index)
index = mesh.addVertex(-1,1,0)
indices.append(index)

mesh.addFace(indices[0],indices[1],indices[2],indices[3])

Blender.connect(object,mesh)
Blender.connect(scene,object)
mesh.leaveEditMode()

################## END CODE ###################

I would like to create a primitives module to create 3D solids I would
like to wrap the above code (or similar) into a module and call it
'Primitives' with a 'plane' function so my code would look like:

import Blender
import Primitives
Primitive.plane()

Of course I would like the 'plane' function to take args for the size,
rotation, scaling, etc.., 
I have begun to read the docs on extending and embedding, and since
the above does not use any C functions I imagine I would be doing
mostly embedding. How would I go about creating a Windows DLL or Linux
.so that would do this? I don't need help on compiling, just on how to
get the module to directly connect to the 'Blender' module that is
builtin the Blender executable. I am having a hard time getting
started with the limmited knowledge I have in creating simple external
modules that just do things like divide 2 numbers and return a result.
How do I attach to the Blender module from within my module and make
use of its functions like:

mesh = Blender.Mesh("plane")
object = Blender.Object("plane")
mesh.enterEditMode()

I know I sound desparate, but the 2 books I bought and the online docs
barely scratch the surface of embedding.

TIA


From fredrik at pythonware.com  Wed Feb 28 08:18:20 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 07:18:20 GMT
Subject: Intersesting Questions involving extending and embedding
References: <3a9c6f89.9613214@news.digital.net>
Message-ID: <0D1n6.15927$Qb7.2600092@newsb.telia.net>

"Dev_NuLL" wrote:
> I would like to create a primitives module to create 3D solids I would
> like to wrap the above code (or similar) into a module and call it
> 'Primitives' with a 'plane' function so my code would look like:
>
> import Blender
> import Primitives
> Primitive.plane()

umm.  any reason you cannot just put your code in a
Primitives.py file?

    http://www.python.org/doc/current/tut/node8.html

Cheers /F




From dev_null at email.com  Wed Feb 28 15:27:09 2001
From: dev_null at email.com (Dev_NuLL)
Date: Wed, 28 Feb 2001 14:27:09 GMT
Subject: Intersesting Questions involving extending and embedding
References: <3a9c6f89.9613214@news.digital.net> <0D1n6.15927$Qb7.2600092@newsb.telia.net>
Message-ID: <3a9f0af7.37387355@news.digital.net>

>umm.  any reason you cannot just put your code in a
>Primitives.py file?

The writers of Blender modules release the modules compiled to protect
the source code, and that is what I am also interested in doing. This
Primitives.py is just an example, I have a much more complicated
project involving Catmul-Clark subdivision libs which are not open
source.


From fredrik at pythonware.com  Wed Feb 28 15:34:41 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 14:34:41 GMT
Subject: Intersesting Questions involving extending and embedding
References: <3a9c6f89.9613214@news.digital.net> <0D1n6.15927$Qb7.2600092@newsb.telia.net> <3a9f0af7.37387355@news.digital.net>
Message-ID: <508n6.15966$Qb7.2607706@newsb.telia.net>

Dev_NuLL wrote:
> >umm.  any reason you cannot just put your code in a
> >Primitives.py file?
>
> The writers of Blender modules release the modules compiled to protect
> the source code, and that is what I am also interested in doing.

compiled as in DLL/SO, or compiled as in PYC?

Cheers /F




From sfriedman at bewellnet.com  Wed Feb 28 04:48:32 2001
From: sfriedman at bewellnet.com (Steven Friedman)
Date: Tue, 27 Feb 2001 20:48:32 -0700
Subject: Newbie needs help on Python CGI scripts
Message-ID: <97k893$hr$1@news.bewellnet.com>

I am a complete newbie when it comes to CGI and scripting. I am trying to
test out some scripts running Savant Server on a Windows 98 machine. Since I
only want to learn CGI scripting this seems a perfectly acceptable soltion.
Unfortunately everything I read seems to assume a UNIX server. Could someone
PLEASE help me get things set right.

1. What and where do I need to set a path and where does my Python20.exe
need to be.
2. How do I configure my server to tell it how to execute the file.
3. from what in my browser do I actually call the script? An HTML doc?

Please be patient with me, its my first time at CGI.

Thanks in advance

sfriedman at bewellnet.com




From sjuranic at u.washington.edu  Wed Feb 28 07:39:36 2001
From: sjuranic at u.washington.edu ('Steve' Stephen W. Juranich)
Date: Tue, 27 Feb 2001 22:39:36 -0800
Subject: Changes in __getitem__ in Python 2.0?
Message-ID: <Pine.A41.4.33.0102272228370.18474-100000@dante36.u.washington.edu>

It's been a while since I've been playing with my Python toys (I've been
banished to C++ land for a while <retch>).  I have a class with the
following definition for __getitem__:

def __getitem__(self, key):
    # In some cases, I'm getting requests for data that are
    # past the end of the list.  Given the sloppy nature of
    # the hand-labelled data, I don't think this is too much
    # of a problem, so I'll try this scheme.
    try:
        retval = self.data[key]
    except IndexError:
        retval = self.data[-1]

    return retval

I'm now encountering a problem with Python 2.0 that I never had with earlier
versions (1.5 and 1.6 specifically).

I used to be able to do something like this:

for foo in MyClass:
	print foo.member,

But now when I do this, Python just starts spinning out of control in an
infinite loop (it's like a car wreck... I want to press ^C, but I just can't
look away).  I am assuming that the problem is because of the __getitem__
definition above.  Is there a way around this, or am I going to have to
change a bunch of code to make this work on a new Python setup?

I apologize if this is a topic that has been beaten to death.  I tried
looking for somthing relevant on Deja-cum-Google, but I didn't see any
relevant hits and I didn't feel like looking through the next 13,000
possibilities.

Thanks in advance for any help.  I appreciate it.

----------------------------------------------------------------------
Stephen W. Juranich                         sjuranic at ee.washington.edu
Electrical Engineering         http://students.washington.edu/sjuranic
University of Washington                 http://ssli.ee.washington.edu




From fredrik at pythonware.com  Wed Feb 28 08:14:14 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 07:14:14 GMT
Subject: Changes in __getitem__ in Python 2.0?
References: <Pine.A41.4.33.0102272228370.18474-100000@dante36.u.washington.edu>
Message-ID: <az1n6.15926$Qb7.2600009@newsb.telia.net>

"'Steve' Stephen W. Juranich" wrote:
> def __getitem__(self, key):
>     # In some cases, I'm getting requests for data that are
>     # past the end of the list.  Given the sloppy nature of
>     # the hand-labelled data, I don't think this is too much
>     # of a problem, so I'll try this scheme.
>     try:
>         retval = self.data[key]
>     except IndexError:
>         retval = self.data[-1]
>
>     return retval
>
> I'm now encountering a problem with Python 2.0 that I never had with earlier
> versions (1.5 and 1.6 specifically).
>
> I used to be able to do something like this:
>
> for foo in MyClass:
>    print foo.member,
>
> But now when I do this, Python just starts spinning out of control in an
> infinite loop (it's like a car wreck... I want to press ^C, but I just can't
> look away).

it spins out of control in 1.5.2 too: for-in expects getitem to
throw an index exception when there are no more items.

afaik, for-in has always worked that way.

the real problem is probably somewhere else...

Cheers /F




From tim.one at home.com  Wed Feb 28 08:43:26 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 28 Feb 2001 02:43:26 -0500
Subject: Changes in __getitem__ in Python 2.0?
In-Reply-To: <Pine.A41.4.33.0102272228370.18474-100000@dante36.u.washington.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEKDJCAA.tim.one@home.com>

[Stephen W. Juranich]
> It's been a while since I've been playing with my Python toys (I've been
> banished to C++ land for a while <retch>).  I have a class with the
> following definition for __getitem__:
>
> def __getitem__(self, key):
>     # In some cases, I'm getting requests for data that are
>     # past the end of the list.  Given the sloppy nature of
>     # the hand-labelled data, I don't think this is too much
>     # of a problem, so I'll try this scheme.
>     try:
>         retval = self.data[key]
>     except IndexError:
>         retval = self.data[-1]
>
>     return retval
>
> I'm now encountering a problem with Python 2.0 that I never had
> with earlier versions (1.5 and 1.6 specifically).

Sorry, but this is hard to believe.  You Changed Something.

> I used to be able to do something like this:
>
> for foo in MyClass:
> 	print foo.member,

I'm assuming you meant to say that you're iterating over an *instance* of the
class that contains the __getitem__ shown above.

> But now when I do this, Python just starts spinning out of control in an
> infinite loop

Under the assumption above, it *should* be an infinite loop.  Also in 1.6,
1.5, ..., and 0.9.6.  Nothing relevant has ever changed here.  The "for"
protocol doesn't stop until __getitem__ raises an IndexError.  The only way
your __getitem__() above can raise an IndexError is if self.data is empty:
then the first iteration will try self.data[0], that will raise IndexError,
but then you'll catch IndexError, then the code will try self.data[-1], and
that will raise an uncaught IndexError (which terminates the loop).  So your
loop should terminate if and only if self.data is empty.

> ...
> I am assuming that the problem is because of the __getitem__ definition
> above.  Is there a way around this, or am I going to have to
> change a bunch of code to make this work on a new Python setup?

Again, it has nothing to do with 2.0, but you're not going to believe that
until you accept my challenge to show us a complete piece of executable code
that demonstrates the problem under 2.0 but not under earlier Pythons <wink>.
If your __getitem__ never raises IndexError, your "for" loop will never end;
that's always been the case.

or-the-problem-isn't-where-you-think-it-is-ly y'rs  - tim




From sjuranic at u.washington.edu  Wed Feb 28 17:44:14 2001
From: sjuranic at u.washington.edu ('Steve' Stephen W. Juranich)
Date: Wed, 28 Feb 2001 08:44:14 -0800
Subject: Changes in __getitem__ in Python 2.0?
In-Reply-To: <mailman.983346250.29347.python-list@python.org>
References: <mailman.983346250.29347.python-list@python.org>
Message-ID: <Pine.A41.4.33.0102280841240.122840-100000@dante27.u.washington.edu>

Thanks to Frank and Tim for the help.

After further review on the play, the ruling on the field is that I'm a
dunderhead.

You guys were both right (as if there was ever any doubt).  The same kind of
thing chokes in Python 1.6 as well.

hoping-that-the-code-is-robust-because-the-project-\
is-due-in-two-weeks-ly y'rs...

----------------------------------------------------------------------
Stephen W. Juranich                         sjuranic at ee.washington.edu
Electrical Engineering         http://students.washington.edu/sjuranic
University of Washington                 http://ssli.ee.washington.edu




From grobe+news at netins.net  Wed Feb 28 09:21:44 2001
From: grobe+news at netins.net (Jonathan Grobe)
Date: 28 Feb 2001 08:21:44 GMT
Subject: alt.internet.p2p (peer-to-peer) newsgroup created
Message-ID: <slrn99pd8o.u86.grobe+news@worf.netins.net>

Subject: alt.internet.p2p (peer-to-peer) newsgroup created
Followup-To: alt.internet.p2p

alt.internet.p2p has just been created for discussion of p2p
(peer-to-peer) applications on the internet. 

Quoting Clay Shirky:

"P2P is a class of applications that takes advantage of resources --
storage, cycles, content, human presence -- available at the edges of
the Internet. Because accessing these decentralized resources means
operating in an environment of unstable connectivity and unpredictable
IP addresses, P2P nodes must operate outside the DNS system and have
significant or total autonomy from central servers.
...
If you're looking for a litmus test for P2P, this is it: 1) Does it
treat variable connectivity and temporary network addresses as the
norm, and 2) does it give the nodes at the edges of the network
significant autonomy? If the answer to both of those questions is yes, 
the application is P2P. If the answer to either question is no, it's not
P2P."

These applications include file sharing such as Napster and Gnutella,
instant messaging such as ICQ, distributed computing such as SETI at home.

Many news administrators only add new alt.* groups on user request. So if
it is not available at your site ask your news administrator to add it:
Write to him at the address news or usenet at your site or to the
technical support people there (address: support)). Because of the
poor propagation of new alt groups it will be a while before a 
significant amount of traffic shows up.


-- 
Jonathan Grobe



From jafo at tummy.com  Wed Feb 28 10:09:25 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 28 Feb 2001 02:09:25 -0700
Subject: Swalow: Where to get packages?
Message-ID: <20010228020925.E29199@tummy.com>

First of all, let me take care of this:
<RIGHT FIST UP>Distutils RULES!</RIGHT FIST UP>

[Oops, the next 5 paragraphs are rambling with a status update.  Skip
them to get to what I mentioned in the subject]

My goal for tonight was to understand distutils and get a couple of my
packages converted over to use it and just get familiar with it a bit.
Very well thought-out.  Works beautifully and required very little time
for me to wrap my mind around.

So, I released Distutiled versions of my netstring and CHAP modules
(the latter has a dependency on the former), and updated my sample data
to include that information.  Then I set to work on the prototype code
for the server and client.

It's reasonably primitive right now, but I've got it to the point where
the command "swalow download ITEMNAME=chap PKGVERS=1.13" will pick up the
file from one of the listed locations (selected randomly), try the
the list of locations until the size and checksum match the what the
server said they should be, and return the path and file name it was
downloaded to.

I also added "swalow install ITEMNAME=chap PKGVERS=1.13", which does the
download, extracts the tar.gz file, creates a binary RPM and installs it.
Works wonderfully for distutils packages.  Oh, I've also asked one of
our guys who is familiar with both RPM and .debs about doing a
"bdist_deb" target for Distutils, and have asked him to work on it.

My goal for tomorrow is to figure out how to deal with dependencies.

Dang, I got off topic there...

One of my questions now is, where do we get the packages?  Is there a
collection of nice, juicy distutils packages out there that I just
don't know about?  What about the Vaults, any way to get information from
them?  At the very least I need to know:

	Package name
	Package version
	Does it include a Distutils setup.py
	Where to download it

If somone can recommend some locations where I can pick up at least a few
nice dist-utils packages to include in the setup, that'd be great as some
additional testing.

Sean
-- 
 "The great thing about Meat Loaf is he's his own special effect." -- J.S.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From amk at mira.erols.com  Wed Feb 28 13:47:17 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 28 Feb 2001 12:47:17 GMT
Subject: Swalow: Where to get packages?
References: <mailman.983351415.11444.python-list@python.org>
Message-ID: <slrn99ptm0.go.amk@mira.erols.com>

On Wed, 28 Feb 2001 02:09:25 -0700, Sean Reifschneider <jafo at tummy.com> wrote:
>don't know about?  What about the Vaults, any way to get information from
>them?  At the very least I need to know:

Write Timothy Middleton, VoP's maintainer.  Some HTML dumps of the
data are available, which could be parsed, but I can't recall the URL
and also don't know if Timothy wants their location to be handed
around.  

Most of my code now uses Distutils -- poke around www.amk.ca/files/ --
but don't use the zodb package, because it has a weird special
setup.py.

--amk


From phd at phd.pp.ru  Wed Feb 28 10:30:49 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 28 Feb 2001 12:30:49 +0300 (MSK)
Subject: Bug in __contains__
Message-ID: <Pine.LNX.4.30.0102281229250.7324-100000@fep132.fep.ru>

Hello!

Python 2.0 (#1, Jan 26 2001, 15:34:20)
[GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> 10 in xrange(10)
1

   Is it a known bug?

Python 1.5.2 (#2, Dec 12 2000, 17:07:14)  [GCC 2.95.2 20000220 (Debian
GNU/Linux) on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> 10 in xrange(10)
0

   It was ok...

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From thomas.heller at ion-tof.com  Wed Feb 28 11:08:06 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Wed, 28 Feb 2001 11:08:06 +0100
Subject: Bug in __contains__
References: <mailman.983352719.16794.python-list@python.org>
Message-ID: <97iim7$ogn8k$1@ID-59885.news.dfncis.de>

"Oleg Broytmann" <phd at phd.pp.ru> wrote in message news:mailman.983352719.16794.python-list at python.org...
> Hello!
>
> Python 2.0 (#1, Jan 26 2001, 15:34:20)
> [GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> 10 in xrange(10)
> 1
>
>    Is it a known bug?
Yes, I reported this to sourceforge some months ago. Should
be fixes in 2.1.

Thomas Heller




From othello at javanet.com  Wed Feb 28 18:51:47 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Wed, 28 Feb 2001 12:51:47 -0500
Subject: Bug in __contains__
References: <mailman.983352719.16794.python-list@python.org>
Message-ID: <3A9D3AB3.981DD5D9@javanet.com>

Oleg Broytmann wrote:

> Hello!
>
> Python 2.0 (#1, Jan 26 2001, 15:34:20)
> [GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> 10 in xrange(10)
> 1
>
>    Is it a known bug?

How did you run across this bug?  I have never had reason to
check for membership in xrange().

Just curious.
Raymond



From phd at phd.pp.ru  Wed Feb 28 19:07:35 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 28 Feb 2001 21:07:35 +0300 (MSK)
Subject: Bug in __contains__
In-Reply-To: <3A9D3AB3.981DD5D9@javanet.com>
Message-ID: <Pine.LNX.4.30.0102282104400.19078-100000@fep132.fep.ru>

On Wed, 28 Feb 2001, Raymond Hettinger wrote:
> > >>> 10 in xrange(10)
> > 1
>
> How did you run across this bug?  I have never had reason to
> check for membership in xrange().

   A collegue asked me if it's a known bug :) While he wrote the Russian
translation of Python tutorials (http://nd.spb.ru/dialect/book.asp?ID=60) he
found some bugs in python 2.0. So expect some more questions from me. If he
found some unknow bugs either he or I will file bug reports.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From bsb at winnegan.de  Wed Feb 28 11:15:34 2001
From: bsb at winnegan.de (Siggy Brentrup)
Date: 28 Feb 2001 11:15:34 +0100
Subject: Bug in __contains__
In-Reply-To: <Pine.LNX.4.30.0102281229250.7324-100000@fep132.fep.ru>
References: <Pine.LNX.4.30.0102281229250.7324-100000@fep132.fep.ru>
Message-ID: <87pug3glkp.fsf@winnegan.de>

Oleg Broytmann <phd at phd.pp.ru> writes:

> Hello!
> 
> Python 2.0 (#1, Jan 26 2001, 15:34:20)
> [GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> 10 in xrange(10)
> 1
> 
>    Is it a known bug?

Python 2.1a2 (#6, Feb 28 2001, 10:30:37) 
[GCC 2.95.3 20010219 (prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> 10 in xrange(10)
0

Seems to be fixed :)

Siggy




From phd at phd.pp.ru  Wed Feb 28 11:27:41 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 28 Feb 2001 13:27:41 +0300 (MSK)
Subject: Bug in __contains__
In-Reply-To: <87pug3glkp.fsf@winnegan.de>
Message-ID: <Pine.LNX.4.30.0102281326570.15197-100000@fep132.fep.ru>

On 28 Feb 2001, Siggy Brentrup wrote:
> > Python 2.0 (#1, Jan 26 2001, 15:34:20)
> > [GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
> > Type "copyright", "credits" or "license" for more information.
> > >>> 10 in xrange(10)
> > 1
> >
> >    Is it a known bug?
>
> Python 2.1a2 (#6, Feb 28 2001, 10:30:37)
> [GCC 2.95.3 20010219 (prerelease)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> 10 in xrange(10)
> 0
>
> Seems to be fixed :)

   Thanks!

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From tyler at tylereaves.com  Wed Feb 28 15:20:27 2001
From: tyler at tylereaves.com (Tyler Eaves)
Date: Wed, 28 Feb 2001 14:20:27 GMT
Subject: Annoucing the CyberBio DIrectory System!
Message-ID: <LO7n6.1811$%4.214080@newsread1.prod.itd.earthlink.net>

Annoucing the CyberBio DIrectory System!

http://www.ultracoastercentral.com/cyberbio/

The CyberBio is a revolutionary idea, an easy
to use system for community building. I'm sure
you've all heard that before. However, CyberBio
is REALLY unique. There is literally nothing else
like it!

CyberBio allows the building of an online 'whitepages'.

CyberBio is great for:

*Schools
*Churchs
*Community Orginizations
*Military Groups
*Anyone!

Thanks for your time.

Tyler Eaves




From cg at gaia.intranet.cdegroot.com  Wed Feb 28 15:59:19 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 28 Feb 2001 15:59:19 +0100
Subject: [JOB-Netherlands] Zope/Python developer wanted
Message-ID: <97j3o7$jmf$1@gaia.intranet.cdegroot.com>

The InternetOne is a young ISP located in Amersfoort, The Netherlands,
doing wholesale hosting through an international dealer network. We
develop our on-line e-business/commerce software in Zope and Python, and
need a skilled developer for this environment on short notice, either as
a sub-contractor or as an employee. A temporary assignment is possible,
we're looking at at least a three months' job.

Necessary skills: Zope, Python, simple HTML design, PostgreSQL, Linux. We are
a small shop, so you'll have a broad task range: developing software for
on-line webhosting transactions (our dealers need to be able to create 
new virtual hostings, domain mgt., etcetera through-the-web), 3rd line
support, integrating other systems (we host Intershop, Apache, IIS). We
welcome foreigners, although chances are that non-EC citizens won't
qualify due to time delays in work permit applications. You don't need to
speak Dutch, English is fine. Compensation will be according to industry
standards, for longer-term commitments we are working on a stock option
program.

If you are interested or want more information, drop me an email at
cg at theinternetone.net. You can also ring me at +31 628 960861.

Regards,

Cees de Groot
CTO
The InternetOne

-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From embed at geocities.com  Wed Feb 28 16:20:46 2001
From: embed at geocities.com (Warren Postma)
Date: Wed, 28 Feb 2001 10:20:46 -0500
Subject: Combinations of n Lists 
Message-ID: <NI8n6.2979$TW.15640@tor-nn1.netcom.ca>

I was just wondering if anyone has a more general version of this little
helper function:

def combinations(list1,list2):
    return [ (i,j) for i in list1 for j in list2 ]

print combinations( [1,2,3], ['a','b','c'] )

[(1, 'a'), (1, 'b'), (1, 'c'), (2, 'a'), (2, 'b'), (2, 'c'), (3, 'a'), (3,
'b'), (3, 'c')]

So, what if I want

def combinations(*lists):
        ....
print combinations( [1,2,3], [4,5,6], [7,8,9], .... )
    [ [1,4,7,....],  .... [1,4,8, ... ], ..... ]

Any takers!? ;-)

Warren




From aleaxit at yahoo.com  Wed Feb 28 17:14:32 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 28 Feb 2001 17:14:32 +0100
Subject: Combinations of n Lists
References: <NI8n6.2979$TW.15640@tor-nn1.netcom.ca>
Message-ID: <97j85e02ked@news1.newsguy.com>

"Warren Postma" <embed at geocities.com> wrote in message
news:NI8n6.2979$TW.15640 at tor-nn1.netcom.ca...
> I was just wondering if anyone has a more general version of this little
> helper function:
>
> def combinations(list1,list2):
>     return [ (i,j) for i in list1 for j in list2 ]
    [snip]
> def combinations(*lists):
>         ....
> print combinations( [1,2,3], [4,5,6], [7,8,9], .... )
>     [ [1,4,7,....],  .... [1,4,8, ... ], ..... ]

As usual, the easiest approach is through recursion, e.g.:

def combinations(*lists):
    if not lists: return [ [] ]
    more = combinations(*lists[1:])
    return [ [i]+js for i in lists[0] for js in more ]


Alex





From shaleh at valinux.com  Wed Feb 28 17:20:41 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Wed, 28 Feb 2001 08:20:41 -0800 (PST)
Subject: Combinations of n Lists
In-Reply-To: <NI8n6.2979$TW.15640@tor-nn1.netcom.ca>
Message-ID: <XFMail.20010228082041.shaleh@valinux.com>

On 28-Feb-2001 Warren Postma wrote:
> I was just wondering if anyone has a more general version of this little
> helper function:
> 
> def combinations(list1,list2):
>     return [ (i,j) for i in list1 for j in list2 ]
> 
> print combinations( [1,2,3], ['a','b','c'] )
> 
> [(1, 'a'), (1, 'b'), (1, 'c'), (2, 'a'), (2, 'b'), (2, 'c'), (3, 'a'), (3,
> 'b'), (3, 'c')]
> 

try zip() in python2 (-:



From fredrik at pythonware.com  Wed Feb 28 18:05:35 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 17:05:35 GMT
Subject: Combinations of n Lists
References: <mailman.983377210.26087.python-list@python.org>
Message-ID: <zdan6.15979$Qb7.2610993@newsb.telia.net>

Sean 'Shaleh' Perry wrote:
> > def combinations(list1,list2):
> >     return [ (i,j) for i in list1 for j in list2 ]
> >
> > print combinations( [1,2,3], ['a','b','c'] )
> >
> > [(1, 'a'), (1, 'b'), (1, 'c'), (2, 'a'), (2, 'b'), (2, 'c'), (3, 'a'), (3,
> > 'b'), (3, 'c')]
> >
>
> try zip() in python2 (-:

>>> print zip( [1,2,3], ['a','b','c'] )
[(1, 'a'), (2, 'b'), (3, 'c')]

Cheers /F




From tim.hochberg at ieee.org  Wed Feb 28 19:09:08 2001
From: tim.hochberg at ieee.org (Tim Hochberg)
Date: Wed, 28 Feb 2001 18:09:08 GMT
Subject: Combinations of n Lists 
References: <NI8n6.2979$TW.15640@tor-nn1.netcom.ca>
Message-ID: <89bn6.1121$DY.186465@news1.rdc1.az.home.com>

"Warren Postma" <embed at geocities.com> wrote in message
news:NI8n6.2979$TW.15640 at tor-nn1.netcom.ca...
> I was just wondering if anyone has a more general version of this little
> helper function:
>
> def combinations(list1,list2):
>     return [ (i,j) for i in list1 for j in list2 ]
>
> print combinations( [1,2,3], ['a','b','c'] )
>
> [(1, 'a'), (1, 'b'), (1, 'c'), (2, 'a'), (2, 'b'), (2, 'c'), (3, 'a'), (3,
> 'b'), (3, 'c')]
>
> So, what if I want
>
> def combinations(*lists):
>         ....
> print combinations( [1,2,3], [4,5,6], [7,8,9], .... )
>     [ [1,4,7,....],  .... [1,4,8, ... ], ..... ]
>
> Any takers!? ;-)

zip([1,2,3], [4,5,6], [7,8,9], .... )

(New in 2.0 I believe).

or equivalently

map(None, [1,2,3], [4,5,6], [7,8,9], .... )

for you 1.5.2 types out there....

-tim




From rich_somerfield at tertio.com  Wed Feb 28 16:33:39 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Wed, 28 Feb 2001 15:33:39 -0000
Subject: Problems compiling Tkinter
Message-ID: <WW8n6.3515$sT1.701218@news2-win.server.ntlworld.com>

has anybody tried compiling tkinter on linux (suse 6.4, have already
downloaded and compiled Python 2.0 and Pil 1.1.1) ?  I downloaded the source
and couldnt find any info as to how to compile it.

i tried the command "python setup.py install" but got the following error.
I know that the last path is invalid (my install is in /usr/bin/python) but
cant find where it is specified wrongly.  Should I cheat and create a
softlink so that the install can pick up the relavent files?, or is there
something I can do to solve this problem properly?

Any help would be appreciated.

Regards
Rich

bash-2.03# python setup.py install
running install
Traceback (most recent call last):
  File "setup.py", line 25, in ?
    long_description = __doc__
  File "/usr/bin/python-2.0/Lib/distutils/core.py", line 138, in setup
    dist.run_commands()
  File "/usr/bin/python-2.0/Lib/distutils/dist.py", line 829, in
run_commands
    self.run_command(cmd)
  File "/usr/bin/python-2.0/Lib/distutils/dist.py", line 848, in run_command
    cmd_obj.ensure_finalized()
  File "/usr/bin/python-2.0/Lib/distutils/cmd.py", line 112, in
ensure_finalized
    self.finalize_options()
  File "/usr/bin/python-2.0/Lib/distutils/command/install.py", line 255, in
fina
lize_options
    (prefix, exec_prefix) = get_config_vars('prefix', 'exec_prefix')
  File "/usr/bin/python-2.0/Lib/distutils/sysconfig.py", line 368, in
get_config
_vars
    func()
  File "/usr/bin/python-2.0/Lib/distutils/sysconfig.py", line 280, in
_init_posi
x
    raise DistutilsPlatformError, my_msg
distutils.errors.DistutilsPlatformError: invalid Python installation: unable
to
open /usr/local/lib/python2.0/config/Makefile (No such file or directory)
bash-2.03#




From marvind at Glue.umd.edu  Wed Feb 28 17:43:22 2001
From: marvind at Glue.umd.edu (Arvind Mani)
Date: Wed, 28 Feb 2001 11:43:22 -0500
Subject: c thread routines from python
Message-ID: <3A9D2AAA.8B8E3116@Glue.umd.edu>

hi,

I have a multi-threaded C program that I need to interface with a
Tkinter
GUI. I used Swig to create a module containing all the thread routines.
Then I created
a thread in python corresponding to each thread in the C program, and
called the C
routines in them.

The problem is the that the python interpreter executes these
instructions
sequentially. I will include a small test program to illustrate my
problem.
The problem is the that the python interpreter executes these
instructions
sequentially. I will include a small test program to illustrate my
problem.

I added the code posted by Mark in the interface file but I think I got
something messed up. I would be really grateful for any help.

Thanks,
arvind

hello.c
-----
void print_hello1(void)
{
...... inf loop
}

void print_hello2(void)
{
.....inf loop
}

.i file
----

%typemap (python, except) void {
        PY_BEGIN_ALLOW_THREADS
        $function
        PY_END_ALLOW_THREADS
}

extern void print_hello1(void);
extern void print_hello2(void);

.py
---

thread.start_new_thread(hello.print_hello1, ())
thread.start_new_thread(hello.print_hello2, ())





From slhath at home.com.nospam  Wed Feb 28 17:44:48 2001
From: slhath at home.com.nospam (Scott Hathaway)
Date: Wed, 28 Feb 2001 16:44:48 GMT
Subject: Pmw ComboBox question
Message-ID: <4W9n6.369$xT4.93987904@newssvr10-int.news.prodigy.com>

How can I change the length of the entry portion of the combox widget?

Thanks,
Scott




From bernie at pacific.net.hk  Wed Feb 28 19:04:50 2001
From: bernie at pacific.net.hk (bernie)
Date: Thu, 01 Mar 2001 02:04:50 +0800
Subject: Pmw ComboBox question
References: <4W9n6.369$xT4.93987904@newssvr10-int.news.prodigy.com>
Message-ID: <3A9D3DC2.5387B37F@pacific.net.hk>


Scott Hathaway wrote:

> How can I change the length of the entry portion of the combox widget?
>

_width = 4                                        # width of entry field
_widget = Pmw.ComboBox( parent)
_widget._entryfield._entryFieldEntry.config( width = _width)


>
> Thanks,
> Scott



From slhath at home.com  Wed Feb 28 23:03:52 2001
From: slhath at home.com (Scott Hathaway)
Date: Wed, 28 Feb 2001 22:03:52 GMT
Subject: Pmw ComboBox question
References: <4W9n6.369$xT4.93987904@newssvr10-int.news.prodigy.com> <3A9D3DC2.5387B37F@pacific.net.hk>
Message-ID: <cBen6.124025$bb.1370419@news1.rdc1.tx.home.com>

Thanks!

Scott

"bernie" <bernie at pacific.net.hk> wrote in message
news:3A9D3DC2.5387B37F at pacific.net.hk...
>
>
> Scott Hathaway wrote:
>
> > How can I change the length of the entry portion of the combox widget?
> >
>
> _width = 4                                        # width of entry field
> _widget = Pmw.ComboBox( parent)
> _widget._entryfield._entryFieldEntry.config( width = _width)
>
>
> >
> > Thanks,
> > Scott
>




From spamfranke at bigfoot.de  Wed Feb 28 18:49:12 2001
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Wed, 28 Feb 2001 17:49:12 GMT
Subject: Is += on strings a thread-safe operation?
Message-ID: <3a9d3833.737414@news.btx.dtag.de>

What about augmented assignments of other built-in types?  Being able to thread-safely append 
to a string buffer  (or other sequences) with += would be a big advantage of the somewhat infamous
augmented assignment operators in my eyes.

Stefan




Stefan Franke
www.meso.net




From aahz at panix.com  Wed Feb 28 19:43:22 2001
From: aahz at panix.com (Aahz Maruch)
Date: 28 Feb 2001 10:43:22 -0800
Subject: Is += on strings a thread-safe operation?
References: <3a9d3833.737414@news.btx.dtag.de>
Message-ID: <97jgsa$9it$1@panix6.panix.com>

In article <3a9d3833.737414 at news.btx.dtag.de>,
Stefan Franke <spamfranke at bigfoot.de> wrote:
>
>What about augmented assignments of other built-in types?  Being able
>to thread-safely append to a string buffer (or other sequences) with +=
>would be a big advantage of the somewhat infamous augmented assignment
>operators in my eyes.

Using the dis module, we see that the only difference between "x+='b'"
and "x=x+'b'" lies in substituting INPLACE_ADD for BINARY_ADD.  If the
latter is thread-safe, so is the former; I'm not entirely confident
enough in my byte-code skills to make an assertion.  But it looks to me
that it's not thread-safe because you're reading and assigning the same
variable in the same statement.

I *do* know that x.append('b') is a thread-safe operation -- and that's
probably a better bet, anyway, even if threading isn't involved.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Nostalgia just ain't what it used to be


From embed at geocities.com  Wed Feb 28 21:25:05 2001
From: embed at geocities.com (Warren Postma)
Date: Wed, 28 Feb 2001 15:25:05 -0500
Subject: Is += on strings a thread-safe operation?
References: <3a9d3833.737414@news.btx.dtag.de>
Message-ID: <3adn6.3087$TW.15865@tor-nn1.netcom.ca>

"Stefan Franke" <spamfranke at bigfoot.de> wrote in message
news:3a9d3833.737414 at news.btx.dtag.de...
> What about augmented assignments of other built-in types?  Being able to
thread-safely append
> to a string buffer  (or other sequences) with += would be a big advantage
of the somewhat infamous
> augmented assignment operators in my eyes.
>
> Stefan

The sad truth is that everything in Python is thread safe, because
ironically, nothing in python is thread safe.

What I mean to say is that when Threading support is enabled in Python,
which on most systems it now is, everything the Python interpreter runs
occurs while it is holding onto a Global Interpreter Lock which is only
released explicitly when it is safe to do so.

So that means in effect that all Python Bytecode operations are explicitly
atomic, thus all Python builtin C functions are atomic, except those which
choose not to be (such as time.sleep()) by letting go of and then
re-acquiring the global interpreter lock.  It is not possible for the
interpreter to let go of that lock while in the middle of a string "+=".

In other words, anything that resolves to a single bytecode in Python
interpreter is done while the current thread holds the lock. That makes
invocation of any builtin C function atomic as far as I can see.

Am I right? Where is Timbot when you really need him?  <grin>

Warren




From tim.one at home.com  Wed Feb 28 21:49:42 2001
From: tim.one at home.com (Tim Peters)
Date: Wed, 28 Feb 2001 15:49:42 -0500
Subject: Is += on strings a thread-safe operation?
In-Reply-To: <3adn6.3087$TW.15865@tor-nn1.netcom.ca>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEMOJCAA.tim.one@home.com>

[Warren Postma, correctly explains Python's peculiarly forgiving notion
 of thread-safety as follows from the global interpreter lock, but
 falls for a flawed example]

> ...
> So that means in effect that all Python Bytecode operations are explicitly
> atomic, thus all Python builtin C functions are atomic, except those which
> choose not to be (such as time.sleep()) by letting go of and then
> re-acquiring the global interpreter lock.  It is not possible for the
> interpreter to let go of that lock while in the middle of a string "+=".

There's the rub:  += generates more than one bytecode, so there actually are
chances for the interpreter to let go of the GIL "in the middle" of *any* +=
(and for multiple spellings of "+=").

> In other words, anything that resolves to a single bytecode in Python
> interpreter is done while the current thread holds the lock.

Yes.

> That makes invocation of any builtin C function atomic as far as
> I can see.

Your eyes are fine -- at least for long-term viewing <wink>.

> Am I right? Where is Timbot when you really need him?  <grin>

I only show up when I'm not really needed -- like here.

it's-too-stressful-to-take-a-stand-when-i-am-needed-ly y'rs  - tim




From jp at ulgo.koti.com.pl  Wed Feb 28 19:34:55 2001
From: jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=)
Date: 28 Feb 2001 18:34:55 GMT
Subject: Python + MySQL + qt
Message-ID: <slrn99qh6f.7h0.jp@localhost.localdomain>

I have a school project: create database and write application which will use
it (I am not sure what exactly it will do, yet), it should have graphical menu
and be user friendly. I can choose tool. I am thinking about MySQL or
Postgresql (probably MySQL) and qt/gtk+ as toolkit (probably qt). 
I am much more experienced in C than in Python, but I like this language and I
think it will be next step to learn it more. 
And now questions:
- is Python good choice here? what troubles will I have?
- what libraries will I need except pygtk or pyqt?
- I am programming only in Linux, is it possible to run my application in
  Windows, without any code fixes? are needed libraries available for Win?

-- 
Look all around
Can't you open your eyes
Voices are calling                                         "Master of Insanity"
Killing rain falling down from the sky                      - Ronnie James Dio


From hgg9140 at seanet.com  Wed Feb 28 22:14:12 2001
From: hgg9140 at seanet.com (Harry George)
Date: 28 Feb 2001 13:14:12 -0800
Subject: Python + MySQL + qt
References: <slrn99qh6f.7h0.jp@localhost.localdomain>
Message-ID: <m3n1b6zf17.fsf@wilma.localdomain>

jp at ulgo.koti.com.pl (Jacek =?iso-8859-2?Q?Pop=B3awski?=) writes:

> I have a school project: create database and write application which will use
> it (I am not sure what exactly it will do, yet), it should have graphical menu
> and be user friendly. I can choose tool. I am thinking about MySQL or
> Postgresql (probably MySQL) and qt/gtk+ as toolkit (probably qt). 
> I am much more experienced in C than in Python, but I like this language and I
> think it will be next step to learn it more. 
> And now questions:
> - is Python good choice here? what troubles will I have?

Python is an excellent choice.  Python is probably the easy part.

DBMS's can be tricky.  Are you already a DBA or do you already have a
DBMS running (with your database defined and loaded)?  I can give you
some install scripts if you are starting fresh.

Next, you need a DBI 2.0 python binding.  I'm using MySQLdb, and it
works fine.


> - what libraries will I need except pygtk or pyqt?

pygtk, pygnome, pyqt, and pykde are all do-able, and interesting
exercises in their own right.  But if you are on a tight deadline,
consider using cgi's (if you are comfortable with forms).  Also,
python and wxWindows is a nice cros-platform combination.


> - I am programming only in Linux, is it possible to run my application in
>   Windows, without any code fixes? are needed libraries available for Win?
> 

If you use cgi's, you can run of course against a Linux server.  Even
on Windows, the cgi's may be ok without code changes.  Or you can use
wxWindows and be cross-platform.  MySQL and PostgreSQL can also be
served from Linux even if the GUI app is on Windows.  If you must run
fully in Windows, you might have to do odbc to Access or something
like that.

In any case, I'm not aware of any code changes due to different
platforms, but I'm not much of a Windows programmer.

> -- 
> Look all around
> Can't you open your eyes
> Voices are calling                                         "Master of Insanity"
> Killing rain falling down from the sky                      - Ronnie James Dio

-- 
Harry George
hgg9140 at seanet.com


From jbeaulie at softimage.com  Wed Feb 28 20:34:45 2001
From: jbeaulie at softimage.com (motif)
Date: Wed, 28 Feb 2001 14:34:45 -0500
Subject: How do I load a .png graphic file into a dialog in Irix?
Message-ID: <97jjqh$ct3$1@paperboy.CORP.Softimage.COM>

Im try to load a .png file in Irix,. Im habe to load it, he program don't
display it.
whats the problem?

where a bit of my code.

void CBaseDialog::LoadBitmap(wxString bitmapname,unsigned int nbcontrol)
 {
 #if wxUSE_LIBPNG
  wxImage::AddHandler( new wxPNGHandler );
 #endif

  wxBitmap bit;
  if (bit.LoadFile(bitmapname,wxBITMAP_TYPE_PNG))
  {
   //wxMessageBox(bitmapname,"Quit setup?", wxICON_EXCLAMATION | wxYES_NO |
wxYES_DEFAULT , NULL);
   wxStaticBitmap* bitmap =
static_cast<wxStaticBitmap*>(GetDlgItem(m_BaseDialog,nbcontrol));
   bitmap->SetBitmap(bit);
  }
 }




From leo at iems.nwu.edu  Wed Feb 28 21:46:52 2001
From: leo at iems.nwu.edu (Leonardo B Lopes)
Date: Wed, 28 Feb 2001 14:46:52 -0600
Subject: docs for mysqldb
Message-ID: <3A9D63BC.C3EB9246@iems.nwu.edu>

Sorry for the crossposting, but this actually interests both lists...

Does anyone use the python module mysqldb? Does anyone have any decent
docs for it? In particular, how do you do cursor.execute() with
different parameters?
-- 
=======================================================================
Leonardo B. Lopes                                      leo at iems.nwu.edu 
Ph.D. Student                                             (847)491-8470
IEMS - Northwestern University             http://www.iems.nwu.edu/~leo


From jfontain at free.fr  Wed Feb 28 22:17:55 2001
From: jfontain at free.fr (Jean-Luc Fontaine)
Date: Wed, 28 Feb 2001 21:17:55 GMT
Subject: ANNOUNCE: moodss-14.0
Message-ID: <3A9D6B01.83F13FB2@free.fr>

Hi everybody: here is a new version of moodss.


Check it out! (or at least the screenshots at
http://jfontain.free.fr/moodss3.gif or
http://jfontain.free.fr/moodss4.gif :).


Notes:
- moodss modules can now be written in Python! Sample randpy and
minipy modules are provided. Documentation (see randpy.py file)
needs improving. Feedback is highly welcomed.
(modules can also be written in Perl, and Tcl of course)
- please let me know if you are interested in writing modules in
Ruby (I'm on a roll :-)

### CHANGES ###

--- version 14.0 ---
modules can now be written in the Python language (the tclpython
  package is required, available in my homepage)
added Python based randpy and minipy modules

### README ###

This is moodss (Modular Object Oriented Dynamic SpreadSheet) version
14.0.

Moodss won in the Best System Admin Technology category (Tcl Tips and
Tricks, Valuable Real World Programming Examples) at the O'Reilly
Tcl/Tk 1999 Conference.
Linux Magazine calls it a "lifesaver".
Tucows gives it 5 stars (cows or penguins :-).

Moodss is a modular application. It displays data described and
updated in one or more modules, which can be specified in the command
line or dynamically loaded or unloaded while the application is
running. Data is originally displayed in tables. Graphical views
(graph, bar, 3D pie charts, ...), summary tables (with current,
average, minimum and maximum values) and free text viewers can be
created from any number of table cells, originating from any of the
displayed viewers. Thresholds can be set on any number of cells.

Specific modules can easily be developed in the Tcl, Perl and Python
scripting languages or in C.

A thorough and intuitive drag'n'drop scheme is used for most viewer
editing tasks: creation, modification, type mutation, destruction,
... and thresholds creation. Table rows can be sorted in increasing or
decreasing order by clicking on column titles. The current
configuration (modules, tables and viewers geometry, ...) can be saved
in a file at any time, and later reused through a command line switch,
thus achieving a dashboard functionality.

The module code is the link between the moodss core and the data to be
displayed. All the specific code is kept in the module package. Since
module data access is entirely customizable (through C code, Tcl,
Perl, Python, HTTP, ...) and since several modules can be loaded at
once, applications for moodss become limitless.
For example, thoroughly monitor a dynamic web server on a single
dashboard with graphs, using the Apache, MySQL, cpustats, memstats,
...  modules. If you have replicated servers, dynamically add them to
your view, even load the snmp module on the fly and let your
imagination take over...

Along with a core trace module, random, ps, cpustats, memstats,
diskstats, mounts, route, arp, kernmods, netdev, pci, system, MySQL
(myquery, mystatus, myprocs, myvars) modules for Linux, ping, snmp and
snmptrap for UNIX, apache and apachex modules are included (running
"wish moodss ps cpustats memstats" mimics the "top" application with a
graphic edge and remote monitoring capability).

Thorough help is provided through menus, widget tips, a message area,
a module help window and a global help window with a complete HTML
documentation.

Moodss is multi-langual thanks to Tcl internationalization
capabilities. So far only English and partially French are
supported. Help with other languages will be very warmly welcomed.

Development of moodss is continuing and as more features are added in
future versions, backward module code compatibility will be maintained.

I cannot thank the authors of the tkTable, BLT, MIME/SMTP and the HTML
libraries enough for their great work.

In order to run moodss, you need to install the following packages
(unless you can use the rpm utility, see below):
Tcl/Tk 8.3.1 or above, at (or at a mirror near you)
  http://dev.scriptics.com/ or ftp://ftp.scriptics.com/
the latest tkTable widget library at:
  http://www.hobbs.wservice.com/tcl/main.html
the latest BLT library at:
  ftp://tcltk.sourceforge.net/pub/tcltk/blt/
eventually the latest tclperl library for writing modules in Perl, or
the latest tclpython library for writing modules in Python at:
  http://jfontain.free.fr/
(see the INSTALL file for complete instructions, for UNIX and also
Windows platforms).

You also have the option of using the moodss rpm file (also in my
homepage), if you are using a Redhat Linux system (6.0 or above).
You can find the required tcl, tk, tktable, blt, tcpperl and other
rpms at: http://jfontain.free.fr/

Whether you like it (or hate it), please let me know. I would like to
hear about bugs and improvements you would like to see. I will correct
the bugs quickly, especially if you send me a test script (module code
with a data trace would be best).

###

you may find it now at my homepage:

http://jfontain.free.fr/moodss-14.0.tar.gz
http://jfontain.free.fr/moodss-14.0-1.i386.rpm
http://jfontain.free.fr/moodss-14.0-1.spec

and a bit later at:

ftp://contrib.redhat.com/ in libc6 sub-directory.

Enjoy and please let me know what you think.


--
Jean-Luc Fontaine  mailto:jfontain at free.fr  http://jfontain.free.fr/





From jfontain at free.fr  Wed Feb 28 22:20:04 2001
From: jfontain at free.fr (Jean-Luc Fontaine)
Date: Wed, 28 Feb 2001 21:20:04 GMT
Subject: ANNOUNCE: tclpython-1.1
Message-ID: <3A9D6B7F.D7F73316@free.fr>

### CHANGES ###

--- version 1.1 ---
include interpreter name in error messages

### README ###

tclpython version 1.1: a Python package for Tcl

This package allows the execution of Python code from a Tcl
interpreter, as in:


    package require tclpython
    set interpreter [python::interp new]
    $interpreter eval {x = 3/2.0}
    set result [$interpreter eval {print x}]
    puts $result


Sending data from the Python interpreter to the Tcl interpreter
requires the print Python command. This will hopefully no longer be
required in future versions.


This library is free software; you can redistribute it and/or modify
it under the terms of the GNU Library General Public License as
published by the Free Software Foundation; either version 2 of the
License, or (at your option) any later version.

This library is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

###

you may find it now at my homepage:

http://jfontain.free.fr/tclpython-1.1.tar.gz
http://jfontain.free.fr/tclpython-1.1-1.i386.rpm
http://jfontain.free.fr/tclpython-1.1-1.spec
http://jfontain.free.fr/tclpython.htm

Enjoy and please let me know what you think.


--
Jean-Luc Fontaine  mailto:jfontain at free.fr  http://jfontain.free.fr/





From ljohnson at resgen.com  Wed Feb 28 22:44:56 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Wed, 28 Feb 2001 15:44:56 -0600
Subject: ANN: FXPy-0.99.160, Python bindings for FOX GUI toolkit
Message-ID: <t9qs6fi6g2qq3c@corp.supernews.com>

I posted this earlier this morning but it hasn't shown up yet. Apologies in
advance if it appears multiple times...

This is an update of FXPy, a Python extension module which provides
an interface to the FOX cross-platform GUI library:

  Home Page:        http://fxpy.sourceforge.net

  Download Source:
http://download.sourceforge.net/fxpy/FXPy-0.99.160.tar.gz

  Windows Binaries:
http://download.sourceforge.net/fxpy/FXPy-0.99.160-win32.exe

To build FXPy from the sources, you will also need to download version
0.99.160 of the FOX library:

  FOX Home Page:    http://www.cfdrc.com/FOX/fox.html

  Download Source:  ftp://ftp.cfdrc.com/pub/FOX/fox-0.99.160.tar.gz
                    ftp://ftp.cfdrc.com/pub/FOX/fox-0.99.160.zip

As you may have guessed by now, the latest release of FXPy is
version 0.99.160.

What is it?
-----------

FXPy is a Python extension module which provides an interface to the FOX
cross-platform GUI library. With a few minor exceptions, FXPy provides a
complete interface to FOX. FOX is a C++-based toolkit for developing
graphical user interfaces easily and effectively. Some of the
significant features of FOX include:

 *  A rich set of widgets (including dials, shutters, tree lists, and
    many other "modern" widgets).
 *  Powerful but easy-to-use layout managers.
 *  Extensive support for 3-D modeling using OpenGL or Mesa.
 *  Supports the XDND (v4) protocol for drag-and-drop.
 *  Registry for persistent application settings.
 *  Runs natively under Unix/X and Microsoft Windows.
 *  And much, much more!

Please see the FOX home page (URL listed above) for more details about
the FOX library's features. Also consider subscribing to the foxgui-users
mailing list for the latest news about FOX development; again, see the
FOX home page for information on this.

FXPy is (c) 1999, 2000, 2001 Lyle Johnson (lyle at users.sourceforge.net) and
is
released under the GNU Lesser General Public License.

<P><A HREF="http://fxpy.sourceforge.net">FXPy-0.99.160</A> - interface to
the FOX cross-platform GUI library.  (28-Feb-01)





From Cayce at actzero.com  Wed Feb 28 23:41:21 2001
From: Cayce at actzero.com (Cayce Ullman)
Date: Wed, 28 Feb 2001 14:41:21 -0800
Subject: Best way to string compare a python object in a getattrofunc?
Message-ID: <F0D64494733BD411BB9A00D0B74A02640F1B1D@MAIL01>

Hi,  I'm writing an extension class that has a getattro method implemented
in C.  The method takes a python object as it's name, not a string.  My
question is what is the best way to do this.  For example, if I always want
my extension class to return "eggs" when attribute "spam" is requested, no
matter what is in the instance dictionary, what's the fastest way to compare
the python string object with my string constant.  Currently I'm doing it
this way:

static PyObject *
myec_getattro(myec_object *self, PyObject *name)
{
    if (strcmp(PyString_AsString(name),"spam")==0)
    {
        return PyString_FromString("eggs");
    }

    return Py_FindAttr((PyObject *)self, name);
}

This obviously requires that the every time an attribute is requested it is
converted into a char* then strcmped (obviously some type checking would
eventually be appropriate here).  This seems like a lot of work.  I am
thinking about interning the string "spam" on module initialization, then
saving that pointer somewhere, and then just comparing the pointers on every
fetch.  This approach would save me type checking, string convertion, and a
strcmp.  However, I think it might not be fool proof.  Ie, if someone
computed the attribute name "sp" + "am", this would no longer work (I don't
think that is very likely, though).  Also, if the attribute name came from
some other C module, it seems like that too could have problems.  I can't be
the first person wondering about this, is there a universally accepted best
practice here?

TIA,

Cayce



From davidlam at cableinet.co.uk  Wed Feb 28 23:41:45 2001
From: davidlam at cableinet.co.uk (bruce)
Date: Wed, 28 Feb 2001 22:41:45 GMT
Subject: python performance
Message-ID: <J8fn6.756$me1.378895@news2.cableinet.net>

Can  anyone comment on the speed of Python programs under Windows?

I am interested in using Python as a embedded game scripting language.
Amongst other things, I want to doing AI, but not the low level routines
like path finding.


TIA
Bruce




From guido at digicool.com  Wed Feb 28 23:54:18 2001
From: guido at digicool.com (Guido van Rossum)
Date: Wed, 28 Feb 2001 17:54:18 -0500
Subject: Python webring is dead
Message-ID: <200102282254.RAA31089@cj20424-a.reston1.va.home.com>

There used to be a Python webring.  But from the following exchange, I
see it is dead.  If someone wants to create a new one, please go ahead
-- if you let webmaster at python.org know, we'll add python.org to the
ring.

--Guido van Rossum (home page: http://www.python.org/~guido/)


------- Forwarded Message

Date:    Wed, 28 Feb 2001 23:02:19 +0000
From:    Manuel Gutierrez Algaba <algaba at gmx.net>
To:      Guido van Rossum <guido at digicool.com>
Subject: Re: Random python site

On Mon, 26 Feb 2001, Guido van Rossum wrote:

> > I'm the current maintainer of the Python ring. I've lost control of 
> > my Yahoo account. That is, I've lost any control over the python
> > ring. I'd suggest you to create a new ring.
> > 
> > If you don't trust me try to add a new site to the ring. Or just
> > watch when was the last adding to the ring ? 
> > Much more than 3 months. 
> >  
> > Sorry for my incompetence.
> 
> Sorry, I'm not interested in taking up more work.  If you want to
> create a new ring, that's fine.  Otherwise, we'll just drop the link.
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)
> 


Then, please, drop the link, forward this to comp.lang.python to try
if anybody is interested or just leave it, but knowing it's a 
no-end road.

- --- 
Regards/Saludos
Manolo
http://spain.50g.com

------- End of Forwarded Message




From duncan at rcp.co.uk  Thu Feb  8 11:11:58 2001
From: duncan at rcp.co.uk (Duncan Booth)
Date: Thu, 8 Feb 2001 10:11:58 +0000 (UTC)
Subject: urllib.open('http://www.redherring.com/')
References: <95th67$36j$1@nnrp1.deja.com>
Message-ID: <Xns9042677DFF798duncanrcpcouk@127.0.0.1>

apederse at my-deja.com wrote in <95th67$36j$1 at nnrp1.deja.com>:

>I have some problems with urllib. Any URL works
>just fine except http://www.redherring.com/ which
>returns nothing, no headers, no nothing. (Red
>Herring works fine from any browser).

It works fine when I try it (using Python 2.0). Changing things like the 
accept header which might be expected to make a difference don't seem to 
have any effect.


From fredrik at pythonware.com  Thu Feb 22 15:57:52 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 22 Feb 2001 14:57:52 GMT
Subject: Newbie Error? Compiling _tkinter from Source on Win32
References: <mailman.982851663.3862.python-list@python.org>
Message-ID: <QN9l6.19328$AH6.2619674@newsc.telia.net>

Chris Withers wrote:
> > --------------------Configuration: _tkinter - Win32 Release--------------------
> > Compiling...
> > _tkinter.c
> > E:\Python-2.0\Modules\_tkinter.c(44) : fatal error C1083:
> > Cannot open include file: 'tcl.h': No such file or directory
> > tkappinit.c
> > E:\Python-2.0\Modules\tkappinit.c(15) : fatal error C1083:
> > Cannot open include file: 'tcl.h': No such file or directory
> > Error executing cl.exe.
> >
> > _tkinter.pyd - 2 error(s), 0 warning(s)
>
> Any ideas what I'm doing wrong?

Does "newbie" mean C newbie?  The error message pretty much
means what it says: the compiler cannot find and open that file.

Make sure you have tcl.h somewhere, and check the compiler's
include settings.

(iirc, the distributed build file expects to find the Tcl/Tk include
files in "<python-sources>\..\tcl\include")

Cheers /F




From sheila at spamcop.net  Wed Feb 14 15:44:04 2001
From: sheila at spamcop.net (Sheila King)
Date: Wed, 14 Feb 2001 14:44:04 GMT
Subject: smtplib problem
References: <966mnl$2dl$1@uranium.btinternet.com> <xUBh6.15619$AH6.2202087@newsc.telia.net> <nk1f8tgclg4qrth40e0nd86tv08jt5l8u6@4ax.com> <k98k8tkp9k6undl9t71ogvlu0vse4pqa1k@4ax.com> <87pugl1x2c.fsf@psyche.dnsalias.org>
Message-ID: <bc6l8tod6qgtrbq8sha4qh30kst34isaeo@4ax.com>

On 14 Feb 2001 21:35:39 +1300, Carey Evans <c.evans at clear.net.nz> wrote in
comp.lang.python in article <87pugl1x2c.fsf at psyche.dnsalias.org>:

:Tim Roberts <timr at probo.com> writes:
:
:> There is no such thing.  The SMTP protocol does not include authorization.
:
:RFC2554 specifies an extension to SMTP, that adds authentication on
:message submission to an SMTP server.  BCP46 (RFC3013) section 5.4
:recommends its use.
:
:smtplib.py doesn't implement SMTP AUTH, although you could do it
:yourself using the docmd(), send() and getreply() methods.

Thank you! This is the type of tip I was looking for. Maybe in a few weeks,
when I have time, I'll mess around with it.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From rob at netzilient.com  Wed Feb 21 23:21:04 2001
From: rob at netzilient.com (Rob Eden)
Date: Wed, 21 Feb 2001 16:21:04 -0600
Subject: Interrupting blocking reads
Message-ID: <3a943dcb@news.advancenet.net>

I have an application that is reading from a number of different files
(well, external process' output streams to be precise). I need to be able to
interrupt the threads, but I don't see a way to do this other than with a
signal.  The problem is, I want to be able to interrupt an individual thread
while possibly leaving the others running.

So can you answer any of these questions:

  1) Is there a way to interrupt a blocking file read other than with a
signal?
  2) Is there a non-blocking file read?  read(0) seems to return no data.
     But it returns right away! (This is useful why?)
  3) Is there a way to get signals to work with multiple threads?
  4) Any other solutions to the problem?

Thanks!


Rob Eden




From shaleh at valinux.com  Mon Feb 19 19:52:44 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Mon, 19 Feb 2001 10:52:44 -0800
Subject: Weird Language Features
In-Reply-To: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>; from dave@dave.org.uk on Sun, Feb 18, 2001 at 01:16:49PM +0000
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <20010219105244.B29164@valinux.com>

On Sun, Feb 18, 2001 at 01:16:49PM +0000, Dave Cross wrote:
> 
> [Please watch the replies on this message as it's heavily
> cross-posted]
> 
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?
>

python will throw an exception, the code could then decide that the exception
was due to a missing function and decide how to deal with it.  I suspect this
could be useful in really restricted areas, but in general code it strikes me
as horribly confusing.
 
> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could
> you put enough intelligence into the translator so that it could
> handle certain strings differently depending on where they appeared in
> the source code?

perligata makes a LOT of assumptions for you.  it is not really latin.  I
presume it is one of the poster children for this in perl.  its fun, but a toy.

even if you could code essential language words in natural language Foo, things
like if and for, there is a wealth of code out there NOT in Foo.  So even if
I use perligata for the base of my code, when I use any other module I have to
be able to speak the language the module was written in.  So if I try to
import a module written by your friend in French, I now have to know French to
figure out the function and variable names.  As much as we hate it, a common
language is good.

As for python, the interpreter, keyword and parser are available, I suspect
this could be done with a little work.




From latlong at css2.com  Wed Feb 14 10:34:31 2001
From: latlong at css2.com (LatLong)
Date: Wed, 14 Feb 2001 09:34:31 GMT
Subject: ANN: Latitude longitude database
Message-ID: <Hisi6.5549$hN2.1055418@news3.cableinet.net>

Latitude longitude database of over 2.8 million locations worldwide.

Download from:
http://www.css2.com/latlong.htm

Ideal for sales and marketing applications and travel related websites.
Includes sample code.





From dalke at acm.org  Tue Feb 13 03:26:00 2001
From: dalke at acm.org (Andrew Dalke)
Date: Mon, 12 Feb 2001 19:26:00 -0700
Subject: Graph (cyclic kind, not pretty picture kind) rendering algorithm?
References: <rw_g6.54$q04.78888@news.pacbell.net> <3A87A0EC.BA802E3E@student.gu.edu.au>
Message-ID: <96a61a$9ln$1@slb1.atl.mindspring.net>

s713221 at student.gu.edu.au wrote:
>I don't know of any, but I'd be interested if anyone else could direct
>us to one, I need exactly this for a chem editor I'm building at the
>moment.
>
>Joal Heagney/AncientHart

Having been involved in several of these discussions,
don't use that approach.  There is a lot of chemistry
involved in making a chemical layout algorithm look
correct.

Your best bet is to look at how JMDraw
  http://vanilla.ice.mpg.de/~stein/projects/SmilesViewer/
or JChemPaint
  http://jchempaint.sourceforge.net/
do their layout.  The latter is LGPL'ed.  There isn't
much other publically available source code to look
at for development of new layout algorithms.

Even they won't be exactly what a chemist wants, because
chemists disagree with what they expect.  Eg, many
companies have guidelines on how different families
of compounds should be laid out, so the only real
solution involves a lot of substructure matching against
a set of templates.  Or you hire someone to draw
every molecule, which is done!

                    Andrew
                    dalke at acm.org





From aahz at panix.com  Tue Feb 20 19:42:16 2001
From: aahz at panix.com (Aahz Maruch)
Date: 20 Feb 2001 10:42:16 -0800
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982467140.8836.python-list@python.org> <slrn98vrqv.i0.amk@mira.erols.com> <96os85$1qe$1@panix3.panix.com> <slrn993luj.24c.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <96udq8$mh$1@panix3.panix.com>

In article <slrn993luj.24c.ssthapa at ntcs-ip45.uchicago.edu>,
 <s-thapa-11 at NOSPAM.alumni.uchicago.edu> wrote:
>Aahz Maruch <aahz at panix.com> wrote:
>>
>>Is there any chance this can be treated as a bug and fixed for the
>>release of 2.1?  Alternatively, given that we're already breaking code
>>with the change in the way complex numbers are handled, should cmp() now
>>raise an exception *every* time the type/class differs?
>
>  If you are referring to the comparision of complex numbers, I'm not
>sure how the ordering would be done.   For example consider 1 and 1j, 
>they have the same magnitude so mathematically speaking there is no
>way to consistently order them like the real numbers where magnitude
>is a more useful measure. IIRC, for any given magnitude n there are an
>infinite number of complex numbers with that magnitude, namely all n*e^(ix)
>for real x.

True enough.  Question is, in the Real World [tm], should everyone who
calls list.sort() be forced to wrap in try/except on the chance that a
pair of complex numbers will sneak in?
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

The problem with an ever-changing .sig is that you have to keep changing it


From clayberg at instantiations.com  Mon Feb 19 03:40:35 2001
From: clayberg at instantiations.com (Eric Clayberg)
Date: Sun, 18 Feb 2001 21:40:35 -0500
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <96q137$q3d$1@nntp9.atl.mindspring.net>

"Dave Cross" <dave at dave.org.uk> wrote in message
news:nnhv8t4obq30ljfgseplp7pi4khhsf9g9n at 4ax.com...
>
> [Please watch the replies on this message as it's heavily
> cross-posted]
>
> I'm doing some comparisons on programming language features and I'd be
> very interested to know how you would handle the following scenarios
> in your programming language of choice.
>
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries). Is the a feature whereby the
> programmer can create a "catch-all" function which is called in cases
> like these? Can this function examine the list of existing functions
> and call the most appropriate one? Or create a new function on the fly
> and install it into the application?

Smalltalk can do all of the above very easily. This general technique is
used to create generic proxies, for example. I have also played around with
"fault tolerant" apps that could automatically detect and correct spelling
errors in function calls (as a lark; never in production code).

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings? Imagine you wanted to allow someone to program your
> language of choice in, say, French. How would you go about translating
> French keywords into ones that the compiler (or interpreter) could
> understand. What if the translation wasn't one-to-one or fixed? Could
> you put enough intelligence into the translator so that it could
> handle certain strings differently depending on where they appeared in
> the source code?

Smalltalk can also do the above very easily. Everything is Smalltalk is an
object - including the parser, the compiler, the compiled methods and the
method source. Setting up a pre-processor or modifying the actual parser is
very easy to do. Since all of Smalltalk's control structures are built in
Smalltalk itself, you can easily alias any function name or operator to
another. Enhancing the "syntax" and adding your own control structures is
also trivial.

> If you're wondering why I'm inventing these bizarre scenarios, it's
> for a paper I'm writing for this year's Perl Conference.  Perl does
> have these features (see the AUTOLOAD function and source filters) and
> I'm interested in seeing how widespread they are in other languages.
>
> Of course, if you'd like to tell me just why you consider it's a good
> thing that your language of choice doesn't have these features, then
> I'd be only too happy to hear that too.
>
> I'd just like to make it clear that I'm not interested in getting into
> "my language is better than your language" types of flamewars. I'm
> certainly not trying to argue that Perl is better than other languages
> for having these features.
>
> Thanks for your time

Thanks for the interesting thread.

-Eric




From thomas.heller at ion-tof.com  Tue Feb 27 10:26:21 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Tue, 27 Feb 2001 10:26:21 +0100
Subject: ANN: eGenix.com mx Extensions -- Version 2.0.0 (mxODBC, mxDateTime, ...)
References: <3A958CED.F7B95C33@lemburg.com> <r7cg9toqrkgklmk6egokarju1g3mc9mei2@4ax.com>
Message-ID: <97frrt$p1a5r$1@ID-59885.news.dfncis.de>

"Paul Moore" <gustav at morpheus.demon.co.uk> wrote in message news:r7cg9toqrkgklmk6egokarju1g3mc9mei2 at 4ax.com...
> On Thu, 22 Feb 2001 23:04:29 +0100, "M.-A. Lemburg" <mal at lemburg.com>
> wrote:
>
> >The download archives and instructions for installing the packages can
> >be found at:
> >
> >     http://www.lemburg.com/files/python/
>
> One thing which would be nice would be a Win32 binary download, which
> was NOT packaged using an installer. Just a zip file to be unzipped in
> an appropriate place...
>

The windows installer (egenix-mx-base-2.0.0.win32-py2.0.exe) is more or
less a self-extracting zip.file, which also can be opened with WinZip
or other tools...

Thomas




From rvprasad at cis.ksu.edu  Mon Feb  5 21:15:05 2001
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 05 Feb 2001 14:15:05 -0600
Subject: os.popen2
Message-ID: <m3snlsrinq.fsf@boss.dreamsoft.com>

Hi,

The following code stalls

import os
i = os.popen2("/bin/bash")
i[0].write("ls -lR")
i[1].read()
-----------stalls here

Am I missing something the way it should behave or isn't it possible to execute
a process using popenX, keep it alive and interact with it?

waiting for reply,

-- 
Venkatesh Prasad Ranganath


From aleaxit at yahoo.com  Sat Feb 17 08:48:23 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 17 Feb 2001 08:48:23 +0100
Subject: Win32 COM: Passing NULL as a value
References: <904A72E79williamwickerspectra@207.126.101.100> <3A8DBDC4.4040705@ActiveState.com>
Message-ID: <96lag10eee@news2.newsguy.com>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A8DBDC4.4040705 at ActiveState.com...
> William Wicker wrote:
>
> > I am trying to work via the COM interface to an object that exposes a
> > method something like this:
> >
> > AnObject.DoSomething(doWhat, toWho)
> >
> > where toWho is either a dispatch pointer, or NULL, if toWho is not
> > significant.
> >
> > When I try to do this with
> >
> > obj = win32com.client.Dispatch("AnObject")
> > obj.DoSomething("kick", None)
>
> None is used to pass a variant with VT_NULL.  It works for every object I
have come across, so I am afraid I have no idea.

If the AnObject used canonical API's to handle its arguments
(VariantChangeType, directly or under the covers) it should
indeed work -- a VariantChangeType from VT_NULL to VT_DISPATCH
should succeed and give a VARIANT with pdispVal of 0.  So, I
guess from the symptoms described that AnObject is not being
'canonical' in its handling of arguments.

Is there a way to prepare a VARIANT directly in Python, with
VT_DISPATCH and pdispVal==0?  When met with such issues
in the past I used a little helper, an in-process COM server I
wrote that supplies a few objects for explicit fine-grained
issues of Automation ("prepare a variant with exactly this vt
tag-part and exactly this bitpattern in its value-union part"
being a good example).  But Pythoncom looks like it's powerful
enough to handle it directly if one digs deep enough, rather
than needing an auxiliary C++ coded library of objects...?


Alex





From m.faassen at vet.uu.nl  Mon Feb 12 00:18:26 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 11 Feb 2001 23:18:26 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <3A85C996.9000504@hccnet.nl>
Message-ID: <9676k2$brc$1@newshost.accu.uu.nl>

Jaap Spies <j.spies at hccnet.nl> wrote:
[snip]
> Grow up! Miracles are of another World! Are you joking?

What's the reason for such a weird response? I'm not used to such
behavior on the Python newsgroup, so please adjust. There are many languages
that more or less fulfill his criterions, and even if there weren't,
you could respond in a more polite fashion. The PSU thanks you.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From porter at et.byu.edu  Thu Feb 22 02:13:00 2001
From: porter at et.byu.edu (C. Porter Bassett)
Date: Wed, 21 Feb 2001 18:13:00 -0700 (MST)
Subject: Curses for Win32?
In-Reply-To: <3A9449BB.284DC79B@iol.ie>
References: <3A9449BB.284DC79B@iol.ie>
Message-ID: <Pine.GHP.4.21.0102211811220.28286-100000@don.et.byu.edu>

I myself would love to have a python2.0 windows curses module, but nobody
with the know-how seems to think that it is important.  

What you could do, though, is create a dummy module called curses.  You
say that you don't mind if it doesn't function.  Just create dummy
functions corresponding to the functions that your program calls.  

--------------------------------------------------------------------------
C. Porter Bassett     porter at et.byu.edu   http://www.cporterbassett.com
--------------------------------------------------------------------------
"Pretend like this is a really witty saying." - Anonymous
--------------------------------------------------------------------------

On Wed, 21 Feb 2001, Adam Lock wrote:

> I have a Python 2.0 program that uses Tkinter and curses for display. It
> works fine on Unix because there is a curses package but not on Win32
> because there isn't. Perversely even MacPython 2.0 has a curses package!
> 
> Does anyone know if a curses package exists for Win32? I would even be
> happy if it compiled but didn't function since I want will be using the
> Tkinter UI anyway.
> 
> Many thanks
> 
> -- 
> Adam Lock - locka at iol.ie
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From jh at web.de  Wed Feb  7 22:56:50 2001
From: jh at web.de (Jürgen Hermann)
Date: Wed, 7 Feb 2001 22:56:50 +0100
Subject: Touch in python?
References: <95sfla$adf$1@bunyip.cc.uq.edu.au>
Message-ID: <95sg8e$mod$00$1@news.t-online.com>

"Chui Tey" <teyc at bigfoot.com> schrieb im Newsbeitrag
news:95sfla$adf$1 at bunyip.cc.uq.edu.au...
> Is there an equivalent of 'touch' in python? I would like to change the
file
> modified time.

os.utime()




From alet at unice.fr  Tue Feb 20 17:05:43 2001
From: alet at unice.fr (Jerome Alet)
Date: Tue, 20 Feb 2001 17:05:43 +0100
Subject: [ANNOUNCE] New version of the JAXML module
References: <3A9128D8.35D0EE22@unice.fr> <slrn992gj2.j5.gerhard.nospam@lilith.hqd-internal> <3A92324B.3F7CD22D@unice.fr> <slrn995336.2pm.gerhard.nospam@lilith.hqd-internal>
Message-ID: <3A9295D7.9B2231AC@unice.fr>

Gerhard H?ring wrote:

> options. I might troll and say the GPL is less free than the BSD license,
> but let's not start this flamewar again.

No, it would be a very bad idea (tm) !-))

> Yes that's my point. If I use a GPL'd module the author forces me to GPL the
> rest of my code. That's why using it restricts me. The LGPL would protect
> *your* work, but wouldn't try to force *mine* under GPL, too. That's why I
> prefer LGPL or BSD licensed libraries.

yes but the LGPL wouldn't encourage you to write free software, since
you'd be allowed
to use it in a non free application.

> Just to be clear: Of course I respect your licensing decision. You perfectly
> know the implications of putting *libary* under the GPL. That's fine. I just
> wanted to be sure.

There's no problem. As mentionned by Martin above, I'm open to
discussion ;-)

bye,

Jerome Alet


From mmiller3 at iupui.edu  Thu Feb 15 18:49:11 2001
From: mmiller3 at iupui.edu (Michael A. Miller)
Date: 15 Feb 2001 12:49:11 -0500
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org>
Message-ID: <874rxv4z1k.fsf@lumen.med.iupui.edu>

    > From: "Chris Richard Adams"

    > I'm trying to create a loop that reads through a simple
    > file with strings on each line.

Take a look at the fileinput module:

    import fileinput
    for line in fileinput.input('filename'):
        process(line)



From joconnor at cybermesa.com  Mon Feb 26 17:59:34 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 26 Feb 2001 09:59:34 -0700
Subject: is and ==
References: <movk9t0una5nuca6upgk9rcco96fbkeu51@4ax.com> <97e17a$o3o$1@panix6.panix.com>
Message-ID: <3A9A8B76.777FCA6A@cybermesa.com>


Aahz Maruch wrote:

> In article <movk9t0una5nuca6upgk9rcco96fbkeu51 at 4ax.com>,
> Daniel Klein  <danielk at aracnet.com> wrote:
> >
> >Is there a functional difference between 'is' and '==' ?  If there is, I can't
> >find it.
>
> >>> a = 'xy'
> >>> b = a
> >>> c = 'x' + 'y'
> >>> d = 'xy'
> >>> a is b
> 1
> >>> a is d
> 1
> >>> a is c
> 0
> >>> a == c
> 1
>
> Essentially, is tests for object identity: are two variables pointing to
> the same location in memory.  == tests to see whether the *values* are
> equal.

Consider the following example

>>> class TestVal:
 def __init__(self):
  self.x = 100
 def __cmp__(self, other):
  return cmp (self.x, other.x)


>>> x = TestVal()
>>> y = TestVal()
>>> x == y
1
>>> x is y
0
>>> x =y
>>> x is y
1

x == y is true because I've defined what == means, a value comparison, but x is y
is false because they are not both the same object.  However by assign the variable
x to be y, I now get true for "x is y" because they are now the same object

Take care,
Jay




From sanner at scripps.edu  Tue Feb 27 17:44:02 2001
From: sanner at scripps.edu (Michel Sanner)
Date: Tue, 27 Feb 2001 08:44:02 -0800
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: Doug Hellmann <doughellmann@home.com>
        "[Distutils] Re: CPAN functionality for python - requirements" (Feb 26, 10:45pm)
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> 
	<01022606501904.15046@branagan> 
	<20010226183606.F7531@tummy.com> 
	<01022623140707.15046@branagan>
Message-ID: <1010227084403.ZM33944@noah.scripps.edu>

Hi,

I followed this discussion (probably not as much as I should have).
We have setup a Python distribution site for the tools we have produced for
computational biology. It is quite primitive but let's us distribute to "ready
to go" packages as well as pre-compiled Python interpreters for some platforms
to "non-wizard" users.

One issue that we are strugling with is versions of packages.Is there a
"standard" way to deal with versions ? how do you plan to find out whether
someone has a compatible version of any given package ?

-Michel

On Feb 26, 10:45pm, Doug Hellmann wrote:
> Subject: [Distutils] Re: CPAN functionality for python - requirements
> On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> > On Mon, Feb 26, 2001 at 06:45:33AM -0500, Doug Hellmann wrote:
> > >What if Python kept up with its own packages and modules like XEmacs does?
 We
> > >could then use the same format on all platforms -- it might even be
something
> > >we roll ourself to make sure we can unpack it with *only* Python installed
on a
> > >platform.
> >
> > My understanding is that that's what distutils is meant to do, to an
extent.
> > The problem is that it's not entirely unreasonable to expect that an
> > "rpm -qa" produce a list of all the software that's installed on your
> > box.  If you don't use an RPM, you can't really make use of that, and
> > packages WILL get dropped after a re-install.
>
> The command "rpm -qa" only produces meaningful results on a system where RPM
is
> the default and primary package installation system.  There are more OSes
where
> that is *not* the case (and where RPM isn't installed at all) than where it
is.
>
> It still isn't clear to me why we would want to use more than one
distribution
> package format.  It appears that one argument is so that sysadmins can use
> those tools to track what is installed on the system.  But isn't that part of
> what this new thing (cyphon?) is supposed to do?  Do we want Python to
require
> an external package management system on each platform where this tool is
> supported, or do we want a Python-esq tool which just works the way it is
> supposed to, in the same way, everywhere Python is available?
>
> But I think we're arguing over a design before we've worked out requirements.
> The basic tasks of this new tool, as I see them, are:
>
> 1. Locate Python packages you want on the net.
> 2. Resolve dependencies between what you want, what you have, and what you
>    need.
> 3. Download the appropriate packages to give you what you want.
> 4. Install those packages.
>
> Do we agree these are the primary features?
>
> I sense a consensus that the "install" part should be handled by distutils.
 Is
> that right?
>
> Other requirements we might lay down up front:
>
> 1. Should run on all platforms where Python runs.
> 2. Must support mirrors on the server side.
> 3. Need to include documentation along with source for packages.
>
> The features related to dependencies and downloads could be handled by a
> platform-specific packaging system, but integrating with all of the different
> options on all of the different platforms where Python runs increases the
> complexity of the new tool.  (How do we handle RPMs on MacOS?  How do we
handle
> HQX files on Win32?  What about for Pippy, where files as such might not even
> be appropriate?  Does Jython have any special requirements?)  Do we deal with
> this complexity by deciding on a per-platform basis what format to use, or do
> we force the user to specify the format they want downloaded?
>
> If we're modeling what we're doing based on CPAN, maybe we should look at
> that design.  What format does CPAN use when downloading Perl packages?  What
> features does CPAN have that we want?  What does it not have that we want?
>
> Doug
>
> _______________________________________________
> Distutils-SIG maillist  -  Distutils-SIG at python.org
> http://mail.python.org/mailman/listinfo/distutils-sig
>-- End of excerpt from Doug Hellmann



-- 

-----------------------------------------------------------------------

>>>>>>>>>> AREA CODE CHANGE <<<<<<<<< we are now 858 !!!!!!!

Michel F. Sanner Ph.D.                   The Scripps Research Institute
Assistant Professor			Department of Molecular Biology
					  10550 North Torrey Pines Road
Tel. (858) 784-2341				     La Jolla, CA 92037
Fax. (858) 784-2860
sanner at scripps.edu                        http://www.scripps.edu/sanner
-----------------------------------------------------------------------




From wayne.izatt at _nospam_myself.com  Tue Feb 27 08:16:22 2001
From: wayne.izatt at _nospam_myself.com (Wayne Izatt)
Date: Tue, 27 Feb 2001 07:16:22 GMT
Subject: list to string
References: <1103_983241679@cc462845-a>
Message-ID: <avIm6.174966$df5.5092180@news1.crdva1.bc.home.com>

"glen mettler" <glen.mettler at home.com> wrote in message
news:1103_983241679 at cc462845-a...
> These commands will turn a string into a list:
> myword="dog"
> mywordlist=list(myword)
> result = ["d","o","g"]
>
> is there a command/function that can take the list ["d","o","g"] and make
it a string "dog"?

>>> lst =["d", "o", "g"]
>>> lst
['d', 'o', 'g']
>>> dog = lst[0]+lst[1]+lst[2]
>>> dog
'dog'

or is there something I'm not getting?

cheers

>
> Glen
>
>
>




From jgraves3 at austin.rr.com  Tue Feb 27 03:50:38 2001
From: jgraves3 at austin.rr.com (jay graves)
Date: Tue, 27 Feb 2001 02:50:38 GMT
Subject: list to string
References: <1103_983241679@cc462845-a>
Message-ID: <3a9b14c9.33237453@news-server.austin.rr.com>

On Tue, 27 Feb 2001 02:42:12 GMT, glen mettler <glen.mettler at home.com>
wrote:

>These commands will turn a string into a list:
>myword="dog"
>mywordlist=list(myword)
>result = ["d","o","g"]
>
>is there a command/function that can take the list ["d","o","g"] and make it a string "dog"?

One way is to join them.

''.join(result)

Some people don't like using the string methods in this way and they
might write:  (at least until the string module goes away.)

import string
string.join(result,'')

Hope this helps.

...
jay


From hgg9140 at cola.ca.boeing.com  Thu Feb  1 15:56:07 2001
From: hgg9140 at cola.ca.boeing.com (Harry George)
Date: Thu, 1 Feb 2001 14:56:07 GMT
Subject: Perl-to-Python converter/translator?
References: <959jsd$njn$1@nnrp1.deja.com>
Message-ID: <xqxofwmv4e0.fsf@cola.ca.boeing.com>

I've done some perl-->python, and found (at first) that I missed some
of the perl idioms.  So I did a python module to support perl-ish
idioms.  See "pyperl" at:

  http://www.seanet.com/~hgg9140/comp/index.html

It lets you pretty much read the perl code and write the python code
on the fly.  More tedious than true converter, but do-able.  

I agree with other posters that after you get into python's own
idioms, you won't miss perl.

Lance Sloan <lsloan00 at my-deja.com> writes:

> I'm an experienced Perl programmer who has just been assigned
> a project that must be done in Python.  I'm looking for a
> Perl-to-Python translator to help me out.  I don't expect to
> write the whole project in Perl and translate it, though.  I'm
> just looking for some automation to help me see how my old work
> could be done in Python.
> 
> I didn't find anything useful via Google or Vaults of Parnassus
> searches.  So, I would appreciate any pointers!
> 
> Thanks in advance.
> 
> --
> Lance Sloan - lsloan00 at my-deja.com
> 
> --
> Lance Sloan - lsloan00 at my-deja.com
> 
> 
> Sent via Deja.com
> http://www.deja.com/

-- 
Harry George                E-mail:  harry.g.george at boeing.com
The Boeing Company          Renton:  (425) 237-6915
P. O. Box 3707  02-CA       Everett: (425) 266-3868
Seattle, WA 98124-2207      Page:    (425) 631-8803  


From andrew at andrewcooke.free-online.co.uk  Sun Feb 18 09:56:06 2001
From: andrew at andrewcooke.free-online.co.uk (Andrew Cooke)
Date: Sun, 18 Feb 2001 08:56:06 +0000
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A8F8D9B.78B37578@andrewcooke.free-online.co.uk>
Message-ID: <3A8F8E26.2D8D948F@andrewcooke.free-online.co.uk>


Andrew Cooke wrote:
[...]
> For something faster, how about C++ (although all the above have
> compilers and will be faster than C - what I mean here is a language
[...]

Sorry - faster than Python!

(and I add "probably")
(and the comment that speed isn't that important - I use and like Python
myself)
(and the comment that Python isn't that slow).

OK?

Andrew


From sandipan at vsnl.com  Tue Feb  6 15:35:10 2001
From: sandipan at vsnl.com (Sandipan Gangopadhyay)
Date: Tue, 6 Feb 2001 20:05:10 +0530
Subject: smtp mail with attachment
References: <n6Tf6.5629$tv5.438444@news.flash.net>
Message-ID: <00e001c0904a$02831aa0$020000c3@node02>

AFAIK, and I could be ignorant, but you just have to construct the mail with
MimeWriter:
With content-type as multi-part (because the email will have more than one
part (itself) - the attachment(s))
And then, in one of the parts, put in the attachment after declaring its
type.
Please note that if you are going to need to send binary attachments, base64
will come in use to encode it first. As in:
Content-Type: whatever/whatever
Content-Transfer-Encoding: base64
You can use smtplib to send the mime you have built then.

Also, you can have other parts of the email refer to this attachment by
giving it an identifier (URL sort of)

Works fine on both Unix and Windows.

HTH. Regards,

Sandipan
----- Original Message -----
From: "Scott Hathaway" <slhath at home.com.nospam>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Tuesday, February 06, 2001 7:07 PM
Subject: smtp mail with attachment


> How do I send an email with an attachment with python (I need a cross
> platform solution for Windows and Unix)?
>
> Thanks,
> Scott
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>




From baasad at qualitynet.net  Sat Feb  3 01:39:56 2001
From: baasad at qualitynet.net (bashar A. Asad)
Date: Sat, 03 Feb 2001 03:39:56 +0300
Subject: help with python
Message-ID: <95g8mv$ne21@news.qualitynet.net>

hello ;
since am new to python I would like to start my experience with a small
project that will my life easier.  I would like to connect to a remote linux
machines and check for a certain process "may be useing a system command is
their something like that in python??" eg. httpd if its up or not if it is
up the program will produce an activity file or whatever... I have no idea
how to start working on that in python..any hints guys :)
thanx

bashar



From othello at javanet.com  Wed Feb 21 01:17:03 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 20 Feb 2001 19:17:03 -0500
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <xqyk6.533$S7.49256@ruti.visi.com>
Message-ID: <3A9308FF.D07FB37F@javanet.com>

>
> >Introducing some strange behavior with '!' may very well not be worth
> >it.  However, while adding '?' as a legal identifier character does
> >not make the Python more expressive, it can make Python programs more
> >self-documenting.  The convention of ending predicate function names
> >with '?' is useful, in my opinion.
>
> Definitely.  I always found that convention in Scheme to be
> very expresive and compact.  The convention of ending
> predicates with "!" when they modified internal state was also
> useful though perhaps not as intuitive as "?".
>

These naming conventions were useful to me iin Forth and Scheme;
however, they're ugly (IMHO).  The plain English forms are much
more expressive and flexible without carrying the risk of confusing
my optical lexer when I read code.

Count my vote:  -1

Raymond

"In theory, there is no difference between theory and practice.
 In practice, there is." -- Yogi Berra




From wware at world.std.com  Sat Feb 24 13:23:09 2001
From: wware at world.std.com (Will Ware)
Date: Sat, 24 Feb 2001 12:23:09 GMT
Subject: where to learn python programing
References: <9778so$725$1@news1.Radix.Net> <mailman.982985707.26846.python-list@python.org>
Message-ID: <G99H2L.A6s@world.std.com>

On Fri, Feb 23, 2001 at 10:35:24PM -0500, Pastor Duke wrote:
> I'm very interested in learning Python programming as I'm a 3D graphics
> designer, and want to develop some games.

William Park (parkw at better.net) wrote:
> Standard answer is Tutorial which is part of standard docs.

You can find this on-line at:
http://www.python.org/doc/current/tut/tut.html

Other useful stuff here, including the Library Reference:
http://www.python.org/doc/current/

-- 
-----------------------------------+---------------------
 22nd century: Esperanto, geodesic | Will Ware
 domes, hovercrafts, metric system | wware at world.std.com


From swun at esec.com.au  Mon Feb  5 23:18:51 2001
From: swun at esec.com.au (Sam Wun)
Date: Tue, 06 Feb 2001 09:18:51 +1100
Subject: sorting on IP addresses
Message-ID: <3A7F26CB.BE85C96F@esec.com.au>

Hi,

Does anyone know what is the quickly way to sort a list of IP addresses?

ie. 203.21.254.89 should be larger than 203.21.254.9

Thanks
Sam



From annis at biostat.wisc.edu  Thu Feb 22 18:53:03 2001
From: annis at biostat.wisc.edu (William Annis)
Date: 22 Feb 2001 11:53:03 -0600
Subject: Module rfc822 - uses file-objects - why?
References: <3A954CA3.72A6D51C@gdp-group.com>
Message-ID: <jtvgq2zjsw.fsf@wazor.biostat.wisc.edu>

Thomas Volkmar Worm <worm at gdp-group.com> writes:

>                                Do I need to store every list/string
> onto disk when I want to use these classes or methods on them?
> 
> To solve the problem I started to write a helper class which implements
> methods like readline for my objects, but I really do not want to write
> an emulation of the file-class again and again for each object I create.

        You have solved a problem other people have had and solved
already.  The StringIO class -- part of standard Python -- can be used
to make strings act like files:

        http://www.python.org/doc/current/lib/module-StringIO.html

-- 
William Annis - System Administrator - Biomedical Computing Group
annis at biostat.wisc.edu                       PGP ID:1024/FBF64031
Mi parolas Esperanton - La Internacian Lingvon  www.esperanto.org


From jschmitt at vmlabs.com  Thu Feb 15 04:46:19 2001
From: jschmitt at vmlabs.com (John Schmitt)
Date: Wed, 14 Feb 2001 19:46:19 -0800
Subject: __contains__()
Message-ID: <008F0A63472BD311AF9800104BCD102561CC8B@minirex.vmlabs.com>

I did something a little cheezy.

>>>class foo:
       def __contains__(self,n):
           return 42

>>>f = foo()

>>>3 in f
1

I thought I would get back 42.  The manual doesn't say that it must strictly
return 0 or 1, just something true or false.

I wonder what the PSU has to say about this.  How does one become a memb



From wilsoncredit at email.com  Fri Feb 16 09:56:53 2001
From: wilsoncredit at email.com (wilsoncredit at email.com)
Date: Fri, 16 Feb 2001 08:56:53 GMT
Subject: ****SPECIAL REPORT****
Message-ID: <pX5j6.39556$Jp2.270876@news6.giganews.com>

--_NextPart_0000904A-000004CC-0388DEEA-5224
Content-Type: text/plain
Content-Transfer-Encoding: 7bit

Get a brand new, clean,
credit file in 15 days or less...
Get the credit you deserve...
For more info, send a email to
wilsoncredit at email.com
with "INFO" in the subject line...
--_NextPart_0000904A-000004CC-0388DEEA-5224--


From robin at jessikat.fsnet.co.uk  Sat Feb  3 12:25:23 2001
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 3 Feb 2001 11:25:23 +0000
Subject: nested scopes
References: <qUQrSUAN$we6EwEu@jessikat.fsnet.co.uk>
 <mailman.981181863.15223.python-list@python.org>
Message-ID: <467PpKAjq+e6EwFJ@jessikat.fsnet.co.uk>

In article <mailman.981181863.15223.python-list at python.org>, Tim Peters
<tim.one at home.com> writes
>[Robin Becker]
....
>WRT scopes, the determination of which names are local is entirely done at
>compile-time.  That wasn't always true, but has been true since version
>0.9.9, and was the single biggest speedup in Python's history.  The
>determination of which names are local is still done at compile-time in the
>presence of "import *" and "exec", but then the results can't be trusted,
>and the compiler generates different code that does all sorts of crap under
>the covers to give "the expected" results -- most of the time.  I doubt that
>crap will get extended to live nice with deeply nested scopes, though.
>
>> ...
>> Optimality is the enemy of stability.
>
>in-america-they-tell-us-it's-drugs<wink>-ly y'rs  - tim
>
>
as I suspected this is about optimality; speed vs clarity/simplicity is
the argument here. I guess that the nested scopes stuff has just brought
it to the fore. Luckily we're not forced to use nested scopes in python,
pascal experience makes me despise them because of all that paging up
and down to find the defining declaration. 
-- 
Robin Becker


From sl0op at my-deja.com  Sat Feb 10 23:26:06 2001
From: sl0op at my-deja.com (sl0op at my-deja.com)
Date: Sat, 10 Feb 2001 22:26:06 GMT
Subject: Python compiler/installer?
Message-ID: <964f5n$oqs$1@nnrp1.deja.com>

Hi. I'm new to Python (which seems like a very nice language to me) and
have a few questions. First, is there any way I can compile it so I
don't have to send out any source code I don't want to? Are there any
installers for either compiled-Python (if that exists) or simple .py
files? And where can I find some examples of Python programs ?(some
that will help me learn Tkinter.)

Any help would be greatly appreciated.

Thank you,
Jason


Sent via Deja.com
http://www.deja.com/


From vasvir at iit.demokritos.gr  Mon Feb 26 06:50:35 2001
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Mon, 26 Feb 2001 07:50:35 +0200
Subject: abusing __builtins__
References: <mailman.983142547.10845.python-list@python.org>
Message-ID: <3A99EEAB.C756064C@iit.demokritos.gr>

Tim Peters wrote:
> Cute!
> 
> 
> The language doesn't guarantee this will always work, so it's certainly
> abuse.

Oups! That's the killer argument here.

> 
> Looking up builtins is slower than looking up module globals (a module's
> global dict is always searched before the builtin dict).

certainly

 
> The language may someday make "GlobalVariable" the name of its own builtin,
> and then your code won't work correctly if you either try to use the new
> builtin or call any other module that does (incl. without limitation std
> library modules).

This I don't get it. Why is that? GlobalVariable is just a name to
indicate the purpose of my variable. If a newer version of python use it
in a later stage as a reversed word (__builtins__ variable or function),
then obviously my old program won't be able to use the new functionality
but it will overwrite it happily. So it would work even then.

Consider the following imaginary example. In a project where there is no
need to open files one replaces the open function. So this project will
never manage to open a file but aside from that it will continue to work
as designed.

 
> People other than you will be unable to understand your code regardless,
> because GlobalVariable will look to them like an unbound variable (i.e.,
> nobody else uses this trick, so nobody else will recognize the usage
> pattern).
> 
> Introspective tools won't think to look in the builtins either (again
> because nobody else etc).
> 

Valid points too.



Thank you and everybody else for your answers. I think I will try to
avoid that trick anyway.

	.Bill


From aahz at panix.com  Mon Feb 26 20:35:15 2001
From: aahz at panix.com (Aahz Maruch)
Date: 26 Feb 2001 11:35:15 -0800
Subject: Splitting comp.lang.python
References: <mailman.983199848.8916.python-list@python.org> <Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com> <79ym6.2288$TW.12847@tor-nn1.netcom.ca>
Message-ID: <97eb5j$cpr$1@panix2.panix.com>

In article <79ym6.2288$TW.12847 at tor-nn1.netcom.ca>,
Warren Postma <embed at geocities.com> wrote:
>
>I mean, doesn't anyone search Deja anymore? Oh yeah, Deja is gone, and
>replaced by the Beta Groups at groups.google.com. Oh well, ironically, now
>that we have no Deja.com, I think we are in for a Lot More of that old time
>Deja Vu.

Well, at least until Google brings up the full DejaNews archive.  I've
heard that the initial estimate is on the order of three months.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

"I used to have a .sig but I found it impossible to please everyone..."  --SFJ


From dsh8290 at rit.edu  Wed Feb 14 19:29:55 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 14 Feb 2001 13:29:55 -0500
Subject: pth information
Message-ID: <20010214132954.A17229@harmony.cs.rit.edu>

I was searching through the documentation on python.org, but couldn't
find anything describing the usage and syntax for .pth files.  All I
managed to find was a brief mention of their existence in some
extension modules in section 3.23 of the Library Reference.

Could someone point me in the right direction? (Or explain here if you
feel inclined to do so)

Thanks,
-D




From j.spies at hccnet.nl  Mon Feb 12 16:26:12 2001
From: j.spies at hccnet.nl (Jaap Spies)
Date: Mon, 12 Feb 2001 16:26:12 +0100
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com> <3A85C996.9000504@hccnet.nl> <9676k2$brc$1@newshost.accu.uu.nl>
Message-ID: <3A880094.6060906@hccnet.nl>

Martijn Faassen wrote:

> Jaap Spies <j.spies at hccnet.nl> wrote:
> [snip]
> 
>> Grow up! Miracles are of another World! Are you joking?
> 
> 
> What's the reason for such a weird response? I'm not used to such
> behavior on the Python newsgroup, so please adjust. There are many languages
> that more or less fulfill his criterions, and even if there weren't,
> you could respond in a more polite fashion. The PSU thanks you.
> 
> Regards,
> 
> Martijn

You'r right! I was a bit short (tempered).

But I don't agree that there are many languages (and environments)
that fulfill the stated criterions. If any. Python will be close.

I followed the thread and I admire all the kind answers.
Please accept my apologies.

Jaap Spies



From fredericbonnet at free.fr  Wed Feb 21 11:19:28 2001
From: fredericbonnet at free.fr (Frederic BONNET)
Date: Wed, 21 Feb 2001 10:19:28 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com> <r5iu8tsqccj65ote3hfgo6dvjphskn15l3@4ax.com> <slrn990n47.382.hs@paradise.nirvananet> <3A90FAD9.B54BCFF6@cs.man.ac.uk> <3A939540.19FB5346@free.fr>
Message-ID: <3A939613.2A32E483@free.fr>

Oh, and this one, too:

General Massu: "One should eliminate all the idiots"
De Gaulle: "Vast program..."

-- 
Fr?d?ric BONNET                          fredericbonnet at free.fr
---------------------------------------------------------------
"Theory may inform, but Practice convinces"
                                                    George Bain


From greg at cosc.canterbury.ac.nz  Fri Feb 16 04:24:14 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 16 Feb 2001 16:24:14 +1300
Subject: __contains__()
References: <mailman.982208777.26211.python-list@python.org>
Message-ID: <3A8C9D5E.9AEBECB7@cosc.canterbury.ac.nz>

John Schmitt wrote:
> 
> The manual doesn't say that it must strictly
> return 0 or 1, just something true or false.

Which is what you're doing, and it's interpreting it as
"true". It also doesn't say that "x in y" will return
the exact value returned by y's __contains()__ method,
either.

The reason is that, internally, it's being funneled
through the C version of the __contains__ method in
the typeobject, which returns a C int, not a general
Python object. The interpreter then converts that into
a canonical Python boolean value.

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From dave at dave.org.uk  Sun Feb 18 18:24:19 2001
From: dave at dave.org.uk (Dave Cross)
Date: Sun, 18 Feb 2001 17:24:19 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <3iTj6.1813$E41.169931@news.uswest.net>
Message-ID: <c7109to8ccq3r4gmpqq1eqf9io8co0vhum@4ax.com>

On Sun, 18 Feb 2001 09:09:20 -0800, "Robert Hand"
<Robert_Hand at notspammehotmail.com> wrote:

>>
>> If you're wondering why I'm inventing these bizarre scenarios, it's
>> for a paper I'm writing for this year's Perl Conference.  Perl does
>> have these features (see the AUTOLOAD function and source filters) and
>> I'm interested in seeing how widespread they are in other languages.
>>
>Really?  Thats great!  If you can do it in perl, you can probably implement
>it in java.
>I'm not a perl expert.  Is the perl source code translation a function of
>perl or a separate app?
>if it's a separate app, handling ascii input, give me a link.

It's a separate library that gets to see the source code before it's
compiled. For an (admittedly extreme) example of what you can do see:

<http://www.csse.monash.edu.au/~damian/papers/HTML/Perligata.html>

Which discusses a filter allowing you to write Perl programs in Latin!

Cheers,

Dave...

-- 
<http://www.dave.org.uk>  SMS: sms at dave.org.uk


From donn at u.washington.edu  Wed Feb 21 19:40:57 2001
From: donn at u.washington.edu (Donn Cave)
Date: 21 Feb 2001 18:40:57 GMT
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu>
Message-ID: <97123p$ojg$1@nntp6.u.washington.edu>

Quoth Joseph Holland King <insanc at cc.gatech.edu>:
| Joseph Holland King <insanc at cc.gatech.edu> wrote:
| : is there anyway to lock a file so that two different programs cannot access
| : the file at the same time. ie, my python module is editting file foo, and
| : at the same time user x (or program x) tries to read from it. is there 
| : anyway that the python module can prevent x from accessing the file for
| : a given amount of time?
|
| to clarify: i am on a unix system, i have tried to use flock however when i
| locked the file i was still able to read, write and copy the file, none of 
| which i want to happen while the file is locked. while using the flock 
| method i was using LOCK_EX. also the third party program is not going to be
| controlled, written, or have anything to do with mine. am i just not using
| flock properly or is there something else that might do the job? thank you.

Basically, that's how file locking works.  It's advisory, for the benefit
of software components that are designed to work together.

But there is a variation that you can find supported on some filesystems,
where in combination with an obscure permission mode on the file the lock
becomes mandatory and effective against any and all I/O.  The permission
trick is setgid & not group execute, i.e., (oldperm | 02000) & ~010.  It
works for me on Digital UNIX local (AdvFS) filesystem.  I didn't try on
NFS, but from the documentation I doubt that it will work there.  It does
not work on NetBSD 1.5 local filesystem.  Where it does work, you must
use fcntl() or lockf() to obtain the lock, not a genuine flock().

While I'm here, let me add a little note about flock(), for the benefit
of those whose UNIX platforms don't have the deluxe man pages that you
get on NetBSD for example.  There are two kinds of locks, one obtained
by Berkeley flock() and the other by POSIX 1003.1 fcntl().  The lockf()
function is a different, X/Open interface that interoperates with fcntl().
Many modern platforms do not support a genuine flock().  Some of them
provide an flock() function that is actually implemented with fcntl(),
which means that the semantics of the lock are per fcntl() and not
flock().  Python follows this trend and does the same thing if there's
no flock(2), so it's up to the programmer to find out what kind of lock
fcntl.flock() really gets, if the difference is important.

	Donn Cave, donn at u.washington.edu


From fredrik at effbot.org  Mon Feb  5 02:03:35 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Mon, 05 Feb 2001 01:03:35 GMT
Subject: How do I know all thrown exceptions of a function?
References: <g3lb49.73s.ln@127.0.0.1> <94d93h$55d$1@troll.powertech.no> <mailman.980059223.25947.python-list@python.org> <Qeya6.7932$AH6.1260668@newsc.telia.net> <3A6B24F0.EA164698@gte.net> <94icr2$6kv$1@animus.fel.iae.nl> <3A6CDBDA.C3120F98@gte.net> <94ipci$afo$1@animus.fel.iae.nl> <3A6CFD8E.BD473DFD@gte.net> <iEab6.8765$Qb7.1233360@newsb.telia.net> <3A6D54CB.2261D535@gte.net> <94jo9u$h45$1@nnrp1.deja.com> <94jt7e01fil@news1.newsguy.com> <Zvjb6.8419$AH6.1356839@newsc.telia.net> <pv8znm7zq2.fsf@agni.research.nokia.com>
Message-ID: <HZmf6.12204$AH6.1891634@newsc.telia.net>

Pekka Pessi wrote:
> >might so be, but I still cannot make any sense of his post.
>
>         He says that according to the docs, "from tkinter include *" should
>         be used instead of "include tkinter".  The latter adds lot of
>         unnecessary text, ie. cruft, into your program.  Cruft makes your
>         code harder to read.

umm.  *I* wrote the part on Tkinter, and I'm pretty sure I
know what I meant with the original text.  still don't know
what Steve meant, and frankly, I don't really care.

Cheers /F




From guido at digicool.com  Wed Feb 28 23:54:18 2001
From: guido at digicool.com (Guido van Rossum)
Date: Wed, 28 Feb 2001 17:54:18 -0500
Subject: Python webring is dead
Message-ID: <200102282254.RAA31089@cj20424-a.reston1.va.home.com>

There used to be a Python webring.  But from the following exchange, I
see it is dead.  If someone wants to create a new one, please go ahead
-- if you let webmaster at python.org know, we'll add python.org to the
ring.

--Guido van Rossum (home page: http://www.python.org/~guido/)


------- Forwarded Message

Date:    Wed, 28 Feb 2001 23:02:19 +0000
From:    Manuel Gutierrez Algaba <algaba at gmx.net>
To:      Guido van Rossum <guido at digicool.com>
Subject: Re: Random python site

On Mon, 26 Feb 2001, Guido van Rossum wrote:

> > I'm the current maintainer of the Python ring. I've lost control of 
> > my Yahoo account. That is, I've lost any control over the python
> > ring. I'd suggest you to create a new ring.
> > 
> > If you don't trust me try to add a new site to the ring. Or just
> > watch when was the last adding to the ring ? 
> > Much more than 3 months. 
> >  
> > Sorry for my incompetence.
> 
> Sorry, I'm not interested in taking up more work.  If you want to
> create a new ring, that's fine.  Otherwise, we'll just drop the link.
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)
> 


Then, please, drop the link, forward this to comp.lang.python to try
if anybody is interested or just leave it, but knowing it's a 
no-end road.

- --- 
Regards/Saludos
Manolo
http://spain.50g.com

------- End of Forwarded Message




From nvithadt at bellsouth.net  Wed Feb  7 14:48:51 2001
From: nvithadt at bellsouth.net (nvithadt at bellsouth.net)
Date: Wed, 7 Feb 2001 08:48:51 -0500
Subject: using external dll's in Python 2.0
Message-ID: <kncg6.892$EA5.74766@news1.atl>

Hey folk,
    I want to be able to use some 3rd party dll in a python 2.0 program I'm writing. I've searched through the message threads and came across a solution using something called calldll that was available on the parnassus(sp?) site. However, this seems to over useful for python 1.5 . After modifying the makefile script and using vc++ 6.0 I can get a calldll.pyd to be created but whenever I try to import it into my project python dies. I'm fairly new to python as well. I was hoping maybe in the win32 extension libraries there would be some info but alas no. There is LoadLibrary and GetProcAddress but seems to be no way to execute a function pointer. I really would hope I do not have to write a python extension module since there are a lot of functions in the dll's that I would like to use.

TIA, Jason
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20010207/1c6cae76/attachment-0001.htm>

From wilson.austin.aj at bhp.com.au  Mon Feb 19 20:36:03 2001
From: wilson.austin.aj at bhp.com.au (Austin Wilson)
Date: Tue, 20 Feb 2001 06:36:03 +1100
Subject: Problem with py2exe
Message-ID: <96rsie$dsc$1@gossamer.itmel.bhp.com.au>

Hi

I am trying to freeze a program using py2exe 0.2.2.  However,
I am getting the following output.  Can anyone help.

BTW:  I am using Python20 and PIL 1.1.1 on Win98.

Thanks
Austin

Resolving binary dependencies:
Traceback (most recent call last):
  File "buildwatcher.py", line 6, in ?
    scripts=["watcher.py"],
  File "c:\program files\python20\lib\distutils\core.py", line 138, in setup
    dist.run_commands()
  File "c:\program files\python20\lib\distutils\dist.py", line 829, in
run_commands
    self.run_command(cmd)
  File "c:\program files\python20\lib\distutils\dist.py", line 849, in
run_command
    cmd_obj.run()
  File "py2exe\py2exe.py", line 354, in run
    dlls, unfriendly_dlls = self.find_dependend_dlls(use_runw, dlls)
  File "py2exe\py2exe.py", line 464, in find_dependend_dlls
    alldlls, warnings = bin_depends(loadpath, images)
  File "py2exe\py2exe.py", line 674, in bin_depends
    for result in py2exe_util.depends(image, loadpath).items():
py2exe_util.bind_error: tcl83.dll






From me at nospam.net  Sat Feb 24 21:13:23 2001
From: me at nospam.net (Scottie)
Date: Sat, 24 Feb 2001 12:13:23 -0800
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <mailman.982608802.22541.python-list@python.org>
Message-ID: <t9g5fd3qc51df@corp.supernews.com>

The "Mesa" language (used as a system language at Xerox Parc in the
70s) allowed this same fully general exception/interrupt structure.
their experience, after a number of years, was that it did not work out
so well.  Finally, they looked at all of their code, found almost all of
the uses of "fix and continue" could be easily converted to the same
kind of mechanism that Python now uses.  Of the few (five, as I recall)
remaining instances, one had a bug.  So, the last few were rewritten
and the feature was removed from the language.

"Sean 'Shaleh' Perry" <shaleh at valinux.com> wrote in message
news:mailman.982608802.22541.python-list at python.org...
> On Sun, Feb 18, 2001 at 01:16:49PM +0000, Dave Cross wrote:
> > ...1/ The programmer calls a function that doesn't actually exist within
> > the application (or libraries). Is the a feature whereby the
> > programmer can create a "catch-all" function which is called in cases
> > like these? Can this function examine the list of existing functions
> > and call the most appropriate one? Or create a new function on the fly
> > and install it into the application?
> ...python will throw an exception, the code could then decide that the
> exception was due to a missing function and decide how to deal with it.
> I suspect this could be useful in really restricted areas, but in general
> code it strikes me as horribly confusing.

-Scott





From bobhicks at adelphia.net  Thu Feb  1 19:27:14 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Thu, 01 Feb 2001 18:27:14 GMT
Subject: function attributes are like function objects
References: <Pine.GSO.4.21.0101242015290.2444-100000@z.glue.umd.edu> <mailman.980400568.16791.python-list@python.org> <3A78AE8D.C9D25694@cosc.canterbury.ac.nz> <m3g0hy7fn9.fsf@atrus.jesus.cam.ac.uk> <FXge6.10628$AH6.1747678@newsc.telia.net>
Message-ID: <B69F14A9.9B8D%bobhicks@adelphia.net>


> From: "Fredrik Lundh" <fredrik at effbot.org>
> Organization: Telia Internet
> Newsgroups: comp.lang.python
> Date: Thu, 01 Feb 2001 17:22:45 GMT
> Subject: Re: function attributes are like function objects
> 
> Michael Hudson wrote:
>>> What's so special about functions that we need to
>>> be able to plonk arbitrary attributes on them, but
>>> not any other builtin types?
>> 
>> The fact that people were using the one attribute they could get at
>> (__doc__) for things far from its original purpose?
> 
> well, if people are too lazy to write
> 
> def f(...):
> ...
> a[f] = "something"
> 
> do you really think they'll find it much easier to write:
> 
> def f(...):
> ...
> f.a = "something"
> 
> (fwiw, I think function attributes is about the dumbest
> thing we've added to python lately -- face it, "because
> we can" isn't a very good argument when deciding what
> to add to the language core...)

Was that the rational? Let's add it because we can? 



From MarkH at ActiveState.com  Wed Feb 14 00:07:29 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Tue, 13 Feb 2001 23:07:29 GMT
Subject: Please Help:  COM Events
References: <96c7de$mb9$1@bob.news.rcn.net>
Message-ID: <3A89BD4F.3040708@ActiveState.com>

Ben Shapiro wrote:


> Seems to me that when the event handler is called, the argument passed in
> loses it's type as a JabberMsg and becomes a generic Dispatch object
> so i can't call properties or methods on it.. Is there any way to cast the
> argument back to a JabberMsg object?

You should be able to say:

Msg = win32com.client.Dispatch(Msg)

However, event handlers are supposed to do this for you automatically.  Hrm.  No time to investigate ATM - I am busy adding support for vtable implemented interfaces to win32com ;-)

Mark.

-- 
Nominate Your Favorite Programmers
for the Perl and Python Active Awards!
http://www.ActiveState.com/Awards/



From jhylton at my-deja.com  Fri Feb  9 20:48:00 2001
From: jhylton at my-deja.com (Jeremy Hylton)
Date: Fri, 09 Feb 2001 19:48:00 GMT
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
Message-ID: <961hhc$gp1$1@nnrp1.deja.com>

In article <3dg0hn8zlr.fsf at ute.cnri.reston.va.us>,
  Andrew Kuchling <akuchlin at mems-exchange.org> wrote:

> Should the full text of PEPs be posted at some point?

I think the text would be too long to post directly.  A URL should be
sufficient.

--
-- Jeremy Hylton, <http://www.python.org/~jeremy/>


Sent via Deja.com
http://www.deja.com/


From aleaxit at yahoo.com  Sat Feb 24 17:37:34 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 24 Feb 2001 17:37:34 +0100
Subject: Suffering For Your Art
References: <n2rl6.1686$v5.5388@newsfeeds.bigpond.com> <a8bd9tkvianmkqtg5c3qg33i8igq47n6kj@4ax.com>
Message-ID: <978oih02b74@news1.newsguy.com>

<mkx at excite.com> wrote in message
news:a8bd9tkvianmkqtg5c3qg33i8igq47n6kj at 4ax.com...
    [snip]
> >IMHO I think Python is a marvellously eloquant and beautiful
> >language because it does not get in the road of programming. It pisses me
> >off if I have to consult a tome of a help file/manual to find some
obscure
> >reference to do something that should have been quite simple ...
>
> Other environments allow the true computer geeks to flex their brain
> muscles, in an attempt to impress the intellectually-poor masses.

So does Python (I'm guilty of my share of "clever" solutions posted
to this group), but in Python's case it's clearer that "being clever" is
a temptation to be *wisely resisted* -- that "the simplest thing that
can possibly work" is the RIGHT way to go (and I try to recall to
point this out just about every time I come up with "cleverness"
that would likely be a very bad idea to use in production code...:-).


Alex





From m.faassen at vet.uu.nl  Wed Feb 28 00:43:32 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 27 Feb 2001 23:43:32 GMT
Subject: ANNOUNCE: Python - The RPG
Message-ID: <97he34$jhc$1@newshost.accu.uu.nl>

Python - The RPG

The Hype

  A Python role playing game! Play your favorite Python community
  member! While said member watches! Travel through time! Save the
  planet!

The Scenario

  The aliens are not expected to unleash their whitespace eating
  nanovirus to Earth for decades, right? There is no Python Secret
  Underground trying to put a stop to this, right?

When? Where?

  This is to be determined. Tentative time is the evening of wednesday
  the 7th of january.
 
Participants

  * Benevolent Dictator For the Game (BDFG)

  * A number of Players playing Characters

  * Members of the public (if sufficiently quiet)

Sounds Great! How can I play?

  Mail the BDFG at m.faassen at vet.uu.nl with a convincing explanation
  on why you want to play a particular character. Don't mail him after
  sunday; it's no use -- talk to him at the conference. Playing
  yourself is illegal; you have to pick someone else. Write-in
  characters are allowed and encouraged, if you can convince the
  BDFG. If you don't get to play a character, never fear; members of
  the public have voting rights if deemed Worthy.

  The use of humor is a good way to convince the BDFG.

Cast of Possible Characters

  Suggestions for new characters and/or character attributes are
  welcome. If people feel offended, sprinkle the following generously
  with smileys and <winks> in your mind. Yes, you can play any of
  these characters and others, if you manage to convince the BDFG.

  The advantages and disadvantages are expected to be roleplayed,
  mainly. They may also influence people's voting patterns during task
  resolution (see the rules section).

  Guido van Rossum (AKA BDFL)
  
    Advantages:
  
      * BDFL. Other Python Community Members will tend to listen to
        Guido. Guido can tell them what to do, and especially what
        they shouldn't do. When team members are deciding on an
        action, Guido can vote -1, thereby overruling the team members
        (note that this does not affect success determination, just
        whether the team will try something).

      * Access to time machine. Years ago Guido received a mysterious
        envelope containing the complete plans to a working time
        machine. He uses it mainly to implement features for Python
        before they're even suggested on comp.lang.python. Of course,
        the PSU uses it as well.

    Disadvantages

      * Dutch Bluntness. His Dutch bluntness may be a disadvantage
        when negotiating with NPCs.
        
  Tim Peters (AKA the timbot)

    Advantages:

      * Channeling Guido. Whenever Guido is absent, Tim Peters can
        Channel Guido. Tim can use this to gain extra authority over
        any Python Community Member. Tim can vote the overruling -1 in
        the absence of Guido.

      * Invisibility. Tim can turn invisible for a short while,
        allowing him to hide his secret identity, and to attend IPC8
        (last year's Python conference) without being seen.

    Disadvantages:

      * Robot. Tim's secret identity is that he's a robot, which may
        shock some people. Not just an android robot like the effbot
        and the martellibot -- those look like real people. No, the
        timbot is a genuine lumbering metallic robot with blinking
        lights, and a heap of Perl scripts running the innards.

      * Distracted by side issues. Tim can go on at length on issues
        which are not really the core of the problem, complicating
        said problem for himself and everybody else.
  
  Fredrik Lundh (AKA the effbot)

    Advantages:

      * Gadgeteer. A problem? The effbot can frequently construct a
        small gadget to solve it, often enlightening people in the
        process.

    Disadvantages:

      * Grouchiness. The effbot does not suffer fools gladly. This may
        may complicate his interaction with some NPCs, as well as
        other team members.

  Alex Martelli (AKA the martellibot)

    Advantages:

      * His erudite, knowledgable and *lengthy* discourse. This will
        either convince NPCs, or otherwise make them fall asleep
        (which in the case of enemies is good).

    Disadvantages:

      * Erupts into berserker rage when reminded of the print >>
        syntax.
  
  Jim Fulton

    Advantages:

      * Engineering. Ability to understand and build amazingly
        complicated machinery.

    Disadvantages:

      * Can inadvertantly make people's heads explode when attempting
        to explain how something works, rendering them unable to think
        very well for a while.

  Christian Tismer

    Advantages:

      * Can accomplish the impossible by changing the meaning of truth.
        Mystic ability influencing the fabric of reality.

      * Inspired Koreans. Christian can inspire hundreds of Koreans to
        the Pythonic cause. This should come in handy for something.

    Disadvantage:

      * A hard time explaining to other people what he's
        doing, or convincing them. Particilarly Guido.  

Rules

  The style of play is very light on actual rules. The only mechanical
  thing that may sometimes occur is task resolution by voting.

  Whenever a character wants to accomplish some task in the game, the
  BDFG decides whether success is trivial or not. For trivial success
  (for instance, adding simple a bug fix to the standard Python
  library) or trivial failure (for instance, adding a Perl style $ to
  the Python language) the BDFG just says whether something succeeds
  or fails, and that's all there is to it.

  In case of non-trivial success or failure, the BDFG has the option
  to put the resolution to a vote. In a vote each player can issue one
  vote, -1, 0, and +1. The BDFG can vote as well.

  If the total vote count is above 0, the action succeeds. If it's
  negative, the action fails. If it's exactly 0, a coin is tossed to
  determine success.

  If there is a public, members of the public who are deemed Worthy by
  the BDFG can participate in the voting process. Members of the
  public who are too loud at critical moments will likely not be
  deemed Worthy.

  Finally, if the BDFG thinks it's necessary, the BDFG can overrule
  the outcome of the vote.

References

   The PSU's Existence Revealed

    http://groups.yahoo.com/group/python-list/message/66070

  Roswell: The Truth

    http://groups.yahoo.com/group/python-list/message/87088

  Python Labs moves to Digital Creations

    http://groups.google.com/groups?q=Roswell&rnum=2&seld=971292152&ic=1

  Status of the PSU

    http://groups.google.com/groups?q=alien+whitespace&seld=941359988&ic=1

do-I-really-want-to-post-this-oh-well-here-goes-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From nyp at njc.ch  Sun Feb 25 07:26:06 2001
From: nyp at njc.ch (Yves Perrenoud)
Date: Sat, 24 Feb 2001 22:26:06 -0800
Subject: Threading problems with SWIG generated code
Message-ID: <3A98A57E.49EBCE99@njc.ch>

I've ran into some problems when trying to use multiple threads which
make use of modules generated with SWIG. Here's the shortest example I
could come up with to illustrate the problem:

Given the following SWIG interface file:

----
%module blip

unsigned int sleep(unsigned int seconds);
----

I generate the C code using "swig -python", compile it and generate the
shared library.

Here's some sample code using the newly generated "blip" module:

----
import threading, blip

def foo():
    blip.sleep(5)

threading.Thread(target=foo).start()
threading.Thread(target=foo).start()
----

You would expect this function to run for a total of approximately 5
seconds, but instead, it runs for 10 seconds! Clearly, the sleep
function calls are running one after the other, when they should be
running in parallel. It's almost as if I was acquiring a lock before the
sleep and releasing it afterwards.

Any idea what's causing this?

 ..Yves


From mix77 at usa.net  Thu Feb  8 10:26:38 2001
From: mix77 at usa.net (Mix)
Date: Thu, 8 Feb 2001 11:26:38 +0200
Subject: CGI-Python Image viewing
Message-ID: <95tp4p$bev$1@ctb-nnrp2.saix.net>

I am new to CGI scripting AND Python. What I am trying to do is,
display an image on an html page that has been retrieved from a Postgre
database
using Python (cgi). My problem is : that image that not show up on the page,
regardless of where it is sitting on my html or cgi directory!
(I am running Apache Web server under Redhat 7.0 with the default linux
kernel).
But, I CAN display the image on a plain html page (i.e without the
scripting).

Please help!
Thanks




From echuck at mindspring.com  Fri Feb  2 19:53:27 2001
From: echuck at mindspring.com (Chuck Esterbrook)
Date: Fri, 02 Feb 2001 13:53:27 -0500
Subject: Comparison oddities
Message-ID: <5.0.2.1.0.20010202134528.05141ad0@mail.mindspring.com>

It appears that tuples and strings are "greater than" lists.

 >python
ActivePython 2.0, build 202 (ActiveState Tool Corp.)
based on Python 2.0 (#8, Oct 19 2000, 11:30:05) [MSC 32 bit (Intel)] on win32
 >>> a = (1, 1)
 >>> b = [1, 1]
 >>> a < b
0
 >>> a > b
1
 >>> a==b
0
 >>> a = '1'
 >>> a < b
0
 >>> a > b
1


Does this seem a little odd to anyone?

Since tuples and lists are both "sequences of anything", you would hope 
that they would compare in "the natural fashion".

Since comparing a string and a list doesn't make sense, would an exception 
be more appropriate?

And yes, I know the manual says "Otherwise, objects of different types 
*always* compare unequal, and are ordered consistently but arbitrarily. ... 
In the future, the comparison rules for objects of different types are 
likely to change."

My question is more of a design question.


-Chuck




From mrq at for.mat.bham.ac.uk  Thu Feb  8 18:38:19 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Thu, 8 Feb 2001 17:38:19 +0000
Subject: Some basic questions about Tkinter (and Python)
In-Reply-To: <5A26DA93165A7B85.352CA80CD738363E.9B52D4C723B787A0@lp.airnews.net>
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102071351520.20781-100000@babbage> <7mdg6.13217$AH6.1992583@newsc.telia.net> <5A26DA93165A7B85.352CA80CD738363E.9B52D4C723B787A0@lp.airnews.net>
Message-ID: <Pine.GSO.4.21.0102081734460.16415-100000@babbage>

On 7 Feb 2001, Cameron Laird wrote:

> So I guess the answer is that we don't have a
> Tkinter intro for those new to Python ...  Per-
> haps it's best to recommend that Mr. Quick
> simply buy John Grayson's book without more
> delay.

You're not the first person to recommend me this book on this
newsgroup.  Unfortunately, we don't seem to have it in the library at the
university where I work, as otherwise it would be in my hands by now.  I
know you say buy the book, but I can't afford that - perhaps you earn more
than I do?

I have a friend who is a computer programmer (and definitely earns more
than I do ;-)  who recommended that I give python a try.  He is going to
lend me his copy of a book on python on Saturday - Hammond & Robinson,
"Python Programming on Win32".  I don't know whether this has sufficient
info on Tkinter to help, but it's about the only non-online source that I
can get hold of.

Martyn

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick





From danielk at aracnet.com  Sun Feb  4 17:20:24 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Sun, 04 Feb 2001 08:20:24 -0800
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com>
Message-ID: <3iuq7t06864beg5hpg1tjcbvo1m1uq81iv@4ax.com>

On Sun, 04 Feb 2001 14:19:04 +0000, Jay O'Connor <joconnor at cybermesa.com>
wrote:

>Smalltalk-style cascade operations would be *very* cool
>
>win = GtkWindow();
>	set_title("Hello World");
>	set_name ("window");
>	set_usize (400,200)
>
>versus
>
>win = GtkWindow()
>win.set_title("Hello World")
>win.set_name ("window")
>win.set_usize (400,200)


That's cos Smalltalk returns 'self' by default when a there is no explicit
return value. This can be done in Python if your 'set' methods return 'self'
instead of 'None'. For example:

>>> class A:
	def m1(self, t1):
		self.t1 = t1
		return self
	def m2(self, t2):
		self.t2 = t2
		return self
	def m3(self, h, w):
		self.h = h
		self.w = w
		return self
	
>>> a = A().m1("Hello World").m2("window").m3(400,200)
>>> a.h
400

pythonic-ly yr's,

Daniel Klein
Portland OR USA



From jwbnews at scandaroon.com  Fri Feb 23 06:16:19 2001
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Thu, 22 Feb 2001 21:16:19 -0800
Subject: bug in string.join()?
References: <4o589tsg0u2irh8mtds3u6kn05mvd35dao@4ax.com> <qe099tcnql5nnlbl4r0enalcq09np11mup@4ax.com> <973bqv$iub$1@m1.cs.man.ac.uk> <3A941827.20C9F0F9@deprince.net> <Thfl6.2686$Qd4.324192@news1.rdc1.az.home.com>
Message-ID: <jwbnews-64D9C0.21161922022001@corp.supernews.com>

In article <Thfl6.2686$Qd4.324192 at news1.rdc1.az.home.com>,
 "Tim Hochberg" <tim.hochberg at ieee.org> wrote:

> Death to map! Use list comprehensions instead.

As soon as my 1940s brain comprehends them.  ;-)  --john

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From bwilk_97 at yahoo.com  Sun Feb 11 02:42:35 2001
From: bwilk_97 at yahoo.com (Bill Wilkinson)
Date: Sun, 11 Feb 2001 01:42:35 GMT
Subject: Python compiler/installer?
References: <964f5n$oqs$1@nnrp1.deja.com>
Message-ID: <f6mh6.12452$lk6.897687@typhoon.midsouth.rr.com>


| Hi. I'm new to Python (which seems like a very nice language to me) and
| have a few questions. First, is there any way I can compile it so I
| don't have to send out any source code I don't want to?

It compiles to byte code. But the persistant can get to it if they want.
It decompiles quite nicely. :)

|Are there any installers for either compiled-Python (if that exists) or
simple .py
| files?

For windows only:
http://starship.python.net/crew/theller/   (Thanks Thomas!)

For windows and Linux:
http://starship.python.net/crew/gmcm/distribute.html


|And where can I find some examples of Python programs ?(some
| that will help me learn Tkinter.)

The best way to learn Tkinter IMHO is to read /F's tutorial, It's righteous.
http://www.secretlabs.com/library/tkinter/introduction/index.htm




From d98aron at dtek.chalmers.se  Fri Feb  9 15:06:52 2001
From: d98aron at dtek.chalmers.se (Fredrik Aronsson)
Date: 9 Feb 2001 14:06:52 GMT
Subject: What is better, JPython or Jython?
References: <981723867.586455890@news.t-online.de>
Message-ID: <960ths$40u$1@nyheter.chalmers.se>

Jython is the mantained, new release of JPython.
They have changed name because CNRI have a trademark
(or something) on the name JPython.

/Fredrik

In article <981723867.586455890 at news.t-online.de>,
	Zamurai <zamurai at gmx.net> writes:
> It seems that JPython and Jython are quite similar, but what are the
> differences between both? And which one is better?

-- 


From clarence at netlojix.com  Sun Feb  4 22:11:57 2001
From: clarence at netlojix.com (Clarence Gardner)
Date: Sun, 4 Feb 2001 13:11:57 -0800
Subject: Please translate this easy snip of C++ to Python
References: <mailman.981317716.4248.python-list@python.org>
Message-ID: <981321451.86433377@news.silcom.com>

On Sun, 04 Feb 2001, chris at onca.catsden.net wrote:
>On 4 Feb 2001, Phlip wrote:
[C debugging macro snipped]

>Try this:
>
>def TRACE ( str ):
>	print "%s: %s" % ( str, eval(str) )
>
>TRACE('y+z')
>TRACE('strAnimals')
>
>Sorry... you /do/ need to put the parameters in quotes, so Python doesnt
>try to evaluate them before calling TRACE

You also need to provide the TRACE function with the caller's local
namespace, or TRACE will, in general, get NameErrors.

def TRACE(str, locals):
    print "%s: %s" % (str, eval(str, globals(), locals))

and call it like
    TRACE('y+z', locals())

(I know I'm stomping on a predefined function name in TRACE, but I
have no problem with it in this case.)
--
Clarence Gardner
Software Engineer
NetLojix Communications
clarence at netlojix.com



From erno-news at erno.iki.fi  Sun Feb  4 19:46:37 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 04 Feb 2001 20:46:37 +0200
Subject: Q: shelve ovewrites for different users
References: <lxk1ytev19i.fsf@ioasun3.epfl.ch>
Message-ID: <kuzog22smq.fsf@lasipalatsi.fi>

In article <lxk1ytev19i.fsf at ioasun3.epfl.ch>, Jan Kybic
<Jan.Kybic at epfl.ch> writes:

| I am sharing the shelve file between two users on Unix.

are your two different users reading and writing it at the
same time? if so, problems are to be expected. you
need to either

1) set up locking manually, there are some functions
   in the fcntl module to do that, or

2) use gdbm for the shelf object (gdbm does locking itself): copy the
   DbfilenameShelf class from shelve.py and change "anydbm" to "gdbm".

  -- erno


From k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m  Wed Feb 14 17:41:01 2001
From: k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m (Keith Ray)
Date: Wed, 14 Feb 2001 16:41:01 GMT
Subject: python interpreter in smalltalk
References: <9579gr$g0jjg$1@ID-65212.news.dfncis.de> <V1Ud6.568$nD1.12282@nnrp1.sbc.net> <95abs4$g1nic$1@ID-65212.news.dfncis.de> <95coka$hp3$1@nnrp1.deja.com> <k1e2i3t4h5r6a7y-84C34A.18574802022001@news> <7MGJOov0nvx+VW3xYgkmsmMz2tDQ@4ax.com>
Message-ID: <k1e2i3t4h5r6a7y-53B089.08410214022001@news>

In article <7MGJOov0nvx+VW3xYgkmsmMz2tDQ at 4ax.com>, sma at 3plus4.de wrote:

> On Sat, 03 Feb 2001 02:57:36 GMT, Keith Ray
> <k1e2i3t4h5r6a7y at 1m2a3c4.5c6o7m> wrote:
> 
> >Has anyone tried to write a Python interpreter in Smalltalk?
> 
> I tried this, although using Squeak Smalltalk not Dolphin Smalltalk.
> I never finished the project but there's a partially working parser
> which I could share.
> 
> bye
> --
> Stefan Matthias Aust____Truth until Paradox!____________________
>      Jobs? ==> jobs at baltic-online.de   www.baltic-online.de

I have the desire to see Python running inside Squeak 
<http://www.squeak.org/>.

I was planning to start porting from the Jython sources, since those are 
already in an object oriented language (Java).  If you could, please 
email your parser to me and it will help me learn how Squeak could host 
a Python language. 

C. Keith Ray
remove spaces to get my email address
  c k e i t h r a y @ h o m e . c o m

-- 
       <http://homepage.mac.com/keithray/resume.html>


From dsh8290 at rit.edu  Wed Feb 21 18:10:06 2001
From: dsh8290 at rit.edu (D-Man)
Date: Wed, 21 Feb 2001 12:10:06 -0500
Subject: A simple (newbie) question.
In-Reply-To: <9VRk6.116$DT7.616555@newsserver.ip.pt>; from man_at_moon@hotmail.com on Mon, Feb 19, 2001 at 08:08:42PM +0000
References: <9VRk6.116$DT7.616555@newsserver.ip.pt>
Message-ID: <20010221121006.A23949@harmony.cs.rit.edu>

On Mon, Feb 19, 2001 at 08:08:42PM +0000, David A. wrote:
| Hello to everyone,
| 
|     maybe I have already sent this question to the group, but I am not sure
| (the first one never reached the group I think). Anyway, I am a newbie, so
| probably my question is a realy simple for the most experienced python
| users.
| 
| Well what I want to is this:
| >>> a='python'
| >>> b='print'
| >>> c=b+a
| >>> eval(c)
| Traceback (innermost last):
|   File "<pyshell#19>", line 1, in ?
|     eval(c)
|   File "<string>", line 1
|     print"python"
|         ^
| SyntaxError: invalid syntax

Try this to find the problem :

>>> print c
printpython


Once you check the data you are sending to eval you know that

printpython

is indeed invalid syntax.  Instead, try this:

a = '"python"'
b = 'print'
c = b + ' ' + a
eval( c )


HTH,
-D





From vasvir at iit.demokritos.gr  Sat Feb  3 15:13:24 2001
From: vasvir at iit.demokritos.gr (Vassilis Virvilis)
Date: Sat, 03 Feb 2001 16:13:24 +0200
Subject: Win32 CreateProcess with In/out Redirection and a Way to kill the 
 job?
References: <mailman.981064983.9908.python-list@python.org>
Message-ID: <3A7C1204.A226B21E@iit.demokritos.gr>

Michael Jonas wrote:
> 
> Seems like I need people with win32 geek factor....
> 
> What I want:
> I'd like to code a extension module with a function popen3 (same as in os)
> and a function is_alive() and terminate().
> 
> What I allready have:
> a module which has an popen3 method which does exactly the same like the
> posixmodule popen3 except that I extracted it to get hold of the process
> Handle (from the normally inaccessible dict). The redirection works fine, I
> can check if the job is still alive with my new is_alive(<file>). But I'm
> really struggling to terminate the job.
> 
> Where my proble is:
> The CreateProcess in the popen3 not only starts the job specified, it first
> starts a cmd.exe which then starts the job. Ergo: the process handle I save
> is the handle of the console app. The started app does not care at all if I
> do a TerminateProcess on the console. The cmd.exe dies and the started job
> stays alive. I tried to find a way to get hold of the Window Handle of the
> console, but did not find a way. (The doughter process seems to die if I
> close the Console Window. So I thought I could send a WM_Close to the
> Window....) Nothing really worked till here...
> 
> Any idea how to implement something like this on Win32?!?
> 
> Thanks Michel
> 
> PS. Please cc the mail also to mj at muc.das-werk.de, cause I'm having trouble
> getting the mail from the list.

I don't know how exactly the popen* functions are implemented but I
think it's better to use CreateProcess/DuplicateHandle to mimic the
popen* behavior in windows and fork/dup in UNIX.

PS
In UNIX a libc popen invocation fires a shell also (bin/sh).

	.Bill


From marco at atmosp.physics.utoronto.ca  Wed Feb 14 19:48:24 2001
From: marco at atmosp.physics.utoronto.ca (marco at atmosp.physics.utoronto.ca)
Date: 14 Feb 2001 18:48:24 GMT
Subject: I don't understand popen2 :(
Message-ID: <96ejto$4q3$1@news.netmar.com>

Hello All,

I'm a python newbie, happy so far, but somewhat bewildered :(

My problem is with popen2.popen3. It sometimes gets stuck when
reading the process std_out/err. Here is an example:

>>> (po, pi, pe) = popen2.popen3('ssh computer "ls"')
>>> po_out = po.read()
>>> pe_err = pe.read()
>>> pe_err
''
>>> po_out
[Directory listing OK]

Alternatively (notice the read order):

>>> (po, pi, pe) = popen2.popen3('ssh computer "ls"')
>>> pe_err = pe.read()
>>> po_out = po.read()
>>> pe_err
''
>>> po_out
[Directory listing OK]

Fine. Now I try something slightly different:

>>> (po, pi, pe) = popen2.popen3('ssh computer "find /home/marco"')
>>> po_out = po.read()
>>> pe_err = pe.read()
>>> pe_err
''
>>> po_out
[snip find results]

which is fine, BUT if I do:

>>> (po, pi, pe) = popen2.popen3('ssh computer "find /home/marco"')
>>> pe_err = pe.read()

[freezes!! I then Ctrl-C]

Traceback (innermost last):
  File "<stdin>", line 1, in ?
KeyboardInterrupt

>>> po_out = po.read()
>>> po_out
[snip find results]

So, in this last case if I pe.read() first
my script freezes, and I have to Ctrl-C. Sure, I
could always po.read() first, but my biggest problem
is that sometimes it's the other way around! That is,
performing po.read() first causes a freeze! Unfortunately,
I cannot reliably reproduce this last behaviour, but here
is the result from a script I had to Ctrl-C:

prompt> ./myprog.py

[frozen -- Ctrl-C]

Traceback (innermost last):
  File "./myprog.py", line 1245, in ?
    get_files()
  File "./myprog.py", line 1102, in get_files
    po_out = po.read()
KeyboardInterrupt

line #
 1101     (po, pi, pe) = popen2.popen3(scp_from_remote_IP)
 1102     po_out = po.read()
 1103     pe_err = pe.read()

Sadly, the scp operation *does* seem to work most of the time,
why does it sometimes get stuck? Thoughts? Ideas?

I'm using:
Python 1.5.2 (#1, Feb  1 2000, 16:32:16)  [GCC egcs-2.91.66 19990314/Linux
(egcs- on linux-i386)]

Thanks for any help!
(Oh, and please CC me a reply if possible -- Deja's undergoing
transition pains to Google right now)

 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse at newsone.net


From fredrik at effbot.org  Sat Feb  3 21:55:12 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Sat, 03 Feb 2001 20:55:12 GMT
Subject: nested scopes
References: <mailman.981230290.31442.python-list@python.org>
Message-ID: <Qe_e6.11370$AH6.1837949@newsc.telia.net>

Tim Peters wrote:
> Python's "2-level" scheme was a deliberate counterpoint to the abuses of
> deep lexical nesting in Pascal programs of the time (according to Guido),
> and it's something I've grown to like a lot.  But newbies eternally stumble
> over that functions nest textually but not lexically in Python, and this
> will make life simpler for them.

I'm beginning to suspect that a new "sorry, you can no longer
nest def statements" might have been an easier sell...

Cheers /F




From fredrik at pythonware.com  Wed Feb 28 05:56:57 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 04:56:57 GMT
Subject: PIL for the Mac?
References: <3A9C6FDF.3F9422C@alcyone.com>
Message-ID: <ty%m6.15923$Qb7.2598968@newsb.telia.net>

Erik Max Francis wrote:
> Is PIL (PythonWare Image Library) available on the Mac?  I don't see any
> reference to the Mac on its main Web page, and the source tarball
> doesn't appear to have any Mac specific files in it ...

from what I can tell, it's shipped with the standard Mac
distribution:

http://www.python.org/download/download_mac.html

    "This is a single installer which gives you the option to
    install classic 68k, cfm68k, and/or PPC. The installer also
    optionally installs NumPy, PIL, img, Tkinter, and an IDE,
    and interactive development environment.

Cheers /F






From sholden at holdenweb.com  Thu Feb  8 15:57:42 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 8 Feb 2001 09:57:42 -0500
Subject: CGI-Python Image viewing
References: <95tp4p$bev$1@ctb-nnrp2.saix.net>
Message-ID: <pxyg6.10982$x4.169982@e420r-atl2.usenetserver.com>

Mix" <mix77 at usa.net> wrote in message
news:95tp4p$bev$1 at ctb-nnrp2.saix.net...
> I am new to CGI scripting AND Python. What I am trying to do is,
> display an image on an html page that has been retrieved from a Postgre
> database
> using Python (cgi). My problem is : that image that not show up on the
page,
> regardless of where it is sitting on my html or cgi directory!
> (I am running Apache Web server under Redhat 7.0 with the default linux
> kernel).
> But, I CAN display the image on a plain html page (i.e without the
> scripting).
>
> Please help!
> Thanks
>
>
Perhaps you should give us a little more information: for example, is it the
HTML you are retrieving from Postgres, or the image as well? An example of
the generated / retrieved HTML would be useful.

Generally speaking you should be sending HTML to the client with an <img
src="..."> tag in it to specify the location of the graphic. If the value of
the "src" attribute is absolute (begins with a /) then the browser will
request by sending a path from the server's root. If, on the other hand,
it's relative (does not begin with a /) then the browser will request it by
generating an absolute path based on the URL which was used to access the
HTML.

All this assumes you haven't put <BASE> tagging in your HTML.

Does this help?

regards
 Steve




From daniel at dittmar.net  Mon Feb 26 20:48:42 2001
From: daniel at dittmar.net (Daniel Dittmar)
Date: Mon, 26 Feb 2001 20:48:42 +0100
Subject: Can't redirect output
References: <mailman.983207049.352.python-list@python.org>
Message-ID: <3A9AB31A.31DC2202@dittmar.net>

> I guess this is a known problem on NT? I have to work with NT, and
> I never found out exactly why it won't redirect. I found that if I create
> a .bat file that does nothing but call the .py file, I can redirect that
> just fine. Sounds like something evil going on under the covers.

There is a known Problem on NT that starting a program through it's
extension will make the redirection impossible.

So 'python hello.py > somefile.txt' should work. But 'hello.py >
somefile.txt' does not. Solutions: don't know, I'm using 4NT as a shell,
so I don't have this problem. Maybe auto-creating batch files for every
Python program is the simplest method.

Daniel


From jfontain at winealley.com  Thu Feb 15 14:56:45 2001
From: jfontain at winealley.com (Jean-Luc Fontaine)
Date: Thu, 15 Feb 2001 14:56:45 +0100
Subject: how to get the output of embedded python?
References: <96b6o6$a6e$1@s1.read.news.oleane.net> <un1bqcu9a.fsf@ctwd0143.fitlinxx.com> <96dkvb$eb8$1@s1.read.news.oleane.net> <uzofod9uz.fsf@ctwd0143.fitlinxx.com>
Message-ID: <96gmq7$p9r$1@s1.read.news.oleane.net>

David Bolen wrote:

> Jean-Luc Fontaine <jfontain at winealley.com> writes:
> 
> > Yes. That is whatever output you would see when in interactive mode, if
> > that makes any sense...
> 
> Oh, I think I understand better.  You don't actually want to trap any
> output generated during the execution of the code - you just want to
> deal with the result of the expression you may be evaluating.
> 
> > That does not seem to matter in other scripting languages, such as Tcl
> > or Perl: calling an internal eval C function with a script as argument
> > returns the result of the script (not what comes out on stdout or
> > stderr), which may be empty.
> > 
> > For example, if I define a function foo that returns a string, invoking
> > eval("foo()") in C would return that string. Is not that the behavior of
> > python in interactive mode, for example?
> 
> On an expression by expression basis, yes, each expression has a
> result object (but not statements such as "print" for example).
> 
> You should be able to handle the same thing from your controlling code
> by dipping a little lower than the "Simple" high level functions.  By
> using either PyRun_File or PyRun_String you can evaluate Python code,
> and the result of the function is the resulting Python object (a
> PyObject *).  Use PyEval_GetGlobals and PyEval_GetLocals for the
> dictionaries to supply to the execution.
> 
> Once you have the resulting object pointer you can do anything you
> want with it (using any of the Python functions available for
> extending/embedding applications), including converting it into a
> string representation if you like.  Note that the result does not
> itself have to be a string object, since a Python expression can
> result in any Python object.  Check out the abstract and concrete
> objects layer in the C/API reference or the Python source for all
> sorts of functions.
> 
> But if you know you want a string representation, you should be able
> to use something like PyObject_Str to return a new Python object with
> a string representation of the result object.  Then, if you just want
> a char*, use PyString_AsString.
> 
> --

Thank you so very much for all this information: that is what I needed to 
get started.

-- 
Jean-Luc Fontaine mailto:jfontain at winealley.com http://www.winealley.com



From scarblac at pino.selwerd.nl  Tue Feb 27 18:40:45 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 27 Feb 2001 17:40:45 GMT
Subject: IP Math anyone?
References: <mailman.983221028.6007.python-list@python.org> <kupug5qgtg.fsf@lasipalatsi.fi> <slrn99mnvn.ia7.scarblac@pino.selwerd.nl> <kuwvacpcft.fsf@lasipalatsi.fi>
Message-ID: <slrn99npha.j09.scarblac@pino.selwerd.nl>

Erno Kuusela <erno-news at erno.iki.fi> wrote in comp.lang.python:
> In article <slrn99mnvn.ia7.scarblac at pino.selwerd.nl>,
> scarblac at pino.selwerd.nl (Remco Gerlich) writes:
> 
> | .0.255 and .1.0 are not IP addresses you could give to a machine.
> | I've never seen existing code for this, seems to need some custom hacking.
> 
> | (I don't know if .255 and .0 are the only exceptions, probably not).
> 
> why do you think .0 or .255 are not usable addresses?

I was confused about netmasks. I didn't expect the Spanish Inquisition!

-- 
Remco Gerlich


From mwh21 at cam.ac.uk  Mon Feb 19 22:32:27 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Feb 2001 21:32:27 +0000
Subject: Language change (was RE: iterators (was: python-dev summary))
References: <mailman.982392863.25850.python-list@python.org> <3A90A2AA.21FB5D05@seebelow.org> <96qh2f$962$1@216.39.151.169> <96s0oj02nsq@drn.newsguy.com>
Message-ID: <m3ofvytl3o.fsf@atrus.jesus.cam.ac.uk>

Grant Griffin <not.this at seebelow.org> writes:

> If that were the _only_ reason, it might not be so great.  But
> you're talking to the guy who actually _likes_ "self.", as a
> replacement for the "m_" notation he uses in C++ to mark class
> member variables.  (True, I sometimes wake up screaming in the
> middle of the night about those extra three characters Python's
> convention makes me type.  But then I tell myself that I could have
> used "m." if I had really wanted to. <wink>)

I still have the mental scars from trying to understand code that
named *local* variables m_blah.  Explicit *is* better than implicit.

Cheers,
M.

-- 
  please realize that the Common  Lisp community is more than 40 
  years old.  collectively, the community has already been where 
  every clueless newbie  will be going for the next three years.  
  so relax, please.                     -- Erik Naggum, comp.lang.lisp


From danielk at aracnet.com  Fri Feb 16 15:15:29 2001
From: danielk at aracnet.com (Daniel Klein)
Date: Fri, 16 Feb 2001 06:15:29 -0800
Subject: equivalent of NULL or OM in Python
References: <2G5j6.1$CF4.251@typhoon.nyu.edu>
Message-ID: <dgcq8t4iahltqlcot0brik37iuf1h8hucq@4ax.com>

In this case, accessing an element outside of the range of values would produce
an IndexError exception. You can 'catch' this exception like this

poly = [(2,3),(0,4)]
try:
    thirdpart = poly[2]
except IndexError:
    print 'There are only 2 parts.'

In Python, when an object has not been initialized, it is set to 'None' (quotes
are for clarity), but I don't think this is useful in this situation.

For question 2:

for element in poly:
    print element[1] # prints the second part of each tuple 

Hope this helps,

Daniel Klein
Portland OR USA



On Fri, 16 Feb 2001 03:37:43 -0500, "cyberian bear" <cyberian_bear at hotmail.com>
wrote:

>1.I'm doing a program in Python which perform +, -, *, / on two polynomials.
>Their coefficient and exponentials are stored in tuples. So for example
>3X^2+4X+4 would be stored like [(2,3),(1,4),(0,4)] but what if the term is
>absent completely from one of the polynomials then i would have to check if
>one of the tuuples is not present at all and not just zero. For  example
>[(2,3),(0,4)]. In Pascal there is NULL in SETL there is OM but is there a
>corresponding statement in Python. I've checked through several sources
>including my textbook but didn't find the satisfactory answer. Maybe I
>should just check if the second term of every tuple(i.e the coefficient) is
>zero which means that the term is not present.
>2. Also can anyone give me a clue how to iterate over all the coefficient
>terms in the tuples in the first polynomials' tuples.
>My guess is should be something like
>for coefficient in polynomial1:
>where polynomial is a list of tuples
>but how do i tell it to iterate specifically over the second term in every
>tuple and not over the whole tuples
>cb
>



From tjg at exceptionalminds.com  Fri Feb  9 19:32:32 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Fri, 9 Feb 2001 10:32:32 -0800
Subject: open("ls -l |", "r") ??
In-Reply-To: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>; from fm_duende@yahoo.com on Fri, Feb 09, 2001 at 05:56:39PM +0000
References: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>
Message-ID: <20010209103232.U17173@trufflehunter.avalongroup.net>

On Fri, Feb 09, 2001 at 05:56:39PM +0000, steveFarris wrote:
> Hi all, is some variation of  the above possible in Python?
> 

>>> import os
>>> cmd = os.popen('ls -l')
>>> for l in cmd.readlines():
...   print l

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.                   (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  23 days 22:41 hours ago<<



From erno-news at erno.iki.fi  Fri Feb 16 20:22:43 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 16 Feb 2001 21:22:43 +0200
Subject: iterators (was: python-dev summary)
References: <mailman.982256267.9885.python-list@python.org> <3A8CAC77.91555B93@seebelow.org> <96ipv7$hug$2@nntp9.atl.mindspring.net> <kud7cjuf4b.fsf@lasipalatsi.fi> <mailman.982340500.13146.python-list@python.org> <ku7l2qvau6.fsf@lasipalatsi.fi> <mailman.982342971.24194.python-list@python.org> <96jt2v$ogk$1@nntp6.u.washington.edu>
Message-ID: <ku3ddev3ek.fsf@lasipalatsi.fi>

In article <96jt2v$ogk$1 at nntp6.u.washington.edu>, Donn Cave
<donn at u.washington.edu> writes:

| I have heard claims that they're faster, and that's a priority I can
| understand.

actually, they are slower (than map + python function). but don't
think that should be important.

| But every time some cool notion gets added to Python, it adds to
| the amount that we all eventually have to know.  I can't indefinitely
| postpone my education in list comprehensions without marginalizing
| myself, just as someone who comes along and learns them right away
| will still have to learn all the alternatives too.  Does that add
| to our ability to write software, or subtract from it?

well said.

  -- erno


From jmarshal at mathworks.com  Thu Feb 22 21:22:16 2001
From: jmarshal at mathworks.com (Joshua Marshall)
Date: 22 Feb 2001 20:22:16 GMT
Subject: Any way for a function to refer to itself?
References: <3A956962.3BBFED86@americasm01.nt.com> <973s4j$2h9$1@news.mathworks.com>
Message-ID: <973sdo$2no$1@news.mathworks.com>

Joshua Marshall <jmarshal at mathworks.com> wrote:
> Lee, Rick <rickylee at americasm01.nt.com> wrote:
>> I can't find any other way for a function or method to refer to itself
>> than something like the following:

>> def foo():
>>     myself = foo   # only if foo is global
>>     mydoc = myself.__doc__
>>     myname = myself.__name__

>> So it can be done this way, but:

>> - only if the function name can actually be accessed from the current
>> name space
>> - if the function name changes, that first line inside this function
>> also has to change

>> Seems to me there should be a more "Pythonic" way of doing this.  Is
>> there?

> It might be unpleasant, but you can do something like:

>   def fib(f, x):
>      if x < 2: return 1
>      return f(f, x-1) + f(f, x-2)

>   print fib(fib, 10)

And I guess if you don't like the idea of always having to pass your
function in to itself, you can wrap it:

  def fib(x):
     def _fib(f, x):
        if x < 2: return 1
        return f(f, x-1) + f(f, x-2)
     return _fib(_fib, x)

  print fib(10)


From dsh8290 at rit.edu  Tue Feb 13 16:35:56 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 13 Feb 2001 10:35:56 -0500
Subject: python 2.0 won't run scripts with \r\n line termination on RH7?
In-Reply-To: <20010213155947.B759@freedom.puma-ag.com>; from stephen_purcell@yahoo.com on Tue, Feb 13, 2001 at 03:59:47PM +0100
References: <t8c7a69cqerq7c@corp.supernews.com> <t8hegl11hkm465@corp.supernews.com> <mailman.982053978.29454.python-list@python.org> <3A894883.61C64508@uol.com.br> <"from csrabak"@uol.com.br> <20010213155947.B759@freedom.puma-ag.com>
Message-ID: <20010213103555.B10906@harmony.cs.rit.edu>

On Tue, Feb 13, 2001 at 03:59:47PM +0100, Steve Purcell wrote:
| Cesar Rabak wrote:
| > 
| > So elaborating, on your reasoning, Steve, I think the 'sollution' would
| > be the OP to have a kind of "header" or macro in his favorite editor
| > which inserts the correct line in whatever OS be the script being
| > written.
| 
| Perhaps, although what's the simplest thing? Save the file in UNIX text
| format (there must be Windows editors that support this).

There are -- I use gvim :-).

:set fileformat=unix


| 
| As Toby Dickenson suggests, you can use a technique like CVS' filters to
| transparently correct line endings according to the system on which files are
| checked out from source code control. In this case, files are usually saved
| in UNIX format in the source repository, and Windowsified as necessary on
| checkout.
| 
| I think there's a utility that ships with Python for changing the #! line of
| a number of scripts -- perhaps that would help. Anybody remember the name?
| 

I don't know if one came with it, but it shouldn't be hard to make
such a thing.  (Just be sure to open the file in _binary_ not text
mode) It would be simplest to use a RandomAccessFile type of interface
(there is a Java class by that name) and simply change the first \r\n
to \n\n.

In (g)vim you could type the following commands :

:1 
:set binary
:s/^V^M//


where ^V means Ctrl-V and ^M means Ctrl-M (or enter, in this case)

-D




From fredrik at pythonware.com  Sun Feb 18 12:47:27 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 11:47:27 GMT
Subject: Remove directories from os.listdir output
References: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>
Message-ID: <jDOj6.18081$AH6.2469355@newsc.telia.net>

Fernando Rodr?guez wrote:
> I need to filter out all the diretories of the list returned by
> os.listdir(). How can I tell if an entry of that list is a dir or a regular
> file? O:-)

os.path.isdir(filename)
os.path.isfile(filename)

e.g.

for file in os.listdir(mydir):
    fullname = os.path.join(mydir, file)
    if os.path.isfile(fullname):
        ...

Cheers /F

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From phlip_cpp at my-deja.com  Sat Feb 17 00:37:44 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 16 Feb 2001 23:37:44 GMT
Subject: string, split, sort, None, huh?
References: <96kadi$f2t@dispatch.concentric.net> <96kb5d$bb0$2@bcarh8ab.ca.nortel.com>
Message-ID: <96kdk8$dtr@freepress.concentric.net>

Glenn W Jackman wrote:

> ...sort() sorts the list in place...

Jay O'Connor wrote:

> ...sort() sorts the sequence in in place...

Oh, man - I gotta lay off that crack pipe, huh?  @-}

Thanks, guys.

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  My opinions are those of your employer  --


From aleaxit at yahoo.com  Tue Feb 27 08:12:26 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 08:12:26 +0100
Subject: list to string
References: <mailman.983245987.15676.python-list@python.org>
Message-ID: <97fjv1028qd@news2.newsguy.com>

"Delaney, Timothy" <tdelaney at avaya.com> wrote in message
news:mailman.983245987.15676.python-list at python.org...
> Indeed.
>
> Personally, I think the join() string method is *really* bad. A method
> should act on the object it is a part of. In this case, there is no
> conceptual way that ''.join(['a', 'b', 'c']) could be considered to be
> acting on '' - it is *using* ''.

And, similarly, it's *using* ['a','b','c'] -- not "acting on" it.  It's not
a mutator.  Not all methods of an object mutate it, and you clearly
accept that, as you write:

> OTOH, ''.length() is perfectly fine. It is obvious.

Since join may use the joiner in a completely polymorphic way, but
only uses the sequence through a sequence-iteration protocol, it's
pragmatically _useful_ to have it be a method of the joiner object
(as Python doesn't do multimethod-like dispatching).  Usefulness
seems to be a good indicator: if something offends your personal
sense of purity, but is pragmatically good, then 'practicality beats
purity' and it might be wise to reconsider your aesthetics:-).


Alex





From chris.gonnerman at usa.net  Fri Feb 23 14:44:39 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Fri, 23 Feb 2001 07:44:39 -0600
Subject: Can't redirect output
References: <3A95F8BA.9CD22C6F@pcug.org.au>
Message-ID: <002401c09d9e$c91eb3c0$a100000a@local>

From: "Robert Olney" <rolney at pcug.org.au>
Subject: Can't redirect output
> I am running Python 2.0 on Windows 98. When I try to redirect output
> from a command prompt, e.g.
> 
> C:\> python hello.py > out.txt
> 
> the output is printed on the screen and not to the file. Why? 

Is the output you are talking about from print commands and/or 
sys.stdout.write(...) calls, or is it error messages?  Unix and Windows 
processes have two standard channels for writing messages, stdout and
stderr, which Python maps to sys.stdout and sys.stderr.  The redirection
you show only redirects stdout/sys.stdout.

On Unix (and reportedly Windows NT) you can write:

    python hello.py >out.txt 2>errors.txt

which separates the output channels, or

    python hello.py >out.txt 2>&1

to combine them in the file.  Unfortunately this does not work on Win9X,
but there is a program available called stderr which combines the two 
channels.  Get it from:

    http://www.teaser.fr/~amajorel/stderr/

IF NONE OF THIS HELPS YOU... please post your hello.py so we can
inspect it.

>The FAQ
> mentions that this is a problem under Linux.

I'll have to look at the FAQ, I'm not sure why Linux would have a problem 
with it.





From ssthapa at classes.cs.uchicago.edu  Tue Feb 13 04:39:34 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Tue, 13 Feb 2001 03:39:34 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <T1Ah6.3842$n4.29046@e420r-atl2.usenetserver.com> <slrn98g6r6.3eo.ssthapa@ntcs-ip45.uchicago.edu> <969l15$8cm$1@panix6.panix.com>
Message-ID: <slrn98h9sj.523.ssthapa@ntcs-ip45.uchicago.edu>

Aahz Maruch <aahz at panix.com> wrote:
>I suggest:
>
>    ... [[depend1, minVersion, maxVersion], ...
>
>with None a valid value for minVersion and maxVersion.

  That sounds like a great idea.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From yartz at imaginet.fr  Mon Feb 19 08:06:02 2001
From: yartz at imaginet.fr (Yann Schwartz)
Date: Mon, 19 Feb 2001 08:06:02 +0100
Subject: Why ASP <%@ %> does not work
References: <3a909e7f.5108679@News.CIS.DFN.DE>
Message-ID: <47F7E45B059247C9.6C3B173F0635E5FD.35F50673E3AFCE68@lp.airnews.net>

On Mon, 19 Feb 2001 04:22:34 GMT, costas at springmail.com (Costas
Menico) wrote:

>
>Hi,
>
> I am trying to execute the following page in PWS which contains
>Python ASP. However it does not execute to display the "text".
> I have installed the latest  ActivePython .  Is there some registry
>setting that I should be dong? I know <SCRIPT language=Python> works.
>
><html>
><head></head>
><body>
><h1>Querying database</h1>
><%@ LANGUAGE = Python%>
><%
>text="abc"
>%>
><%= text%>
></body>
></html>
>
>Any ideas?
>Thanks

The compilation directive
<%@LANGUAGE=FOO%>
must be the first line of the ASP script (even before any HTML). Using
Python or not has nothing to do with it.

Your page should look like :

><%@ LANGUAGE = Python%>
><html>
><head></head>
><body>
><h1>Querying database</h1>
><%
>text="abc"
>%>
><%= text%>
></body>
></html>


From donn at u.washington.edu  Mon Feb  5 22:44:51 2001
From: donn at u.washington.edu (Donn Cave)
Date: 5 Feb 2001 21:44:51 GMT
Subject: os.popen2
References: <m3snlsrinq.fsf@boss.dreamsoft.com>
Message-ID: <95n6sj$dei$1@nntp6.u.washington.edu>

Quoth Venkatesh Prasad Ranganath <rvprasad at cis.ksu.edu>:
| Hi,
|
| The following code stalls
|
| import os
| i = os.popen2("/bin/bash")
| i[0].write("ls -lR")
| i[1].read()
| -----------stalls here
|
| Am I missing something the way it should behave or isn't it possible to execute
| a process using popenX, keep it alive and interact with it?

Yes, you're missing a couple of things, yes, you can interact with bash
this way, no, you can't interact this way with just any process.

To get a better sense of what's going on in your program, it might help
you to use a shell command that has some effect besides output to the
other pipe.  For example, have it write to a file - instead of 'ls -lR',
for example, try 'ls -lR > xxx' (or 'date > xxx', you really don't need
much output to start with.)  Now you can see specifically whether the
shell command executed, before you waste a lot of time asking why you
don't see its output.  You can also use "ps" to see whether the shell
process is still alive and waiting for something.

Now the first thing you need for your shell command lines is a trailing
newline:  'ls -lR\n', just like any file.  The shell doesn't execute
a command until it has read it, and it hasn't finished reading it until
it encounters the trailing newline.  Next, you have to actually write
the command line to the pipe, from the C library stdio buffer it's
sitting in now  - i[0].flush().  

At this point, I expect you will see evidence that the command actually
ran, and you should now change your command back to write to output -
but you still won't have any output.  This is because i[1].read()
is trying to read up everything, and the pipe is still open so there
could be more coming.  Another C library notion.  I personally would
just ditch the file object / C stdio stuff and use the UNIX file
descriptor directly - see fileobject fileno(), posix read() and write().

All right, now if you have been able to fix all this up, you should
be able to write and read one shell command after another.  It may
be a bit tricky to figure out how much output to expect from a shell
command, and when it's finished and ready for the next command, but
you can probably figure something out.  But if you take bash out of
there and replace it with almost any other application, you'll find
that you're back where you started, because most of these applications
use the same C library output and they're back there at step 1.b.,
without any way to get them to flush their buffer.

If there's any solution at all, it's to use a "pty" device instead
of pipes.  The pty library module has support for that, especially
in Python 2.0 and on the more common platforms.  The pty master fd
you get back will be a UNIX file descriptor that handles both reading
and writing.

	Donn Cave, donn at u.washington.edu


From **MailFreeAmerica**  Wed Feb  7 21:24:23 2001
From: **MailFreeAmerica** (OpenMind)
Date: Wed, 7 Feb 2001 12:24:23 -0800
Subject: Is Python "Compiled"?
Message-ID: <3a81ad07_3@goliath2.newsfeeds.com>

Recently started reading about Python.  The language looks very promising!

I am confused, however, about the impressions I have received.   If I write
an application in Python, would it be distributed in the form of text
modules?  Or would they be compiled into bytecode like Java and Visual
Basic?

Too, is there a native-code compiler or front end available for Python yet,
something like the GNU C preprocessor I seem to recall?

Thanks

___




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----==  Over 80,000 Newsgroups - 16 Different Servers! =-----


From n8gray at caltech.edu.is.my.e-mail.address  Wed Feb 21 04:49:05 2001
From: n8gray at caltech.edu.is.my.e-mail.address (Nathan Gray)
Date: Tue, 20 Feb 2001 19:49:05 -0800
Subject: Proposal: allow '?' and '!' in identifiers
References: <96sh22$jq5$1@news.mathworks.com> <Vklk6.6049$pS4.2369984@typhoon2.ba-dsg.net> <96sp4s$ni7$1@news.mathworks.com> <96t2oi$4ce@gap.cco.caltech.edu> <igcu69.95h.ln@wh4-112.st.uni-magdeburg.de>
Message-ID: <3A933AB1.108@caltech.edu.is.my.e-mail.address>

Christoph Horst wrote:

> Nathaniel Gray <n8gray at caltech.edu.is.my.email.address> wrote:
> 
>> IMHO it's not worth adding "!", but if "?" doesn't break anything then it's 
>> a net gain.  I assert that it _would_ make Python more expressive to add an 
>> elegant way of indicating that a function returns only boolean values.
> 
> 
> What's wrong with prefixing the function name with 'is'?

Nothing.  I just find it pleasingly clear to read languages where 
predicates end with '?'.  It's simple and elegant.

Like I said, though, I'd be very surprised if this proposal ends up 
being accepted.

--
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
              Nathaniel Gray
    California Institute of Technology
      Computation and Neural Systems
      n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From dan at eevolved.com  Fri Feb 16 18:21:30 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 16 Feb 2001 17:21:30 GMT
Subject: segfault. which module to blame? (entire traceback)
References: <%lZi6.127299$Pm2.2356545@news20.bellglobal.com>	<mailman.982285338.32590.python-list@python.org>	<fM%i6.239159$f36.9599321@news20.bellglobal.com> <mailman.982340600.13592.python-list@python.org>
Message-ID: <ukdj6.263124$JT5.9137465@news20.bellglobal.com>

Jeremy Hylton wrote:

> Are there any C extensions being used with this process?  My first
> guess would be bad reference counting in a third-party extension.
> Just in case that guess is wrong, what are you pickling?
> 
> Jeremy
> 
Unfortunately everything I use came with BeOpen Python 2.0.1's rpm on 
python.org.

I write normal python dictionaries and strings to a shelve db. I also use 
locks (threading.Lock) to implement row-level locking in shelve. I've had 
segfaults before while coding normal python scripts. I think it was the 
dictionaries because that's pretty much all I used. Could it be bad 
reference counting in those? I would have to read the code, but could there 
be a problem if two keys hashed to the same value in the dictionary and 
that one is deleted before the other or something similar?

thanks Jeremy,
Dan


From ken at lt.com  Mon Feb 19 14:31:32 2001
From: ken at lt.com (Kenneth Loafman)
Date: Mon, 19 Feb 2001 07:31:32 -0600
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com>
Message-ID: <3A912034.BD25C502@lt.com>

Rainer Deyke wrote:
> 
> "Kenneth Loafman" <ken at lt.com> wrote in message
> news:mailman.982538961.27262.python-list at python.org...
> > Please also teach the C Standard Library along with C++.  I've seen some
> > very poor examples of "reinventing the wheel" happen because the student
> > did not even know the basics of the C Standard Library and its
> > interaction with the basic elements of C/C++.  Consider one case I found
> > where the goal was to remove the last character of a line.  Instead of
> > using something like:
> >
> >    if (strlen(s)) s[strlen(s)-1] = 0
> 
> Argh!
> 
> The following is the correct ways of doing this in C++:
> 
> s = std::string(s.begin(), s.end() - 1);
> 
> Not using 'std::string' in C++ is analogous to not using the built-in string
> type in Python.

They were using CString under M$VC.  I don't think std::string existst
on the M$ compilers.  Would make it too easy to port to the heathen UNIX
boxen <grin>.

...Ken



From fredrik at pythonware.com  Thu Feb 15 08:11:46 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Feb 2001 07:11:46 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net> <slrn98l4rh.2s8.jepler@potty.housenet>
Message-ID: <SiLi6.17192$AH6.2345642@newsc.telia.net>

Jeff Epler wrote:
> This is one reason that 2.1 will include "xreadlines", which will work
> just like the above:
>         for z in f.xreadlines():
>                 print z
> but will never hold the entire file in memory.

never?  not even if it's a single byte? ;-)

Cheers /F




From bobhicks at adelphia.net  Fri Feb  9 22:40:59 2001
From: bobhicks at adelphia.net (Robert L Hicks)
Date: Fri, 09 Feb 2001 21:40:59 GMT
Subject: CGI
Message-ID: <B6A9CE15.A3FD%bobhicks@adelphia.net>

I would like to have Python handle all my forms via cgi. I have access to
everything to make changes. Where is a good resource to get information on
this?

Right now I am just looking to get a mailer working...I have formmail.py but
I am getting an error so I need more info on how to set everything up.

- Bob



From tdickenson at devmail.geminidataloggers.co.uk  Mon Feb 26 13:12:10 2001
From: tdickenson at devmail.geminidataloggers.co.uk (Toby Dickenson)
Date: Mon, 26 Feb 2001 12:12:10 +0000
Subject: fileobj.write(buffer(...))
References: <mailman.982795168.21257.python-list@python.org> <972ip3$juj$1@news.tpi.pl> <973mhu$1r2e$1@news1.simtel.ru> <uhf1k9ao7.fsf@ctwd0143.fitlinxx.com>
Message-ID: <f3hk9tsek2smno687h0kmrjf1jcsceu32n@4ax.com>

David Bolen <db3l at fitlinxx.com> wrote:

>"Alexander Semenov" <sav at ulmen.mv.ru> writes:
>
>> No, python python implements reference counting. Then file object loose all
>> references it will del-ed. I think file object handles del correctly and
>> closes
>> corresponding handle.
>
>Just as a general point, you can't guarantee when _del_ will be
>called.  While the current CPython implementations (even in 2.x I
>think) trigger that when the RC hits zero, there's no guarantees, and
>it might not be called until later (or even until Python's process
>terminates).  I'm not sure if JPython/Jython does it immediately even
>in their current releases.
>
>So depending on the criticality of controlling the close, you may want
>to handle it yourself explicitly, particularly if you are writing data
>that might otherwise not get flushed until (possibly) much later.

That is less true than it was before 2.0...

2.0 takes care not to be recursive when deallocating recursive data
structures. This is achieved by moving dead objects onto a separate
list, and deleting things iteratively.

If you have more than one thread manipulating deeply recursive data
structures then it is possible for deallocations to be delayed, and
then actually handled by a thread other than the one that released the
last reference.

Im still suprised noone else finds this as much of a problem as I do.


Toby Dickenson
tdickenson at geminidataloggers.com


From root at rainerdeyke.com  Mon Feb  5 16:56:33 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Mon, 05 Feb 2001 15:56:33 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com>
Message-ID: <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3A7E3C04.9EB02932 at engcorp.com...
> def trace(expr = None):
>     # evaluate expression in callers namespace
>     import sys
>     try:
>         raise None
>     except:

Better make that 'except None'.  Unqualified 'except' clauses are generally
a bad idea.  In this case, you really don't want to catch 'MemoryError',
'KeyboardInterrupt', or 'ComputerExplodingError'.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From m.faassen at vet.uu.nl  Wed Feb  7 00:47:00 2001
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 6 Feb 2001 23:47:00 GMT
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com> <GDBf6.175479$ge4.60994110@news2.rdc2.tx.home.com>
Message-ID: <95q2dk$hji$4@newshost.accu.uu.nl>

Rainer Deyke <root at rainerdeyke.com> wrote:
> "Jay O'Connor" <joconnor at cybermesa.com> wrote in message
> news:3A7ECEAB.DB6F8648 at cybermesa.com...
>> Rainer Deyke wrote:
>>
>> > Better make that 'except None'.  Unqualified 'except' clauses are
> generally
>> > a bad idea.  In this case, you really don't want to
> catch...'ComputerExplodingError'.
>>
>> I hadn't seen that exception raised yet....is it new?  What do you
>> usually do to handle it?

> ComputerExplodingError is one of the "hidden" exceptions of the Underground
> of Pythonic

Something must be wrong with some Usenet feed somewhere. Your post got
cut off in the middle. I suspect that this was due to


From akuchlin at mems-exchange.org  Mon Feb  5 16:46:39 2001
From: akuchlin at mems-exchange.org (Andrew Kuchling)
Date: Mon, 05 Feb 2001 10:46:39 -0500
Subject: Alpha release of ZODB programming guide
Message-ID: <E14Pnqh-0006eL-00@ute.cnri.reston.va.us>

For what seems like half an eternity, I've been working on a ZODB
programming guide, written for Python programmers who want to use the
ZODB to store data.  The document is placed under the GNU Free
Documentation License.  It's still far from being finished, but while
on vacation I filled it out enough to make it alpha-worthy, though
note that there are still empty sections.  Please take a look at:

http://www.amk.ca/zodb/guide/

What I'd like right now:
     * Suggestions for topics that should be covered.  This would be
the most helpful.

     * Comments on the organization (ordering of chapters, topics, &c.)

     * Reporting errors and inaccuracies would also be helpful, but
are less important at the moment.  Getting the outline correct
is my primary concern at the moment.

--amk

Table of contents:
     * 1 Introduction
          + 1.1 What is ZODB?
          + 1.2 OODBs vs. Relational DBs
          + 1.3 What is ZEO?
          + 1.4 About this guide
     * 2 ZODB Programming
          + 2.1 Installing ZODB
               o 2.1.1 Requirements
               o 2.1.2 Installing the Packages
          + 2.2 How ZODB Works
          + 2.3 Opening a ZODB
          + 2.4 Writing a Persistent Class
          + 2.5 Rules for Writing Persistent Classes
               o 2.5.1 Modifying Mutable Objects
               o 2.5.2 Some Special Methods Don't Work
               o 2.5.3 Fixing isinstance and issubclass
               o 2.5.4 __getattr__, __delattr__, and __setattr__
          + 2.6 Writing Persistent Classes
               o 2.6.1 Changing Class Attributes
     * 3 ZEO
          + 3.1 How ZEO Works
          + 3.2 Installing ZEO
               o 3.2.1 Requirements
               o 3.2.2 Installation
          + 3.3 Configuring and Running a ZEO Server
          + 3.4 Testing the ZEO Installation
          + 3.5 Sample Application: chatter.py
     * 4 Transactions and Versioning
          + 4.1 Subtransactions
          + 4.2 Undoing Changes
          + 4.3 Versions
          + 4.4 Multithreaded ZODB Programs
     * 5 Storages
          + 5.1 FileStorage
          + 5.2 BerkeleyStorage
          + 5.3 OracleStorage
     * 6 Related Modules
          + 6.1 ZODB.PersistentMapping
          + 6.2 ZODB.PersistentList
          + 6.3 B-tree Modules
     * A. Resources
     * B. GNU Free Documentation License



From cg at gaia.intranet.cdegroot.com  Mon Feb 19 21:12:46 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 19 Feb 2001 21:12:46 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com>
Message-ID: <96runu$hvg$1@gaia.intranet.cdegroot.com>

Sheila King  <sheila at spamcop.net> said:
>I must say, that I am shocked at the number of apparent *groans* over C++
>language, in this thread.
>
*groan*. I learnt C++ by installing a, err, not-so-official AT&T cfront source
tape on a Xenix box and hacking away until the damn thing ran, and really did
my best to learn it. It's /not/ a real OO language. But it is sort of nice for
applying OO to low-level/system programming, and because it was meant to be
just that, it's OK. Just don't go writing apps in it (but the same holds
for C).

>I think that
>the trick, is to use a carefully selected subset of the C++ language.
>
Yup. And that might be workable in a classroom project, but forget it for
commercial development. Better start with a good language right away. For
teaching OO, I'd use Smalltalk. For teaching general programming, Python is a
nice choice because it is the most "neutral" language I know (neutral w.r.t.
to OO, procedural, functional). Although chances are I'd use Smalltalk
anyway :-)



-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From fredrik at pythonware.com  Thu Feb  8 19:14:33 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 08 Feb 2001 18:14:33 GMT
Subject: problems with PIL
References: <95uemq$7vr$1@pietro.crs4.it>
Message-ID: <dmBg6.13588$AH6.2041509@newsc.telia.net>

Davide Carboni wrote:
> I'm fighting against PIL, I put _imaging20.dll and _imagingTK.dll in DLLs
> but when I try the piece of code below I get the error

looks like you forgot to rename the DLL.  try doing:

    import _imaging20

if that works, you've installed it correctly, but under
the wrong name.

renaming it to "_imaging.dll" should help.

(and yes, it's "_imagingtk.dll", not "_imagingTK.dll".
Python is case-sensitive, also on Windows)

Cheers /F




From ivnowa at hvision.nl  Mon Feb 12 13:56:43 2001
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 12 Feb 2001 12:56:43 GMT
Subject: Permissions or mutable objects... Read only file contents template to write only?
References: <9uPh6.16771$o85.85546@news-server.bigpond.net.au>
Message-ID: <3A87DBA4.710B@hvision.nl>

Gavin Tomlins wrote:

> #open files for various operations
> fileHdr = open(cTemplateHdr, "r")
> fileFooter = open(cTemplateFooter, "r')
> fileOutput = open(cOutput, "w")
> 
> cTempStr = ""
> cTempStr = fileHdr.readlines()
> 
> fileOutput.write(cTempStr)
> 
> at this point I am presented with a type error indicating that I'm trying to
> write a 'read buffer'.

fileHdr.readlines() returns a list. So, you are trying to write a list
of strings to fileOutput, while the .write method wants a string.

You can loop over the list, writing it line by line:

    for line in cTempStr:
        fileOutput.write(line)

or use the .writelines method:

    fileOutput.writelines(cTempStr)

Maybe you wanted to read only one line from the input file; in that case
you can use the .readline method.

> What is the process for rectifying this problem. Any assistance, reference
> would be greatly appreciated. I've searched the Python documentation but was
> unable to find appropriate documentation. If possible, could people cite the
> reference for my understanding. I can only put it down to lack of comprehension
> or poor choice of search criteria on my behalf.

See the Python library reference, 2.1.7.9.

HTH,


From phd at phd.russ.ru  Mon Feb 26 11:10:44 2001
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 26 Feb 2001 13:10:44 +0300 (MSK)
Subject: Spamcopped
In-Reply-To: <Pine.LNX.4.30.0102221313560.17279-100000@fep132.fep.ru>
Message-ID: <Pine.LNX.4.30.0102261305570.24293-200000@fep132.fep.ru>

Hello!

On Thu, 22 Feb 2001, Chris Withers wrote:
> >    If SpamCop does the same, whay it's better? (I really interesting...)
>
> SpamCop is also very well respected. In fact, I've read the some ISPs value
> SpamCop reports a lot more than hand written emails to their abuse addresses.

   Thank you very much! The SpamCop.NET service proved to be very good!!!
I took their perl script and rewrote it to Python. Julian gladly accepted
the script on http://spamcop.net/fom-serve/cache/166.html.
   I'm attaching the script here too, for archiving purpose.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.
-------------- next part --------------
#! /usr/local/bin/python -O
"""Spamcop reporter. This program is a translation of a perl script
published on http://spamcop.net/fom-serve/cache/166.html to Python.
Author: Oleg BroytMann <phd2 at mail.com>"""


# The programs expects UCE with full headers to be fed into its stdin.
# After it forwards the message to SpamCop, it opens additional filehandle
# to read from /dev/tty and waits for Enter. This is to report that the spam
# was successfully forwarded.


### IMPORTANT! IMPORTANT! IMPORTANT!
### Put you *real* address or forwarder here
### Because some "abuse departments" may collect such addresses to
### flood you with additional spam (UCE) you may register additional
### addres specially for this. There are many free e-mail providers to
### help you.

my_real_address = "user at host.domain"


# open a pipe to sendmail
import os
sendmail = os.popen("/usr/lib/sendmail -oi -t", 'w')

# write headers and simulate MIME attachments in the body of the message
sendmail.write("""From: %s
To: spamcop at spamcop.net
Subject: reply anyway
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="DeathToSpamDeathToSpamDeathToSpam"

This is a multi-part message in MIME format.
--DeathToSpamDeathToSpamDeathToSpam
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit


--DeathToSpamDeathToSpamDeathToSpam
Content-Type: message/rfc822
Content-Disposition: attachment

""" % my_real_address)


# copy stdin to the pipe
import sys
while 1:
   line = sys.stdin.readline()
   if not line: break

   sendmail.write(line)


# finish MIME part
sendmail.write("""

--DeathToSpamDeathToSpamDeathToSpam--
""")


# report success
print
print "Message forwarded"
print "Press Enter to continue"

# open /dev/tty for reading and wait for Enter key; this is needed because
# stdin is not available (there was UCE fed into stdin, do you remember?)
tty = open("/dev/tty", 'r')
line = tty.readline()
tty.close()

From grante at visi.com  Tue Feb 20 19:07:12 2001
From: grante at visi.com (Grant Edwards)
Date: Tue, 20 Feb 2001 18:07:12 GMT
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com> <3A90240D.384ABA3E@alcyone.com> <m6h09t422jim1jsp2gu2d9cisavvao3h8a@4ax.com> <d1219tg7f00icu4p4as621u3dlsb2r94ue@4ax.com> <tm619t0luql0612vi1jrftdpts9amrrg67@4ax.com> <slrn99334p.2fk.qrczak@qrnik.zagroda> <slrn993auo.2bo.grante@isis.visi.com> <mailman.982636942.2852.python-list@python.org> <jwbnews-081BEF.21114119022001@news.olympus.net>
Message-ID: <knyk6.500$S7.49088@ruti.visi.com>

In article <jwbnews-081BEF.21114119022001 at news.olympus.net>, John W. Baxter wrote:
>
>> So why does closing the pipe to signal EOF not work?  I've
>> written a great many programs based on that.  The writer closes
>> the pipe and terminates, and the reader gets EOF and continues
>> operation.  What am I missing here?
>
>He wants to signal the end of each chunk of data, leaving the pipe open 
>for the next chunk.  I think he needs to define and use an end-of-chunk 
>sentinal.

Often, the easiest thing to do is to pass information in ASCII
and use '\n' as the chunk delimiter.  That way all of the
normal library routines just "do the right thing".

If you don't want to use ASCII, you can use some sort of binary
protocol (e.g. STX/ETX framing).

-- 
Grant Edwards                   grante             Yow!  Look into my eyes and
                                  at               try to forget that you have
                               visi.com            a Macy's charge card!


From jepler at inetnebr.com  Wed Feb 14 15:14:30 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 14 Feb 2001 14:14:30 GMT
Subject: setproctitle [was: How do I change a process name or even a thread name in python!]
References: <mailman.982109894.26467.python-list@python.org>
Message-ID: <slrn98l4m5.2s8.jepler@potty.housenet>

On Tue, 13 Feb 2001 19:16:53 -0500 (EST), Steven D. Majewski
 <sdm7g at virginia.edu> wrote:
>The actual args to (C) setproctitle is a format string and optional
>args, but it makes more sense to do any string formatting in Python,
>and just pass that string. 

Actually, this is unsafe, and has been the cause of a number of bugs 
recently exposed on Bugtraq.

Imagine that you are writing an ftp server which uses 'setproctitle' to
change its process name to give information including the e-mail address an
anonymous user gave when he logged in.  The user might pass a string like
"%x%x%x%x", which will be treated as printf-style specifiers by setproctitle.

So, rather than
	setproctitle(new_title);
you must
	setproctitle("%s", new_title);
to be safe.

See bugtraq or search for "setproctitle format vulnerability" for more
details on this potential security risk.  One URL:
	http://www.hert.org/papers/format.html

Unfortunately, that means the convenience function in the posix module
can't be used.

Jeff
jepler at inetnebr.com


From aleaxit at yahoo.com  Mon Feb 19 23:39:05 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 19 Feb 2001 23:39:05 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <kR_j6.257786$ge4.89529176@news2.rdc2.tx.home.com> <mailman.982589543.7737.python-list@python.org> <96rgs602drn@news1.newsguy.com> <3A9154F1.4351E80D@alcyone.com>
Message-ID: <96s78k0a7a@news1.newsguy.com>

"Erik Max Francis" <max at alcyone.com> wrote in message
news:3A9154F1.4351E80D at alcyone.com...
> Alex Martelli wrote:
>
> > (HP, IBM, Dec, and Sun, at least; those were the ones we tried -- and
> > that's when I learned that Microsoft's C++ was *MUCH* cleaner, and far
> > closer to the standard, than _many_ other proprietary C++ compilers by
> > various other operating system vendors -- before then, the only C++
> > compiler I had used on Unix and Unix-like boxes was gcc, but that was
> > not supported by 'mainsoft' [if that's the right name]).
>
> Which is ironic, since that's the best, most Standard compliant C++
> compiler for UNIX out there.  (It certainly has some problems, but
> they're all have workarounds and are being addressed in gcc 3.0.)

I think KCC (KAI's compiler) is/was better (not sure about its status
now that Intel [?] has acquired it).  But mainsoft didn't support that
one, either.


Alex





From andreas at digicool.com  Mon Feb 26 13:30:16 2001
From: andreas at digicool.com (Andreas Jung)
Date: Mon, 26 Feb 2001 07:30:16 -0500
Subject: HowTo Fix illegal chars in xml-documents??
References: <rCom6.7781$X_2.129881@news1.oke.nextra.no>
Message-ID: <000001c09ff6$84f53ed0$d0063bd0@SUXLAP>

I remember there is a tool "xp" in the SP package (a tool for SGML
conversion....).
This utility allows to make SGML documents XML compliant by fixing several
problems. I have used in a project with about 10.000.000 SGML documents
where my own "SGMl2XML" converter had problems due to invalid tagging
or wrong entities.

Andreas
----- Original Message -----
From: "Thomas Weholt" <thomas at cintra.no>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Monday, February 26, 2001 3:38 AM
Subject: HowTo Fix illegal chars in xml-documents??


> Hi,
>
> I'm trying to "convert" alot of legacy information into collection of
> xml-documents. In the old data there are alot of characters that are
illegal
> if used in the character part of the XML, like &, ;, < and >.  I've tried
to
> make a quick-fix for this, to swap these characters with the proper
> HTML-alternative, like &#59; for ; etc. but haven't found a really good
way
> of doing this. I need to be able to swap the old chars back too, but
that's
> not very important now. Does anybody have a clue or a piece of code to get
> me going? As of now, a great deal of the converted documents are not
proper
> xml-documents and will not get parsed at all.
>
> Thanks in advance,
> Thomas
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list




From dsh8290 at rit.edu  Tue Feb  6 19:27:12 2001
From: dsh8290 at rit.edu (D-Man)
Date: Tue, 6 Feb 2001 13:27:12 -0500
Subject: Standalone applications
In-Reply-To: <95p5ok$7cu$1@nnrp1.deja.com>; from paulus132@my-deja.com on Tue, Feb 06, 2001 at 03:37:57PM +0000
References: <95c000$8nc$1@ryu.it> <p6qlmrooik8.fsf@informatik.hu-berlin.de> <95p5ok$7cu$1@nnrp1.deja.com>
Message-ID: <20010206132712.B1814@harmony.cs.rit.edu>

On Tue, Feb 06, 2001 at 03:37:57PM +0000, paulus132 at my-deja.com wrote:
| In article <p6qlmrooik8.fsf at informatik.hu-berlin.de>,
|   Martin von Loewis <loewis at informatik.hu-berlin.de> wrote:
| > "Ryujin" <ryu-jin at bigfoot.com> writes:
| >
| > > Any suggestion???
| >
| > I personally prefer freeze. It requires a C compiler, but gives you
| > best control and reliability. Plus it allows to create *truly*
| > stand-alone binaries, i.e. binaries that are not even dependent on
| > shared libraries.
| >
| 
| I also have some good experience with freeze, but the binaries are big
| (1.8M on Sun, 0.5M after strip), even for very simple *.py scripts !!
| 
| Do you have some tips to reduce it ??

Tell the users to install the interpreter on their own.  Then your
package will only consist of your code and be much smaller. ;-)

-D




From neelk at alum.mit.edu  Sun Feb 18 22:15:37 2001
From: neelk at alum.mit.edu (Neelakantan Krishnaswami)
Date: 18 Feb 2001 21:15:37 GMT
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <mtk86n6928.fsf@astron.berkeley.edu>
Message-ID: <slrn990ime.a2b.neelk@alum.mit.edu>

[Crossposting trimmed]

On 18 Feb 2001 12:16:31 -0800, Johann Hibschman <johann at physics.berkeley.edu>
wrote:
>Dave Cross writes:
>
>> 1/ The programmer calls a function that doesn't actually exist within
>> the application (or libraries).
>
> It's easy to trap this error in Common Lisp, but it's hard to set up a
> "general replacement" routine.  You can, generally, trap the error for
> a given block of code, then attempt to guess a replacement function,
> but I don't think you can then resume the calculation.

I'm not a serious CL user, but aren't its exceptions restartable? That
would permit you to do this. It would be really nice if Python's
exception system supported this. Then I wouldn't have to choose
between raising an exception and having my constructor functions
behave algebraically.

> Have fun.  I do suspect that Common Lisp whups Perl in this respect,
> though.  :-)

Yeah. :)


Neel


From stephen_purcell at yahoo.com  Wed Feb 21 18:41:27 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 21 Feb 2001 18:41:27 +0100
Subject: newbie - concatanating 2 lists
References: <970sgj+rrf7@eGroups.com>
Message-ID: <20010221184127.A25981@freedom.puma-ag.com>

Gnanasekaran Thoppae wrote:
> i am just beginning to use python.

Good choice!

> li1 = ['a', 'b', 'c']
> li2 = ['x', 'y', 'z']
> 
> i want:
> 
> li3 = ['ax', 'by', 'cz']

There may be more readable ways, but the shortest is probably this:

>>> li1 = ['a', 'b', 'c']
>>> li2 = ['x', 'y', 'z']
>>> li3 = map(lambda a,b: a+b, ['a','b','c'], ['x','y','z'])
>>> li3
['ax', 'by', 'cz']


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From shaleh at valinux.com  Sat Feb 17 06:20:02 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 21:20:02 -0800
Subject: why the string function / module change
In-Reply-To: <20010216201223.A14948@valinux.com>; from shaleh@valinux.com on Fri, Feb 16, 2001 at 08:12:23PM -0800
References: <20010216201223.A14948@valinux.com>
Message-ID: <20010216212002.A14970@valinux.com>

On Fri, Feb 16, 2001 at 08:12:23PM -0800, Sean 'Shaleh' Perry wrote:
> 
> split(foo) worked just as well.  I understand there were problems making the
> string module do this, but still.  Could they have not been _split() or
> something to dissuade their use?
> 

example of recent code I have seen:

keys[keyname] = ": ".join(s.split(": ")[1:])

foo.join(sequence) # that logic is just odd



From digitig at cix.co.uk  Tue Feb 13 19:29:00 2001
From: digitig at cix.co.uk (Tim Rowe)
Date: Tue, 13 Feb 2001 18:29 +0000 (GMT Standard Time)
Subject: XML processing in Python -- changing horses in mid-stream
References: <3a8563ae.3383090887@news.iol.ie>
Message-ID: <memo.20010213182932.64417A@digitig.compulink.co.uk>

The scripts are not very big. The result files could be. The sparse tree 
processing might be the thing.

Thanks,

Tim

In article <3a8563ae.3383090887 at news.iol.ie>, sean at digitome.com (Sean Mc 
Grath) wrote:

> Tim,
> 
> How big are the test scripts / test results? If the sizes are
> manageable I would suggest looking at Pyxie Tree processing.
> While processing a tree T, you can create another tree, T1 with
> the results of running the second test script, you can then 
> use the Paste functions to paste the new tree into the
> existing one.
> 
> If size is a problem, you might like to look that the Sparse
> Tree Processing stuff in Pyxie that allows you to switch
> from event processing to tree processing as needed.
> 
> Sean
> 
> On Sat, 10 Feb 2001 14:50 +0000 (GMT Standard Time), digitig at cix.co.uk
> (Tim Rowe) wrote:
> 
> >I have a Python program that parses a test script written in XML and 
> puts >the test results to stdout in XML (with a different DOCTYPE).
> >
> >I also have a Python program that takes the test result XML and 
> processes >it for graphical presentation.
> >
> >Here's my challenge. If the second program were given a test script 
> >instead of results, I would like it to recognise the different 
> doctype, >run the tests and process the results all in one.
> >
> >When my result parser spots a test script I should be able to pass the 
> >script to the test program via a pipe, put the results of the test 
> through >a new result parser, and copy anything relevant from the new 
> result parser >to the old one. But can I stop the old result parser 
> parsing the rest of >the test script?
> >
> >I plan to use either expat or pyxie, whichever is easiest.
> >
> >TIA.
> >
> >
> 
> 



From aahz at panix.com  Sat Feb 17 03:42:29 2001
From: aahz at panix.com (Aahz Maruch)
Date: 16 Feb 2001 18:42:29 -0800
Subject: missing modules in BeOpen-Python-2.0?
References: <96kar0$bb0$1@bcarh8ab.ca.nortel.com>
Message-ID: <96koel$fms$1@panix2.panix.com>

In article <96kar0$bb0$1 at bcarh8ab.ca.nortel.com>,
Glenn W Jackman  <glennj at nortelnetworks.com> wrote:
>
>I fetched and installed BeOpen-Python-2.0-1.i386.rpm from python.org,
>and was unable to import:  crypt, dbm, dl, nis
>
>Why is this release missing these modules (that are documented in the
>Global Module Index)?

You need to compile them.  Check the README.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Why doesn't "Just Say NO" include caffeine, nicotine, alcohol, Prozac,
and Ritalin?  --Aahz


From qrczak at knm.org.pl  Sat Feb 24 15:03:18 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 24 Feb 2001 14:03:18 GMT
Subject: An alternative approach to bound methods
References: <E14UvH4-00054g-00@mail.python.org> <20010220155508.4879.qmail@web6403.mail.yahoo.com> <mailman.982689503.3821.python-list@python.org> <slrn998h48.1te.qrczak@qrnik.zagroda> <972opb0ngg@news2.newsguy.com> <slrn99a93s.vn8.qrczak@qrnik.zagroda> <973er001d39@news2.newsguy.com> <slrn99ap97.lm7.qrczak@qrnik.zagroda> <975hpa05qm@news1.newsguy.com>
Message-ID: <slrn99ffp6.ep3.qrczak@qrnik.zagroda>

Fri, 23 Feb 2001 12:33:00 +0100, Alex Martelli <aleaxit at yahoo.com> pisze:

> class Aclass:
>     def amethod(x, *args):
>         # body snipped
> 
> is this 'amethod' a "class method" or an "instance method"?

Both.

> Does calling it via
> 
>     x=Aclass();x.amethod(y,z)
> 
> or via
> 
>     x=Aclass();Aclass.amethod(x,y,z)
> 
> change anything?

No.

> > Are you asking about the semantics of a particular piece of code?
> > If so, which code? My proposal is unambiguous.
> 
> And it does unambiguously mean that mentioning 'amethod' within
> the above amethod's code body will in fact refer to Aclass.amethod,
> not self.amethod, then?

Yes.

> People will often and erroneously make recursive calls to
> "amethod(a,b,c)" when they should be going through self, because
> "going through self" IS the normal, most frequently desired case.

Both are equally easy. Why would anybody write
    amethod(self,b,c)
instead of
    self.amethod(b,c)
when he means the latter?

> And this is checked at runtime, and gives a very clear
> error message if the user makes a mistake about it:
> 
> TypeError: unbound method must be called with class instance 1st argument

It's not necessarily a mistake.

> class Derived(Base):
>     def foo(self, *args):
>         print "Derived.foo",args
>         Base.foo(*args)
> 
> I don't want this error to become a silent, mysterious one; we know
> have clear diagnostics for it, and losing them would be a serious
> step backwards.

I admit that this error would have a worse diagnostics (usually:
too few arguments, i.e. not that bad). The same can be said about
writing 'self' in method definitions: when one forgets it, the error
is equally mysterious.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From mkx at excite.com  Tue Feb 20 23:52:23 2001
From: mkx at excite.com (mkx at excite.com)
Date: Tue, 20 Feb 2001 17:52:23 -0500
Subject: ZLIB decompressing only portion of data
References: <s1s59tojmev7822ndpmuhmg4d5gepipq39@4ax.com>
Message-ID: <h5t59tsb1bgg6j7ids19vnljabs0mme9ds@4ax.com>

Of course, there is a better way to implement my test script (see
below), but I still get the same results. The input sample is 9,859
bytes long, where the string returned by the decompress method is only
546 bytes in length.


###

import os, zlib

fd=os.open('d:/temp/zlb/snippets.zlb', os.O_RDONLY | os.O_BINARY )
dco=zlib.decompressobj(15)

str=dco.decompress(os.read(fd,16384))

print len(str)
#print str



From lreilly at cs.strath.ac.uk  Thu Feb 22 23:40:10 2001
From: lreilly at cs.strath.ac.uk (Lee Reilly CS1997)
Date: Thu, 22 Feb 2001 22:40:10 +0000
Subject: which IDE?
Message-ID: <3A95954A.72621882@cs.strath.ac.uk>

Hi,

I'm relatively new to Python and have been developing a product using
Zope and experimenting using IDLE. I was wodnering if anyone could tell
me the name of a good IDE - something similar to Symantec Visual Cafe
perhaps, where I could develop my code and it can spice it up
(text/variable highlighting, etc.).

WingIDE looks good http://archaeopteryx.com/wingide but AFAIK it's not
available on Win32.

Anyone


From cobrien at Radix.Net  Fri Feb  2 02:26:15 2001
From: cobrien at Radix.Net (Cary O'Brien)
Date: 1 Feb 2001 20:26:15 -0500
Subject: Data management products (was: How do you lock your web app data?)
References: <956uqd$6st$1@newsreaderg1.core.theplanet.net> <958j810o2e@news2.newsguy.com> <D62094CDC48C0FCA.44A0CB1CEB4E1F76.B8EA27DB5535EBAD@lp.airnews.net> <95a49u11uao@news1.newsguy.com>
Message-ID: <95d2bn$s7b$1@saltmine.radix.net>

In article <95a49u11uao at news1.newsguy.com>,
Alex Martelli <aleaxit at yahoo.com> wrote:
>"Cameron Laird" <claird at starbase.neosoft.com> wrote in message
>news:D62094CDC48C0FCA.44A0CB1CEB4E1F76.B8EA27DB5535EBAD at lp.airnews.net...
>    [snip]
>> Actually, all these Alex has mentioned have
>> capabilities that'll knock the socks off vic-
>> tims habituated to Access, for example.
>
>Just one attempt at terminology clarification...:
>
>Access is a possibly-handy GUI IDE front-end for cobbling together
>possibly-intricated DB operations; at least if you don't mind

[snip]
>
>Fortunately, Access of course works perfectly well with MSDE (or
>other editions of SQL Server), Oracle, and no doubt other good
>RDBMS's yet. 

Well...  The PostgreSQL people had one heck of a time dealing with
the sql that access generated.  But that was a year or two ago.

-- cary

>               Even if one IS habituated to Access (and there may
>be perfectly good reasons for choosing it to develop certain DB
>apps, I guess), there is no reason not to switch to a better engine
>underneath the chrome (given that MSDE is freely redistributable

[snip]


From bwilk_97 at yahoo.com  Sun Feb 11 02:42:35 2001
From: bwilk_97 at yahoo.com (Bill Wilkinson)
Date: Sun, 11 Feb 2001 01:42:35 GMT
Subject: Is Python for me?
References: <20010210175216.26827.00000474@ng-mh1.aol.com>
Message-ID: <f6mh6.12451$lk6.897622@typhoon.midsouth.rr.com>

| I'm investigating at M, Python, BETA, Scheme, Haskell, Icon, and Common
Lisp.
| Which would you use? Why?

An older comparison by Guido on python.org.
http://www.python.org/doc/essays/comparisons.html

A cool comparison by Norvig of Lisp and Python.
http://www.norvig.com/python-lisp.html






From boswell at caltech.edu  Fri Feb 16 00:47:07 2001
From: boswell at caltech.edu (Dustin Boswell)
Date: Thu, 15 Feb 2001 15:47:07 -0800
Subject: how to compile https support into Python2.0
Message-ID: <3A8C6A7B.42C809C4@caltech.edu>

I originally posted a question asking how to get Python to support
urls like https://secure-server.com so you could use urllib.urlopen()
transparently.

After a bunch of searching on the web, here's what I did (on Redhat 6.2)

to get the Python Distribution to compile correctly with SSL...

(If it doesn't work, or there's an easier/better way, please correct
me.)

OPENSSL
This is a package that you need to have installed before you
can build the Python source distribution (in the SSL way).

1) go to http://www.openssl.org/source/
 and download openssl-0.9.6.tar.gz
2) gunzip openssl-0.9.6.tar.gz
3) tar -xvf openssl-0.9.6.tar (make sure you are in the
directory you want to put the source in; this can be in
any temporary place, since the build will automatically
place it in the right spot later)
4) cd openssl-0.9.6
5) read the INSTALL file and...
6) ./config ; make ; make test ; make install
7) if that didn't work, I can't help you

This should have installed files into a /usr/local/ssl
directory (along with various others)


PYTHON
Here's where and how to build a Python interpreter..

1) Go to http://www.python.org/2.0/#download
   and download BeOpen-Python-2.0.tar.gz
2) gunzip BeOpen-Python-2.0.tar.gz
3) tar -xvf BeOpen-Python-2.0.tar ( same deal as step 3 above )
4) goto
http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/python/dist/src/?cvsroot=python#dirlist

        and make sure you have fresh copies of:
        - Modules/socketmodule.c
        - Lib/urllib.py
        - Lib/httplib.py
        - (possilby others, although I didn't, and haven't (yet) run
        into problems
5) in Python-2.0/Modules/ do:
     cp Setup.in Setup
6) emacs (or, gulp, vi if you prefer) Setup
7) uncomment the four lines so that it looks like:
# Socket module compiled with SSL support; you must edit the SSL
variable:
SSL=/usr/local/ssl
_socket socketmodule.c \
        -DUSE_SSL -I$(SSL)/include -I$(SSL)/include/openssl \
        -L$(SSL)/lib -lssl -lcrypto

8) in the top directory ( Python-2.0/ ) do ...
        ./configure ; make ; make test (took a while) ; make install


Dustin Boswell
boswell at caltech.edu



From joconnor at cybermesa.com  Sun Feb  4 08:39:21 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Sun, 04 Feb 2001 07:39:21 +0000
Subject: NameError in cross imports of modules
References: <Pine.LNX.4.21.0102032128320.9418-100000@kev.ioc.ee> <mailman.981253934.10580.python-list@python.org>
Message-ID: <3A7D0729.C7BF5EF1@cybermesa.com>

Thomas Wouters wrote:

> and as far as I'm concerned there are only two viable uses: the way
> os.py uses it (import any and all symbols from a platform-dependant C
> module) and the way Mailman uses it (import all symbols from a template
> config file into the user-editable config file, so it only needs to contain
> overriding values.) I still owe Fred a "From module import * considered
> harmful" tutorial-section ;-P If prefixing all module variables with the
> module name is too painful because the module name is too long or unclear,
> consider using 'import my_long_module as _a' or some such.


I've followed the convention (maybe I made it up for myself :)  That if
I'm importing classes, I use "from package import *" such as "from gtk
import *", but if I'm importing a module of functions, I just use
"import package name" such as "import string" and use the module name
prefix for all function calles.

Don't know if it's good or not...


Take care,
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

"God himself plays the bass strings first when He tunes the soul"


From aleaxit at yahoo.com  Tue Feb 27 14:23:13 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 14:23:13 +0100
Subject: win32com.client - parameter types
References: <slrn99n5be.8f8.tom@tnt.teikom.no>
Message-ID: <97g9or029n3@news1.newsguy.com>

"Tom Danielsen" <tom at teikom.no> wrote in message
news:slrn99n5be.8f8.tom at tnt.teikom.no...
>
>
> I have a COM server with the following method:
>
>         [id(0x60020016)]
>         HRESULT getDict([out] IDispatch** pDictionary);
>
> which makepy made into:
>
>         def getDict(self, pDictionary=defaultNamedNotOptArg):
>                 return self._ApplyTypes_(0x60020016, \
>                         1, (24, 0), ((16393, 2),), \
>                         'getDict', None,pDictionary)

Not to worry -- this may be slightly misleading, but it DOES
in fact work just right.


> This is, the method returns its value in parameter 1.

That's what it does COMwise, but, for Python, this output
parameter IS transformed into a return value!


> VB and VBscript handles this, but I cannot get Python (2.0) to work.
>
> Is there any way to get at this value in Python?  Is there a way I
> can change the [out] to a [out, retval] and make Python return the
> value as the return value of the method call?  If so, how?

Nothing more than the makepy you've already done.

To try it out, I built a toy (ATL) project (inproc server, just
one object, project name toyatl, object name Toy) with IDL:

// toyatl.idl : IDL source for toyatl.dll
//

// This file will be processed by the MIDL tool to
// produce the type library (toyatl.tlb) and marshalling code.

import "oaidl.idl";
import "ocidl.idl";

 [
  object,
  uuid(565B5BFD-0CB2-11D5-9E52-0060B0EB1D67),
  dual,
  helpstring("IToy Interface"),
  pointer_default(unique)
 ]
 interface IToy : IDispatch
 {
  [id(1), helpstring("method getDict")] HRESULT getDict([out]IDispatch**
pDictionary);
 };

[
 uuid(565B5BF1-0CB2-11D5-9E52-0060B0EB1D67),
 version(1.0),
 helpstring("Toy ATL project TLB")
]
library TOYATLLib
{
 importlib("stdole32.tlb");
 importlib("stdole2.tlb");

 [
  uuid(565B5BFE-0CB2-11D5-9E52-0060B0EB1D67),
  helpstring("Toy Class")
 ]
 coclass Toy
 {
  [default] interface IToy;
 };
};

and the 'obvious' implementation (FinalConstruct of the CToy class builds an
instance member that's a CComPtr<IDispatch> with a CoCreateInstance on
ProgId "Scripting.Dictionary"; method getDisp just calls a CopyTo on that
member).  After a makepy...:

D:\pybridge>python
Python 2.0 (#8, Oct 16 2000, 17:27:58) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> from win32com.client import Dispatch as disp
>>> x=disp("toyatl.Toy")
>>> a=x.getDict()
>>> a
<win32com.gen_py.clareg 1.0 Type Library.IDictionary>
>>>


Alex







From bill.xu at newlix.com  Wed Feb 14 15:43:02 2001
From: bill.xu at newlix.com (Bill Xu)
Date: Wed, 14 Feb 2001 14:43:02 GMT
Subject: Anybody know about pulldom in PyXML?
Message-ID: <WPwi6.5804$HG.1720272@news4.rdc1.on.home.com>

Hi, Gang:

Will u please tell me how to use the pulldom.startElement() in the Python
XML DOM (PyXml 0.6.3, Python 1.5, Redhat Linux 6.2) ?
 Following is my simple test program and result.
 Will u please tell me how to correct this bug?
"AttributeError: 'None' object has no attribute 'createElement'";

Thanks in advance!

Bill

=====================================================
#!/usr/bin/python
#

import os, sys, string, StringIO
import xml.dom
from xml.dom import minidom
from xml.dom.pulldom import PullDOM
import traceback



# Save the new message
def test_meth():
  print "Am i here"
  builder = PullDOM()
  builder.startElement("SET", {"PATH" : "a path"})
  builder.endElement("SET")
  print "Am i here"
  confirm(builder)
 #result = self.Print(builder)

def confirm(test, testname = "Test"):
    if test:
        print "Passed " + testname
    else:
        print "Failed " + testname
        raise Exception

print "? am i herer"
test_meth()

=======================================================
Output:

? am i herer
Am i here
Traceback (innermost last):
  File "./testdom.py", line 34, in ?
    test_meth()
  File "./testdom.py", line 20, in test_meth
    builder.startElement("SET", {"PATH" : "a path"})
  File "/usr/lib/python1.5/site-packages/xml/dom/pulldom.py", line 85, in
startElement
    node = self.document.createElement(name)
AttributeError: 'None' object has no attribute 'createElement'






From gradha at iname.com  Sun Feb 25 13:51:57 2001
From: gradha at iname.com (gradha at iname.com)
Date: Sun, 25 Feb 2001 13:51:57 +0100
Subject: Can I popen a command with a spawnv alike interface?
Message-ID: <d5va79.k5e.ln@127.0.0.1>

Hello.

I need to popen a command from a python script, and I will always be using
names which expanded at the shell won't make it to the called program. I
have solved this in other scripts using os.spawnv, where I can pass
directly file names to the process and know it will get it right.

Is there a way I could use popen giving file names as arguments rather
than all the command in a single string which will get it wrong due to
spaces and such things in file names?

-- 

 Grzegorz Adam Hankiewicz   gradha at iname.com - http://gradha.infierno.org
 Other web pages:  http://glub.ehu.es/  -  http://welcome.to/gogosoftware/


From johngrayson at home.com  Sun Feb 11 01:01:38 2001
From: johngrayson at home.com (John Grayson)
Date: Sun, 11 Feb 2001 00:01:38 GMT
Subject: Tkinter idioms (was: Some basic questions about Tkinter (probably v easy for experts!))
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net> <Pine.GSO.4.21.0102081029430.20775-100000@babbage> <ADBg6.13592$AH6.2042423@newsc.telia.net> <50B7C024B0AA609B.8CB8E2F63E2849E7.2783498FE2A00925@lp.airnews.net>
Message-ID: <964kov$tbl$1@nnrp1.deja.com>

In article
<50B7C024B0AA609B.8CB8E2F63E2849E7.2783498FE2A00925 at lp.airnews.net>,
  claird at starbase.neosoft.com (Cameron Laird) wrote:
> In article <ADBg6.13592$AH6.2042423 at newsc.telia.net>,
> Fredrik Lundh <fredrik at effbot.org> wrote:
>
	.
>
	.
>
	.
> >not directly -- there's something called lambdas (in Python, not
> >in Tkinter), but it's usually better to use a plain function:
> >
> >    def myfunc():
> >        otherfunc(1, 2, 3)
> >
> >    b = Button(master, command=myfunc)
> >
> >Cheers /F
> >
> >
>
> Tell me if this is pertinent--I find myself writing a
> lot of formulaic lambdas in my Tkinter compositions.
> Example:  a Button command has a no-arg signature, but
> when I want to gang some other binding to the same com-
> mand, I have to write a trivial lambda to ignore the
> supplied event.  Is there something I'm missing?
> --
>
> Cameron Laird <claird at NeoSoft.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html
>

It is a common practice to put a default arg as the first
argument of a function, then you can call the function
from an event or a command:

def buttonAction(e=None):
    ....

or:

    def buttonAction(self, e=None):
        ....


(Page 106 PTkP)

    John


Sent via Deja.com
http://www.deja.com/


From guido at digicool.com  Fri Feb 23 03:59:26 2001
From: guido at digicool.com (Guido van Rossum)
Date: Thu, 22 Feb 2001 21:59:26 -0500
Subject: Nested scopes resolution -- you can breathe again!
In-Reply-To: Your message of "Thu, 22 Feb 2001 16:45:00 EST."
             <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> 
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> 
Message-ID: <200102230259.VAA19238@cj20424-a.reston1.va.home.com>

We (PythonLabs) have received a lot of flak over our plan to introduce
nested scopes despite the fact that it appears to break a small but
significant amount of working code.  We discussed this at an
PythonLabs group meeting today.  After the meeting, Tim posted this
teaser:

> PS: At the internal PythonLabs mtg today, I voted against nested
> scopes.  But also for them.  Leaving that to Jeremy to explain.

After the meeting Jeremy had a four hour commute home due to bad
weather, so let me do the honors for him.  (Jeremy will update the
PEP, implement the feature, and update the documentation, in that
order.)

We have clearly underestimated how much code the nested scopes would
break, but more importantly we have underestimated how much value our
community places on stability.  At the same time we really like nested
scopes, and we would like to see the feature introduced at some point.

So here's the deal: we'll make nested scopes an optional feature in
2.1, default off, selectable on a per-module basis using a mechanism
that's slightly hackish but is guaranteed to be safe.  (See below.)

At the same time, we'll augment the compiler to detect all situations
that will break when nested scopes are introduced in the future, and
issue warnings for those situations.  The idea here is that warnings
don't break code, but encourage folks to fix their code so we can
introduce nested scopes in 2.2.  Given our current pace of releases
that should be about 6 months warning.

These warnings are *not* optional -- they are issued regardless of
whether you select to use nested scopes.  However there is a command
line option (crudest form: -Wi) to disable warnings; there are also
ways to disable them programmatically.  If you want to make sure that
you don't ignore the warnings, there's also a way to turn warnings
into errors (-We from the command line).

How do you select nested scopes?  Tim suggested a mechanism that is
used by the ANSI C committee to enable language features that are
backwards incompatible: they trigger on the import of a specific
previously non-existant header file.  (E.g. after #include
<imaginary.h>, "imaginary" becomes a reserved word.)

The Python equivalent of this is a magical import that is recognized
by the compiler; this was also proposed by David Scherer for making
integer division yield a float.  (See
http://mail.python.org/pipermail/edu-sig/2000-May/000499.html)  You
could say that Perl's "use" statement is similar.

We haven't decided yet which magical import; two proposals are:

    import __nested_scopes__
    from __future__ import nested_scopes

The magical import only affects the source file in which it occurs.
It is recognized by the compiler as it is scanning the source code.
It must appear at the top-level (no "if" or "try" or "def" or anything
else around it) and before any code that could be affected.

We realize that PEP 5 specifies a one-year transition period.  We
believe that that is excessive in this case, and would like to change
the PEP to be more flexible.  (The PEP has questionable status -- it
was never formally discussed.)

We also believe that the magical import mechanism is useful enough to
be reused for other situations like this; Tim will draft a PEP to
describe in excruciating detail.

I thank everybody who gave feedback on this issue.  And thanks to
Jeremy for implementing nested scopes!

--Guido van Rossum (home page: http://www.python.org/~guido/)



From matthias.oberlaender at daimlerchrysler.com  Mon Feb 12 10:53:06 2001
From: matthias.oberlaender at daimlerchrysler.com (matthias.oberlaender at daimlerchrysler.com)
Date: 12 Feb 2001 09:53:06 GMT
Subject: How to copy class objects?
References: <960b75$a2e$1@news.sns-felb.debis.de> <961103$a2e$2@news.sns-felb.debis.de> 
	<9615nm0sjd@news2.newsguy.com>
Message-ID: <968bq2$dm7$1@news.sns-felb.debis.de>

In <9615nm0sjd at news2.newsguy.com> "Alex Martelli" wrote:
< deleted my stuff>
> 
> new.classobj, followed by setting the __bases__ attribute of the
> new class object to an empty tuple, should do what you're asking.
> 
> Of course, the new class will not inherit anything from anybody,
> since it will have no base classes whatsoever.  Alternatively,
> you can recursively 'copy-class' up the inheritance DAG to
> effect a (somewhat) 'deep-copy' of an entire DAG of related
> classes, if that is what you actually want (not copying a [1]
> class-object, but a whole graph of them).  Similarly, you will
> have to copy (with the copy module, the new module, whatever)
> any object to which class X has references, which you don't
> want Xnew to share -- depending on what those "I don't want
> this to be shared" objects are... it's hard to tell without
> having any idea of the problem you're trying to solve this way
> (e.g., the "not sharing any base classes" item is totally new
> to me here, nor is it fully clear what it implies, as in,
> emptying out the new class's __bases__ vs recursive copies).
> 
> 
> Alex
> 
> 
Ok, here's a more detailed example. I want a stack containing only ints. Then 
I want a stack containing only floats. Extending the pattern to other types 
should be straightforward. Of course, neither should FloatStack be a base 
class of IntStack, nor vice versa. Indeed, new.classobj seems to offer a 
possible solution.

from types import IntType, FloatType

class IntStack:
  """ Stack of int's. Type is checked during push."""
 
  elementType = IntType
  
  def __init__(self):
    self.data = []
    
  def push(self, elem):
    assert type(elem) == self.elementType
    self.data.append(elem)
  
  def pop(self):
    return self.data.pop()
    
# Now let's make a float stack. IntStack serves as a template

from copy import copy
#floatStack = copy(IntStack) # That was my original idea. But, well, it 
doesn't work!

# After a little experimenting I found that I could do it with module 'new' 
like this:
import new

FloatStack = new.classobj("IntStack", IntStack.__bases__, 
IntStack.__dict__.copy()) 

# It is very import do use a copy of IntStack's dictionary.
# If IntStack had base classes, FloatStack would share them. That's ok, since
# we can still edit the __bases__ attribute later.  

FloatStack.__name__ = "FloatStack"
FloatStack.elementType = FloatType
FloatStack.__doc__ = """ Stack of floats's. Type is checked during push."""

print IntStack.__doc__
print FloatStack.__doc__




From kalle at gnupung.net  Tue Feb 13 00:02:20 2001
From: kalle at gnupung.net (Kalle Svensson)
Date: Tue, 13 Feb 2001 00:02:20 +0100
Subject: How do I change a process name or even a thread name in python!
In-Reply-To: <3A885AAB.863BC0B8@lt.com>; from ken@lt.com on Mon, Feb 12, 2001 at 03:50:35PM -0600
References: <3a884bc2@si-nic.hrz.uni-siegen.de> <3A885AAB.863BC0B8@lt.com>
Message-ID: <20010213000220.A13868@apone.network.loc>

[reformatted included text]
Sez Kenneth Loafman:
> Fionn Behrens wrote:
> > 
> > I am sick of seeing 10x"python" in top, without being able to say which is
> > which.
> 
> Try putting a '#! /usr/bin/python' as the first line, then running it as
> '$ ./foo.py'.  Under Linux/UNIX, that picks up the right name and
> arguments for ps aux.

As long as the magic line isn't
#!/usr/bin/env python

Peace,
  Kalle
-- 
Email: kalle at gnupung.net     | You can tune a filesystem, but you
Web: http://www.gnupung.net/ | can't tune a fish. -- man tunefs(8)
PGP fingerprint: 0C56 B171 8159 327F 1824 F5DE 74D7 80D7 BF3B B1DD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20010213/64d8fc80/attachment-0001.pgp>

From mwh21 at cam.ac.uk  Thu Feb  1 13:28:50 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Feb 2001 12:28:50 +0000
Subject: Unicode pb in Python 2.0 solved - no thanks to WinPython
References: <95b65r$l27$1@wanadoo.fr>
Message-ID: <m3d7d27fjx.fsf@atrus.jesus.cam.ac.uk>

"Jurie Horneman" <jhorSPAMBGONneman at wanadoo.fr> writes:

> I would very much like to "RTFM", and believe me I tried, but there isn't
> much M to FR. How could I have avoided this problem with less agony? Should
> I report the bug to the authors of WinPython?

Do you mean PythonWin?  I would, yes.

Cheers,
M.

-- 
31. Simplicity does not precede complexity, but follows it. 
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From matt at mondoinfo.com  Thu Feb 22 23:12:04 2001
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Thu, 22 Feb 2001 22:12:04 GMT
Subject: os.path.join("/a","/b")
Message-ID: <slrn99b3lk.vh.matt@happy-hour.mondoinfo.com>

I was a little surprised today to notice that

os.path.join("/a","/b") returns '/b'

(os.path is posixpath in this case of course) while

os.path.join("/a/","b") returns "/a/b"

which I expected from both. Poking back into my dusty archives, I find
that it has been that way since at least 1.4 so I presume that it's a
feature but I'll be darned if I can figure out the logic. It sure
doesn't seem to have much to do with joining. The module's docstring
just says "Join two or more pathname components, inserting '/' as
needed".

If someone could explain the logic to me, I'd be grateful.

Regards,
Matt


From tchur at optushome.com.au  Tue Feb 20 20:41:43 2001
From: tchur at optushome.com.au (Tim Churches)
Date: Wed, 21 Feb 2001 06:41:43 +1100
Subject: Optimisation problem - in C?
Message-ID: <3A92C877.C67AAD8@optushome.com.au>

After a bit of profiling I have worked out exactly which part of a data
crunching class on which I am working needs to be optimised - maybe in
C? The abstracted function which needs to run faster is as follows - not
surprisingly it is itself a loop which lives within two other loops -
hence it is called rather frequently:

############
def makevector(sourcelist,elementlist):
	resultvector = []
	for element in elementlist:
		resultvector.append(sourcelist[element])
	return resultvector
###########

My test data has about 300,000 elements in sourcelist and between about
10,000 and 70,000 elements in elementlist, but larger lists are
envisaged, hence the earnest desire for maximum speed.

Questions:
Can this function (which essentially does the fairly generic operation
of "fetch the elements of sequence a using the element indexes contained
in sequence b") be optimised without resorting to a C module?

If not, is anyone willing to hold my hand (it's clean, but rather
sweaty) while I fumble with trying to replace the function with a C
module? (Note: after years of unremitting effort, I have finally
mastered all aspects of hello_world.c)

Tim Churches
Sydney, Australia



From insanc at cc.gatech.edu  Wed Feb 21 15:48:34 2001
From: insanc at cc.gatech.edu (Joseph Holland King)
Date: 21 Feb 2001 14:48:34 GMT
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu>
Message-ID: <970kg2$cn0$1@solaria.cc.gatech.edu>

Joseph Holland King <insanc at cc.gatech.edu> wrote:

: is there anyway to lock a file so that two different programs cannot access
: the file at the same time. ie, my python module is editting file foo, and
: at the same time user x (or program x) tries to read from it. is there 
: anyway that the python module can prevent x from accessing the file for
: a given amount of time?

to clarify: i am on a unix system, i have tried to use flock however when i
locked the file i was still able to read, write and copy the file, none of 
which i want to happen while the file is locked. while using the flock 
method i was using LOCK_EX. also the third party program is not going to be
controlled, written, or have anything to do with mine. am i just not using
flock properly or is there something else that might do the job? thank you.

-- 
---
Joseph Holland King  | "God whispers to us in our pleasures, speaks in our
                     |  conscience, but shouts in our pains: it is His
                     |  megaphone to rouse a deaf world." C. S. Lewis


From wmiller at mediaone.net  Tue Feb 20 23:03:22 2001
From: wmiller at mediaone.net (Walter Miller)
Date: Tue, 20 Feb 2001 14:03:22 -0800
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com>
Message-ID: <YQBk6.80$3e2.62458@news.pacbell.net>

The "InternetExplorer.Application.1" com object has a HWND property from
which you can get the window handle.  It seem reasonable to expect that one
could get a device context based on a valid window handle.

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:96ujcc12r6k at news1.newsguy.com...
> "Walter Miller" <wmiller at mediaone.net> wrote in message
> news:NVxk6.15$uj7.15036 at news.pacbell.net...
> > Anyone know how to get a device context for a com object?  For example:
>
> There is no general way: the COM object might have/use any
> number of windows, from 0 upwards, at any given time, and/or
> other devices, so it's not clear at all what 'a device context for
> the object' might mean (what does it mean to get a device
> context for a process, or a thread, or a DLL, or...?).
>
> Some specific COM interface for which it makes sense will
> implement a GetDC method, of course (offhand, though, only
> IDirectDrawSurface3 comes to mind -- which doesn't help
> you much with an internet explorer application object!-)
>
> Other will let you get a handle to a window object (again
> only one comes to mind, IOleWindow, but that one has
> several derivatives, and is used heavily in the in-place
> activation protocol, so it _may_ be possible, with a little
> bit of effort, to get *that* one for internet explorer -- if
> you activate it in-place within a container of yours, or use
> it as the container for your own activex control's in-place
> activation).
>
>
> Alex
>
>
>
>




From pfaffben at msu.edu  Sun Feb 18 18:44:28 2001
From: pfaffben at msu.edu (Ben Pfaff)
Date: 18 Feb 2001 12:44:28 -0500
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <874rxrj37n.fsf@pfaffben.user.msu.edu>

Dave Cross <dave at dave.org.uk> writes:

> I'm doing some comparisons on programming language features and I'd be
> very interested to know how you would handle the following scenarios
> in your programming language of choice.
> 
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries).

Undefined behavior in C.

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings?

#define keyword replacement
Also causes undefined behavior in at least some circumstances.
-- 
"The expression isn't unclear *at all* and only an expert could actually
 have doubts about it"
--Dan Pop


From ryu-jin at bigfoot.com  Wed Feb  7 17:40:30 2001
From: ryu-jin at bigfoot.com (Ryujin)
Date: Wed, 07 Feb 2001 16:40:30 GMT
Subject: SSL Client-Server
Message-ID: <95ru3l$2hk$1@ryu.it>

I have Python2.0 and OpenSSL-0.9.6 installed
Anybody knows where to find a tutorial showing how to create
a client-server ssl socket connection? I'd like to realize a simple
client connecting and sending something to the server using SSL
I could only find out this:                

s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)                
s.connect((HOST, PORT))
ssl = socket.ssl(s, keyfile, certfile)

whatabout the server? and whatbout the documentation?
Any help would be really appreciated.

matte



-- 
---------------------------------------
Memelli Matteo					
Could u imagine a world without Windoz?	 
ryu-jin at bigfoot.com			 
---------------------------------------



From dan at eevolved.com  Sun Feb 11 21:28:28 2001
From: dan at eevolved.com (Dan Parisien)
Date: Sun, 11 Feb 2001 20:28:28 GMT
Subject: CPAN functionality for python
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu>
Message-ID: <MBCh6.101338$Pm2.1931128@news20.bellglobal.com>

ssthapa at cs.uchicago.edu wrote:

> If there is enough
> positive feedback, I'll open up a project on Sourceforge and put my
> code up there.  Any comments, suggestions, or help on where to go from
> here would also be appreciated.

POSITIVE FEEDBACK!!
POSITIVE FEEDBACK!!
POSITIVE FEEDBACK!!

is that enough? ;)

I'd be glad to pitch in as well!

peace,
Dan


From db3l at fitlinxx.com  Mon Feb 26 22:52:58 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 26 Feb 2001 16:52:58 -0500
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <V0yJ0qA4Ejl6EwIt@jessikat.fsnet.co.uk>
Message-ID: <ulmqt158l.fsf@ctwd0143.fitlinxx.com>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> Whilst helping my colleague migrate from 1.5.2 to 2.0 we came up against
> the dreaded python15.dll not found message. The irritating thing about
> the message is that it didn't tell us which .pyd it was attempting to
> load.

Note that if you actually have a python15.dll on your system (so that
the Windows system loader can resolve it), then the Python import
mechanism will complain about the mismatch as an ImportError - at
least if the dependency is directly from the module being imported -
which may be of more help.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From qrczak at knm.org.pl  Sun Feb 18 22:41:55 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 18 Feb 2001 21:41:55 GMT
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <slrn990gd3.tc5.qrczak@qrnik.zagroda>

Thu, 15 Feb 2001 16:56:54 +0000 (GMT), Michael Hudson <mwh21 at cam.ac.uk> pisze:

>  Discussion went round and round for a while and moved
>  on to more general iteration constructs, prompting Ka-Ping Yee to
>  write a PEP entitled "iterators":
> 
>   <http://python.sourceforge.net/peps/pep-0234.html>

What about mp_iterkeys trying to use __iteritems__ after __iterkeys__ is
not available, before failing back to __iter__? Similarly __itervalues__.
It would allow a mapping to implement only __iteritems__, obtaining correct
iteration over keys and values for free.

What about introducing a "black hole" or "wildcard pattern" for
assignments?
    e0, e1, _, _ = some_4_tuple
    for k:_ in mapping:
    for _:v in mapping:
If there is code using _ as a variable name, maybe some different name.
But it's _ in some other languages (ML, Haskell, Prolog, Clean).

After these two things increasing the complexity :-( let me propose
a simplification. Have _ as above, but don't have mp_iterkeys nor
mp_itervalues. The mapping iteration protocol uses only mp_iteritems
and __iteritems__ / __iter__. Iteration over keys or values is
expressed as iteration over items, while throwing half of each pair
into a black hole.

For mp_iteritems implemented in terms of __iter__ (i.e. sequences
treating as mappings from integers) and for builtin types the
interpreter can generate as efficient code for k:_ and _:v iteration
as if there were separate iterkey and itervalue protocols. For user
classes it's a bit worse (an unnecessary 2-tuple is created), but
the simplification of concepts certainly justifies it.

So finally I propose adding black holes, and having only the iteritems
protocol and syntax. iterkeys and itervalues are just instances of
these concepts put together, not separate concepts.

-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From peter at engcorp.com  Fri Feb  2 14:31:51 2001
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 02 Feb 2001 08:31:51 -0500
Subject: Diamond x Jungle Carpet Python
References: <mailman.981100230.20774.python-list@python.org>
Message-ID: <3A7AB6C7.386D14D@engcorp.com>

Rod wrote:
> 
> I have several Diamond x Jungle Capret Pythons for SALE.
> 
> Make me an offer....
> 
> Go to: www.qnet20.com

Which platform?  Version 1.5.2 or earlier?  Doesn't
ActiveState already make these available as a free
download provided you don't need to redistribute them?

Jeez, why don't people ever provide enough detail in 
their posts?!

:-)


From tim.one at home.com  Thu Feb  1 10:34:32 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 04:34:32 -0500
Subject: . Python 2.1 function attributes
In-Reply-To: <95432j014on@news2.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEIIIMAA.tim.one@home.com>

[Tim]
>> ...
>> The testimony of one person who tried a thing and found it
>> lacking is more compelling than 100 virgins speculating about
>> why they're certain they would despise it if only they could
>> bear stooping low enough to try it <wink>.

[Alex Martelli]
> Actual experience with using some tool may be a _part_ of what
> gives 'impressiveness' to reports that said tool is inferior,
> but it's generally not a major component thereof -- e.g., the
> tool may have been misunderstood, misused, etc.

That's important too, though.  That is, if people misunderstand or misuse,
that may point to serious flaws in the docs or the design.  There's just no
substitute for experience.  Not that that's *all* there is, but it's
important.  In this case, we learned that *nobody* misunderstood what it
did.  I suspect some people still take it way too seriously, though (like
reduce, map and lambda, it's meant to be a minor convenience, not a new
foundation on which to build all your file output!  maybe that's a lesson we
failed to learn from reduce & friends).

>> telling fact that Guido rejected Barry's print>> proposal
>> before giving it a test drive in his own code -- pure argument
>> never had much to do with this one.

> Interesting -- I don't recall hearing about this detail previously.

There's a bit of it in the SourceForge patch history:

http://sourceforge.net/patch/index.php?func=detailpatch&patch_id=
    100970&group_id=5470

although there are many gaps in the discussion.  The Python-Dev archives may
or may not have more, but I've really got no interest in reliving this one.

> It _does_ interestingly contradict Guido's posts at the time,
> requesting _arguments_ (NOT "experiences") -- which, to me and I
> guess to other respondents, implied that the requested arguments
> *would* have a bit to do with something; I'm less surprised, now
> that I hear this, that said arguments mattered not a whit.

We need some context here.  Stephen Figgins wrote a great (IMO) "outsider"
summary of the brouhaha at the time:

  http://www.oreillynet.com/pub/a/python/2000/08/30/pythonnews.html

Against the background of "a flurry of posts that accused the Python
development team of creeping featurism, selling out the language to
corporate interests, moving too fast, and turning a deaf ear to the Python
community", Guido was trying to insist on some sanity.  Don't take it
personally, I don't think it was ever directed at you.  But the overall
debate had gone nuts.  As Moshe reminded already, what Guido asked NOT to
get was "votes", not "experience".  A killer pure argument would have been
great too.  However, in his estimation he didn't get one, much as in your
estimation he didn't make one.  So it goes.

> ...
> Oh, I do realize this -- Python *will* have this wart forevermore.
> That doesn't mean I intend to suffer it *silently*.

No indeed, so long as it's understood that the right to speak is not also
the right to be listened to -- don't you think there are more *interesting*
things to do in the Python world than beat this dead horse?  I do.

> ...
> I thought I had posted (must have been 4-5 months ago) about my
> "print>>" experiments?

I could well have missed that -- sorry.  In the crunch to get Python 2.0 out
the door, we were all pulling 20-hour days for a while, and facing the
prospect of being out of work anyway (indeed, as things turned out, we
finished the release on our own time).  It was darned hectic.

> Briefly -- I took one module that was using about 20 print
> statements in two functions (and used the sys.stdout setting for
> redirect when needed, from a separate wrapping function) and
> recoded it using '>>blotch' on every print (with 'blotch' a
> function parameter defaulting to None on the operating functions).

I don't think I would have liked ">>blotch" in that context either:  if I
find myself using print>> in more than one or two lines in a function, I
don't like it.  It's simply too wordy for me then.  It's the bee's knees,
though, for an infrequent print to a rarely-used file, or for echo'ing the
same msg to multiple files (like a log and stdout).  Then it's pleasant.

BTW, I see you also picked up on my dislike for sys.stderr.  In the doctest
context, yes, that's 100% my baby and I wrote it to please me.  I don't have
that luxury when working on other peoples' code, though, and if someone is
already printing to sys.stderr, I'll play along for the sake of keeping
their code true to *their* vision.  And then

   print >> sys.stderr, "Hey! I got", `got`, "but expected", `want`

1 line in 250 is pure pleasure.

> The small wrapper was not bad:
>
> from [before 'print>>']:
>
> def foobarRedirect(fileob, *args):
>     save = sys.stdout
>     try: return foobar(*args)
>     finally: sys.stdout = save
>
> to the simpler:
>
> def foobarRedirect(fileob, *args):
>     return foobar(*args, blotch=fileob)
>
> However, function foobar itself was cluttered by a more
> substantial amount than the wrapper was simplified, what with
> each 'print x' becoming 'print>>blotch,x' (plus the extra
> blotch=None argument to both foobar and the other function,
> also exposed from the module, that foobar calls).

As above, I don't believe I would use print>> here either.

> This has as little 'scientific' value as any other similar
> experiment, of course (which is part of why I find your
> 'more-impressed' somewhat unlikely) -- it's not a controlled
> ergonomics experiment:

We have no funding for that, so anecdotal evidence-- in the absence of a
killer argument --is the best we're going to get.

> in the end, the evaluation of "how readable and maintainable is
> the resulting code" is about as subjective and aesthetic as one's
> original appreciation of the construct.  But then, actual human-
> factors controlled experiments are just as scarce on the side
> _supporting_ 'print>>', aren't they?

There are none in either direction that I know of, and I doubt there ever
will be.  The realm of the subjective is where a BDFL is *most* needed!

> I don't think I posted on the similar experience I did with
> code previously using flob.write, but it was even worse from
> my point of view.  Changing the calls, typically originally
> all in the form:
>     flob.write(format % (a,b,c))
> to print-with-redirect:
>     print >> flob, format % (a,b,c),
> not only had no observable benefit, but actually gave a
> slight problem in a case where an unwanted blank space got
> inserted by two back-to-back 'print>>' (while the .write
> calls, of course, did not insert it).

The space-inserting gimmicks of plain "print" are best for output intended
to be read by humans, and I think that's just as true of "print>>".  I
*suspect* (but don't know) that one reason Guido found print>> attractive is
the number of msgs (over the years) directed to sys.stderr that are
*missing* crucial spaces when done in the sys.stderr.write(msg) form -- and
since they're "error msgs", sometimes years go by before we see that one of
those is damaged.

Now a space-eager printing gimmick could have been done via a function too,
but Python already *had* a stmt for that purpose.  From that POV, it was
natural to extend it.

> But then, I had even lower expectations this time -- it seems
> that 'print>>' is studied to substitute for print-plus-
> sys.stdout-changes, not for .write (which is what I most often
> use for output).

I don't think print>> is ever a good substitute for .write(), except in
those cases where .write() is being used for human-readable output and the
author didn't want to do the stdout redirection dance around plain "print".
Else automagical space-insertion gets in the way more than it helps.  Don't
know whether Guido would agree, but that's my take on it.

> The "lower expectations" may be important -- but I can't
> brainwash myself (and wouldn't if I could) to expect the
> sky and stars from a construct that strikes me so badly:
> giving it "a fair chance" may thus be impossible for me.

Try thinking of it as a minor convenience?  I do!  I could live without it;
I'd rather not.

> Still, we're all humans (or bots), and thus, this is no
> doubt a widespread issue.  I've spent an hour or so on it,
> confirmed the preconceived ideas I had before I did, and
> my opinions about 'print>>' are, as could be expected,
>
> Be that as it may, I'm still waiting to see real-life
> examples, in my code *or others'*, where the use of
> 'print>>' will strike me as "ah yes, clearly the one
> right, obvious way to do it".

A quick regexp search turns up about 135 uses in the current CVS tree.  Some
I don't like.  Some I do.  I'll mention two of the latter because I expect
to hear about all the others from you <wink>.

1. The last line of Lib/test/test_largefile.py:

    print >>sys.stderr, name, "exists:", os.path.exists(name)

A human-readable msg and the only use of sys.stderr in the module.  There is
no (IMO) alternative as obvious or convenient as this way of writing it.

2. Lib/warnings.py:  3 instances in 280 lines of code, again all trying to
write something human-readable to sys.stderr.

And I'll toss in one marginal one:  Lib/smtpd.py has 12 instances of

    print >> DEBUGSTREAM, some, human, readable, msg

That's in over 531 lines of code, and it never appears often in a single
function, so it's not being overused by my lights.  But I prefer using a
debug *function* for this kind of thing:  it's more flexible.  For example,
the module's default value for DEBUGSTREAM is

DEBUGSTREAM = Devnull()

where the author wrote a Devnull class with a do-nothing .write() method:

class Devnull:
    def write(self, msg): pass
    def flush(self): pass

While just a few lines of code, it's on the edge of being strained.  I'd
also rather see these lines as

    DEBUG(some, human, readable, msg)

because it's the DEBUG part that's important, not the "print" part (and,
indeed, the *normal* use of these lines is *not* to print!).

Ack -- it's after 4AM again, and we've got a release to get out tomorrow
(today?  who can tell anymore <wink/sigh>).  One quickie:

> ...
> I guess, when he comes around to it, since he has no C background
> to understand the rules for the format string; I'd better write up
> something for him -- or can somebody help with an URL about %-
> formatting already written up for somebody with zero C background?).

I don't, but there's a very *good* writeup at:

    http://www-ccs.ucsd.edu/c/lib_prin.html

Lots of examples, explanation, and even a railroad diagram for the syntax of
a format code (which I'm afraid is more confusing than helpful!).

At least the docs for 2.1 contain *a* description of each format code.  The
contributed patch Fred started from can be found here:

http://sourceforge.net/patch/index.php?func=detailpatch&
    patch_id=103412&group_id=5470

good-topic-for-a-"howto"!-dense-little-languages-aren't-obvious-ly
    y'rs  - tim




From mwh21 at cam.ac.uk  Sun Feb 18 18:52:07 2001
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 18 Feb 2001 17:52:07 +0000
Subject: [Q] how to protect python program from decompilation
References: <newscache$8sny8g$bdk$1@lnews.actcom.co.il> <mailman.982514238.21934.python-list@python.org> <newscache$ixqy8g$mwo$1@lnews.actcom.co.il>
Message-ID: <m37l2nev5k.fsf@atrus.jesus.cam.ac.uk>

Leonid Gluhovsky <gleonid at actcom.co.il> writes:

> Right now we are considering the following scheme: build a Python
> with opcodes in Include/opcode.h reshuffled, and without the dis module
> in it; use this interpreter to byte compile our code; use its Tools/freeze
> to package our program into executable and ship it to customers. 
> 
> What are the holes in this approach?  What is a better approach?

Sounds like a reasonable (effort on your part)/(effort on their part
tradeoff).  It would still probably be crackable (by statistical
analysis of byteocde of decompiling the assembly of your ceval.c), but
it would be a lot of work.

Cheers,
M.

-- 
  I also feel it essential to note, [...], that Description Logics,
  non-Monotonic Logics, Default Logics and Circumscription Logics can
  all collectively go suck a cow. Thank you.
              -- http://advogato.org/person/Johnath/diary.html?start=4


From max at alcyone.com  Sun Feb 18 02:03:57 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 17 Feb 2001 17:03:57 -0800
Subject: wildcard in tuples
References: <8XDj6.13$sR5.356@typhoon.nyu.edu>
Message-ID: <3A8F1F7D.6781AF14@alcyone.com>

cyberian bear wrote:

> lets say i have a tuple in one list and i need to check it against
> every
> tuple in second list.  So (e,c) --and i want to find every tuple in
> second
> list which has tuples who are (e, *). What is the proper syntax for
> doing
> this

I'm not sure exactly what you're asking.  If all you want is to select
all 2-tuples that, say, have 2 as the first element, then you can do it
easily enough with list comprehensions:

>>> l = [(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]
>>> [ x for x in l if x[0] == 2 ]
[(2, 3), (2, 4)]

This can obviously be trivially modified for matching on the second
list.

You could do something similar using None as a wildcard:

>>> def selectTuple(list, tuple):
...     return [ x for x in list if (x[0] == tuple[0] or tuple[0] is \
...              None) and (x[1] == tuple[1] or tuple[1] is None) ]
... 
>>> l = [(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]
>>> selectTuple(l, (2, 3))
[(2, 3)]
>>> selectTuple(l, (2, None)) # tuples beginning with 2
[(2, 3), (2, 4)]
>>> selectTuple(l, (None, 4)) # tuples ending with 4
[(2, 4), (4, 4)]
>>> selectTuple(l, (None, None)) # all tuples
[(1, 1), (2, 3), (2, 4), (3, 5), (3, 6), (4, 4)]

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Society attacks early when the individual is helpless.
\__/ B.F. Skinner
    Physics reference / http://www.alcyone.com/max/reference/physics/
 A physics reference.


From vonWedel at lfpt.rwth-aachen.de  Thu Feb 22 15:43:58 2001
From: vonWedel at lfpt.rwth-aachen.de (Lars von Wedel)
Date: Thu, 22 Feb 2001 15:43:58 +0100
Subject: XML Module problems
References: <mailman.982845604.24626.python-list@python.org> <j47l2iua5d.fsf@informatik.hu-berlin.de>
Message-ID: <3A9525AE.D92CBBC4@lfpt.rwth-aachen.de>

Hello Martin,

> > First I tried Python on a Solaris machine, finding out that the
> > python distribution installed does not contain _any_ XML parsers!?
> 
> It sure does include pyexpat - you have to have the expat library to
> compile it, though. Of course, I don't know how you had installed
> Python, so it is certainly possible that there is no XML parser in
> your installation.

As far as I know we installed a binary package from some site, don't
know
from where exactly, though. However, in xml/parsers there is expat.py
but no pyexpat.

> > Hence I chose to install PyXML -- it installs fine but running it
> > gives me a segmentation fault in sax.make_parser(), so probably it
> > doesn't succeed in loading the pyexpat.so module. Are there any
> > binary versions of the module?
> 
> What version of PyXML? It works fine for me. If you need binary
> releases: what version of Solaris?
It shows the same behavior for both 0.6.2 and 0.6.3. 'uname -a'
says 'SunOS kazul 5.7 Generic sun4m sparc'. I just tried to recompile
the package on a more recent machine (SOS 5.7, but type sun4u instead
of sun4m) and this one seems to work (on both machines!), I can create
an Expat after installing the package. Strange, but anyway.

> >     self.feed(buffer)
> >   File "d:\programme\python20\lib\xml\sax\expatreader.py", line 81, in
> > feed
> >     self._parser.Parse(data, isFinal)
> > TypeError: too many arguments; expected 1, got 2
 
> That is a confusing traceback. It is not an error in the call to
> Parse, but in the callback invoked from pyexpat. It would be good if
> you could post your ContentHandler.

Now, that's indeed confusing! With this hint I could quickly trace the
problem down.

Thanks for your quick help,
Lars



From rvprasad at cis.ksu.edu  Thu Feb  1 23:42:43 2001
From: rvprasad at cis.ksu.edu (Venkatesh Prasad Ranganath)
Date: 01 Feb 2001 16:42:43 -0600
Subject: PyXML
References: <m37l3f76i7.fsf@boss.dreamsoft.com> <3A776B59.FDEC3975@ogbuji.net>
Message-ID: <m3snlyxbx8.fsf@boss.dreamsoft.com>

Hi,

The example is 

from xml.dom.ext.reader.Sax2 import Reader
from xml.dom.ext import PrettyPrint
import sys

i = open(sys.argv[1])
j = Reader(1).fromStream(i)
i.close()
PrettyPrint(j)

print "++++", j.doctype.__systemId, "----"

Nothing is printed between ++++ and ----

Also, shouldn't Line 61 in _xmlplus/dom/NamedNodeMap.py be 
	return self.get(('', name))
and Line 65 in _xmlplus/xpath/XPathParserBase.py be
	if XPath.cvar.g_errorOccured == 1:

-- 
Venkatesh Prasad Ranganath


From Pieter_Claerhout at CreoScitex.com  Mon Feb 12 16:04:27 2001
From: Pieter_Claerhout at CreoScitex.com (Pieter Claerhout)
Date: Mon, 12 Feb 2001 16:04:27 +0100
Subject: Command prompts on a Windows station
Message-ID: <2B1262E83448D211AE4B00A0C9D61B030130191A@msgeuro1.177.115.149.IN-ADDR.ARPA>

Hello all,

is there a way in Python to catch that a command prompt was closed by
clicking on the close button in the title bar? I have a script that is
launching several other scripts, and I would like to kill those processes
again once the main script is closed again. I can catch this when he is
pressing CTRL+BREAK, but I can seem to catch is when the window is closed
using the close button? Any ideas? I remember that an email about this was
going around, but I can't seem to find it back...

Thanks,


Pieter



From mlh at idi.ntnu.no  Fri Feb  2 14:03:41 2001
From: mlh at idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 2 Feb 2001 14:03:41 +0100
Subject: "in" operator for strings
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com>
Message-ID: <95ebcv$bn$1@tyfon.itea.ntnu.no>

"Alex Martelli" <aleaxit at yahoo.com> wrote in message
news:95birk0r0d at news1.newsguy.com...
> "Magnus Lie Hetland" <mlh at idi.ntnu.no> wrote in message
> news:95beco$e0u$1 at tyfon.itea.ntnu.no...
[...]
> Right, and an extension of this is basically what's being
> asked for (though the original poster may not have thought
> of this 'obvious' generalization, specialcasing string would
> surely not be warranted).  Unfortunately, for general cases
> it doesn't scale well -- i.e., now:

Well... Then "in" would test for both membership and
subsequenceness... Quite a strong case of ambiguity if you
ask me. In my opinion a really *bad* idea. (And would it
only allow contiguous subsequences, or any subsequence?
If we ever get built-in sets contiguity would be meningless,
as would it be if we get "in" tests for dictionaries.)

[...]
> and having it return 1 in the second case too would be making
> this 'in' very ambiguous and confusing, alas.

Right.

> Also, of course, this would throw any parallel between
> "x in y" and "for x in y" out of the windows unless the
> latter starts looping on all *subsequences* -- eeep!-)

Actually, that sounds very interesting ;-)

How about having a power-sequence-function (like the standard
power-set function in mathematics)? Then pow(seq) or seq.pow()
would return a (perhaps lazy) sequence containing all
subsequences... And one could have seq.pow(contiguous=1) if
one only wanted contiguous subsequences... Then one would
have:

  >>> "Waldo" in "Ralph Waldo Emerson".pow(contiguous=1)
  1

An ingenious implementation of the laziness would be needed
I guess... Or maybe not. A simple string-matching algorithm
would be needed for the contiguous case, and a O(n*m)
exhaustive search for the non-contiguous case. (Or perhaps
something better would be possible by putting the elements
in a hash table... O(n+m)? Might have quite some overhead,
perhaps... Oh, well)

>
> > But you can't do what you ask for, just like you can't write
> >
> >   [1, 2] in [1, 2, 3, 4]
>
> Sure you can, it's a well-formed test and returns 0 since
> [1,2] is not an item in the right-hand operand sequence.

Right. That's essentially what I meant. You can't write the
above to find out whether [1, 2] is a subsequence of
[1, 2, 3, 4].

> > Probably a better idea.
>
> Only if you're looking for words, not for any substring,
> which is at least as frequent.

You are right. But it's a better idea than doing something
that doesn't work, at least <0.8 wink>

> >    "Waldo" in split("Ralph Waldo Emerson")
> >
> > It might be old-fashioned, but... So what :-)
>
> So it doesn't work unless you "from string import *" (horrid
> idea), "from string import split" (doubtful), or rewrite it
> using an explicit string.split (probably best,

Why? This would in my opinion clearly depend on the size of
the script... If you don't expect another split function to
appear, I think it's quite OK to use

  from string import split

I mean -- I'm not against string methods. They're nice. I
just get a bit woozy seeing them called on literals. It
just gives me flashes of weird stuff like

  1.plus(2)

And... In my mind methods seem to be for modifying an
objects state first of all. Oh, well. Here we go -- another
pointless discussion. Sorry :)

> For general substring-matching, a class wrapper is not
> too bad:

[...]
> this only works for strings, as written, AND only to
> enable such idioms as
>
>     if 'ald' in subsOf("Waldo"):
>         print 'yep!'

I should have read the entire posting before replying, I
see. <wink>

This is the same as my power-sequence idea I guess.

> Alex

--

  Magnus Lie Hetland      (magnus at hetland dot org)

 "Reality is what refuses to disappear when you stop
  believing in it"                 -- Philip K. Dick





From chris.gonnerman at usa.net  Sun Feb  4 22:03:46 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sun, 4 Feb 2001 15:03:46 -0600
Subject: Paging Richard Jones...
Message-ID: <009501c08eee$1489a6e0$a100000a@local>

Richard, if you are reading this list, I've been trying to reach you
regarding the Python
GD module you say you aren't maintaining anymore.  I've updated it
dramatically and
am willing to take it over from you.  If you have changed your email address
you need
to update your website.

To all:  If you know how to contact Richard, please let me know (or forward
this to him).





From phrxy at csv.warwick.ac.uk  Tue Feb 27 19:29:18 2001
From: phrxy at csv.warwick.ac.uk (John J. Lee)
Date: Tue, 27 Feb 2001 18:29:18 +0000 (GMT)
Subject: [Distutils] Re: CPAN functionality for python - requirements
In-Reply-To: <20010227110803.Z1781@tummy.com>
Message-ID: <Pine.SOL.4.30.0102271823190.22762-100000@mimosa.csv.warwick.ac.uk>

Followup to Catalog-sig only (pine doesn't seem to let me set that
in the headers...)?  I don't think this has much to do with Distutils.

On Tue, 27 Feb 2001, Sean Reifschneider wrote:

> On Tue, Feb 27, 2001 at 09:30:13AM -0700, Evelyn Mitchell wrote:
> >But it will also discover and resolve dependences in your perl
> >site-packages, and automatically fetch them from your closest
> >CPAN archive.
>
> Not according to my tests the night before last.  I did a test CPAN
> install of "News::Newsrc", which failed because the "make test" was
> failing.  I then installed the "Set::BitSet" (?  Something like that)
> module and tried News::Newsrc again and it worked...
[...]

Did you use CPAN.pm?  CPAN the archive doesn't do dependencies.
Dependencies are listed in the Makefile.PL (BTW, I have no idea why the
capitalised PL), and after the module CPAN.pm downloads what you asked
for, it checks that to see if there are dependencies, then downloads them
if you don't already have them installed.


John




From aleaxit at yahoo.com  Tue Feb 27 17:17:46 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Tue, 27 Feb 2001 17:17:46 +0100
Subject: How to exit Python nicely after error in embedded C code?
References: <3A9B3B7D.66E31386@erols.com> <97fjl9028ir@news2.newsguy.com> <3A9BBD28.8F537F75@northwestern.edu>
Message-ID: <97gjv80147@news1.newsguy.com>

"Louis Luangkesorn" <lluang at northwestern.edu> wrote in message
news:3A9BBD28.8F537F75 at northwestern.edu...
    [talking about "exceptions" from C vs C++]
> (I'm asking this out of complete ignorance, obviously)
> In this situation (error occurs in a C function that is called by other C
> functions which are called by Python), would having the original C call
from
> Python in a try...except block catch the exception?  (I'm just learning
how to
> do this, in both C++ and Python, but I have not done this in real life
yet.  It
> seems like a nice, cheap way of using exceptions in C, just call
everything from
> Python :-)

The C language does not define nor use exceptions.

The Python C API, therefore, does not use them either.  Python is
entirely coded in C, to the C89 standard of ANSI and later ISO, not
in C++.

_Python_ exceptions are not the same thing as _C++_ exceptions.
A "Python Exception" is "raised" by a C-coded function, in an
extension, that is called by Python, by returning 0 (aka NULL)
as its PyObject* (and some previous Python C API calls to give
the exceptions' details).

There are several frameworks that make it easier to extend Python
by harnessing (some of) the extra power of C++ wrt C: CXX, SCXX,
the Boost Python Library (BPL, part of the free Boost collection,
see www.boost.org), and others yet.  Some of them do make use of
the power of C++ exceptions, somewhat in the way you suggest, and,
also, may bridge/translate between C++ exceptions and Python ones.

But there is no clean, portable way to "raise an exception" if you
choose to use C, rather than C++, to code your Python extensions,
and you find yourself in a C function called by a C function [repeat
down a few nesting levels...] called by Python.  Either switch to
C++ (and some appropriate C++/Python framework such as the BPL), or
live with patiently returning an error indication from the deeply
nested function, and checking-and-propagating-upwards in each
caller.

That is, of course, IF you care about portability; if you don't, you
may experiment to see what your platform[s] of interest will allow
in a setjmp/longjmp pair, which is the closest thing that C offers
to "exceptions" (just take a *LOT* of care about memory leaks &c,
as you get absolutely no protection, no equivalent of C++'s automatic
execution of destructors as the stack is unwound due to exceptions).


Alex





From claird at starbase.neosoft.com  Fri Feb 23 15:35:38 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 23 Feb 2001 08:35:38 -0600
Subject: Languages from within Tcl [Was: Re: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr> <975irc$16g$1@srv38.cas.org>
Message-ID: <E43EEA3E68CAF21B.712DA4A67EDE6509.58FFB479E492D2F2@lp.airnews.net>

In article <975irc$16g$1 at srv38.cas.org>,  <lvirden at cas.org> wrote:
>
>Okay, we now have 
>
>o	tclblend
>o	tclperl
>o	tclpython
>o	embedded c
>
>What other programming languages have been integrated into Tcl?
			.
			.
			.
Ada, Forth, Prolog, ...  Occasionally I convulse into an effort to catalogue 'em
<URL: http://starbase.neosoft.com/~claird/comp.lang.misc/polyglot.html >
I welcome help.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From paul at fxtech.com  Sat Feb 17 19:55:21 2001
From: paul at fxtech.com (paul at fxtech.com)
Date: Sat, 17 Feb 2001 18:55:21 -0000
Subject: Python 2.1 - tokenizer fgets crash (Windows)
In-Reply-To: <LNBBLJKPBEHFEDALKOLCIEANJAAA.tim.one@home.com>
Message-ID: <96mhep+u7oa@eGroups.com>

> You do:  "crashes" could mean anything on Windows.  What does it 
mean to
> you?  Is it dying inside MS's implementation of fgets(), in Python; 
is it a
> NULL-pointer error, an illegal instruction; etc etc.  Try it under a 
debug
> build, if you haven't already (single-stepping can be highly 
misleading if
> you're doing that with the non-debug build).

Yes, I'm using the debug build (as I mentioned) and I have 
single-stepped down into the code. It's crashing in Microsoft's 
fgets() function. The FILE * looks correct - the only thing I can 
think of is I'm doing the fopen() in my code and it is getting used in 
the DLL. Loading files from within the DLL (using Import) work fine.

I've gotten around it for now - I'm now importing my scripts as 
modules rather than running them.





From fredrik at effbot.org  Fri Feb  2 09:10:25 2001
From: fredrik at effbot.org (Fredrik Lundh)
Date: Fri, 02 Feb 2001 08:10:25 GMT
Subject: sys.exit and exceptions (newbie question)
References: <3a7998cf.869169328@localhost> <mailman.981087063.16577.python-list@python.org>
Message-ID: <RXte6.11068$AH6.1772830@newsc.telia.net>

Thomas Wouters wrote:
> You can use sys._exit(). This exits in a very abortive way -- right away,
> dropping everything it's doing, without doing any kind of cleanup.

only if you're not catching AttributeError, that is ;-)

>>> import sys
>>> sys._exit()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: _exit

(it's "os._exit(code)", not "sys._exit()")

Cheers /F




From patfourc at libertysurf.fr  Thu Feb 22 19:41:23 2001
From: patfourc at libertysurf.fr (Liberty)
Date: Thu, 22 Feb 2001 19:41:23 +0100
Subject: File browser
Message-ID: <973mgo$l3o$1@news2.isdnet.net>

Hello, I'm a french student looking for a file browser written in Python to
include it in my study project.

Does anyone got it ?


Thanks





From dwndgg at opti.cgi.net  Tue Feb 27 07:23:04 2001
From: dwndgg at opti.cgi.net (Dan)
Date: Tue, 27 Feb 2001 06:23:04 GMT
Subject: Breaking out of nested loop
Message-ID: <3a9b44f1.226505016@news.okc1.ok.home.com>

How might one go about breaking out of a multi level loop?

For instance, the following code will go into an endless loop.

while(1):
    while(1):
        break


Is there a statement that one can use to break out of multiple levels
of loop control?  I really hate to start littering my code with flag
variables and if statements.

-Dan



From ngps at madcap.dyndns.org  Sat Feb  3 04:43:07 2001
From: ngps at madcap.dyndns.org (Ng Pheng Siong)
Date: 3 Feb 2001 03:43:07 GMT
Subject: SSL EOF
References: <mailman.981087116.16733.python-list@python.org>
Message-ID: <95fuob$t3q$1@clematis.singnet.com.sg>

According to Tim Peters <tim.one at home.com>:
> The good news is that it will in Python 2.1a2 also (to be released
> tomorrow).  However, the more we (Jeremy, Guido, me) looked at Python's SLL
> support, the more questionable code we found.  Alas, not being experts in
> this protocol, we're unlikely to fix it.  So we would be grateful for a
> patch against CVS Python for doing this stuff more intelligently, including
> corresponding repairs to the httplib.py code that uses it.

I'll see what I can do. (I recall saying that when 2.0 came out... ;-)

Cheers.

-- 
Ng Pheng Siong <ngps at post1.com> * http://www.post1.com/home/ngps



From Dan.Rolander at marriott.com  Fri Feb  2 18:41:01 2001
From: Dan.Rolander at marriott.com (Rolander, Dan)
Date: Fri, 2 Feb 2001 12:41:01 -0500 
Subject: Four problems with Gordon McMillan's Installer
Message-ID: <6176E3D8E36FD111B58900805FA7E0F80CCF6360@mcnc-mdm1-ex01>

Gordon (Williams):

I am using Installer with wxPython 2.2.2, and Python 2.0 under Win2K. I had
some initial problems but everything is fine now. Here are my comments:

1) Yes, the problem with wxPython.wxc.pyd is known. Apparently this started
happening with Python 2.0. You'll see this often-- files will be named with
the full package path name. As far as I know the only workaround is to
rename them after the build.

2) Yes, I also include wx22_2.dll on the misc line, like this--
misc= MYSTANDALONE, wxPython.wx22_2

3) I see the fixcase messages also but they don't seem to affect anything.

4) I have an application that is using wxPython and the time module and I
haven't had a problem with it. BUT... I do have a couple of updated files
from Gordon McMillan that are newer than the 3i distribution and did fix A
LOT of problems I was having before. He's advised they will be part of the
3j distribution whenever that is released. I'm attaching them. I recommend
you replace your existing versions with these.

5) Yes, Gordon's starship pages are out of date and I don't know why they
are still there. Be sure to use http://www.mcmillan-inc.com instead.

I hope this helps,
Dan

-------------- MEInc\dist\bindepend.py ----------------
# copyright 1999 McMillan Enterprises, Inc.
# license: use as you please. No warranty.
#
# use dumpbin.exe (if present) to find the binary
# dependencies of an extension module.
# if dumpbin not available, pick apart the PE hdr of the binary
# while this appears to work well, it is complex and subject to
# problems with changes to PE hdrs (ie, this works only on 32 bit Intel
# Windows format binaries)
#
# Note also that you should check the results to make sure that the
# dlls are redistributable. I've listed most of the common MS dlls
# under "excludes" below; add to this list as necessary (or use the
# "excludes" option in the INSTALL section of the config file).

import os
import time
import string
import sys
import tempfile
import finder

seen = {}
excludes = {'KERNEL32.DLL':1, 
      'ADVAPI.DLL':1, 
      'MSVCRT.DLL':1,
      'ADVAPI32.DLL':1,
      'COMCTL32.DLL':1,
      'CRTDLL.DLL':1,
      'GDI32.DLL':1,
      'MFC42.DLL':1,
      'NTDLL.DLL':1,
      'OLE32.DLL':1,
      'OLEAUT32.DLL':1,
      'RPCRT4.DLL':1,
      'SHELL32.DLL':1,
      'USER32.DLL':1,
      'WINSPOOL.DRV':1,
      'WS2HELP.DLL':1,
      'WS2_32.DLL':1,
      'WSOCK32.DLL':1,
      'WINMM.DLL':1,
      'COMDLG32.DLL':1,
      'ZLIB.DLL':1,
      'ODBC32.DLL':1,
      'VERSION.DLL':1}     

def getfullnameof(mod, xtrapath = None):
  """Return the full path name of MOD.
  
      MOD is the basename of a dll or pyd.
      XTRAPATH is a path or list of paths to search first.
      Return the full path name of MOD.
      Will search the full Windows search path, as well as sys.path"""
  epath = finder.getpath()
  if mod[-4:] in ('.pyd', '.PYD'):
    epath = epath + sys.path
  if xtrapath is not None:
    if type(xtrapath) == type(''):
      epath.insert(0, xtrapath)
    else:
      epath = xtrapath + epath
  for p in epath:
    npth = os.path.join(p, mod)
    if os.path.exists(npth):
      return npth
  return ''
  
def getImports1(pth):
    """Find the binary dependencies of PTH.
    
        This implementation (not used right now) uses the MSVC utility
dumpbin"""
    rslt = []
    tmpf = tempfile.mktemp()
    os.system('dumpbin /IMPORTS "%s" >%s' %(pth, tmpf))
    time.sleep(0.1)
    txt = open(tmpf,'r').readlines()
    os.remove(tmpf)
    i = 0
    while i < len(txt):
        tokens = string.split(txt[i])
        if len(tokens) == 1 and string.find(tokens[0], '.') > 0:
            rslt.append(string.strip(tokens[0]))
        i = i + 1
    return rslt
    
def getImports2(pth):
    """Find the binary dependencies of PTH.
    
        This implementation walks through the PE header"""
    import struct
    rslt = []
    try:
      f = open(pth, 'rb').read()
      pehdrd = struct.unpack('l', f[60:64])[0]
      magic = struct.unpack('l', f[pehdrd:pehdrd+4])[0]
      numsecs = struct.unpack('h', f[pehdrd+6:pehdrd+8])[0]
      numdirs = struct.unpack('l', f[pehdrd+116:pehdrd+120])[0]
      idata = ''
      if magic == 17744:
          importsec, sz = struct.unpack('2l', f[pehdrd+128:pehdrd+136])
          secttbl = pehdrd + 120 + 8*numdirs
          secttblfmt = '8s7l2h'
          seclist = []
          for i in range(numsecs):
              seclist.append(struct.unpack(secttblfmt,
f[secttbl+i*40:secttbl+(i+1)*40]))
              #nm, vsz, va, rsz, praw, preloc, plnnums, qrelocs, qlnnums,
flags \
              # = seclist[-1]
          for i in range(len(seclist)-1):
              if seclist[i][2] <= importsec < seclist[i+1][2]:
                  break
          vbase = seclist[i][2]
          raw = seclist[i][4]
          idatastart = raw + importsec - vbase
          idata = f[idatastart:idatastart+seclist[i][1]]
          i = 0
          while 1:
              vsa =  struct.unpack('5l', idata[i*20:i*20+20])[3]
              if vsa == 0:
                  break
              sa = raw + vsa - vbase
              end = string.find(f, '\000', sa)
              nm = f[sa:end]
              if nm:
                  rslt.append(nm)
              i = i + 1
    except IOError:
      print "bindepend cannot analyze %s - file not found!" % pth
    except struct.error:
      print "bindepend cannot analyze %s - error walking thru pehdr" % pth
    return rslt
    
def Dependencies(lTOC):
  """Expand LTOC to include all the closure of binary dependencies.
  
     LTOC is a logical table of contents, ie, a seq of tuples (name, path).
     Return LTOC expanded by all the binary dependencies of the entries
     in LTOC, except those listed in the module global EXCLUDES"""
  for (nm, pth) in lTOC:
    fullnm = string.upper(os.path.basename(pth))
    if seen.get(string.upper(nm),0):
      continue
    print "analyzing", nm
    seen[string.upper(nm)] = 1
    dlls = getImports(pth)
    for lib in dlls:
        print " found", lib
        if excludes.get(string.upper(lib),0):
          continue
        if seen.get(string.upper(lib),0):
          continue
        npth = getfullnameof(lib)
        if npth:
          lTOC.append((lib, npth))
        else:
          print " lib not found:", lib, "dependency of", 
  return lTOC
  
        
##if getfullnameof('dumpbin.exe') == '':
##    def getImports(pth):
##        return getImports2(pth)
##else:
##    def getImports(pth):
##        return getImports1(pth)
        
def getImports(pth):
    """Forwards to either getImports1 or getImports2
    """
    return getImports2(pth)



-------------- support\archive_rt.py ----------------
#
# Gordon McMillan (as inspired and influenced by Greg Stein)
#

# subclasses may not need marshal or struct, but since they're
# builtin, importing is safe.
#
# While an Archive is really an abstraction for any "filesystem
# within a file", it is tuned for use with imputil.FuncImporter.
# This assumes it contains python code objects, indexed by the
# the internal name (ie, no '.py').
# See carchive.py for a more general archive (contains anything)
# that can be understood by a C program.

#archive_rt is a stripped down version of MEInc.Dist.archive.
#It has had all building logic removed.
#It's purpose is to bootstrap the Python installation.

_verbose = 0
_listdir = None
import marshal
import struct
import imp
import sys
import imputil

_c_suffixes = filter(lambda x: x[2] == imp.C_EXTENSION, imp.get_suffixes())

for nm in ('nt', 'posix', 'dos', 'os2', 'mac'):
  if nm in sys.builtin_module_names:
    mod = __import__(nm)
    _listdir = mod.listdir

class Archive(imputil.Importer):
  """ A base class for a repository of python code objects.
      The extract method is used by imputil.ArchiveImporter
      to get code objects by name (fully qualified name), so
      an enduser "import a.b" would become
        extract('a.__init__')
        extract('a.b')
  """
  MAGIC = 'PYL\0'
  HDRLEN = 12        # default is MAGIC followed by python's magic, int pos
of toc
  TOCPOS = 8
  TRLLEN = 0        # default - no trailer
  TOCTMPLT = {}     #
  os = None
  _bincache = None
  def __init__(self, path=None, start=0):
    "Initialize an Archive. If path is omitted, it will be an empty
Archive."
    self.toc = None
    self.path = path
    self.start = start
    import imp
    self.pymagic = imp.get_magic()
    if path is not None:
      self.lib = open(self.path, 'rb')
      self.checkmagic()
      self.loadtoc()
      if Archive._bincache is None:
        Archive._bincache = {}
        for fnm in _listdir(sys.path[0]):
          for ext, mode, typ in _c_suffixes:
            sz = len(ext)
            if fnm[-sz:] == ext:
              Archive._bincache[fnm[:-sz]] = sys.path[0] + '/' + fnm, (ext,
mode, typ)
              if _verbose:
                print "Archive caching", fnm[:-sz]

  ####### Sub-methods of __init__ - override as needed #############
  def checkmagic(self):
    """ Overridable.
        Check to see if the file object self.lib actually has a file
        we understand.
    """
    self.lib.seek(self.start) #default - magic is at start of file
    if self.lib.read(len(self.MAGIC)) != self.MAGIC:
      raise RuntimeError, "%s is not a valid %s archive file" \
% (self.path, self.__class__.__name__)
    if self.lib.read(len(self.pymagic)) != self.pymagic:
      raise RuntimeError, "%s has version mismatch to dll" % (self.path)

  def loadtoc(self):
    """ Overridable.
        Default: After magic comes an int (4 byte native) giving the
        position of the TOC within self.lib.
        Default: The TOC is a marshal-able string.
    """
    self.lib.seek(self.start + self.TOCPOS)
    (offset,) = struct.unpack('=i', self.lib.read(4))
    self.lib.seek(self.start + offset)
    self.toc = marshal.load(self.lib)

  ######## This is what is called by FuncImporter #######
  ## Since an Archive is flat, we ignore parent and modname.

  def get_code(self, parent, modname, fqname):
    if _verbose:
      print "get_code(%s, %s, %s, %s)" % (self, parent, modname, fqname)
    rslt = self.extract(fqname) # None if not found, (ispkg, code) otherwise
    if rslt is None:
      if parent:
        importer = getattr(parent, "__importer__", None)
        if importer:
          func = getattr(importer, "func", None)
          if func == self.get_code:
            file, desc = Archive._bincache.get(fqname, (None, None))
            if file:
              try:
                fp = open(file, desc[1])
              except IOError:
                pass
              else:
                module = imp.load_module(fqname, fp, file, desc)
                if _verbose:
                  print "#found", file
                return 0, module, {'__file__':file}
      if _verbose:
        print "#not found"
      return None
    ispkg, code = rslt
    if ispkg:
      return ispkg, code, {'__path__' : [fqname]}
    return ispkg, code, {}

  ####### Core method - Override as needed  #########
  def extract(self, name):
    """ Get the object corresponding to name, or None.
        For use with imputil ArchiveImporter, object is a python code
object.
        'name' is the name as specified in an 'import name'.
        'import a.b' will become:
        extract('a') (return None because 'a' is not a code object)
        extract('a.__init__') (return a code object)
        extract('a.b') (return a code object)
        Default implementation:
          self.toc is a dict
          self.toc[name] is pos
          self.lib has the code object marshal-ed at pos
    """
    ispkg, pos = self.toc.get(name, (0,None))
    if pos is None:
      return None
    self.lib.seek(self.start + pos)
    return ispkg, marshal.load(self.lib)

  ########################################################################
  # Informational methods

  def contents(self):
    """Return a list of the contents
       Default implementation assumes self.toc is a dict like object.
       Not required by ArchiveImporter.
    """
    return self.toc.keys()

  ########################################################################
  # Building
  
  ####### Top level method - shouldn't need overriding #######
##  def build(self, path, lTOC):
##    """Create an archive file of name 'path'.
##       lTOC is a 'logical TOC' - a list of (name, path, ...)
##       where name is the internal name, eg 'a'
##       and path is a file to get the object from, eg './a.pyc'.
##    """
##    self.path = path
##    self.lib = open(path, 'wb')
##    #reserve space for the header
##    if self.HDRLEN:
##      self.lib.write('\0'*self.HDRLEN)
##
##    #create an empty toc
##
##    if type(self.TOCTMPLT) == type({}):
##      self.toc = {}
##    else:       # assume callable  
##      self.toc = self.TOCTMPLT()
##
##    for tocentry in lTOC:
##      self.add(tocentry)   # the guts of the archive
##
##    tocpos = self.lib.tell() 
##    self.save_toc(tocpos)
##    if self.TRLLEN:
##      self.save_trailer(tocpos)
##    if self.HDRLEN:
##      self.update_headers(tocpos) 
##    self.lib.close()
##
##
##  ####### manages keeping the internal TOC and the guts in sync #######
##  def add(self, entry):
##    """Override this to influence the mechanics of the Archive.
##       Assumes entry is a seq beginning with (nm, pth, ...) where
##       nm is the key by which we'll be asked for the object.
##       pth is the name of where we find the object. Overrides of
##       get_obj_from can make use of further elements in entry.
##    """
##    if self.os is None:
##      import os
##      self.os = os
##    nm = entry[0]
##    pth = entry[1]
##    ispkg = self.os.path.splitext(self.os.path.basename(pth))[0] ==
'__init__'
##    self.toc[nm] = (ispkg, self.lib.tell())
##    f = open(entry[1], 'rb')
##    f.seek(8) #skip magic and timestamp
##    self.lib.write(f.read())
##
##  def save_toc(self, tocpos):
##    """Default - toc is a dict
##       Gets marshaled to self.lib
##    """
##    marshal.dump(self.toc, self.lib)
##
##  def save_trailer(self, tocpos):
##    """Default - not used"""
##    pass
##
##  def update_headers(self, tocpos):
##    """Default - MAGIC + Python's magic + tocpos"""
##    self.lib.seek(self.start)
##    self.lib.write(self.MAGIC)
##    self.lib.write(self.pymagic)
##    self.lib.write(struct.pack('=i', tocpos))
   
##############################################################
#
# ZlibArchive - an archive with compressed entries
#

class ZlibArchive(Archive):
  MAGIC = 'PYZ\0'
  TOCPOS = 8
  HDRLEN = 12
  TRLLEN = 0
  TOCTMPLT = {}
  LEVEL = 9

  def __init__(self, path=None, offset=0):
    Archive.__init__(self, path, offset)
    # dynamic import so not imported if not needed
    global zlib
    import zlib
   
  def extract(self, name):
    (ispkg, pos, lngth) = self.toc.get(name, (0, None, 0))
    if pos is None:
      return None
    self.lib.seek(self.start + pos)
    return ispkg, marshal.loads(zlib.decompress(self.lib.read(lngth)))

##  def add(self, entry):
##    if self.os is None:
##      import os
##      self.os = os
##    nm = entry[0]
##    pth = entry[1]
##    ispkg = self.os.path.splitext(self.os.path.basename(pth))[0] ==
'__init__'
##    f = open(pth, 'rb')
##    f.seek(8) #skip magic and timestamp
##    obj = zlib.compress(f.read(), self.LEVEL)
##    self.toc[nm] = (ispkg, self.lib.tell(), len(obj))
##    self.lib.write(obj)
## 

if type(__import__) == type(open):
    imputil.ImportManager().install()
    sys.path.append(imputil.BuiltinImporter())




From max at alcyone.com  Thu Feb 15 02:03:53 2001
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 14 Feb 2001 17:03:53 -0800
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org> <mailman.982186876.10323.python-list@python.org>
Message-ID: <3A8B2AF9.4E2A4029@alcyone.com>

Ben Ocean wrote:

> LOL! No, sorry that I misled you! I want to contact webmasters
> individually
> with individualized emails and invite them to swap links. So... any
> ideas
> on a script?

So then, yes, you _do_ want build a spam database.  Unsolicited bulk
email is unsolicited bulk email, regardless of whether or not _you_
think it's relevant.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ War is cruelty, and you cannot refine it.
\__/ Karl Shapiro
    Erik Max Francis' bookmarks / http://www.alcyone.com/max/links/
 A highly categorized list of Web links.


From flognat at flognat.myip.org  Sun Feb 25 21:30:31 2001
From: flognat at flognat.myip.org (Andrew Markebo)
Date: Sun, 25 Feb 2001 20:30:31 GMT
Subject: Easing up (client) handling of html-forms.
Message-ID: <m3zofa5wwm.fsf@master.athome>

I am working on reading, posting forms (web) and so on, and wonder if
there are any work done to ease up this??

Basically what I am looking for is something to throw a html-document
into, and out comes some kind of data representation of the forms,
that can be used when wanting to post the form back to the server.. 

Assume that we threw it at hotmail.. we would do something like:

<>> document=http.get("www.hotmail.com")
<>> forms=FORMS(document)
<>> loginform=forms["passwordform"]
<>> print loginform.fields
["login", "password", "svc", "name", "RemoteDAPost"]
<>> loginform["login"]="user"
<>> loginform["password"]="pass"
# the rest of the fields in the hotmail form is with their default
<>> loginform.submit()... 

or similar.. Anyone out there doing stuff like this?? Or should I
continue to do some hack.. anyone else interested who can shout 'Go
Andy Go! :-)'

        /Andy

p.s. BTW Any new revolutionarizing thoughts about the prompt for
python?? I am starting to get borde by the >>>, why not turn them
around <<<? :-) 


From erno-news at erno.iki.fi  Tue Feb  6 02:57:46 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 06 Feb 2001 03:57:46 +0200
Subject: cgi.FieldStorage() problem
References: <3A7F20D9.4BE57828@cs.utwente.nl>
Message-ID: <kuelxc3751.fsf@lasipalatsi.fi>

In article <3A7F20D9.4BE57828 at cs.utwente.nl>, Richard van de Stadt
<stadt at cs.utwente.nl> writes:

| OSError: [Errno 24] Too many open files: '/var/tmp/@24739.61'

you probably just need to increase the resource limits for the account
that your http daemon runs under. i don't know what kind of startup
procedure you're using for the daemon so it's hard to say exactly how
to fix it, but the /bin/sh command for manipulating resource limits is
"ulimit" ("limit" with csh descendants).

if it still blows up after you've upped the limit, it might
be a problem with the cgi module going haywire and opening
new files boundlessly, but that seems pretty unlikely.

  -- erno


From whisper at oz.net  Fri Feb 23 20:39:21 2001
From: whisper at oz.net (Dave LeBlanc)
Date: 23 Feb 2001 19:39:21 GMT
Subject: Protecting Python Application Distributions - URGENT
Message-ID: <976e99$7pj$0@216.39.170.247>

Hi;

I have the possible chance to do a consulting project in Python, but
the customer has raised concerns about (not) doing a source
distribution.

I've read some stuff about python->C etc. but I haven't been able to
find any concrete information.

Any information about creating "exe" or "shrouded" or some other form
of protected distribution on Windows would be greatly appreciated. (I
realize that 100% protection is impossible given sufficient
motivation; I'd just like to make it irritating enough that most
people wouldn't bother trying to decompile/reconstruct the code.)

Cross-emails to whisper at oz.net greatly appreciated!

Regards,

Dave LeBlanc

P.S. Any solution that created stand-alone .exe files (no separate
python.exe needed) would be of special interest.



From bshapiro at funnygarbage.com  Tue Feb 13 22:03:04 2001
From: bshapiro at funnygarbage.com (Ben Shapiro)
Date: Tue, 13 Feb 2001 16:03:04 -0500
Subject: Please Help:  COM Events
Message-ID: <96c7de$mb9$1@bob.news.rcn.net>

Hi all
I am having a problem with COM events.
I am using the JabberCOM library.  This library contains a bunch of classes
such as JabberSession and JabberMsg.
I create the object with:   JabberCtl =
win32com.client.DispatchWithEvents('JabberCOM.JabberSession',JabberEvts)
the JabberEvts class is registered to handle events for this object....

The problem is this.... in the JabberEvts Class i have the following method
that gets triggered every time a message arrives:
def OnMessage(self, Msg, Tag):
        print "I made it this far ---" #---this works
        print Msg.MsgType #--- never executes

In this method, Msg is an object of type JabberMsg
when i do a print Msg, python reports that it is an IDispatch

Seems to me that when the event handler is called, the argument passed in
loses it's type as a JabberMsg and becomes a generic Dispatch object
so i can't call properties or methods on it.. Is there any way to cast the
argument back to a JabberMsg object?

thx for any help

Ben




From topmind at technologist.com  Wed Feb 28 16:55:36 2001
From: topmind at technologist.com (Topmind)
Date: Wed, 28 Feb 2001 15:55:36 GMT
Subject: Collection interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <97fjla128ir@news2.newsguy.com> <MPG.1505a76831c34b749896b3@news.earthlink.net> <97h5s40vhc@news2.newsguy.com> <MPG.1506007591cf818b9896bc@news.earthlink.net> <97ifd60sen@news1.newsguy.com>
Message-ID: <MPG.1506c16edd0cb6159896bf@news.earthlink.net>

> "Topmind" <topmind at technologist.com> wrote in message
> news:MPG.1506007591cf818b9896bc at news.earthlink.net...
> > > "Topmind" <topmind at technologist.com> wrote in message
> > > news:MPG.1505a76831c34b749896b3 at news.earthlink.net...
> > >     [snip]
> > > > Why can't the nodes of a multi-set ALSO have time-stamps
> > > > or a sequantial record/node number in them?
> > >
> > > Because that would be unwarranted overhead on 99.947%
> > > of the uses of a multiset...?!
> > >
> > > Why doesn't insertion into the multiset compute "priority"
> > > value according to _all_ possible formulas, as long as it's
> > > at it, so I can later decide that my multiset is actually a
> > > priority queue, too...?
> >
> > I am only saying that IF there are timestamps. I did NOT say
> > it was a prerequisite for a multi-set. You misunderstood me.
> 
> IF there are to be timestamps on insertion of items into a
> collection, a rare but far from unthinkable requirement, I do
> not want the collection to do it behind my back -- just as I
> do not want "prioritizing" to be happening without me knowing
> that (and thus being able, if need be, to customize the way
> items are timestamped, prioritized, etc).
> 


No no no! I am not advocating forced time-stamps. (Although,
a sequential unique record is quite often needed. But nothing
has to be forced.)

However, timestamping *is* a common need in my domain.


> 
> > > > Requirements change and morph and merge.
> > >
> > > And when they do, one refactors appropriately
> >
> > Refactor is a great PHB-directed euphemism for
> > code rework. Call it whatever you want, but it is
> > still unnecessary work.
> >
> > The divisions of collections into types *is*
> > arbitrary for the most part.
> 
> There is definitely a lot of arbitrariness if the division
> is made into a _taxonomy_ -- as single-inheritance thinking
> may suggest.  I'm definitely not advocating or defending
> the exclusive use of taxonomy, which IMHO tends to breed a
> somewhat warped worldview -- the natural view of abstract
> algebra does not lead to that, for example; in chains of
> specialization such as groupoid->semigroup->monoid->group,
> one may add a condition such as 'commutative' at different
> steps, so there is no taxonomic 'single-inheritance' chain
> leading to 'Abelian group'.  However, reacting to the
> modeling over-restrictiveness of taxonomy with a bid to
> drop all _classification_ is, IMHO, an over-reaction; 

When in doubt, leave it out.

Unless one can see and show some natural law of the Universe that
forces things into mutually-exclusive taxonomies, it is
bad, arbitrary modeling, and OO sould be ashamed of itself for
over promoting it. I hope the backlash is as strong as
the hype that propelled it onto the packaging buzzword
standards.

Some domains like chemistry, math, etc. might be able
to take advantage of natural divisions of "sub-things".
However, many domains don't, and OO seems like nothing
but added syntactical and modeling clutter without
sub-divisioning. Two out of the 3 OO Cononical Triads
depend on it. There is no polymorphism if there is
only one "type" with a bunch of independant attributes -
HAS-A reigns and sub-IS-A is rare.
(which is how it should be in biz modeling IME.)

> ....
> a general discussion of collection
> classes, IMHO, gets us closer to domains such as algebra, than
> to real-world domain modeling, anyway).

Algebra? Programming paradigms are mostly for human
communiction, not machine-to-machine. If many humans
don't model the world via algebra, then I don't see
what that gets one.


> It's unclear whether, here, you're talking about an object
> (keeping its _identity_) dynamically changing its _class_ 

No! There is no class to change to. There would be *one*
and only one collection protocol. Not all engines would
implement every feature of the protocol, but the idea
is that you don't pick a protocol that STICKS you with
only say stack-centric protocols or commands.

I don't know about your domain, but I observe collections
changing and growing in my domain. Starting out with
just a dictionary (hash) can be a real pain if you later
need more columns or more indexes, for example.

Look at even a call stack in an interpreter. It would
be nice to "X-ray" it and view it with different
sorts and filters for complex debugging. This is what
I like about tables. Studying the "state" of all kinds
of angles is a sinch: you don't have to re-invent 
micro querie languages, write save-to-file utilities,
etc., because they are usually built in. I like
plug-and-play power. Perhaps some bit-diddling personalities
like to re-invent little wheels, but I like to push simple
buttons and get powerful results.


> and I haven't read
> anything from you that would change my mind on that (surely
> not the lava-flow-prone approach in the URL you gave:-).

It seems because you are USED TO refactoring when your
collection needs change. I am spoiled. I know better.

Or, your domain may be different. 

> the indisputable success
> of Smalltalk does show that ....

I would hardly call it "indisputed". Smalltalk is sort of
the Frank Zappa of programming languages. It gets rave
reviews and it is intellectually stimulating. However,
it never quite catches on with general programmers.

I think because its power does not live up to it's haughtiness.

Using procedural and table techniques, I have matched
Smalltalk "can-you-do-this" challenges in the past,
and usually with more K.I.S.S.

Thanks for your feedback
-tmind-




From stephen_purcell at yahoo.com  Fri Feb  9 15:35:53 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Fri, 9 Feb 2001 15:35:53 +0100
Subject: Closing a file before or after return?
In-Reply-To: <960t2j$9n9$1@cubacola.tninet.se>; from gustafl@algonet.se on Fri, Feb 09, 2001 at 02:58:39PM +0100
References: <960t2j$9n9$1@cubacola.tninet.se>
Message-ID: <20010209153553.A18498@freedom.puma-ag.com>

Gustaf Liljegren wrote:
> Have a look at this function. It checks if a particular user (mail adress)
> exist in a .htpasswd file. The user name is what comes before ':' on each
> line.
> 
> # Check if user exists
> def user_exist(user):
>   f = open('.htpasswd', 'r')
>   for line in f.readlines():
>     i = string.find(line, ':')
>     if line[:i] == user:
>       return 1
> 
> I'd like to close the file too, but I don't know where to put that
> statement. Of course, the file should be closed in both cases -- not just
> when the user is found.

How about try/finally:

  def user_exist(user):
    f = open('.htpasswd', 'r')
    try:
      for line in f.readlines():
        i = string.find(line, ':')
        if line[:i] == user:
          return 1
    finally:
      f.close()

A neater way would just be:


  def user_exist(user):
    f = open('.htpasswd', 'r')
    found = 0
    for line in f.readlines():
      i = string.find(line, ':')
      if line[:i] == user:
        found = 1
        break
    return found


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From joconnor at cybermesa.com  Sat Feb 17 00:01:35 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Fri, 16 Feb 2001 16:01:35 -0700
Subject: string, split, sort, None, huh?
References: <96kadi$f2t@dispatch.concentric.net>
Message-ID: <3A8DB14E.5B4D39FB@cybermesa.com>


Phlip wrote:

> Thy Pon:
>
> Gonna split a document by linefeeds into strings, then sort the strings.
> Child's play, huh?
>
>         lines = split ( """Mary
> had
> a
> little
> lamb""", "\n" )
>
>         print repr(lines)
>         print repr(lines.sort())
>
> Now the output:
>
>         ['Mary', 'had', 'a', 'little', 'lamb']
>         None
>
> So where'd that 'None' come from?

It came from the fact that  lines.sort() sorts the sequence in in place and
returns None, which is what your second repr() is getting

Try instead:

    print repr (lines)
    lines.sort()
    print repr(lines)

Take care,
Jay O'Connor
joconnor at cybermesa.com



From jwbnews at scandaroon.com  Tue Feb 20 06:14:41 2001
From: jwbnews at scandaroon.com (John W. Baxter)
Date: Mon, 19 Feb 2001 21:14:41 -0800
Subject: Python STL ?
References: <mailman.982619303.32151.python-list@python.org> <3A919DBB.77FF09CB@andrewcooke.free-online.co.uk>
Message-ID: <jwbnews-C459C9.21144119022001@news.olympus.net>

All the joy of STL with none of the unending pain.  ;-)   --john


In article <3A919DBB.77FF09CB at andrewcooke.free-online.co.uk>, Andrew 
Cooke <andrew at andrewcooke.free-online.co.uk> wrote:

> There's no need for templates in a dynamically typed language and Python
> already has arrays, strings, hash functions and generic "sequence"
> operations.  There's discussion on iterators elsewhere in this group. 
> So what do you want?
> 
> Andrew
> 
> Nahuel Greco wrote:
> > 
> > There is something like the C++ STL for python (i dont mean CXX), or 
> > any project
> > aimed to this?
> > 
> > ----------------------------------------------------------
> > Nahuel Greco                 Web Development - Open Source
> > http://www.codelarvs.com.ar  Game Programming - Research
> > Freelance coding / sysadmin  Networking. The answer is 42.
> > ----------------------------------------------------------

-- 
John W. Baxter   Port Ludlow, WA USA  jwbnews at scandaroon.com


From dbrueck at edgix.com  Thu Feb  1 20:26:11 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Thu, 1 Feb 2001 12:26:11 -0700
Subject: overloading ==
In-Reply-To: <t7hbgnlmu55l42@corp.supernews.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHIECKPIAA.dbrueck@edgix.com>

Add a __cmp__ method to your class:

def __cmp__(self, obj):
  # return 0, a negative, or a positive number

(you can also define __rcmp__ to handle cases where the object on the left
side of the conversion doesn't have a cmp method)

Note that if you define __cmp__ but not a __hash__ then your objects can't
act as dictionary keys (which is correct if they are mutable).

-Dave

> [mailto:python-list-admin at python.org]On Behalf Of Amritansh Raghav
>
> can it be done?
> i looked at the operator documentation under Python Runtime Services - but
> didnt find anything to suggest it could be done.




From cfelling at iae.nl  Fri Feb 16 16:24:36 2001
From: cfelling at iae.nl (Carel Fellinger)
Date: 16 Feb 2001 16:24:36 +0100
Subject: Can methods and functions be overloaded?
References: <3A8C4CEA.B418B56E@mailzone.com> <PA%i6.14755$GA5.232956@e420r-atl1.usenetserver.com>
Message-ID: <96jgnk$33i$1@animus.fel.iae.nl>

Steve Holden <sholden at holdenweb.com> wrote:
...
> Does this help, or did you alredy know about it? I'm afraid Python doesn't
> allow the same kind of method overloading by signature that, say, Java and
> C++ do.

Right, but you can build your own:)

class Overload:
    def __init__(self):
        self.fun = {}

    def __call__(self, *args):
        try:
            fun = self.fun[self.signature(*args)]
        except KeyError:
            raise TypeError, "invalid overloading signature"
        return fun(*args)

    def signature(self, *args):
        '''smart up if you need to play with class hierarchies'''
        return tuple([type(x) for x in args])

    def overload(self, fun, *args):
        self.fun[self.signature(*args)] = fun


class Overloaded:
    def __init__(self):
        f = self.f = Overload()
        f.overload(self.f_None)
        f.overload(self.f_num, 1)
        f.overload(self.f_num, 1.0)
        f.overload(self.f_string, "1")
        f.overload(self.f_more, 'spam', 1)

    def f_None(self):
        return None

    def f_num(self, n):
        return n * 3

    def f_string(self, s):
        return 'string is %s' % s

    def f_more(self, s, n):
        return s * n

o = Overloaded()

print o.f()
print o.f(10)
print o.f(5.4)
print o.f("spam", 9)
print o.f(9, "spam")


this yields:

None
30
16.2
spamspamspamspamspamspamspamspamspam
Traceback (most recent call last):
  File "t.py", line 48, in ?
    print o.f(9, "spam")
  File "t.py", line 9, in __call__
    raise TypeError, "invalid overloading signature"
TypeError: invalid overloading signature

-- 
groetjes, carel


From fredrik at pythonware.com  Mon Feb 19 09:24:03 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 19 Feb 2001 08:24:03 GMT
Subject: Way to return a dictionary from re.findall?
References: <nqs09t040ckdut25j311ibi19i6dsouukg@4ax.com>
Message-ID: <DK4k6.18241$AH6.2510468@newsc.telia.net>

"Veronica Esquire" wrote:
> Is there a way to get back a dictionary of findall results that is
> based on the regex group names in python's re module?

no.  use a loop.

Cheers /F




From bulbul at ucla.edu  Mon Feb 19 01:45:48 2001
From: bulbul at ucla.edu (Leston Buell)
Date: Sun, 18 Feb 2001 16:45:48 -0800
Subject: can't open older anydbm file
Message-ID: <3A906CBC.1060009@ucla.edu>

Under Python 1.4 which was running on my web host's server, i created a 
database file using anydbm. Using whichdb.whichdb on it, i learn that it 
is a dbhash file, which i understand is what bsddb uses. Now my account 
has been moved to a new server, which has python 1.5.2. Now i can no 
longer open the file. When i try in an interactive session, i get this:

 >>> import anydbm
 >>> db = anydbm.open( 'address.db', 'r' )
Traceback (innermost last):
   File "<stdin>", line 1, in ?
   File "/usr/lib/python1.5/anydbm.py", line 86, in open
     return mod.open(file, flag, mode)
   File "/usr/lib/python1.5/dbhash.py", line 8, in open
     return bsddb.hashopen(file, flag, mode)
bsddb.error: (22, 'Invalid argument')

?Is there anything i can do to open this file or to recover my data?

Thanks,

Leston



From alberik at gmx.de  Wed Feb 14 11:08:19 2001
From: alberik at gmx.de (Thomas Off)
Date: Wed, 14 Feb 2001 11:08:19 +0100
Subject: PHP und remote files
Message-ID: <96dlf3$kms4b$1@ID-66991.news.dfncis.de>

Wie kann ich remote files, in denen wiederum PHP-Code steckt in meine
PHP-Files einbinden, so dass der Code aus dem remote file auch verarbeitet
wird?

MfG Thomas Off




From topmind at technologist.com  Mon Feb 26 02:32:51 2001
From: topmind at technologist.com (Topmind)
Date: Mon, 26 Feb 2001 01:32:51 GMT
Subject: Collection interfaces (Was: New to OO concepts - re-usability)
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com>
Message-ID: <MPG.1503548620e632919896a4@news.earthlink.net>

> "Topmind" <topmind at technologist.com> wrote
> in message news:MPG.150312ceb68f1ada98969f at news.earthlink.net...
> >
> > This is a sore point with me. To me it seems silly (artificial)
> > to split collection handling into categories (stacks, sets, trees, etc.).
> > If needs change from one to another or a combo, then you are hosed
> > with a non-fitting interface.
> 
> Yes, and so OO language provide an advantage here. Collection classes
> in these languages are accessible through common interfaces. You don't
> care how they are implemented, you only care about the interface.
> 


You mean something like SQL or ODBC?

And, this crap that only OO can do 
implimentation-hiding interfaces is
pure propaganda.

(Note that I am talking about using
the interfaces, not making DB
engines or whatnot.)


> > Shape up!
> 
> Yeah!
> 
> --
> Patrick Logan
> mailto:patrickdlogan at home.com
> 

-tmind-


From n8gray at caltech.edu.is.my.email.address  Sat Feb 10 06:50:02 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Fri, 9 Feb 2001 21:50:02 -0800
Subject: I love Python
References: <mailman.981731063.856.python-list@python.org> <vg388tsp5g580mroauk6d2ulvj8vo5s8cg@4ax.com> <3A842931.8C46014C@iit.demokritos.gr>
Message-ID: <962l55$i9i@gap.cco.caltech.edu>

Vassilis Virvilis wrote:

> Daniel Klein wrote:
> > 
> > On Fri, 9 Feb 2001 09:02:21 -0600, Kemp Randy-W18971
> > <Randy.L.Kemp at motorola.com> wrote:
> > 
> > >I noticed this message was sent three times.  There is an ancient
> > >legend that if everyone in the world says "I love Python" three times,
> > >Perl will disappear.
> > 
> > What's Perl? Hey, it worked!
> 
> I don't want to offend anybody but this is the best off-topic thread I
> have ever read.
> .Bill

Can I just add that this is the first time I've ever been on television?

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From joconnor at cybermesa.com  Mon Feb 19 19:38:26 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Mon, 19 Feb 2001 18:38:26 +0000
Subject: Monetary package?
Message-ID: <3A916822.8E2B7573@cybermesa.com>

All,

	Are there any packages available that handle monetary units?

Take care,
-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython


From sheila at spamcop.net  Sun Feb 18 05:01:13 2001
From: sheila at spamcop.net (Sheila King)
Date: Sun, 18 Feb 2001 04:01:13 GMT
Subject: Linda, this is cool!
References: <33Z6UBPU36939.943275463@frog.nyarlatheotep.org> <96mv3102l3d@news2.newsguy.com> <21gu8tsoknv01eegt4jak2lo7180kijno5@4ax.com>
Message-ID: <r5iu8tsqccj65ote3hfgo6dvjphskn15l3@4ax.com>

On Sun, 18 Feb 2001 03:27:19 GMT, Mindfield
<mindfield at badgers.emulationnet.com> wrote in comp.lang.python in article
<21gu8tsoknv01eegt4jak2lo7180kijno5 at 4ax.com>:

:It's really quite a loathesome little business,
:and I've never figured out why people are still stupid enough to think
:that it works, but I guess they still abide by the old motto, "There's
:no such thing as bad publicity."

It costs them next to nothing, and they must end up selling some of their
product or generating some revenue from the spam. If they didn't, the spamming
would decrease, but instead it is on the increase. Some people are stupid
enough to buy from spammers.

:Ah, but there is if you lose your ISP for violating their TOS/AUP.
:And there are plenty of people -- me, for instance -- who are
:perfectly happy to send a nice note to their ISP's abuse department.

Me, too.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From joconnor at cybermesa.com  Tue Feb  6 16:44:06 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 06 Feb 2001 15:44:06 +0000
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com>
Message-ID: <3A801BC6.DF764D5E@cybermesa.com>

"David C. Ullrich" wrote:
> 
> On Tue, 06 Feb 2001 01:35:02 GMT, "Don Tuttle" <tuttledon at hotmail.com>
> wrote:
> 
> >
> >"Jay O'Connor" <joconnor at cybermesa.com> wrote in message
> >news:3A7DB154.87E04245 at cybermesa.com...
> >> I'm not familiar with the 'with' keyword in Delphi.
> >
> >Here's a vbscript example.
> >With MyLabel
> >   .Height = 2000
> >   .Width = 2000
> >   .Caption = "This is MyLabel"
> >End With
> >Personally, I'd love to see Python include the 'with' keyword! It improves
> >readability and should execute a bit faster as well.
> 
> Delphi's "with" is more or less the same except without the
> dots at the start of the field names. For a long time I've
> seen Delphi people say that with is bad, the reason
> being it leads to bugs
> 
> with AnObject do
>   begin
>     OneThing
>     AnotherThing
>   end
> 


Eeewwww....that's ugly :)

Smalltalk cascading really isn't the same.  "with" seems to be
eastablishing a sort of local scope f reference.  Smalltalk cascading
just says "send this message to the same object you sent the last
message."

One point that may have gotten missed.  Smalltalk doesn't use ";" as a
delimiter between statements, it uses "." as a delimiter and a ";" to
indicate a cascade.  So the code

	anObject
		messageOne;
		messageTwo.

It's syntactically obvious that messageTwo is sent to anObject; there
can be no other confusion.

> when AnObject does not have a OneThing method
> but there _is_ a OneThing in the surrounding scope.

Oh, yes...I could see that being a big issue.  Smalltalk doesn't have
that problem because a) the syntax is explicit that you are cascading a
message send and b)  Smalltalk doesn't have the concept of functions not
bound to objects so you don't run into a problem of "this method doesn't
apply in this scope, let me try an outer scope".

> For years I thought this was just stupid - you should
> know what sort of object you're dealing with. The
> Python slogans about how explicit is better than
> implicit just about have me convinced that
> people are right, with is bad.

I agree.  Explicit is better than implicit (I still have Perl
nightmares...), and the Pascal 'with' statement seems wrought with
potential disaster.  Fortunately, the simularity between 'with' and
Smalltalk cascading is only superficial, at best.

Take care,

-- 
Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discusssion Forum -
http://pub1.ezboard.com/fobjectorienteddevelopmentpython


From dan at eevolved.com  Fri Feb  9 00:28:18 2001
From: dan at eevolved.com (Dan Parisien)
Date: Thu, 08 Feb 2001 23:28:18 GMT
Subject: Python developers of the world unite ;)
References: <ClFg6.214447$f36.8613204@news20.bellglobal.com> <3dn1bw6an9.fsf@ute.cnri.reston.va.us>
Message-ID: <mYFg6.214568$f36.8619306@news20.bellglobal.com>

Andrew Kuchling wrote:

> Dan Parisien <dan at eevolved.com> writes:
> > Doesn't anyone feel a sort of CPAN for Python should exist? What about a
> 
> See the Catalog-SIG:  http://www.python.org/sigs/catalog-sig/
> 
> --amk

Thank you for pointing me in the right direction. I have just joined the 
mailing list.

Dan



From tjg at exceptionalminds.com  Mon Feb 26 21:59:03 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Mon, 26 Feb 2001 12:59:03 -0800
Subject: IP Math anyone?
Message-ID: <20010226125903.X29494@trufflehunter.avalongroup.net>

Hi all,

I checked Parnassus, but couldn't find anything, so I'll ask
here. Has anyone done anything with IP math (e.g.,
192.168.0.254 + 1 = 192.168.1.1).



-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.         <><       (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  41 days  1:10 hours ago<<



From peter at engcorp.com  Mon Feb  5 06:10:39 2001
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 05 Feb 2001 00:10:39 -0500
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D69DF.99352687@engcorp.com> <BNmf6.53327$Pm2.1068813@news20.bellglobal.com>
Message-ID: <3A7E35CF.65F1C91F@engcorp.com>

Dan Parisien wrote:
> 
> Peter Hansen wrote:
> 
> > Jay O'Connor wrote:
> > >
> > > Smalltalk-style cascade operations would be *very* cool
> > >
[snip]
> >
> > Why?  'Cause it saves a bit of typing?
> 
> YES!!!
> 
> That is not a good enough reason not to add a feature like this :)

Absolutely not.  Look at http://www.pauahtun.org/TYPython/PythonWay.html
for some concise reasons why (scroll to the bottom).

As Erik Max wrote in another response, clarity of code is more 
important than quantity of keystrokes.


From claird at starbase.neosoft.com  Mon Feb  5 15:06:23 2001
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Feb 2001 08:06:23 -0600
Subject: help with process monitoring using python
References: <95g62e$ne31@news.qualitynet.net> <mailman.981218524.27316.python-list@python.org>
Message-ID: <3ADDADEF1AF041A6.5188C6AE0188D1C8.788BE319FE924227@lp.airnews.net>

In article <mailman.981218524.27316.python-list at python.org>,
Sean Reifschneider  <jafo at tummy.com> wrote:
			.
			.
			.
>I'm actually working on the ultimate process restarter right now.  The
>code is done, but I don't have any docs for it yet.  It'll probably
>be under ftp.tummy.com:/pub/tummy/ called something like "trestarter"
>when it's finally released (hopefully in the next day or so).
			.
			.
			.
What's the relation between trestarter and init(8)?
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From kdahlhaus at yahoo.com  Sun Feb 18 17:17:48 2001
From: kdahlhaus at yahoo.com (Kevin Dahlhausen)
Date: Sun, 18 Feb 2001 16:17:48 GMT
Subject: vim and python
References: <mailman.982268898.28556.python-list@python.org>
Message-ID: <MASj6.2250$ws2.139240@ord-read.news.verio.net>

'Pyf' is attempting to source the file using the embedded Python
interpreter in Vim.  I don't know offhand if it contains different import
hooks or if it plays with the PYTHONPATH.   I think of it as more for
scripting Vim with Python that debugging Python code, although in theory
what you're doing should work fine.    The first thing to try is running
that python file from outside of Vim and see what happens.   

There was a short series of articles on Python / Vim here a few weeks
ago.  When google puts the archives back up, you might check for 'Vim and
Python Efficiently'


By the way, I commend your choice of editor and language!  :)



In article <mailman.982268898.28556.python-list at python.org>, "Timothy
Grant" <tjg at exceptionalminds.com> wrote:

> Since this is more python related than vim related I'll post it here. I
> have recently had contact with a vim missionary (came to my door one
> Saturday morning and started telling me that I needed to be saved from
> emacs). 
> 
> Since I had been feeling a bit frustrated with emacs, and since I do
> suffer from tendonitis in one of my wrists, I thought I'd give it a
> shot. Surprisingly, I quite like it, and it is certainly easier on my
> bad wrist.
> 
> I've been muddling through for a couple of days now, and discovered that
> there is quite nice builtin python support. I played with it and got it
> sort of working, but have run into what look like environmental
> problems.
> 
> I start vim from the directory where my source files are located, and
> attempt to do a :pyf %, and I started getting import errors. Not a
> problem, I simply added my source directory to the end of sys.path and
> tried again. Now the import works correctly, but I am getting more
> errors and I can't explain them.
> 
> Traceback (innermost last):
>   File "<string>", line 1, in ? File "wlautoconf.py", line 20, in ?
>     class acOptions(wlOptions):
> NameError: wlOptions Press RETURN or enter command to continue
> 
> wlOptions is a class that is imported from a file in my source
> directory.
> 
> So, does anyone have any tips on setting up a nice python/vim
> environment?
> 
> Thanks.
>


From SBrunning at trisystems.co.uk  Fri Feb  9 10:08:52 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 9 Feb 2001 09:08:52 -0000 
Subject: None assigment
Message-ID: <31575A892FF6D1118F5800600846864D5B17D6@intrepid>

> From:	Michael Hudson [SMTP:mwh21 at cam.ac.uk]
> >>> del __builtins__ 
> 
> This makes Python, erm, interesting, to use:
> 
> >>> import sys
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: __import__ not found
 
Eeek!
 
> but yet:
> 
> >>> None 
> >>> 
> 
> i.e. None still gets found.
 
Always something new to learn on c.l.py. *Totally* useless of course, but
fascinating nevertheless. ;-)

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk
There is nothing so absurd that it has not been said by philosophers. -
Cicero




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From linux at thewebsons.com  Sun Feb  4 08:48:08 2001
From: linux at thewebsons.com (Ben Ocean)
Date: Sat, 03 Feb 2001 23:48:08 -0800
Subject: HTTPD Writable
Message-ID: <5.0.2.1.0.20010203234754.00a4d0d0@thewebsons.com>

Hi;
I'm trying to install a program that requires I create an *httpd writable* 
file. How do I do this? What does it mean to have httpd write to a file? Do 
I need to assign httpd a passwd? Do I need to group it with root? Please 
advise.
BenO 




From amk at mira.erols.com  Sun Feb 18 16:36:31 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 18 Feb 2001 15:36:31 GMT
Subject: A suspected bug
References: <mailman.982467140.8836.python-list@python.org> <Z6Oj6.18070$AH6.2468100@newsc.telia.net>
Message-ID: <slrn98vrqv.i0.amk@mira.erols.com>

On Sun, 18 Feb 2001 11:12:57 GMT, Fredrik Lundh <fredrik at pythonware.com> wrote:
>except that it doesn't always work:

And as Andrew Dalke pointed out:
>>> L=[1j, 2j]
>>> L.sort()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: cannot compare complex numbers using <, <=, >, >=
>>>

--amk



From jepler at inetnebr.com  Wed Feb 14 15:17:22 2001
From: jepler at inetnebr.com (Jeff Epler)
Date: Wed, 14 Feb 2001 14:17:22 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org> <96d492$88c@dispatch.concentric.net>
Message-ID: <slrn98l4rh.2s8.jepler@potty.housenet>

On 14 Feb 2001 05:15:14 GMT, Phlip
 <phlip_cpp at my-deja.com> wrote:
>Came to the thread late can't read all posts don't know if this was covered 
>so sue me but...
>
>...what's wrong with this?
>
>        for z in f.readlines():
>                print z
>
>That's how I done it since I was a chile'. 

If the contents of f are larger than memory, this won't work.  If the
contents of f are large compared to memory, then it won't work very well.

This is one reason that 2.1 will include "xreadlines", which will work
just like the above:
        for z in f.xreadlines():
                print z
but will never hold the entire file in memory.

Jeff


From sholden at holdenweb.com  Wed Feb 28 00:11:49 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 27 Feb 2001 18:11:49 -0500
Subject: Stream interfaces
References: <3a90fa97.0@news1.mweb.co.za> <3A933830.7F98D118@ix.netcom.com> <CgJk6.515$v77.66568@newsread2.prod.itd.earthlink.net> <220220010813598042%cmh@bDistributed.com> <ud7cakvi8.fsf@roundpoint.com> <MPG.150312ceb68f1ada98969f@news.earthlink.net> <8Gem6.473353$U46.14105062@news1.sttls1.wa.home.com> <MPG.1503548620e632919896a4@news.earthlink.net> <MCjm6.474652$U46.14163288@news1.sttls1.wa.home.com> <MPG.150394195f7f04a19896a7@news.earthlink.net> <3A9A86B1.9A8C69A2@ix.netcom.com> <MPG.1504d14563dce6499896af@news.earthlink.net> <3A9BC3F9.1929EB9A@ix.netcom.com> <MPG.1505ab2efd31b1639896b4@news.earthlink.net>
Message-ID: <JyWm6.834$zL.28252@e420r-atl1.usenetserver.com>

"Topmind" <topmind at technologist.com> wrote in message
news:MPG.1505ab2efd31b1639896b4 at news.earthlink.net...
>
[ ... ]
> I want to flatten GUI's and Collections in a similar way.
>
> Flatten flatten flatten!
>
> -flatmind-
>
> (But I am not flat-footed. I have a very prominent arch.)
>
Troll, troll, troll (but I am not a troll).

regards
 Steve





From cg at gaia.intranet.cdegroot.com  Sun Feb 18 19:53:18 2001
From: cg at gaia.intranet.cdegroot.com (Cees de Groot)
Date: 18 Feb 2001 19:53:18 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <96p5mu$3eq$1@gaia.intranet.cdegroot.com>

Jim Eaton  <eatonalive at qwestinternet.net> said:
>I am learning Python as a first language, and I have been wondering what
>is a good language to learn after Python?  I'm thinking of either going
>into C or Java but I'm not sure which one because I've heard many
>arguments either way.  Will going into Java first be any detriment to
>learning C later?  Any suggestions would be greatly appreciated.
>
You should learn at least a new programming language per year to keep
your mind fresh, so it doesn't really matter. 

Python is nice for a general view of programming: it does OO, procedural
and functional programming reasonably well. I'd advise C as a second
language because it forces you to deal with the computer behind all 
that code, which tends to give you a good perspective for other
language. Smalltalk is probably the best general language around if
you think OO is cool - watch out, lot of people never want anything
else after learning it. I'd sniff at Haskell and/or Scheme as well,
but you probably want to do quite a bit of programming and reading
before these languages start to make sense, and even then... 

If you want a job, decide whether you are pro or contra M$ and sign
up for a VB or Java course depending on the outcome. Needless to say
that these languages won't teach you a lot of new stuff compared to
the ones I mentioned above. You probably shouldn't take them as second
language because they are likely to teach you a lot of bad habits.



-- 
Cees de Groot               http://www.cdegroot.com     <cg at cdegroot.com>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B


From MarkH at ActiveState.com  Fri Feb  2 00:21:39 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Thu, 01 Feb 2001 23:21:39 GMT
Subject: win32com: passing object to threads
References: <3A79690D.563A7248@ira.uka.de>
Message-ID: <3A79EE92.2000908@ActiveState.com>

Markus von Ehr wrote:


 > I know that I have to use the: pythoncom.CoInitialize() and
 > pythoncom.CoUninitialize()  in a way, but don't know
 > why.   Can anybody help me?

COM has different behaviour depending on the COM threading model.  This 
threading model is controlled by CoInitialize/CoInitializeEx(), hence 
you have to understand the different behaviours and make an appropriaate 
choice.  Just to scare you even more, if you pass objects between 
different "threading apartments", you have to use the scarily named 
pythoncom.CoMarshalInterThreadInterfaceInStream() to pass COM objects 
between threads.

The only reasonable documentation on this for Python is "Python 
Programming on Win32" (by me and Andy, published by OReilly.)  However, 
generic documentation on COM's threading models will also help.

Mark.



From phd at phd.pp.ru  Tue Feb 13 15:29:15 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 13 Feb 2001 17:29:15 +0300 (MSK)
Subject: CPAN functionality for python
In-Reply-To: <buai6.64783$m5.3620776@news1.rdc1.ga.home.com>
Message-ID: <Pine.LNX.4.30.0102131727520.2534-100000@fep132.fep.ru>

On Tue, 13 Feb 2001, Doug Hellmann wrote:
> The server is going to have to keep up with what modules are available,
> and all the meta data about those modules (version, dependencies,
> description, location, etc.).
>
> That information needs to be available on all mirror sites in order for
> the mirror sites to be useful.  It is not enough to mirror some web
> pages, the database needs to be mirrored to support the custom client
> search mechanism.

   This metadata can be put in static files. Thus, it can be easily
mirrored/reached by www/ftp/rsync/etc. In case of Zope, mirror site must
run Zope, so mirroring will be limited to probably 2-3 sites in the world.

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From fimafeng at yahoo.com  Wed Feb 21 05:21:42 2001
From: fimafeng at yahoo.com (Chris Nelson)
Date: Wed, 21 Feb 2001 04:21:42 GMT
Subject: 1st install on Linux (RH 7.0) I hit a snaggg !!!
References: <3A93280C.2080802@yahoo.com> <28Gk6.499$c7.3553@newsfeed.slurp.net>
Message-ID: <3A93422B.3040605@yahoo.com>

bowman wrote:

> Chris Nelson <fimafeng at yahoo.com> wrote in message
> 
> news:3A93280C.2080802 at yahoo.com...
> 
> 
>> "Tkinter-2.0-8.0-3 requires libtcl8.0.so"
> 
>> "Tkinter-2.0-8.0-3 requires libtk8.0.so"
> 
>> I show the following RPMs as beeing installed...
> 
>> tcl 8.3.1-46
> 
>> tcl 8.2.0-46
> 
>> tk  8.3.1-46
> 
> 
> 
> welcome to rpm hell. tcl8.0 is not tcl8.3. unless you have some
> 
> dependencies, you might remove the 8.3 rpms and install 8.0. I don't know if
> 
> the problems were resolved, but there were some issues with 8.3 that caused
> 
> people building packages that required tcl/tk to revert back to 8.0. rpm's
> 
> are fine things, but they imply your machine has all the bits and pieces
> 
> that the build machine had, which is not always the case.

I removed the 8.3 libraries (even though it said there were 
dependencies..installed the 8.0.5 series of tcl and tk....and TKinter 
installed all right. I was goint to then upgrade the 8.0's back to 8.3 
but I got the warning about those libraries mentioned above so I decided 
to wait.

BTW.. How do I start the IDLE in Linux???

Thanks,
Chris



From jafo at tummy.com  Thu Feb  8 07:44:55 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Wed, 7 Feb 2001 23:44:55 -0700
Subject: anydbm can't find type of dbm file
In-Reply-To: <95s05c$njd$1@nnrp1.deja.com>; from jon_warner@my-deja.com on Wed, Feb 07, 2001 at 05:20:57PM +0000
References: <95s05c$njd$1@nnrp1.deja.com>
Message-ID: <20010207234455.G505@tummy.com>

>  File "/usr/lib/python1.5/anydbm.py", line 83, in open
>    raise error, "db type could not be determined"
>anydbm.error: db type could not be determined

Please see http://www.tummy.com/radiuscontext/errata.html for a discussion
of this and possible ways to fix it.

Sean
-- 
 The politician was gone but unnoticed, like the period after the Dr on a
 Dr Pepper can.
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From cjc26 at nospam.cornell.edu  Thu Feb  1 16:00:24 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Thu, 01 Feb 2001 15:00:24 GMT
Subject: curly-brace-aphobic?
References: <3A74EBD6.3C87FD15@seebelow.org> <slrn97a9fu.7sf.clgonsal@keeshah.penguinpowered.com> <3A757FBE.E54C5ECD@seebelow.org> <mailman.980789494.4239.python-list@python.org> <slrn97c6bs.bjo.scarblac@pino.selwerd.nl> <mailman.980870616.378.python-list@python.org> <95718r018s8@drn.newsguy.com> <3A78E3BE.AED8CABF@cosc.canterbury.ac.nz>
Message-ID: <cSee6.1237$o91.133183@typhoon.nyroc.rr.com>

* Greg Ewing <greg at cosc.canterbury.ac.nz> menulis:
| Grant Griffin wrote:
| > 
| > However, the rule that I seem to want to apply is "All indexers are indexed
| > using the style of brace used to bind them."
| 
| But tuples and strings aren't constructed using any
| kind of brackets at all:
| 
|    t = 1,2,3
|    s = "parrot"
| 
| How do you propose to index those, using your rule?

Well, >obviously<, you would do the following:

>>> t 2
2
>>> s"4"
r

Of course, you have to be careful with tuples, because sometimes
parentheses are used when creating them:

>>> t2 = (4, 5, 6)
>>> t2(3)
6


-- 
Cliff Crawford               http://www.people.cornell.edu/pages/cjc26/
                             print "Just another Python hacker"


From gmunsey at adobe.com  Thu Feb  1 15:46:06 2001
From: gmunsey at adobe.com (Grant Munsey)
Date: Thu, 1 Feb 2001 06:46:06 -0800
Subject: Registering LOTS of C level functions
Message-ID: <OEee6.80706$Wq1.34166930@nnrp5-w.sbc.net>

I am covering several libraries ... each of which has a couple of thousand
calls (don't ask!).
I would really like some way to be able to register all of them to Python
with one actual
C function and an ID so that inside the C function I can generate the actual
call by using
a data structure that tells the actual functions address and what parameters
it requires.

As it stands I don't see any way to do this without an excessive amount of
trickery. Looks
to me like there is no way to specify an extra parameter to the low level
registration process
for each function ... There is the "self" parameter but that would require
me to allocate a Python
object for each procedure I wanted to cover.

The only thing I can think of at the moment is to generate a bunch of
re-direction procedures ....
Tiny procedures that just get called from Python, add the ID code ... and
then pass on to the
"real" command handler.

Anyone have any suggestions?




From mikael at isy.liu.se  Thu Feb 22 14:37:56 2001
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 22 Feb 2001 14:37:56 +0100 (MET)
Subject: Rounding Question
In-Reply-To: <972jev0huk@news2.newsguy.com>
Message-ID: <XFMail.010222143756.mikael@isy.liu.se>

I corrected Remco:
 >  I'm sorry, but this also behaves badly. It rounds 10 up to 20. To get
 >  Jacob's desired functionality, you could modify your idea to
 >  
 >    rounded = number-((number-1)%10)+9
 >  
 >  still assuming only integer input.


Alex Martelli wrote:
 >  ...which of course is totally equivalent to the earlier
 >  proposed
 >      rounded = (number+9) - ((number+9)%10)
 >  since addition is commutative and associative (you can
 >  move the +9 from the right of the expression) AND so
 >  is addition in modulo-arithmetic, so that:
 >  
 >  (X-1)%N==(X%N)+((-1)%N)==(X%N)+((N-1)%N)==(X+N-1)%N

Of course! I just couldn't stay silent when Remco stumbled again.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael               
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    22-Feb-01
Time:    14:35:06

         /"\
         \ /     ASCII Ribbon Campaign
          X      Against HTML Mail
         / \

This message was sent by XF-Mail.
-----------------------------------------------------------------------


From pj at sgi.com  Thu Feb  1 00:46:59 2001
From: pj at sgi.com (Paul Jackson)
Date: 31 Jan 2001 23:46:59 GMT
Subject: Perl-to-Python converter/translator?
References: <959jsd$njn$1@nnrp1.deja.com> <3A7879FC.EA879C9B@bigfoot.de>
Message-ID: <95a85j$p1ce$1@fido.engr.sgi.com>

Gerhard wrote:
|> Maybe the Perl/Python phrasebook at
|> http://starship.python.net/~da/jak/cookbook.html is helpful
|> to you.  Intra-document links don't work. Does anybody know a
|> better version of this document?

Try a google.com search for "perl python phrasebook" for pointers
to such Perl-Python translators, and to other comparisons of
Perl and Python.
-- 
                          I won't rest till it's the best ...
                          Manager, Linux System Software
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From dan at sol.control.com  Fri Feb  9 19:57:42 2001
From: dan at sol.control.com (Dan L. Pierson)
Date: 09 Feb 2001 13:57:42 -0500
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <m3ofwdp1u7.fsf@chinon.cnrs-orleans.fr>
Message-ID: <m34ry3lm55.fsf@daystar.control.com>

Konrad Hinsen <hinsen at cnrs-orleans.fr> writes:

> The fact that there are people who read neither the PEPs nor the
> summaries doesn't prove that the summaries aren't useful. I have
> always read them with much interest, since I don't have the time at
> the moment to follow the development in detail. And I would have
> shouted out loud if I had noticed a tendency that looks dangerous to
> me - it just hasn't happened.
> 
> So personally I hope that someone will take up the job of providing
> development summaries in the future. And I'd like to thank Andrew
> for having done so for many months already.

I agree.  Thank you, Andrew,  for all of the work you've put into
creating on the consistent high spots of comp.lang.python for the
last serveral months.

One of the reasons that I haven't followed up on the python-dev stuff
as much as I should is that I'm almost exclusively working with Zope
these days and it's still stuck in 1.5.2 land.  Looking through your
window at 2.0 and 2.1 one makes me drool a lot, but I haven't got any
experience with it, just opinions.

Thanks again,

Dan Pierson


From chris.gonnerman at usa.net  Thu Feb 22 06:02:45 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Wed, 21 Feb 2001 23:02:45 -0600
Subject: Confused about extending.
References: <3A943557.B9893C03@olen.to>
Message-ID: <018b01c09c8c$ba95ae20$a100000a@local>

----- Original Message -----
From: "Joonas Paalasmaa" <joonas at olen.to>
Subject: Confused about extending.


> I have made the following C-function and now I would like to
> use it via Python in that way, that it would
> manipulate Python lists smoothedValues and originalValues.
> But I am just a beginner with C and I got confused with extending HowTo.
> I would be glad if someone would explain how to manipulate the code to
> make it extendingable.
>
> Joonas.

I'm puzzled as to why you have implemented this function in C instead of
Python.  Is it for
speed?  Just how big ARE your lists?

> void SmoothedList(double *originalValues, double *smoothedValues,
>         int listSize, int filterSize)
> {
>         double tmpResult;
>         int i, j;
>         int iterations = listSize - filterSize + 1;
>
>         for(i = 0; i < iterations; i++)
>         {
>                 for(j = 0, tmpResult = 0.0; j < filterSize; j++)
>                         tmpResult += originalValues[i + j];
>                 smoothedValues[i] = tmpResult / (double)filterSize;
>         }
> }

def SmoothedList(originalValues, smoothedValues, listSize, filterSize)
    iterations  =  listSize - filterSize + 1;
    for i in range(0,iterations):
        tmpResult = 0.0
        for j in range(0, filterSize):
            tmpResult  =  tmpResult + originalValues[i + j]
            smoothedValues[i] = tmpResult / filterSize

------------------------------------
Have I missed something here?







From dan at eevolved.com  Wed Feb 14 03:48:42 2001
From: dan at eevolved.com (Dan Parisien)
Date: Wed, 14 Feb 2001 02:48:42 GMT
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com> <3A89E46C.A081293F@tampabay.rr.com>
Message-ID: <emmi6.243735$JT5.8705881@news20.bellglobal.com>

try doing

# grep "font" * 

inside your idle/ directory

Dan


From doughellmann at bigfoot.com  Tue Feb 13 14:17:59 2001
From: doughellmann at bigfoot.com (Doug Hellmann)
Date: Tue, 13 Feb 2001 13:17:59 GMT
Subject: CPAN functionality for python
References: <mailman.982069514.23859.python-list@python.org>
Message-ID: <buai6.64783$m5.3620776@news1.rdc1.ga.home.com>

In article <mailman.982069514.23859.python-list at python.org>, "Oleg
Broytmann" <phd at phd.pp.ru> wrote:

> On Tue, 13 Feb 2001, Doug Hellmann wrote:
>> >> Why use standard CGI? This is a perfect application for Zope, the
>> >> ZODB,
>> >
>> >    The big advatnage of CPAn is that it's distributed and has many
>> >    mirrors.
>> > Zope-based solution will be hard to mirror.
>> >
>>
>> I don't follow that at all.  To support mirrors, the data needs to be
>> replicated from one server to another. That problem exists independant
>> of any tool.
> 
>    What is "data" in this case?

The server is going to have to keep up with what modules are available,
and all the meta data about those modules (version, dependencies,
description, location, etc.).

That information needs to be available on all mirror sites in order for
the mirror sites to be useful.  It is not enough to mirror some web
pages, the database needs to be mirrored to support the custom client
search mechanism.

Doug


From phd at phd.russ.ru  Wed Feb 21 02:03:49 2001
From: phd at phd.russ.ru (Oleg Broytmann)
Date: 20 Feb 2001 19:03:49 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Feb 20)
Message-ID: <D0D42E6D25BC1A27.425AE269661E6091.0F4BD16934D5BD2A@lp.airnews.net>

    Michael Hudson continues Python-dev Summaries after AMK. Thanks, Mike!
        http://lwn.net/daily/pydev-20010215.php3
    
    Pete Shinners announces new version (0.9) of pygame
        http://groups.google.com/groups?q=ANN:+pygame-0.9&hl=en&lr=&safe=off&rnum=1&ic=1
    
    LAMP (Linux, Apache, MySql, and Python) in action
        http://linuxpr.com/releases/3326.html
    
    Suchandra S Thapa moves Siphon project to SourceForge
        http://groups.google.com/groups?q=Suchandra+S+Thapa+Re:+CPAN+functionality+for+python&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Mark Hammond announces new version of win32com that allows you to implement
    any COM interface decribed in a type library, either IDispatch based or not
        http://groups.google.com/groups?q=Mark+Hammond+COM+support+for+vtable+based+interfaces&hl=en&lr=&safe=off&rnum=1&ic=1
    
    ActiveState invites nominations for scripting community heroes
        http://www.activestate.com/Corporate/Communications/Releases/Awards

    Moop project (MUD, Object Oriented, Pythonic) is now hosted at SourceForge
        http://moop.sourceforge.net/
    
    /F, as usual, shows an excellent snippet of code. This time it is
    GlobDirectoryWalker for doing recursive globbing
        http://groups.google.com/groups?q=Fredrik+Lundh+Re:+recursing+glob&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Dustin Boswell explains how to compile https support into Python2.0
        http://groups.google.com/groups?q=Dustin+Boswell+how+to+compile+https+support+into+Python2.0&hl=en&lr=&safe=off&rnum=1&ic=1
    
    David Mertz: Book Roundup
        http://gnosis.cx/cgi/txt2html.cgi?source=http://gnosis.cx/publish/programming/charming_python_12.txt
    
    Fredrik Lundh points to TryType (freetype) support for PIL
        http://groups.google.com/groups?q=Fredrik+Lundh+PIL+Fonts&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Carel Fellinger show how to handle overloading in OO manner
        http://groups.google.com/groups?q=Carel+Fellinger+Re:+Can+methods+functions+overloaded&hl=en&lr=&safe=off&rnum=1&ic=1
    
    Steven D. Arnold announces updated version of Pure Python Encryption with
    fixed IDEA implementation
        http://groups.google.com/groups?q=Steven+D.+Arnold+ANN:+Pure+Python+Encryption+%5BUPDATE%5D&hl=en&lr=&safe=off&rnum=1&ic=1


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the traditional
      center of Pythonia
          http://www.python.org
      Notice especially the master FAQ
          http://www.python.org/doc/FAQ.html

      PythonWare complements the digest you're reading with the
      daily python url
           http://www.pythonware.com/daily  

      comp.lang.python.announce announces new Python software.  Be
      sure to scan this newly-revitalized newsgroup at least weekly.
	  http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

      Michael Hudson continues Andrew Kuchling's marvelous tradition
      of summarizing action on the python-dev mailing list twice a
      month.
	  http://starship.python.net/crew/mwh/summaries/
	  http://www.amk.ca/python/dev

      The Vaults of Parnassus ambitiously collect Python resources
          http://www.vex.net/~x/parnassus/   

      Much of Python's real work takes place on Special-Interest Group
      mailing lists
          http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Cetus does much of the same
          http://www.cetus-links.de/oo_python.html

      Python FAQTS
          http://python.faqts.com/

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
	  http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird <claird at NeoSoft.com>
Business:  http://www.Phaseit.net
Personal:  http://starbase.neosoft.com/~claird/home.html


From tim.one at home.com  Fri Feb  2 05:43:16 2001
From: tim.one at home.com (Tim Peters)
Date: Thu, 1 Feb 2001 23:43:16 -0500
Subject: . Python 2.1 function attributes
In-Reply-To: <95cc26$k0k$1@nntp6.u.washington.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCIELEIMAA.tim.one@home.com>

[Donn Cave, stderr's best buddy]
> ...
> Since stderr is line buffered, it's really much better to go
> there with diagnostic outputs, in fact that's what it's for and
> why it works the way it does.

Maybe on your OS this is helpful.  On mine output to stderr scrolls off the
shell window irretrievably, and can't be redirected from the command line.
Not much fun to see it interleaved in seemingly random order with stdout
either.

> Likewise, the one line in test_*.py should be about every
> instance of print in any such file.

Sorry, this one's wrong independent of OS.  Python's std regression tests
(test_*.py) work by capturing stdout and comparing it to canned "expected
output" files.  stderr *may* have a marginal role there, but only for
exceedingly rare "what the f**k?!  this isn't just wrong, it's an utterly
unexpected disaster!" cases.  But even those would be better directed to
stdout, because the stdout comparison mechanism would point them out if they
happen, and won't forget to look for them.  People do forget, and, as is,
when one of these stderr cases crops up, regrtest.py doesn't realize the
test failed either.

and-no-i-had-no-part-in-writing-regrtest.py<wink>-ly y'rs  - tim




From ssartor at bellatlantic.net  Fri Feb  9 14:06:36 2001
From: ssartor at bellatlantic.net (Steven Sartorius)
Date: Fri, 09 Feb 2001 13:06:36 GMT
Subject: How to write integers to a file?
Message-ID: <wXRg6.3723$bu4.792238@typhoon2.ba-dsg.net>

I'm trying to write a series of whitespace delimited integers to a file.
The 'write' method only accepts strings and using arrays produces binary
data when I open the file with less.  What I want is somehting like
printf for python.  Is there any easy way to do this?  (there must
be...this is python!)

Any help much appreciated...

Thanks,

Steve Sartorius


From aleaxit at yahoo.com  Thu Feb 22 10:19:34 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 22 Feb 2001 10:19:34 +0100
Subject: Comparison with Ruby ?
References: <9716n5$2lln$1@news.okay.net> <3A94489D.5B65ED19@gmx.net>
Message-ID: <972lk10k0n@news2.newsguy.com>

"Juergen Bocklage" <juergen.bocklage at gmx.net> wrote in message
news:3A94489D.5B65ED19 at gmx.net...
> Aehmmm, I'm just a beginner, but some remarks on the ruby-python
> comparision.
    [snip]
>      Python separates types and classes, while Ruby treats them the
> same. Python types are more limited (no
>      inheritance/subclassing; cannot add methods to existing types).
> JB> Don't know

A Ruby win here.


>      Ruby has a better (or "real") closure feature.
> JB> what does this mean

Nothing, actually.  Some people are SO keen on nested lexical
scopes, that they'll (mistakenly) argue that a closure is not
one unless it supports _nested_ lexical scopes.  Until 2.0,
Python's scopes don't nest, so, of course, they keep not
nesting in closures; at 2.1, they will (alas) nest, so, of
course, they'll keep nesting in closures.

>      Ruby converts small integers and long integers automatically.
> JB> Okay, nice for you

No way.  Explicit is better than implicit.  I'd rather NOT
have such 'automatic' (automagick) conversions anywhere!

>      Ruby does not have tuples.
> JB> Bad for Ruby. They are fast.

Disagree.  'Fast' should be handled by the compiler and runtime.
Tuples should be lists with an "immutable" setting (and there
should be similar "immutable" settings for dictionaries) but
otherwise indistinguishable.

>           def +(x)
>            self.to_i + x
>           end
>
> JB> I know this from C++, (Syntactical sugar or a good source for
> errors)

It's not more error-prone than using __add__ as the magic name
for overloading addition, which is why Python does.  Looks like
a total wash to me.

>      Ruby has a loop abstraction using block; e.g.
>
>           10.times do
>             ...
>           end
>
> JB> What's this?

A method 'times' on integer objects, that takes a code-block.


>      You can define your own arbitrary iterator.
> JB> Why?

To iterate arbitrarily.  Why else?

>      Ruby provides method combination using `super'.
> JB> this is a good feature, in python I think you have to use the name
> of parent class.

Disagree -- I _like_ it that the SPECIFIC parent-class you
mean has to be explicitly mentioned (multiple-inheritance,
remember).  I don't like the complex rules of Dylan to allow
magick implicitness in this, nor constraining inheritance to
single just to allow 'super' to work:-).


> Python is for me the language for beginners and designers.

Not just for them.  Lots of Hackers and eXtreme Programmers
are also pretty happy with it.


Alex





From aleaxit at yahoo.com  Fri Feb  9 12:58:13 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Fri, 9 Feb 2001 12:58:13 +0100
Subject: class scope questions
References: <3a83bea8$0$16388$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <960m0p0d3r@news2.newsguy.com>

"Ben de Luca" <c941520 at alinga.newcastle.edu.au> wrote in message
news:3a83bea8$0$16388$7f31c96c at news01.syd.optusnet.com.au...
> if i do this
>
> class a:
>          dog='woof'
>
>           def blah():
>                     something

...then you're doing it wrong, because a method such as
blah should always have at least one argument.  The first
argument of a method is invariably called 'self', by
near-universal Python convention; it is bound to the
object on which the method is being called.

(Another popular convention is to name classes with a
name starting with an uppercase letter).

> how do i call a

You call a this way:

x = a()

this instantiates an instance of class a, and binds
variable x to said instance.


> how do i get soemthing in blah to refernce dog? either to read it or write
> to it

The normal approach to _reading_ a class-attribute is to
reference it as an attribute of self:

class A:
    dog = 'woof'
    def blah(self):
        print self.dog

x = A()
x.blah()

this will print 'woof'.  Since the instance does not directly
have an attribute named dog, the class's dictionary is looked
up for it, so the class-attribute is found.

However, if you *assign* self.dog, then that will bind a
(possibly new) attribute called 'dog' *in the instance*, and
NOT affect the _class_ attribute similarly named.  If you
do want to ensure you're accessing the class attribute,
you can use A.dog for reading, A.dog='buh' for writing;
or, getattr(A,'dog') for reading, setattr(A,'dog','buh')
for writing.


Alex





From sandysj at asme.org  Wed Feb 14 17:48:20 2001
From: sandysj at asme.org (Jeff Sandys)
Date: Wed, 14 Feb 2001 16:48:20 GMT
Subject: math through programming
References: <c88j8ts3gvtt0itgcpc4gp5omcfvh2ntna@4ax.com>
Message-ID: <3A8AB6D4.3216FC3E@asme.org>

I am cross posting this to comp.lang.logo to expose the 
Logo community to python and "Computer Programming for 
Everyone" (CP4E).

My view of Logo is a mathematical playground.  I still 
believe that it is a better language for young students 
(I like working with 9 year old and above), and it 
embraces different learning styles with a cleaner and 
more forgiving syntax.

The advantage of python is object programming, which 
some Logo languages have, and a skill that can be used 
later.  While the learning processes developed with Logo 
translate to other programming languages and to thinking 
in general, I have never seen any commercial applications 
or job openings for Logo.

What python is missing is a course outline that is age 
appropriate and accommodates different learning styles. 
I am eager to try python on some middle school students 
in an after school programming club, but without a plan 
I will stick to tried and proven Logo.

Thanks,
Jeff Sandys

Kirby Urner wrote:
> 
> [
> 
> below is my most recent post to math-learn, clearly a
> response to something earlier -- but with enough standalone
> content to give the idea.
> 
> Earlier posts in the same thread include Python source
> code for computing Pascal's Triangle and comparing the
> results with balls randomly falling, pachinko-style,
> through a triangulated grid of pins (some of you may have
> seen the well-known science museum demo showing the bell
> curved result).
> 
> Thanks again to Tim Peters for cluing me re 'Concrete
> Mathematics' which I went out and grabbed some months ago
> (if edu-sig had a search feature, I'd go find where he
> first posted about it, as an easier intro than Knuth's
> "telegraphic" 'Art of Computer Programming' -- and indeed,
> that's how this text evolved, from notes around the
> mathematical preliminaries leading into Knuth's magnum
> opus).
> 
> I'm cross-posting this here as a snap-shot of my ongoing
> effort to combine the math learning process with a
> strong programming element, to create a hybrid more
> adapted to life in the 21st century.  The mass-education
> juggernaut has too much inertia to steer in this new
> direction any time soon, but there's always the hope
> that more alert passengers will jump overboard and swim
> for their lives :-D (good to have this alternative more
> seaworthy craft in the vicinity, given the more
> conventional curriculum simply isn't).
> 
> complete thread:
> http://www.mathforum.com/epigone/math-learn/snoipumkhee
> 
> Kirby
> 
> ]
> 
> Subject:      [math-learn] Re: Intro to Probability
> Author:       pdx4d at teleport.com
> Date:         Tue, 13 Feb 2001 16:33:03 -0000
> 
> > So the short way to write (n-k+1)! is just n!/(n-k)! as
>                             ^^^^^^^^^^^^^^^^^^^^^^^^^^
> 
> > that cancels all terms except n(n-1)(n-2)..(n-k+1).
> 
> Right, a typo, thanks to a close reader for catching that.  What I
> wanted to express with (n-k+1)! were the k terms in n(n-1)...(n-k+1),
> but of course these aren't equivalent. What I need is notation to
> expressing a "falling factorial" running for k terms, and, indeed,
> such notation is offered in 'Concrete Mathematics' by Graham Knuth
> and Patashnik (Addison-Wesley, 1994), pp 47-48.
> 
> The notation can't easily be written in ascii unfortunately, but
> "n to the m falling" is written [n to the m-underbar].  With "n to
> the m rising" = [n to the m_overbar].  "These functions are also
> called 'falling factorial powers' and 'rising factorial powers',
> since they are closely related to the factorial function n! =
> n(n-1)...(1).  In fact n! = [n to the n-underbar] = [1 to the n-
> overbar]."
> 
> So anyway, what I should have written was [n to the k-underbar].
> 
> This book 'Concrete Mathematics' I've mentioned before on math-teach,
> as exemplary of what I'm aiming for at the college level with my
> K-12 curriculum writing.  It's used for computer science students
> a lot, but it's also just a fine introduction to a lot of concrete
> (not too abstract) mathematics, including discrete probability,
> binomial coefficients and elementary number theory.
> 
> I'm also trying to boost geometry content, by moving 'Beyond
> Flatland' (off the plane) -- which links to a 'Beyond Calculators'
> push as well (as I haven't seen many rotating polyhedra on
> calculator screens -- though maybe these have appeared in Japan
> by this time, a market typically ahead of the USA's in the
> calculator department).
> 
> As I wrote to a math educator recently by email:
> 
>    You may remember from emails past that I'm one of those
>    looking for ways to incorporate more of Bucky Fuller's
>    contributions into early math ed.  I also push a "math
>    through programming" approach, and converge it with the
>    Fuller-informed spatial geometry.  "Beyond Flatland" and
>    "Beyond Calculators" summarize my two pronged strategy
>    to revamp math ed pedagogy.
> 
>    I've just put up a new essay, well-illustrated, which
>    looks at the "Beyond Flatland" approach:
>    http://www.inetarena.com/~pdx4d/ocn/overview.html
> 
>    And here's a review of my "Beyond Calculators" initiative:
>    http://www.oreillynet.com/pub/a/python/2000/10/04/pythonnews.html
> 
> Kirby


From aleaxit at yahoo.com  Mon Feb 19 13:26:31 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Mon, 19 Feb 2001 13:26:31 +0100
Subject: Newbie's Questions
References: <mailman.982581805.9060.python-list@python.org>
Message-ID: <96r3dn01u60@news1.newsguy.com>

"Thomas Guettler" <guettli at interface-business.de> wrote in message
news:mailman.982581805.9060.python-list at python.org...
> Yesterday I had time to start learning python. I like it, after a half

Bravo!-)

> hour I wasn't missing the "{..}" anymore. Here are some questions I
> have:
>
> Is there a way of checking the program when I start it.  So typos are
> detected emidiately, and not when I enter the function.

The syntax of the whole script IS checked when you start it,
since it's internally compiled then; this will catch many
typos -- e.g., if you write:
    fi x>0:
rather than
    if x>0:
a SyntaxError will be generated immediately -- but not, of
course, all, e.g., if you write:
    x=y=z
rather than
    x=y+z
(also a frequent occurrence, since + and = are shifted
version of each other on many keyboards), this will not be
detected until you run suitable tests, as both forms happen
to be syntactically correct.

There is no way to even check for typos _that will
generate exceptions when the function is run_, without
actually running the function; e.g.,
    a=b/c
where one meant to write
    a=b/d
and c happens to be 0, will be caught when the function
is run (the division-by-zero attempt will raise an
exception), but not before then.  And similarly if c
happens to refer to a string (you can't divide by a
string), or if it happens to be unbound at that point;
only actual execution, not compilation, will reveal it.


> Is there something like "use strict" in perl? I like to declare my
> variables before I use them. (Again, to recognize typos)

There are no declarations in Python (the 'global' statement
_might_ be thought of as an exception to this assertion...).
If you're using a variable that's not bound at all, at run
time, or that is bound to a string where a number is needed
at that point, etc, etc, you'll get runtime exceptions; no
way to have the compiler diagnose it beforehand.


> Is there a way of reading a single character from stdin without
> waiting for the return-key? (At best without curses)

On Windows, only, the msvcrt module offers functions getch
(to read a single keypress, without waiting for a return, and
without echoing) and getche (ditto, with echoing); but they
don't work on standard-input (i.e., they are immune to stdin
redirections), but, rather, on the actual keyboard.

On Unix, you can use the low-level module termios, or the
higher-level one tty that in turn uses termios, to set the
mode to 'raw' for file-descriptor 0 (standard input).


> Is there something like "exit" in C? Yes, I could just raise an
> exception, but I prefere a way without a stacktrace.

Module sys contains function exit, which in turn raises a
SystemExit exception.  Either calling sys.exit(), or raising
SystemExit directly, will NOT generate a stacktrace -- but
the exception can be caught, so that try/finally and
try/except semantics are preserved.  If you need a way
to exit in an emergency, *without* proper/normal actions
on termination (no flushing of buffers, no finally blocks,
etc, etc), you can do that with function _exit of module
os, but that's a pretty-rare need ("should normally only
be used in the child process after a fork()" is what the
docs say, but os.fork is only available on Unix platforms,
while os._exit is also available on Windows).


Alex





From johann at physics.berkeley.edu  Sun Feb 18 21:03:48 2001
From: johann at physics.berkeley.edu (Johann Hibschman)
Date: 18 Feb 2001 12:03:48 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net>
Message-ID: <mtofvz69nf.fsf@astron.berkeley.edu>

Jim Eaton writes:

> Thank you very much for your advice.  I was also wondering if learning Perl
> sometime later would be needed.  I've heard people talk of Python as the
> natural successor to Perl and I'd like to know what people feel about this.

Heh.  I didn't even get a chance to get started.

Clearly, this depends on what you want to do with your new-found
programming power.  If you're interested in web-related work, then
Perl would be a good thing to know, simply because it's very common.
I don't do web-stuff, I do numerics, so I've never been able to make
myself learn it.  (Well, I know enough to use it as a "better awk",
but that's about it.)

If you want to become a better general programmer, Perl will not help
you at all, IMHO.  I don't really think that Java would help, either,
since I'm not sure what it would teach you.

C or C++ would at least help you understand how to manage your own
memory and perhaps give you a better sense of what goes on in the guts
of the machine.  Out of those, I'd recommend learning the basics of C
first, because I find C++ very confusing.  Now, that being said, my
main production code is a Python/SWIG/C++ system, so I like some parts
of C++, but it's an awful mess to learn.

This may be heresy in a c.l.python, but I really wish I'd written the
whole thing in Common Lisp to start with.  Too late now, but my next
project will be entirely CL.  I'm tired of the wall between the
scripting language and the speed language; CL can be both (I get
numeric performance as good as Fortran in Common Lisp using the free
CMUCL compiler), so I'd just use that.

(Python is great for things that I think I will never need that
to-the-metal performance in.  If I can stay in pure Python, I use
that.  I'll just never try to maintain a dual Python/C++ project
again.  Arg.)

One warning about this lisps; you might want to be careful with them.
I never like advocating them.  I absolutely *hated* Scheme when I
first had to deal with it in an AI class, but I've grown fonder of it
with time.  I had to do far too many things in a completely backwards
way; if you're just trying to learn new ways to do things, it's great,
but if you're trying to implement alpha-beta pruning in your checkers
game for the assignment due the next day, you'll come to hate it.
Common Lisp is a lot more practical.

Other languages.  Hm.  I've heard prolog is cool, but I've done all of
my constraint-based programming in lisp.  I looked at Haskel, but it
seemed over-pure.  OCaML is nice, but I never really got into it.
Sather is a great language, but it's dying.  It's like Eiffel, in that
it teaches good OO design, but it has really cool Icon-like iterators
(I would so love it if python got those some how), and it's
contravariant rather than covariant, so it doesn't have to do nearly
as many type-checks and is thereby much faster.

That's about it for my experience.  Sorry for the brain dump.  I just
started out by saying that learning Perl won't teach you anything that
you can't get from Python, so don't bother with it, unless you just
want to make yourself more marketable.  And I stand by that statement.

--Johann

-- 
Johann Hibschman                           johann at physics.berkeley.edu


From boud at rempt.xs4all.nl  Wed Feb 21 17:57:13 2001
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 21 Feb 2001 16:57:13 GMT
Subject: custom gui?
References: <xw2k6.1848$zw6.48239@wagner.videotron.net> <bLik6.3517$zw6.74776@wagner.videotron.net> <1F1B8A7CFF73AFD4.03A50AAFC4AC473B.6EA500FCB7E4FD46@lp.airnews.net> <96vmdq$f5$1@news1.xs4all.nl> <33AD0B12D8AC8640.93E4CFDD593CF2E8.A6B598462FA61951@lp.airnews.net>
Message-ID: <970s19$3bo$1@news1.xs4all.nl>

Cameron Laird <claird at starbase.neosoft.com> wrote:

> Yes.  What facilities are and will be available for
> those who want windows with funny geometry?  The im-
> mediate answer, of course, is that motivated people
> can synthesize windows as canvas polygons.  I bet
> that satisfies few, though.


I doubt that there are cross-platform facilities available; but
perhaps the shape extensions for X could be wrapped? Myself, I really
loathe windows without decent borders.

-- 

Boudewijn Rempt  | http://www.valdyas.org


From stevena at neosynapse.net  Sun Feb 18 07:17:29 2001
From: stevena at neosynapse.net (Steven D. Arnold)
Date: Sun, 18 Feb 2001 01:17:29 -0500
Subject: ANN: Pure Python Encryption [UPDATE]
Message-ID: <5.0.2.1.2.20010218011156.03cacbe8@phear.dementian.com>

We've posted an update to the Neosynapse Encryption Tools' IDEA 
implementation.  There was a bug that caused incorrect encryption and 
decryption, even though the test passed.  If you downloaded the previous 
version, you should get the new copy.  Apologies for the inconvenience; 
enjoy!  The software may be downloaded from:

         http://www.neosynapse.net/free_encrypt.htm


--
Steven D. Arnold                   President, Neosynapse Consulting Group
stevena at neosynapse.net                          http://www.neosynapse.net




From aahz at panix.com  Mon Feb 12 23:21:03 2001
From: aahz at panix.com (Aahz Maruch)
Date: 12 Feb 2001 14:21:03 -0800
Subject: python-dev summary, Jan. 16-31
References: <3d4ry56vzg.fsf@ute.cnri.reston.va.us> <95v3mt$lnh$1@nntp9.atl.mindspring.net> <95v5t3$d3q$1@panix3.panix.com> <962j8r$i9i@gap.cco.caltech.edu>
Message-ID: <969nkf$5nd$1@panix3.panix.com>

In article <962j8r$i9i at gap.cco.caltech.edu>,
Nathaniel Gray  <n8gray at caltech.edu.is.my.email.address> wrote:
>
>A moderated newsgroup seems like a much better vehicle if you intend for 
>the public to participate in the discussion of Python development without 
>sending the SNR plummetting.  I imagine most people would hesitate to post 
>to a moderated newsgroup without having given their message a modicum of 
>thought.  (Unlike *most* newsgroups ;^)

Actually, I was thinking that this would be a one-directional gateway
from mailing list to newsgroup, no posting allowed.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

'Gender' isn't a definition, it's a heuristic.  --Aahz


From othello at javanet.com  Tue Feb 27 18:22:28 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Tue, 27 Feb 2001 12:22:28 -0500
Subject: Breaking out of nested loop
References: <3a9b44f1.226505016@news.okc1.ok.home.com> <97gj16$d5d@dispatch.concentric.net> <97gkqh$lsp$1@news.mathworks.com>
Message-ID: <3A9BE254.459265C4@javanet.com>

jcm wrote:

> Phlip <phlip_cpp at my-deja.com> wrote:
> > Proclaimed Dan from the mountaintops:
>
> >>
> >> How might one go about breaking out of a multi level loop?
> >>
> >> For instance, the following code will go into an endless loop.
> >>
> >> while(1):
> >>     while(1):
> >>         break
>
> ...
>
> > Your best bet (just like in a language that has a 'goto') is 'return'.
>
> > Otherwiset you can 'raise' (irritating style that reminds one of VB), or
> > you can add a flag to the 'while' conditions and reset it at the 'break'
> > point.
>
> Something I wouldn't mind seeing added to Python is a generalization
> of break/continue statement[s], allowing you to break out of multiple
> loops:
>
>   break_or_continue_stmt: 'break'
>                         | 'continue'
>                         | 'break' break_or_continue_stmt
>
> So you can do something like:
>
>   while a:  # loop A
>      ...
>      while b:  # loop B
>         ...
>         while c:  # loop C
>            ...
>            break break  # breaks loops C and B, bringing
>                         # you back to the body of loop A

Adding a feature like is begging for hard to find errors.
A factored out function with a return almost always does
the trick and does it clearly.  And if not, then the needed
functionality is provided by try, except, and raise.

Raymond



From emile at fenx.com  Tue Feb 20 15:20:17 2001
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 20 Feb 2001 06:20:17 -0800
Subject: Where is CASE?????
References: <BD674CDC0A68C642B60038986FBB56DF013870@mailsrv.ASPATECH.COM.BR> <mailman.982676904.20104.python-list@python.org>
Message-ID: <96tuh8$mocjg$1@ID-11957.news.dfncis.de>

Or you may consider wrapping it in a dictionary:

Type "copyright", "credits" or "license" for more
information.
IDLE 0.6 -- press F1 for help
>>> def caseA(*args):
 print "CaseA ", args


>>> def caseB(*args):
 print "CaseB ", args


>>> def caseOther(*args):
 print "CaseOther ", args


>>> switch = {'A':caseA, 'B':caseB}
>>> switch.get('A',caseOther)
<function caseA at 0103153C>
>>> switch.get('A',caseOther)(1,2,3)
CaseA  (1, 2, 3)
>>> switch.get('B',caseOther)(1,2,3)
CaseB  (1, 2, 3)
>>> switch.get('C',caseOther)(1,2,3)
CaseOther  (1, 2, 3)
>>>

--

Emile van Sebille
emile at fenx.com
-------------------


"Steve Purcell" <stephen_purcell at yahoo.com> wrote in message
news:mailman.982676904.20104.python-list at python.org...
> Chris Richard Adams wrote:
> > Someone please tell me why I cannot find a CASE
statement in the Python
> > documentation index...
>
> Because there isn't one.
>
> Use if/elif/else instead.
>
> It doesn't sting as much as you might expect.
>
> -Steve
>
> --
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Get servlets at http://pyserv.sourceforge.net/
> "Even snakes are afraid of snakes." -- Steven Wright
>




From bpederson at geocities.com  Tue Feb 13 04:50:43 2001
From: bpederson at geocities.com (Barry Pederson)
Date: Mon, 12 Feb 2001 21:50:43 -0600
Subject: QuakeWorld Python 1.0
References: <3A8844D7.1CE2CC58@geocities.com> <mailman.982030514.9176.python-list@python.org>
Message-ID: <3A88AF13.7B0B5275@geocities.com>


Chris Gonnerman wrote:
> 
> ----- Original Message -----
> From: "Barry Pederson" <bpederson at geocities.com>
> Subject: QuakeWorld Python 1.0
> > Version 1.0 includes sourcecode which builds and runs on both Windows
> > and Unix platforms, and Win32 binaries.
> 
> Where ARE the Win32 binaries?  I downloaded the Win32 package but got no
> binaries
> (no .exe or .dll files).

Under Windows, Python uses .pyd instead of .dll for native-code
extensions.  

In the QWPython Win32 distribution, qwpython\qwsv.pyd is what you're
looking for.  To actually run a game, you run the ctf.py or
deathmatch.py scripts, which take care of loading the native-code
module, along with the pythonized game code (which should be covered in
the README.html file)

	Barry


From craigf at ilid.com.au  Fri Feb  9 05:01:54 2001
From: craigf at ilid.com.au (Craig Findlay)
Date: Fri, 09 Feb 2001 15:01:54 +1100
Subject: How to return from shutdown command
Message-ID: <rjq68tkm4ksalsr7oohnffbo5am6jpgh8m@4ax.com>

I want to execute a Unix shutdown command from a python script, but
have control return to the script, and I'm having trouble doing it.

I've tried:

commands.getstatusoutput('shutdown -h +1')
and
commands.getstatusoutput('/bin/sh -c "shutdown -h +1" \&')

but neither of these commands return.

Any help greatly appreciated!

Craig



From jkraska1 at san.rr.com  Mon Feb 26 07:48:08 2001
From: jkraska1 at san.rr.com (Courageous)
Date: Mon, 26 Feb 2001 06:48:08 GMT
Subject: abusing __builtins__
References: <mailman.983142547.10845.python-list@python.org> <3A99EEAB.C756064C@iit.demokritos.gr>
Message-ID: <bbvj9ts78mki9fcb5pa3ale4uij64bk3ih@4ax.com>

>> The language doesn't guarantee this will always work, so it's certainly
>> abuse.
>
>Oups! That's the killer argument here.

Well, not to worry. It works now, as you've noticed. :)

C//



From tony at lsl.co.uk  Thu Feb 22 10:02:51 2001
From: tony at lsl.co.uk (Tony J Ibbs (Tibs))
Date: Thu, 22 Feb 2001 09:02:51 -0000
Subject: [Python-Dev] RE: Update to PEP 232
In-Reply-To: <200102220012.TAA12047@cj20424-a.reston1.va.home.com>
Message-ID: <001b01c09cae$3c3fa360$f05aa8c0@lslp7o.int.lsl.co.uk>

Guido responded to my points thus:
> > 1. Clarify the final statement - I seem to have the
> > impression (sorry, can't find a message to back it up)
> > that either the BDFL or Tim Peters is very against
> > anything other than the "simple" #f.a = 1# sort of
> > thing - unless I'm mischannelling (?) again.
>
> Agreed.

That's a relief - I obviously had "heard" right!

> > 2. Reference the thread/idea a little while back that ended
> > with #def > f(a,b) having (publish=1)# ...
>
> Sure, reference it.  It will never be added while I'm in charge
> though.

Well, I'd kind of assumed that, given my "memory" of the first point.
But of the schemes that won't be adopted, that's the one *I* preferred.

(my own sense of "locality" means that I would prefer to be placing
function attributes near the declaration of the function, especially
given my penchant for long docstrings which move the end of the function
off-screen. But then I haven't *used* them yet, and I assume this sort
of point has been taken into account. And anyway I definitely prefer
your sense of language design to mine).

Keep on trying not to get run over by buses, and thanks again for the
neat language,

Tibs

--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.co.uk/
"Bounce with the bunny. Strut with the duck.
 Spin with the chickens now - CLUCK CLUCK CLUCK!"
BARNYARD DANCE! by Sandra Boynton
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From johngrayson at home.com  Wed Feb  7 13:57:23 2001
From: johngrayson at home.com (John Grayson)
Date: Wed, 07 Feb 2001 12:57:23 GMT
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net>
Message-ID: <95rgng$8ul$1@nnrp1.deja.com>

In article <95q8tj$1c8$1 at news.nuri.net>,
  "June Kim" <junaftnoon at nospamplzyahoo.com> wrote:
>
>
>

I received a package from my publisher at Manning last night:
Inside was the Korean translation of Python and Tkinter Programming!
It is published by Infobook ISBN 89-8054-424-3.

This is great news for Python -- my only fear is that the frequent
Monty Python quips in the book translate to something offensive
in Korean...

    John Grayson


Sent via Deja.com
http://www.deja.com/


From nobody at home.net  Wed Feb 28 23:34:04 2001
From: nobody at home.net (Bernard Delmée)
Date: Wed, 28 Feb 2001 23:34:04 +0100
Subject: Read INI-File
References: <97gcng$oquj3$1@ID-59219.news.dfncis.de> <97gd97$931$1@rex.ip-plus.net> <97ggqk$p53ii$1@ID-59219.news.dfncis.de> <983288899.644740@newsmaster-04.atnet.at> <97ia2t$p9da2$1@ID-59219.news.dfncis.de>
Message-ID: <97jub8$1th5$1@rivage.news.be.easynet.net>

Your example works fine for me, on W2K, (prints 'timer.drv')
ActivePython 2.0, build 202

I tried both SetProfileFileName('C:\\WINNT\\system.ini')
and SetProfileFileName('C:/WINNT/system.ini')


"Marc Vollmer" <marc.vollmer at mahr.de> wrote in message
news:97ia2t$p9da2$1 at ID-59219.news.dfncis.de...
> Thanks,
>
> but no solution, perhaps my computer don't understand me.
>
> So, I test it with the system.ini in the c:\winnt\.
>
> Please test it, too. Perhaps it is a windows 2k problem? I don't know,
what
> I still could do.
>
> --
> import win32ui
> win32ui.SetProfileFileName('C:\\WINNT\\system.ini')
> k=win32ui.GetProfileVal('drivers','timer','xxx')
> print k
> --
>
> My result is: xxx and yours?
>
> Regards
> Marc
>
>
> "Werner Schiendl" <ws-news at gmx.at> schrieb im Newsbeitrag
> news:983288899.644740 at newsmaster-04.atnet.at...
> >
> > Marc Vollmer <marc.vollmer at mahr.de> wrote in message
> > news:97ggqk$p53ii$1 at ID-59219.news.dfncis.de...
> > > No chance:
> > >
> > > I try:
> > >
> > > win32ui.SetProfileFileName('c:\winnt\test.ini')
> >
> > If you _really_ wrote it this way it will not work.
> > The backslashes are used to escape the values.
> >
> > Use    'c:\\winnt\\test.ini'
> > or    r'c:\winnt\test.ini'
> >
> > hth
> > werner
> >
> >
>
>




From greg at cosc.canterbury.ac.nz  Thu Feb  1 05:26:46 2001
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Thu, 01 Feb 2001 17:26:46 +1300
Subject: "in" for dicts (was: Python 2.1 function attributes)
References: <mailman.980770173.14038.python-list@python.org>
Message-ID: <3A78E586.4638E05B@cosc.canterbury.ac.nz>

Tim Peters wrote:
> 
> Nobody searches phone books or Webster's
> "backwards" in real life.

Indeed! If I were to ask my colleague "Is the word 'bamboozle'
used anywhere in a definition in Webster's dictionary?" the
answer would probably be "Buggered if I know."

Which leads us to what the semantics of
"x in dict" should obviously be:

>>> if x in dict:
...   print "boo:
<stdin>, line 1: BuggeredIfIKnowError

-- 
Greg Ewing, Computer Science Dept, University of Canterbury,	  
Christchurch, New Zealand
To get my email address, please visit my web page:	  
http://www.cosc.canterbury.ac.nz/~greg


From dsavitsk at e-coli.net  Fri Feb 23 00:32:34 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Thu, 22 Feb 2001 17:32:34 -0600
Subject: httplib and buffered data
Message-ID: <mkhl6.124395$Ch.22717962@newsrump.sjc.telocity.net>

i am capturing the data being returned from a script running on a web
server.  The script takes several minutes to run, and trickles back data a
little at a time.
Using httplib (or anything else) is there a way to capture the data line by
line instead of as a lump sum with getfile()? i would like to alert the user
as to the % of the data (a known quantity) that has been processed.

thanks,

doug




From nas at arctrix.com  Fri Feb  9 19:29:43 2001
From: nas at arctrix.com (Neil Schemenauer)
Date: Fri, 9 Feb 2001 10:29:43 -0800
Subject: open("ls -l |", "r") ??
In-Reply-To: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>; from fm_duende@yahoo.com on Fri, Feb 09, 2001 at 05:56:39PM +0000
References: <rbWg6.114913$g6.51400033@news1.elmhst1.il.home.com>
Message-ID: <20010209102943.A20059@glacier.fnational.com>

On Fri, Feb 09, 2001 at 05:56:39PM +0000, steveFarris wrote:
> Hi all, is some variation of  the above possible in Python?

Look at os.popen().  You probably be using the directory methods
in os rather than forking two processes to get this data.

  Neil



From jsteve17 at tampabay.rr.com  Wed Feb 14 02:51:57 2001
From: jsteve17 at tampabay.rr.com (Jeff Stephens)
Date: Wed, 14 Feb 2001 01:51:57 GMT
Subject: Changing Fonts in IDLE-0.5
References: <3A89A78F.C4C6B70A@tampabay.rr.com> <%Oii6.238$Vf.9579@typhoon.aracnet.com>
Message-ID: <3A89E46C.A081293F@tampabay.rr.com>

Daniel Klein wrote:

> Jeff,
>
> You got me curious about this so after a minute of digging, I found two
> files, 'config-unix.txt' and 'config-win.txt' in the '/python20/tools/idle'
> (or C:\python20\tools\idle) directory. It is in these files (depending on
> your platform) where the font size can be changed.
>
> Hope this helps,
>
> Daniel Klein
> Portland OR USA
>

I'm running on Linux, specifically, Redhat 6.2.  I don't have either of these
files.  Here are the
files in my idle-0.5 directory:

AutoExpand.py        IdleHistory.pyc        Separator.py
AutoExpand.pyc       IdlePrefs.py           StackViewer.py
AutoIndent.py        IdlePrefs.pyc          TODO.txt
AutoIndent.pyc       MultiScrolledLists.py  ToolTip.py
Bindings.py          MultiStatusBar.py      TreeWidget.py
Bindings.pyc         MultiStatusBar.pyc     UndoDelegator.py
CallTipWindow.py     NEWS.txt               UndoDelegator.pyc
CallTips.py          ObjectBrowser.py       WidgetRedirector.py
CallTips.pyc         OldStackViewer.py      WidgetRedirector.pyc
ChangeLog            OutputWindow.py        WindowList.py
ClassBrowser.py      OutputWindow.pyc       WindowList.pyc
ColorDelegator.py    PathBrowser.py         ZoomHeight.py
ColorDelegator.pyc   Percolator.py          ZoomHeight.pyc
Debugger.py          Percolator.pyc         __init__.py
Delegator.py         PyParse.py             eventparse.py
Delegator.pyc        PyParse.pyc            extend.py
EditorWindow.py      PyShell.py             extend.pyc
EditorWindow.pyc     PyShell.pyc            extend.txt
FileList.py          README.txt             help.txt
FileList.pyc         ReplaceDialog.py       idle.bat
FormatParagraph.py   ScriptBinding.py       idle.py
FormatParagraph.pyc  ScriptBinding.pyc      idle.pyw
FrameViewer.py       ScrolledList.py        idlever.py
GrepDialog.py        SearchBinding.py       idlever.pyc
IOBinding.py         SearchBinding.pyc      keydefs.py
IOBinding.pyc        SearchDialog.py        keydefs.pyc
Icons                SearchDialogBase.py    pyclbr.py
IdleHistory.py       SearchEngine.py        testcode.py

Any other ideas?


>
> "Jeff Stephens" <jsteve17 at tampabay.rr.com> wrote in message
> news:3A89A78F.C4C6B70A at tampabay.rr.com...
> > I am just starting out with Python and was trying to use idle, but the
> > fonts
> > are so small it's giving me a headache.  I followed the instructions in
> > the
> > help menu and tried editing the 'EditorWindow.py' file to no avail. I
> > commented
> > out the first line and uncommented the second as recommended: no change.
> > I tried
> > changing the sizes of both fonts to 10, 12: no change.  The only thing I
> > can figure is
> > that idle is using some default that is not reflected by either of these
> > lines. Can someone
> > help before I go blind.  Thanks.
> >
> > Regards,
> > Jeff Stephens
> >



From syver at NOSPAMcyberwatcher.com  Wed Feb  7 01:31:45 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Wed, 7 Feb 2001 01:31:45 +0100
Subject: Alpha release of ZODB programming guide
References: <mailman.981388047.29255.python-list@python.org> <95mr3t$ad2$1@troll.powertech.no> <3dg0htymst.fsf@ute.cnri.reston.va.us>
Message-ID: <95q562$fpu$1@troll.powertech.no>

"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3dg0htymst.fsf at ute.cnri.reston.va.us...
> "Syver Enstad" <syver at NOSPAMcyberwatcher.com> writes:
> > This sounds very exciting, I had been looking forward to an object
database
> > on python. I've tried ZEO under MS Win 2000 but it doesn't seem to work.
Is
> > it supposed to?
>
> I don't know, but suspect it should work -- for some value of "work"
> -- on Windows.  But note that I can't make Windows binaries available,
> so you'd need to have a C compiler in order to install it.

I have msdev so that's okay, I just tried to install it and it complained
about various signals and other unix stuff not being available. Maybe I did
something wrong when installing?




From mfletch at tpresence.com  Mon Feb 26 02:47:27 2001
From: mfletch at tpresence.com (Mike Fletcher)
Date: Sun, 25 Feb 2001 20:47:27 -0500
Subject: abusing __builtins__
Message-ID: <CC59868A0A76D311B1F50090277B321C1F37CD@VRTMAIL>

I've used this extensively in one (very large) in-house demo I created, I
used builtins.APPLICATION to refer to a global application object.  Was very
simple and useful.  Yes, a hack, but I considered it a rather nice hack.
Apparently this is another of those "you're wrong, Mike, just plain wrong"
things :) .

I seem to get more and more of those these days,
Mike

-----Original Message-----
From: Thomas Wouters [mailto:thomas at xs4all.net]
Sent: Sunday, February 25, 2001 6:17 PM
To: python-list at python.org
Subject: Re: abusing __builtins__


On Sun, Feb 25, 2001 at 06:08:35PM -0500, Tim Peters wrote:

> People other than you will be unable to understand your code regardless,
> because GlobalVariable will look to them like an unbound variable (i.e.,
> nobody else uses this trick, so nobody else will recognize the usage
> pattern).

Actually, Mailman used this trick for a while. Not in any release, though,
and Barry removed the ugly hack a week or two ago. Not sure if it was
originally Barry's idea, or just part of the internationalization patch he
accepted.

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me
spread!

-- 
http://mail.python.org/mailman/listinfo/python-list



From dbrueck at edgix.com  Fri Feb  2 18:24:26 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Fri, 2 Feb 2001 10:24:26 -0700
Subject: Four problems with Gordon McMillan's Installer
In-Reply-To: <95ele5$gqs15$1@ID-59885.news.dfncis.de>
Message-ID: <BLEOJGEMBNBNEAKMIFDHKEEOPIAA.dbrueck@edgix.com>

> > Maybe there is something that can be done with your program to
> at least hide
> > the python files inside the zip file to provide at least some level of
> > security?  I dont know how others handle this for commercial apps.
> Well, this is on the to-do list for later.
> I want to do this in two stages:
> - first, use a notstandard zip-header so that winzip does
> not recognize it
> - second, maybe crypt the .pyc files in the archive somehow
> and decrypt them in the exe.
>
> Do you think this would be apppropriate?

Yes, that sounds like a good idea... you can use the imp/ihooks (is ihooks
obsolete?) modules to hook into the normal module importer and decrypt them
on the fly. For the zipfile you could still use a normal format but save it
to disk in a munged-up format (something simple like the rotor module would
keep most people out anyway) and then unmunge before reading it.

-Dave

P.S. - I just tried the latest version of py2exe on a wxPython program and
it worked _flawlessly_ and took little time to set up and get going. Very
cool...




From ljohnson at resgen.com  Tue Feb 27 00:03:08 2001
From: ljohnson at resgen.com (Lyle Johnson)
Date: Mon, 26 Feb 2001 17:03:08 -0600
Subject: Packet parsing problem...
References: <97ek6i$o6d@news.or.intel.com>
Message-ID: <t9lo117ns2so54@corp.supernews.com>

> In python, is there any structure that approximates the C++ DWORD?  I need
> to be able to parse incoming packets, as well as properly form messages to
> the server.

I hate to ask, but is this the DWORD that's commonly typedef'd in the
Windows header files? If so, it's just a 32-bit (signed) integer and it
should map to a Python integer.




From gerhard.nospam at bigfoot.de  Wed Feb 14 21:31:56 2001
From: gerhard.nospam at bigfoot.de (Gerhard =?iso-8859-1?Q?H=E4ring?=)
Date: Wed, 14 Feb 2001 21:31:56 +0100
Subject: Robots
References: <mailman.982180157.16563.python-list@python.org>
Message-ID: <3A8AEB3C.4EC31E35@bigfoot.de>

Ben Ocean wrote:
> 
> Hi;
> Does anyone have a good robot script they'd like to share? I need to build
> one that searches the Web for similarly themed Web sites and grabs email
> addresses.

Are you trying to fill a spam database? If so, you deserve a good kick
in your ass. Perhaps this would help you.

Gerhard (who hates being spammed)

PS: In the improbable case that spamming is not your intention, I
apologize. I just cannot imagine any other reason for grabbing emails
off html documents.
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com


From owen at astrono.junkwashington.emu  Mon Feb 26 21:53:15 2001
From: owen at astrono.junkwashington.emu (Russell E. Owen)
Date: Mon, 26 Feb 2001 12:53:15 -0800
Subject: Hiding widgets it Tkinter and shrinking the frame/window?
Message-ID: <97efnv$tqq$1@nntp6.u.washington.edu>

I am trying to generate a window with basic options always showing and 
additional panels that are hidden or displayed if needed. I can easily 
show/hide widgets, but the window does not shrink after a widget is 
hidden. Is there some basic trick to this?

(Also, is there some simple way to get the TopLevel object from a 
widget, perhaps one that is buried many levels deep in enclosing frames 
and etc.? I think I may have to make window manager calls, and I don't 
want to pass the TopLevel widget around if I don't have to.)

-- Russell


From moshez at zadka.site.co.il  Sun Feb  4 21:21:32 2001
From: moshez at zadka.site.co.il (Moshe Zadka)
Date: Sun,  4 Feb 2001 22:21:32 +0200 (IST)
Subject: Please translate this easy snip of C++ to Python
In-Reply-To: <95kcbo$k8s@dispatch.concentric.net>
References: <95kcbo$k8s@dispatch.concentric.net>
Message-ID: <20010204202132.0B13BA840@darjeeling.zadka.site.co.il>

On 04 Feb 2001 19:59:52 GMT, Phlip <phlip_cpp at my-deja.com> wrote:

> Pythonographers:
> 
> Ogle this awesome snip of C++:
> 
>         #define TRACE_(x) cout << #x ": " << x << endl

You can do it in Python the other way around:

>>> def trace_(x):
...         print x+":", eval(x)
...
>>> y = z = 6; animal='lemming'
>>> trace_('y+z')
y+z: 12
>>> trace_('animal')
animal: lemming

HIH!
-- 
Moshe Zadka <sig at zadka.site.co.il>
This is a signature anti-virus. 
Please stop the spread of signature viruses!
Fingerprint: 4BD1 7705 EEC0 260A 7F21  4817 C7FC A636 46D0 1BD6



From loewis at informatik.hu-berlin.de  Mon Feb 19 19:48:24 2001
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 19 Feb 2001 19:48:24 +0100
Subject: What is wrong in my pulldom.startElement()?
References: <R4gi6.1682$HG.377319@news4.rdc1.on.home.com>
Message-ID: <p6qd7cemruv.fsf@informatik.hu-berlin.de>

"Bill Xu" <bill.xu at newlix.com> writes:

> Will u please tell me how to use the pulldom.startElement() in the Python
> XML DOM?
>  Following is my simple test program and result.
>  Will u please tell me how to correct this bug?
> "AttributeError: 'None' object has no attribute 'createElement'";

I think you are using startElement itself quite correctly. However,
you have to emulate exactly a SAX parser, i.e. first you need to
invoke startDocument.

Please note that pulldom is undocumented. I don't know what you are
attempting to do, but using pulldom is most likely the wrong way.

Regards,
Martin



From dsh8290 at rit.edu  Mon Feb  5 18:42:42 2001
From: dsh8290 at rit.edu (D-Man)
Date: Mon, 5 Feb 2001 12:42:42 -0500
Subject: nested scopes
In-Reply-To: <95lu1i$ebs@gap.cco.caltech.edu>; from n8gray@caltech.edu.is.my.email.address on Mon, Feb 05, 2001 at 02:01:58AM -0800
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu>
Message-ID: <20010205124241.A27402@harmony.cs.rit.edu>

On Mon, Feb 05, 2001 at 02:01:58AM -0800, Nathaniel Gray wrote:
| Tim Peters wrote:
| 
| > Whether a name is local to a given scope is, in Python, determined by
| > whether the name is *bound* somewhere in that scope.  "import *" counts as
| > a
| > binding, but the compiler has no idea which names are being bound.  So
| > when compiling code for a nested function N, referencing a non-local name
| > (one that is not bound to within N's body) X, the compiler has no idea
| > which scope X belongs to if an "import *" (or "exec" stmt) appears in an
| > enclosing
| > function E.  Any name whatsoever *may* become local in E then, but it's
| > impossible to know anything about that at compile-time.
| 
| Interesting issue.  What if the following was allowed:
| >>> from tim_bot_module import tb*
| 

Hmm, wouldn't it be better to use :

>>> import tim_bot_module as tb
>>> tb.name_you_want_to_use

?

IMO prefixing names with strings like "tb" is for namespace-less
languages like C.  Since python has namespaces, why not use them?


What I wonder about is whether names defined in a try block or a for
block will still exist outside the block.

Take this Java code for example:

String s ;
try
{
	s = some_func_that_might_throw() ;
}
catch ( ThrownException ex )
{
	s = "Some other value" ;
}


So you look at it and whether the function throws an exception or not,
s has been initialized appropriately.  javac won't compile this since
"s might not be initialized".

In Python, it is much nicer :

try :
	s = some_func_that_might_throw()
except ThrownException :
	s = "Some other value"

and everything works as expected.  If the exception was thrown, s will
have the one value, if not it will have the other.


Will the new nested scopes require me to write:

s = None # make the name exist in this scope first
try :
	...

?


I think it is acceptable if variables like loop counters and temps
inside a loop don't exist in the function's namespace. ie:

for i in range( 10 ) :
	pass
print i # this gives NameError


but :

i = 1
for i in range( 10 ) :
	pass
print i # prints '9' since the name i already existed and was rebound



What I wonder about is how important/useful nested scopes really are.
When I first learned python, I had an Eiffel/C++/C background with
maybe a little Java.  Each of those has nested lexical scopes.  I
thought python was weird and lacking.  I have since learned that
python is quite natural and doesn't require various ugly idioms
(declaring the variable at the top) to make the assignment in the
nested scope persist outside the loop/if/whatever. 

Last quarter I took a class called "Programming Language Concepts"
that discussed the design of programming languages.  It was the first
time I had heard of dynamic scoping.  (Actually, I think I saw it
mentioned in a lisp context, but I didn't know what it meant)  The
professor didn't know of any situations where it would be
advantageous, and I couldn't contrive any either.  I was kind of
surprised when I heard python described as "dynamically scoped"
because its scoping was really quite natural for me.  The [23]-level
scoping makes the dynamic scoping almost static and I don't see that
as a weakness.

I really don't know, at this moment, whether adding nested static
scopes will strengthen or weaken the language.

just-a-few-thoughts-ly y'rs -D




From db3l at fitlinxx.com  Wed Feb 28 02:45:33 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 20:45:33 -0500
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com> <983312067.173004@clam-ext> <pgWm6.169$jv4.27104@e420r-atl2.usenetserver.com> <983315235.473229@clam-ext>
Message-ID: <uofvnmvgi.fsf@ctwd0143.fitlinxx.com>

"Mark Hadfield" <m.hadfield at niwa.cri.nz> writes:

> "Steve Holden" <sholden at holdenweb.com> wrote in message
> news:pgWm6.169$jv4.27104 at e420r-atl2.usenetserver.com...
> > Dunno. Still won't work for me on Windows 98. Neither, come to that, does
> > the parenthesized version
> >
> >     a=(os.popen("dir")).readlines()
> >
> > I get the same error.
> 
> OK so the parentheses were a red herring. It's a difference in the behaviour
> of piping in windowed applications vs console applications under Windows 9x.

I don't recall if the version of Python was noted in this thread yet,
but there are definite issues and differences with console versus
windowed applications in os.popen with Python 1.5.2, and still some
environmental dependencies with Python 2.x.

In Python 1.5.2, os.popen is a direct call to the underlying C popen()
function, which under Windows requires that there be a console to be
reliable under NT, and may not be reliable at all under Windows 9x.

In Python 2.x (and maybe 1.6), os.popen borrowed code from the
win32all package (which you can also use under Python 1.5.2 as
win32pipe.popen()) which works properly in both console and windowed
environments.  But under Windows 9x, it needs a helper application
(win32popenWin9x.exe under win32all, or w9xpopen.exe under Python 2.x)
to function properly.

It's possible that under Pythonwin, the appropriate utility isn't
being found on your path, which may differ from the path you get in a
command prompt or in other tools.  Python should return a runtime
error describing this file, but perhaps that's being absorbed in
Pythonwin or something.

Just as an experiment you might try placing that file (it's in the
Python 2.0 main directory or the win32 subdirectory with win32all) in
a directory known to be in your main path.

I think Mark Hammond may have posted recently on this issue and
suggested copying the executable into the same directory as
PythonWin.exe - yep, that was the suggestion.  I don't know how to
make a canonical reference to Google's newsgroup stuff yet, but if you
search for "w9xpopen pythonwin" on http://groups.google.com it'll
bring it up.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From tim.one at home.com  Fri Feb 23 22:17:40 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 23 Feb 2001 16:17:40 -0500
Subject: Nested scopes resolution -- you can breathe again!
In-Reply-To: <slrn99dg5d.p6b.wtanksle@dolphin.openprojects.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEJHJBAA.tim.one@home.com>

[Guido]
>    import __nested_scopes__
>    from __future__ import nested_scopes

[William Tanksley]
> I like the second version.  That's quite cool.

I'd have a hard time guessing that from the rest of this msg <wink>.

> ...
> And perhaps _old_ code should bear the brunt of the change, with one of
> the features being old_unnested_scoping. This way we don't have to keep a
> default that's deprecated forever.

No.  The point of the scheme is to let motivated people use release N+1
features in release N without anyone else needing to change any of their
code.  To the extent possible, release N will instead produce warnings about
code that *will* break in release N+1.  Nobody is required to endure N+1
semantics before N+1; but people who want to experiment with N+1 features in
N can do so by explicitly asking for them.

> So we'd never see the line "from __features__ import nested_scoping";
> that would become an anachonism, and would quickly get irritating.

Good!  That means people will be motivated to get rid of it when it becomes
an anachronism (in N+1).  This puts the burden where it should be (i.e., on
the pioneers, not on the laggards <0.9 wink>).  BTW, for reasons to be
explained in an upcoming PEP, it will be very easy to find unneeded
__future__ imports as the millennia roll by.

> Instead, we'd see "from __features__ import old_unnested_scoping" at
> the head of modules which won't compile without warning.

In release N+1, release N semantics are *gone*.  There's no going back, else
the Python implementation would become an unboundedly complicated mess as
the years go by, and so would Python programs.  __future__ is for
incompatible changes, not for optional behaviors.  If, at release N+1, you
refuse to move to release N+1 semantics, fine:  stay with release N forever.
That's your choice.




From fabrice.n at home.com  Sun Feb 18 14:01:58 2001
From: fabrice.n at home.com (shaka)
Date: Sun, 18 Feb 2001 13:01:58 GMT
Subject: 3D animation
Message-ID: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com>


Ya, I am planning to write a program to present a project at school.  I need
a language that will allow me to do 3D animations and other animation.  And
I am struggling between writting the program in java or using python.
Can someone help me?  If python is the ultimate choice then can you give me
some instructions on what libraries I should use, because I am quite a
beginner here>

Thanks.

Shaka. Z




From fgeiger at datec.at  Fri Feb 23 18:45:30 2001
From: fgeiger at datec.at (Franz GEIGER)
Date: Fri, 23 Feb 2001 18:45:30 +0100
Subject: Want to build a shop using Zope, but how?
Message-ID: <9767le$82f$1@newsreaderg1.core.theplanet.net>

I want to do the next step in CGI programming and want to use Zope to build
a web shop.

Besides downloading Zope and installing it on my Windows box I printed the
Zope book and read thru it. There is an example all thru the book which is
not of great help for me (at least I think so). While it is good to teach
you the very basics of Zope it does not seem to provide advanced topics.
When it comes to advanced concepts you are very likely to encounter "But
this is beyond this book" (ZClasses e.g.).

I've got the impression, that if one wants to use Zope for anything advanced
he has to download some modules from www.zope.org. But this site really
confuses me. Every day there are new modules released and every few days new
versions of those modules. I don't want to spend hours and days to figure
out what to download so that Zope is of advanced help for me.

The same seems to apply to documentation. Zope still seems to move very
fast. Which of the docs and how-to's are still relevant and which are
outdated?

Yes, I saw one or two Zope products on e-commerce which could be
interesting. But how can I understand those products if I do not understand
Zope yet?

I must admit that I am looking for fast success. Since I know Python I know
fast success is possible ;-). How can I achieve fast success using Zope? My
hope was to find information somewhere on how Zope supports a Pythoneer in
CGI programming: How to deal with cookies in Zope, how to manage session
etc. I am less interested in DTML.

Or did I get something completely wrong here? Anyone here who was in the
same situation? How did you proceed?

Any hint's welcome!

Many thanks in advance and best regards
Franz GEIGER


P.S.: Don't get me wrong because of my clumsy English: I WANT to use Zope.
It's oo approach is too appealing.








From dbrueck at edgix.com  Mon Feb 26 18:00:35 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Mon, 26 Feb 2001 10:00:35 -0700
Subject: is and ==
In-Reply-To: <20010226175023.A7401@freedom.puma-ag.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHIEHPPKAA.dbrueck@edgix.com>

> [mailto:python-list-admin at python.org]On Behalf Of Steve Purcell
> Sent: Monday, February 26, 2001 9:50 AM
> To: Daniel Klein
> Cc: python-list at python.org
> Subject: Re: is and ==

< snip >
> Integers are both equal *and* identical:
>
> >>> a = 1
> >>> b = 1
> >>> a is b
> 1
> >>> a == b
> 1

Not necessarily:

>>> a = 10000
>>> b = 10000
>>> a is b
0
>>> a = 5
>>> b = 5
>>> a is b
1

The interpreter "caches" integer objects from something like -1 to 100, so
'is' evaluates to true for low integers but not higher integers.

(Also, as the docs point out, this caching means that from inside an
extension module you could _redefine_ the value of the integer object 5, for
example:

>>> evilext.causehavoc(5,10) # sets '5' object to have int value of 10
>>> 5 + 5
20  # Whee!!!

)

-Dave




From fredrik at pythonware.com  Sun Feb 18 22:51:12 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 18 Feb 2001 21:51:12 GMT
Subject: Pls Help! Need binary search tree module
References: <OhXj6.4474$Nh7.300636@news1.rdc1.ab.home.com>
Message-ID: <ktXj6.18191$AH6.2496311@newsc.telia.net>

Steve Mak wrote:
> I was wondering if anyone had a simple binary search tree module I could
> use? I need to write a program that counts the number of occurrences of each
> word in a file, and outputs the word and corresponding counts
> alphabetically. And it is required that one of the data structure is a
> binary search tree.

are you sure your teacher doesn't read comp.lang.python?

I'm pretty sure you'll learn more if you do it yourself, but you
could take a look at the bisect.py module (it's in Python's Lib
directory).

Cheers /F




From bryan at eevolved.com  Thu Feb 15 13:35:48 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Thu, 15 Feb 2001 12:35:48 GMT
Subject: Gauging system load
Message-ID: <E2Qi6.237705$f36.9512658@news20.bellglobal.com>

Out of curiosity, does anyone know of a module or a neat trick that will 
permit me to gauge the system load at a given time? For example, in a 
utopian dream world:

import sys, time

# A percentage value
THRESHOLD = 20

while 1:
  if sys.cpuActivity <  THRESHOLD:
    # Do some heavy work
    break
  else
    time.sleep(10)


I've thought of maybe spawning a thread and gauging the amount of time it 
takes for it to be serviced, but that's pretty crude and innaccurate.

Any tips?
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"I want to know God's thoughts; the rest are details."-- Einstein



From slaurent at sonicfoundry.com  Tue Feb 13 19:13:58 2001
From: slaurent at sonicfoundry.com (Sean Laurent)
Date: Tue, 13 Feb 2001 12:13:58 -0600
Subject: os.popen: writing to...?
References: <959pqt$tgp$1@nnrp1.deja.com>
Message-ID: <vvti8t07p3pe5b7b1mrb5tq88knmt81g13@4ax.com>

On Wed, 31 Jan 2001 19:42:23 GMT, James Lockley
<poseidon_cma at my-deja.com> wrote:

>i now want to go to the next step and get rid of the batch files...
>i need to open a pipe to a command and then write the command arguments
>to it.
>
>as far as i could make out, this should work:
>
>>>> abaqus='c:\\abaqus\\5.8-14\\abaqus.exe'
>>>> abq=os.popen(abaqus,'w')
>>>> abq.write('post')
>
>but i get this:
>
>Traceback (innermost last):
>  File "<interactive input>", line 1, in ?
>IOError: [Errno 22] Invalid argument

Hrmmm.  I have been encountering the exact same problem under Windows
2000.

I have tried all the flavors of os.popen ... os.popen(), os.popen2(),
os.popen3() and os.popen4().  I have also tried the popen2 flavors eg.
popen2.popen2(), popen2.popen3(), and popen2.popen4().  None of them
seem to work...

>>> instream, outstream = popen2.popen2("p4 -x - edit")
>>> instream
<open file 'p4 -x edit', mode 'r' at 017A1C88>
>>> outstream
<open file 'p4 -x edit', mode 'w' at 017A1CB0>
>>> outstream.write('d:\\install\\proj\\projects.xml\n')
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
IOError: [Errno 22] Invalid argument

Does anyone have any bright ideas or suggestions?

-Sean-



From jhauser at ifm.uni-kiel.de  Mon Feb 19 12:39:06 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 19 Feb 2001 12:39:06 +0100
Subject: Yet Another Newbie Question (YANQ)
References: <Xlsj6.95$DC1.647178@newsserver.ip.pt>
Message-ID: <878zn2yk9x.fsf@lisboa.ifm.uni-kiel.de>

You should use exec() instead of eval().

HTH,
__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From clickron at webtv.net  Sat Feb 10 22:13:04 2001
From: clickron at webtv.net (clickron at webtv.net)
Date: Sat, 10 Feb 2001 16:13:04 -0500 (EST)
Subject: def problem
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net>
Message-ID: <10844-3A85AEE0-152@storefull-161.iap.bryant.webtv.net>

I write this, save it. and run it.
def hello(): 
??????????????print "hello"
???????print "bye bye" 
hello() 
and I get this message 
hello
Traceback innermost last):
File "<pyshell27>",line 1, in ?
hello 
NameError: There is no variable named 'hello' 
What am I doing wrong? 
It works for me. My guess is that you are putting the function in a
file, saving it, running it, and then typing "hello()" into IDLE (or
PyWin). That would cause the error you report. Is the hello() function
call in the same file as the function? 
_________________________________

That's exactly what I'm doing. I go to IDLE, open a window, write the
function, save it, and run it. Then  I type in hello()
and get the error message. 

Now I retyped it, both with the function call  in the same file as the
function and without it. Did everything else the same and it works both
ways.

I don't know what's going on. How should I do it?

Ron



From clickron at webtv.net  Sun Feb 11 16:41:51 2001
From: clickron at webtv.net (clickron at webtv.net)
Date: Sun, 11 Feb 2001 10:41:51 -0500 (EST)
Subject: def problem (Got it!)
References: <10844-3A8572B2-38@storefull-161.iap.bryant.webtv.net>
Message-ID: <25253-3A86B2BF-12@storefull-163.iap.bryant.webtv.net>

As a newbie you have to know I'm going to make some stupid mistakes, and
this was one of them.

What I was doing was sometimes clicking "run script" and mostly hitting
the "F5 key" which I just realized is not "run script" but "import
module."

I might have another question, but I think I'll go and hang my head in
shame for while. 

Thanks again for taking the time to answer my question.

Ron



From n8gray at caltech.edu.is.my.email.address  Sun Feb 18 08:10:47 2001
From: n8gray at caltech.edu.is.my.email.address (Nathaniel Gray)
Date: Sat, 17 Feb 2001 23:10:47 -0800
Subject: Discussion about PEP 234: iterators
References: <96jonj$i0e$1@nntp6.u.washington.edu> <96k28b01utj@news2.newsguy.com> <3A8DA52A.254849DB@javanet.com> <o_hj6.18$8t2.15688@news1.rdc1.az.home.com>
Message-ID: <96nst0$pe9@gap.cco.caltech.edu>

Tim Hochberg wrote:
> [snip discussion of consistency of "if thing in dict:"]
>
> Unfortunately, people tend to think of a dictionary as containing keys
> more than they think of it as containing definitions (values). So the form
> you propose is better from that point of view. It's also probably more
> useful in a practical sense as well. It is kind of distressing that it's
> not consistent with the correspondence that's set up between
> keys<->indices and values<->values though.

Hmm... I tend to think of a dictionary as containing values more than keys. 
 The keys are just there so you can get to the values, right?  Others tend 
to think dicts contain (key, value) pairs.  Everybody has a different 
opinion on this one.  Rather than supplying an arbitrary default, which 
would lead to confusion, how about we let the programmer specify things 
explicitly.

How about we make it:
if key in dict.keys():
and
if value in dict.values():
and maybe even
if (key, value) in dict.items():

That way it's perfectly clear what the programmer intended.  I think 
there's a really good chance of Guido accepting this proposal, and I've 
already written a fully backwards-compatible patch.  Anybody up for a trip 
back in time?

-n8

-- 
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._
             Nathaniel Gray
   California Institute of Technology
     Computation and Neural Systems
     n8gray <at> caltech <dot> edu
_.~'^`~._.~'^`~._.~'^`~._.~'^`~._.~'^`~._



From cjc26 at nospam.cornell.edu  Mon Feb 19 19:12:44 2001
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: Mon, 19 Feb 2001 18:12:44 GMT
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org> <slrn990gd3.tc5.qrczak@qrnik.zagroda>
Message-ID: <wmdk6.96422$v.7354987@typhoon.nyroc.rr.com>

* Marcin 'Qrczak' Kowalczyk <qrczak at knm.org.pl> menulis:
| 
| What about introducing a "black hole" or "wildcard pattern" for
| assignments?
|     e0, e1, _, _ = some_4_tuple
|     for k:_ in mapping:
|     for _:v in mapping:
| If there is code using _ as a variable name, maybe some different name.
| But it's _ in some other languages (ML, Haskell, Prolog, Clean).

You can already do stuff like

    e0, e1, _, _ = (10, 20, 40, 30)
    print e0  => 10
    print e1  => 20
    print _   => 30

You can use whatever variable name you want for the dummy variable.

As for dict iteration, I don't like the idea of inventing new syntax for
it; I would prefer having dict.items() return a lazy list, or maybe add
a xitems() method to return a lazy list.  (Actually, if there were some
way to signal to a method that it's being called in an iteration
context, then we wouldn't need all these x-methods...)


-- 
Cliff Crawford                 http://www.people.cornell.edu/pages/cjc26/
"huh?  Are you all there?  You missing some key parts up there?  Becoming  
a English major or an minor?   Do you need a annual report?  Are you slow
today, eh?  Did someone sever you parietal lobe?" - Mark


From esr at thyrsus.com  Tue Feb 13 19:41:47 2001
From: esr at thyrsus.com (Eric S. Raymond)
Date: Tue, 13 Feb 2001 13:41:47 -0500
Subject: CPAN functionality for python
In-Reply-To: <20010213102731.B24074@tummy.com>; from jafo@tummy.com on Tue, Feb 13, 2001 at 10:27:31AM -0700
References: <slrn98c5fc.2ah.ssthapa@ntcs-ip45.uchicago.edu> <20010212110204.A4625@harmony.cs.rit.edu> <20010212115117.A7546@tummy.com> <20010213015924.Q1764@thyrsus.com> <20010213102731.B24074@tummy.com>
Message-ID: <20010213134147.B15155@thyrsus.com>

Sean Reifschneider <jafo at tummy.com>:
> On Tue, Feb 13, 2001 at 01:59:24AM -0500, Eric S. Raymond wrote:
> >Trove is dead.  Long live SourceForge.
> >
> >The SourceForge guys picked up many of Trove's core ideas.  SourceForge II,
> >which we're doing a planning meeting for on Wedneday, will pick up more
> >of them.
> 
> No lessons that can be learned from it?

Oh, believe you me, I am carrying forward *lots* of lessons from Trove.
-- 
		<a href="http://www.tuxedo.org/~esr/">Eric S. Raymond</a>

The direct use of physical force is so poor a solution to the problem of
limited resources that it is commonly employed only by small children and
great nations.
	-- David Friedman



From peter at engcorp.com  Sat Feb 10 18:46:56 2001
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 10 Feb 2001 12:46:56 -0500
Subject: None assigment
References: <mailman.981708020.6548.python-list@python.org>
Message-ID: <3A857E90.18EE0634@engcorp.com>

Gregoire Welraeds wrote:
> 
> Maybe None have been added to builtin to avoid special cases. But it
> becomes clear, as we understand it, that None is now a special case of
> builtin.
> 
> So, following that thought, we think that None should be a reserved word.
> 
[snip]
> 
> And we all agree that programmers should not touch to the value of
> __builtin__.None. We think that the ability to override None because None
> is a builtin is a conceptual problem.

No!  Please, why muck with things that work perfectly well?

I can think of at least one situation in which the ability
to do this could be of some _real_ value.  Making None
a reserved word would prevent me from doing that.  It would
gain you nothing except a stricter language which would
be getting ever closer to those "other" languages which
are so restrictive that they place limits on your imagination
and ability to innovate.

Why, I wonder, do some people think that just because
something "shouldn't" happen, it must therefore 
never be allowed to happen?  If "we all agree that 
programmers should not touch None", can't we just
stop there and leave those mature people to do what
they wish?  (People "shouldn't speed", either, but
I don't think many want governors on their car engine 
actually to _prevent_ it.)

Remember, Python is a dynamic language.  The more
special-case restictions you place on it, the less
like Python it's going to look.

likely-starting-a-flame-war-ly yr's  -Peter

P.S. Identifying useful things that could be done with
assigning to None is left as an exercise to the reader.
I wouldn't want to stifle creativity by mentioning my
idea too quickly. :-)


From nem00u at cs.nott.ac.uk  Mon Feb 19 00:29:27 2001
From: nem00u at cs.nott.ac.uk (Neil Madden)
Date: Sun, 18 Feb 2001 23:29:27 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com> <3A8FFF1C.1FFEADE7@cs.nott.ac.uk>
Message-ID: <3A905AD7.152F6626@cs.nott.ac.uk>

BTW. I think that the problems you have posed are very much "scripting"
solutions rather than system programming solutions. In other words,
scripting languages (perl, python, tcl) are likely to be able to solve
these fairly easily, whereas compiled languages are going to find it
harder. If you posted scenarios involving intensive maths computation
the results might be the other way around :-). If you are writing a
paper, compare like to like...

Neil Madden wrote:
> 
> Dave Cross wrote:
> >
> > [Please watch the replies on this message as it's heavily
> > cross-posted]
> >
> > I'm doing some comparisons on programming language features and I'd be
> > very interested to know how you would handle the following scenarios
> > in your programming language of choice.
> >
> > 1/ The programmer calls a function that doesn't actually exist within
> > the application (or libraries). Is the a feature whereby the
> > programmer can create a "catch-all" function which is called in cases
> > like these? Can this function examine the list of existing functions
> > and call the most appropriate one? Or create a new function on the fly
> > and install it into the application?
> 
> Tcl has the "unknown" procedure that is called when the interpreter
> doesn't recognize a command. From the Tcl help pages:
> 
> "If the Tcl interpreter encounters a command name for which there is not
> a defined command, then Tcl checks for the existence of a command named
> unknown. If there is no such command, then the interpreter returns an
> error. If the unknown command exists, then it is invoked with arguments
> consisting of the fully-substituted name and arguments for the original
> non-existent command. The unknown command typically does things like
> searching through library directories for a command procedure with the
> name cmdName, or expanding abbreviated command names to full-length, or
> automatically executing unknown commands as sub-processes. In some cases
> (such as expanding abbreviations) unknown will change the original
> command slightly and then (re-)execute it. The result of the unknown
> command is used as the result for the original non-existent command."
> (<URL: http://dev.scriptics.com/man/tcl8.4/TclCmd/unknown.htm>)
> 
> It is left up to the application programmer to implement this, so you
> could have an unknown procedure that does the things you mention above.
> 
> See also: <URL: http://mini.net/cgi-bin/wikit/495.html>) for some
> examples.
> 
> >
> > 2/ Can ou filter the input source code before compilation (or
> > interpretation) in some way so that language keywords could be changed
> > for other strings? Imagine you wanted to allow someone to program your
> > language of choice in, say, French. How would you go about translating
> > French keywords into ones that the compiler (or interpreter) could
> > understand. What if the translation wasn't one-to-one or fixed? Could
> > you put enough intelligence into the translator so that it could
> > handle certain strings differently depending on where they appeared in
> > the source code?
> 
> There are no reserved words in Tcl, so all keywords can be overwritten
> or renamed. Using the unknown command above, you can put as much
> intelligence as you want into interpreting a command. Although, this
> seems like a fairly dangerous thing to do, as any packages or extensions
> you used would have to know that the keywords had all been changed! But
> once again, you could allow the [unknown] handler to map commands onto
> their new names.
> 
> >
> > If you're wondering why I'm inventing these bizarre scenarios, it's
> > for a paper I'm writing for this year's Perl Conference.  Perl does
> > have these features (see the AUTOLOAD function and source filters) and
> > I'm interested in seeing how widespread they are in other languages.
> >
> > Of course, if you'd like to tell me just why you consider it's a good
> > thing that your language of choice doesn't have these features, then
> > I'd be only too happy to hear that too.
> >
> > I'd just like to make it clear that I'm not interested in getting into
> > "my language is better than your language" types of flamewars. I'm
> > certainly not trying to argue that Perl is better than other languages
> > for having these features.
> >
> > Thanks for your time.
> >
> > Dave...
> >
> > --
> > <http://www.dave.org.uk>  SMS: sms at dave.org.uk
> 
> --
> --------------------------------------------------------------
> Neil Madden.
> neil at tallniel.co.uk                       nem00u at cs.nott.ac.uk
> http://www.tallniel.co.uk     http://www.cs.nott.ac.uk/~nem00u
> --------------------------------------------------------------

-- 
--------------------------------------------------------------
Neil Madden.
neil at tallniel.co.uk                       nem00u at cs.nott.ac.uk
http://www.tallniel.co.uk     http://www.cs.nott.ac.uk/~nem00u
--------------------------------------------------------------


From erno-news at erno.iki.fi  Wed Feb  7 07:48:19 2001
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 07 Feb 2001 08:48:19 +0200
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com> <95qg45$5b0$0@216.39.151.169> <95qim7$i38$1@nnrp1.deja.com>
Message-ID: <kulmrj0z0s.fsf@lasipalatsi.fi>

In article <95qim7$i38$1 at nnrp1.deja.com>, penman <pen_man at my-deja.com>
writes:

| In article <95qg45$5b0$0 at 216.39.151.169>,
|   Donn Cave <donn at oz.net> wrote:
|| Don't mess around with non-blocking.  Even recv(8000) will return
|| more or less immediately if you send any data at all, so it seems
|| kind of likely you're sending no data until the CR.  Substitute a
|| test client that has its data built in, and just send('i') to the
|| server.  I bet recv() will get it right away, Nagel algorithms
|| notwithstanding.

| Thank you, Donn. But the problem is that I have to use ordinary telnet
| clients, such as CRT, Netterm, so let the users access to the server
| without any specialy programs.

the telnet protocol provides a mechanism to negotiate options such
as the character-at-a-time mode. the default is line at a time,
i believe. the rfcs have the details...

| I wonder why nobody ever had the need to recv 1 character "unbuffered"
| in Python.

again: there is no buffering on the python side here.

  -- erno


From syver at NOSPAMcyberwatcher.com  Thu Feb 15 13:40:58 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Thu, 15 Feb 2001 13:40:58 +0100
Subject: COM support for vtable based interfaces.
References: <3A8B0448.1000905@ActiveState.com>
Message-ID: <r7Qi6.195$1%6.1750@news1.oke.nextra.no>

Super, this means that I can implement COM classes support the same
interface as a typical ATL generated dual-interface COM object doesn't it?
That's good news indeed! I've been needing this for a little while.

"Mark Hammond" <MarkH at ActiveState.com> wrote in message
news:3A8B0448.1000905 at ActiveState.com...
> Hi all,
> I have recently checked in a new version of win32com that allows you to
> implement _any_ COM interface decribed in a type library, either
> IDispatch based or not.  This is excellent news :)  You still can't
> _call_ these arbitary interfaces, but no doubt this will follow in good
> time.
>
> If you are keen, please pull the latest sources from CVS and let me know
> what happens :)  win32com\readme.html has a brief example.
>
> Mark.
> --
> Nominate Your Favorite Programmers
> for the Perl and Python Active Awards!
> http://www.ActiveState.com/Awards/
>




From tim at 2wave.net  Wed Feb 21 18:17:03 2001
From: tim at 2wave.net (Tim Couper)
Date: Wed, 21 Feb 2001 17:17:03 -0000
Subject: An alternative approach to bound methods
In-Reply-To: <E14VExO-0004On-00@mail.python.org>
Message-ID: <NDBBKCJHKCHKEGIIANBKKENDCOAA.tim@2wave.net>

Jeremy Hylton wrote
> .... I'm not sure I really followed this argument.  Perhaps some code
> examples would help....
>

What attracted me to Marcin Kowalczy's original posting was that it appeared
to give a way of defining class functions, by merely extending the search
algorithm ("..obj.x, when obj is an instance object, first searches for 'x'
in its __dict__. If not found, it should try
obj.__class__.x ...")

Here's an example of the code I'd like to be able to run, extending a class
attributes example I was using:

class myclass:
    alist=['initial']

    def __init__(self,thelist=['the list']):
        self.__class__.alist=thelist

    def show(self):
        print self.__class__.alist

    """

    # I want to be able to declare a class function: (see unit test element
below)
    # so, as it's a class function, it doesn't need the self argument,
    # but is accessible through __class__. So I think it should take the
form
    #

    def static(args):
        #do something (only with class attributes)
        return SomeValue

    # or perhaps something like (but I think this is indistinguishable from
    # instance methods)

    def static(self,args):
        if not self:
           #do something (only with class attributes)
        else:
           # this is an instance call of this function
        return SomeValue
    """

if __name__=='__main__':

    print
    print 'Usual test'
    print '----------'
    print
    print 'string class attribute: ',myclass.astring

    a=myclass()
    print
    a.show()
    print

    b=myclass(['orig list'])
    a.show()
    b.show()

    print
    print 'Re-Assignment'
    print '-------------'

    a.__class__.alist=['new','list']

    a.show()
    b.show()

    print
    print 'Method'
    print '------'
    print

    # assuming that the static function returns something,
    # I'd like to write, in a completely analogous manner to the
    # class attribute:
    #
    # print myclass.static(args)
    # print a.__class__.static(args)
    # print b.__class__.static(args)






From stephen_purcell at yahoo.com  Sun Feb 18 12:39:10 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Sun, 18 Feb 2001 12:39:10 +0100
Subject: Remove directories from os.listdir output
In-Reply-To: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>; from spamers@must.die on Sun, Feb 18, 2001 at 12:00:35PM +0100
References: <dlav8t0gn2a0g41aaggnic4p7cldml09bj@4ax.com>
Message-ID: <20010218123910.A21374@freedom.puma-ag.com>

Fernando Rodr?guez wrote:
> Hi!
> 
> 	I need to filter out all the diretories of the list returned by
> os.listdir(). How can I tell if an entry of that list is a dir or a regular
> file? O:-)

Use os.path.isdir():

>>> import os
>>> def listdir_files_only(directory):
...     contents = os.listdir(directory)
...     return filter(lambda f,d=directory: not os.path.isdir(os.path.join(d,f)), contents)
... 
>>> listdir_files_only('/tmp')
['python-sHAOTT', 'python-sHASlX', 'python-sHATtS']


-Steve

-- 
Steve Purcell, Pythangelist
Get testing at http://pyunit.sourceforge.net/
Get servlets at http://pyserv.sourceforge.net/
"Even snakes are afraid of snakes." -- Steven Wright



From maderacv at pacbell.net  Wed Feb 14 01:28:34 2001
From: maderacv at pacbell.net (Louie Fimbrez)
Date: Wed, 14 Feb 2001 00:28:34 GMT
Subject: Introduction to Spanish Multimedia Website
Message-ID: <Siki6.7756$bG4.104342@news.pacbell.net>

Madera CineVideo with 15 years of experience would like to announce the opening of its new Website http://www.spanishmultimedia.com, one of the largest distributors of Spanish Educational Videos from Math, to History, Language Arts and Parenting.  In Entertainment we offer something for the whole family, a wide selection of both the old classics and the new releases.  Checkout our Website and let us know how you like it.  All comments are greatly appreciated.  mailto:maderacv at pacbell.net.



From roy at panix.com  Mon Feb 19 02:53:38 2001
From: roy at panix.com (Roy Smith)
Date: Sun, 18 Feb 2001 20:53:38 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <roy-8C4CC0.19021418022001@news.panix.com> <ku4rxrtqq0.fsf@lasipalatsi.fi>
Message-ID: <roy-770A13.20533818022001@news.panix.com>

Erno Kuusela <erno-news at erno.iki.fi> wrote:
> |    if (length = strlen(s))
> |       s[length-1] = 0;
> 
> | and cut your instruction count in half :-)
> 
> a sensible optimising compiler will generate identical code for both.

I will admit that my compiler theory is a bit rusty, but it's hard to see 
how a compiler would know that the strlen function would return the same 
value when called again with the same argument (idempotent?).

In any case, I just tried it on my Linux box (GNU compiler) and on a 
Solaris machine (Sun Workshop compiler).  Neither compiler generated the 
same code for the two versions, regardless of whether optimizing was turned 
on or not.  Both generated two strlen stanzas (see below), with or without 
-O.

The most interesting case was Linux with -O.  I don't know i86 assembler, 
so I'm guessing a bit here, but seems to have in-lined the strlen function, 
eliminating the subroutine call completely.  It still left what looks like 
two code fragments to count the length of the string.


From dante at oz.net  Sat Feb 17 09:34:00 2001
From: dante at oz.net (Michael Esveldt)
Date: 17 Feb 2001 08:34:00 GMT
Subject: ZWiki Web problem
Message-ID: <96ld1o$9a0$0@216.39.162.234>

I'm encountering the following problem in my ZWiki Web. I'm new to Zope 
version 2.3.0 with python 1.5.2. Any attempt to View a Wiki page 
produces the error below. I've tried reinstalling ZWiki from scratch, 
reimporting ZWiki Web and creating a brand new ZWiki Web.

Zope has encountered an error while publishing this resource.

Error Type: KeyError
Error Value: %00

<troubleshooting snipped>

Traceback (innermost last):
  File /home/zope/lib/python/ZPublisher/Publish.py, line 222, in 
publish_module
  File /home/zope/lib/python/ZPublisher/Publish.py, line 187, in publish
  File /home/zope/lib/python/Zope/__init__.py, line 221, in 
zpublisher_exception_hook
    (Object: Traversable)
  File /home/zope/lib/python/ZPublisher/Publish.py, line 171, in publish
  File /home/zope/lib/python/ZPublisher/mapply.py, line 160, in mapply
    (Object: FrontPage)
  File /home/zope/lib/python/ZPublisher/Publish.py, line 112, in 
call_object
    (Object: FrontPage)
  File /home/zope/lib/python/Products/ZWiki/ZWikiPage.py, line 210, in 
__call__
    (Object: FrontPage)
  File /home/zope/lib/python/OFS/DTMLMethod.py, line 189, in __call__
    (Object: standard_wiki_footer)
  File /home/zope/lib/python/DocumentTemplate/DT_String.py, line 538, in 
__call__
    (Object: standard_wiki_footer)
  File /home/zope/lib/python/DocumentTemplate/DT_Util.py, line 334, in 
eval
    (Object: 
bobobase_modification_time().toZone(zwiki_timezone).rfc822())
    (Info: bobobase_modification_time)
  File <string>, line 0, in ?
  File /home/zope/lib/python/DateTime/DateTime.py, line 1111, in toZone
KeyError: (see above)
-- 
dante at oz.net - Michael Esveldt - #FightThePower on Openprojects


From chrisw at nipltd.com  Wed Feb 21 18:50:51 2001
From: chrisw at nipltd.com (Chris Withers)
Date: Wed, 21 Feb 2001 17:50:51 +0000
Subject: Compiling Python 2.0 on Win32
References: <mailman.982772306.7650.python-list@python.org> <MPSk6.14824$Qb7.2368195@newsb.telia.net>
Message-ID: <3A93FFFB.E583F32D@nipltd.com>

Fredrik Lundh wrote:
> 
> Chris Withers wrote:
> > Apparently, that means I need to compile Python from
> > source on my box too.
> >
> > Where should I look to find out how to do this?
> 
> how about reading the README file?

I did look through that, but looked again.

I now have an opticians appointment :-)

*embarrassed (and probably badly spelled) grinz*

Chris



From gtalvola at nameconnector.com  Thu Feb  1 20:40:48 2001
From: gtalvola at nameconnector.com (Geoff Talvola)
Date: Thu, 01 Feb 2001 14:40:48 -0500
Subject: executing JavaScript from Python
References: <959ovu$sqd$1@nnrp1.deja.com> <63E6AD93C013B993.0AA5FD315BC99418.FEB162B59977D6FB@lp.airnews.net>
Message-ID: <3A79BBC0.8F56A389@NameConnector.com>

> In article <959ovu$sqd$1 at nnrp1.deja.com>,  <echuck at mindspring.com> wrote:
> >urllib.urlopen() can be useful for grabbing a web page and checking its
> >contents. However, JavaScript generated content throws a monkey wrench
> >in this technique.
> >
> >Has anyone cobbled together a means to invoke JavaScript from Python,
> >possibly including the DOM and the substitution of output into the HTML
> >file?

On Windows, you can use Internet Explorer via COM:

>>> import urllib
>>> print urllib.urlopen('http://localhost/javascript.html').read()
<html>
<body>
<script>
document.write('Hello, World!')
</script>
</body>
</html>
>>> from win32com.client.gencache import EnsureDispatch
>>> ie = EnsureDispatch('InternetExplorer.Application')
>>> ie.Navigate('http://localhost/javascript.html')
>>> print ie.Document.documentElement.innerHTML
<HEAD></HEAD>
<BODY>
<SCRIPT>
document.write('Hello, World!')
</SCRIPT>
Hello, World! </BODY>
>>>

--


- Geoff Talvola
  Parlance Corporation
  gtalvola at NameConnector.com




From bshapiro at funnygarbage.com  Fri Feb 16 18:04:27 2001
From: bshapiro at funnygarbage.com (Ben Shapiro)
Date: Fri, 16 Feb 2001 12:04:27 -0500
Subject: pywin and wxPython
Message-ID: <96jmi1$jst$1@bob.news.rcn.net>

Hey all...
Here's the predicament..
I have an OCX control that is responsible for playback of mp3 files that i
am trying to
use from python.  This control is wrapped using makepy.  Mark H pointed me
in the right
direction to use pywin to host the OCX..  only problem is that my app uses
wxPython as the
GUI framework.  From what i can tell you need to attach the activex control
to an MFC window for
it to load....  Any example of how to do this would be appreciated greatly!

My existing code is below:

        p = win32ui.CreateWnd()
        win32ui.EnableControlContainer()
        mp3player =
gencache.EnsureModule('{2DF09BC5-B050-11D2-81C1-00C0DFEAA961}', 0, 1, 0)
        self.olectl = activex.MakeControlInstance(mp3player.AMPL3DLite)
        self.olectl.CreateControl("", win32con.WS_TABSTOP , (7,43,500,300),
p , 131)  <---- fails here  / nothing seems to happen if i comment out this
line
        self.olectl.Open("c:\\code\\3LP\\pyfileunder\\music\\test.mp3")
        self.olectl.Play()




From cg at schlund.de  Wed Feb 14 18:34:17 2001
From: cg at schlund.de (Carsten Gaebler)
Date: Wed, 14 Feb 2001 18:34:17 +0100
Subject: Working with DNS via Python
References: <mailman.982169356.5480.python-list@python.org>
Message-ID: <3A8AC199.50A4B5BB@schlund.de>

> 
> Could someone point me to some functions or material that'd explain how
> to work with DNS functions (nslookup, etc) via Python?
> 

I was looking for such functions today, too, and found this on Parnassus:

http://www.interlink.com.au/anthony/python/dns.tar.gz

It works. There is a bug in Lib.py: index.append() takes a 2-tuple
argument, not two single arguments.

cg.


From aleaxit at yahoo.com  Wed Feb 21 00:20:10 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Wed, 21 Feb 2001 00:20:10 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <3A9008A1.6E7DD27F@pop3.qwestinternet.net> <mtofvz69nf.fsf@astron.berkeley.edu> <3A9033B7.D5ADA989@alcyone.com> <8if09tck7vf97h5nm3lm01iigkk40s27pm@4ax.com> <mailman.982538961.27262.python-list@python.org> <3A905F9F.C7D93AA8@alcyone.com> <slrn995hjv.p74.qrczak@qrnik.zagroda>
Message-ID: <96uu2f0j60@news1.newsguy.com>

"Marcin 'Qrczak' Kowalczyk" <qrczak at knm.org.pl> wrote in message
news:slrn995hjv.p74.qrczak at qrnik.zagroda...
> Sun, 18 Feb 2001 15:49:51 -0800, Erik Max Francis <max at alcyone.com> pisze:
>
> > If one is actually writing Standard C++, one doesn't need the
> > str... C functions, since one won't be using NUL-terminated C
> > strings, but rather std::string.
>
> Unless he wants to open a file with a given name, or get command
> line arguments. Not to mention string literals.

These tasks will use C-strings, but Erik's assertions about not
needing "the str... C functions" remains true -- one just builds a
std::string out of every C-string one happens to meet, and, when
a C-string is needed (e.g., in a std::fstream constructor as the
name argument), the .c_str() method of the std::string object
supplies it.  No functionality at all needs strwhatever()...


Alex





From scarblac at pino.selwerd.nl  Thu Feb  1 01:50:02 2001
From: scarblac at pino.selwerd.nl (Remco Gerlich)
Date: 1 Feb 2001 00:50:02 GMT
Subject: Design for slices
References: <959uue$k02$1@cubacola.tninet.se>
Message-ID: <slrn97hc7t.ht5.scarblac@pino.selwerd.nl>

Gustaf Liljegren <gustafl at algonet.se> wrote in comp.lang.python:
> I'm learning Python and have a question about defining ranges/slices in 
> strings. I can't find any defence for what looks like a poor design, so I 
> have to ask. Why is the first character in the string x defined as x[0] and 
> not x[1]? This looks just like the typical geek thinking that I'm trying to 
> avoid with Python.

Because experience has taught generations of programmers that this is the
right thing to do.

For a string x of length n, the following simple equations hold:
x == x[0:i]+x[i:n]   for any i (even negative i)
len(x[i:n]) == n-i   for 0 <= i <= n
x[i:i] == []         for any i

This means that it's always simple to split strings. This is also the reason
that a string slice x[i:j] *does not* include x[j]. It really simplifies
situations that would otherwise call for "-1" and "+1" everywhere.
"Off-by-one errors" have caused lots of problems in the past. The Python
choice avoids these problems for the common cases.

People often say that you have to think of the indexes as marks *between*
the actual values. Say, you have the string "Python", then think of it as

 P y t h o n
0 1 2 3 4 5 6

And now it's obvious that "Python"[2:5] == "tho", a string of length 3,
which is expected since 5-2 = 3.

There are more reasons, but you'll come to appreciate them over time.

-- 
Remco Gerlich


From phlip_cpp at my-deja.com  Sun Feb 18 20:29:56 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 18 Feb 2001 19:29:56 GMT
Subject: What's up with enscript for syntax highlighting?
References: <96p7f2$bqi@dispatch.concentric.net>
Message-ID: <96p7rk$sub@dispatch.concentric.net>

Proclaimed Phlip from the mountaintops:

> Is there a class
> library available somewhere on www.python.org (perhaps under "hilite" or
> something?

It's in ClassModules.py you dumb f**k - can't you tell by the name?

        http://www.lemburg.com/files/python/doc.py.html

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  Have a :-) day  --


From chris at voodooland.net  Tue Feb 13 07:17:54 2001
From: chris at voodooland.net (Chris Watson)
Date: Tue, 13 Feb 2001 00:17:54 -0600 (CST)
Subject: FreeBSD and Py-Kqueue
In-Reply-To: <96ahvl$lo@gap.cco.caltech.edu>
Message-ID: <Pine.BSF.4.21.0102130008240.36475-100000@open-systems.net>

Hi all,

	Has anyone had experience with using python and the python wrapper
to kqueue in FreeBSD? I am authoring a book on postfix the MTA and I have
read posts about an asynchronous syslogd on linux. Allowing for greater
speed for postfix then a normal synchronous syslogd. Well FreeBSD doesn't
have the async option. So I believe I have two choices to test to see how
much this helps, either port Theodore Tso's syslogd to FreeBSD (icky) or
write a cheap syslogd with Python to do writes async. I was wondering if
kqueue would help syslogd out at all for increasing read/write
performance? Any thoughts on this?

--
=============================================================================
-Chris Watson         (316) 326-3862 | FreeBSD Consultant, FreeBSD Geek 
Work:              scanner at jurai.net | Open Systems Inc., Wellington, Kansas
Home:  scanner at deceptively.shady.org | http://open-systems.net
=============================================================================
WINDOWS: "Where do you want to go today?"
LINUX: "Where do you want to go tomorrow?"
BSD: "Are you guys coming or what?"
=============================================================================
GNU = Gnu's Non-portable & Unstable
irc.openprojects.net #FreeBSD -Join the revolution!
ICQ: 20016186




From qrczak at knm.org.pl  Mon Feb 19 21:59:33 2001
From: qrczak at knm.org.pl (Marcin 'Qrczak' Kowalczyk)
Date: 19 Feb 2001 20:59:33 GMT
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
References: <mailman.982310784.27748.python-list@python.org> <96kbdd$j7r$1@news.udel.edu>
Message-ID: <slrn99329l.2fk.qrczak@qrnik.zagroda>

Fri, 16 Feb 2001 18:06:23 -0500, Terry Reedy <tjreedy at udel.edu> pisze:

> One of the hassles with Win9X is that Windows Explorer displays, I believe,
> some files differently from how they are stored.

And stores files differently than was asked to. For example when
you try to create a directory called Spam, it will create SPAM and
display it as Spam. And when asked to create SPAM, it will create it
but display Spam. It behaves correctly when the name has more than
8 characters or when a character outside ASCII is used.

This is Win95 - I don't know if they fixed this in later versions.

Worse: using the Explorer you can create a filename with some non-ASCII
characters (e.g. <<French quotes>> or bullet), which will be silently
converted to other characters, perhaps different in long and short
names (in these cases: '<<' and '>>' character pairs, and '\7' -
a control character which displays as bullet in DOS). Details depend
on the default codepages (the above is for CP-1250 and CP-852). This
can make the file inaccessible under either Windows or DOS or both
(because the converted character is sometimes not legal in a filename,
but the filename was checked for correctness before conversion). You
can't delete or rename such file from either Windows or DOS or both,
until you repair the filesystem. Scandisk is not always able to do it.

not-using-Windows-anymore-ly y'rs
-- 
 __("<  Marcin Kowalczyk * qrczak at knm.org.pl http://qrczak.ids.net.pl/
 \__/
  ^^                      SYGNATURA ZAST?PCZA
QRCZAK


From max at alcyone.com  Sun Feb 18 20:35:41 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 11:35:41 -0800
Subject: Determining EOF character
References: <66409tk72epolackapla0egk4907k9a9hi@4ax.com>
Message-ID: <3A90240D.384ABA3E@alcyone.com>

Daniel Klein wrote:

> When managing child processes (like with popen2.popen2) I have
> occasion to do
> 
>         instream.readlines()
> 
> to get all lines in a list rather than
> 
>         instream.readline()
> 
> which gets the next line terminated with 'os.linesep'.
> 
> On Windows the child process needs to send a ^Z in response to
> readlines() and
> on UNIX it needs to send ^D. I don't know what this is on MAC.

This isn't the way you should signal end of file.  You should signal it
the client closing his end of the pipe.  In fact, what you suggest above
may work on Windows, but it won't work on UNIX -- under UNIX, an end of
file isn't indicated with a sentinel character in the stream, but rather
with an out-of-bounds indicator (the underlying file structure has an
end-of-file flag).  As I recall, Windows/DOS does much the same thing,
but also treats a ^Z as an explicit end-of-file indicator.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ I just don't know why they're shooting at us.
\__/ Capt. Benjamin "Hawkeye" Pierce
    Esperanto reference / http://mirror/alcyone/max/lang/esperanto/
 An Esperanto reference for English speakers.


From Kai.Grossjohann at CS.Uni-Dortmund.DE  Sat Feb 17 17:53:08 2001
From: Kai.Grossjohann at CS.Uni-Dortmund.DE (Kai =?iso-8859-1?q?Gro=DFjohann?=)
Date: 17 Feb 2001 17:53:08 +0100
Subject: Python emacs mode (newbie)
References: <96hu7d$4pq$1@troll.powertech.no> <vafwvaqltd4.fsf@lucy.cs.uni-dortmund.de> <96kmdc$47q$1@troll.powertech.no>
Message-ID: <vafofw1i74b.fsf@lucy.cs.uni-dortmund.de>

On Sat, 17 Feb 2001, Syver Enstad wrote:

> You mean to open modules by parsing import statements in a file?
> Ex.
> module.py
> 
> import sys
> import MyClass
> 
> ......  And then you could place the cursor over let's say MyClass
> and press some key combination to load the py file for the
> module. Is that what you mean?
> 
> This would be cool as an added bonus, but it's the find-file like
> functionality that is crucial.

Oh, I'm sorry.  With M-x find-file-at-point RET, you could type it on
the `import MyClass' statement above and then you'll get a default
filename prompt of /some/path/here/MyClass.py.

But ffap.el does not provide for a facility where you can enter a
module name and then it opens the right file.

Hm.

But it would be trivial to do: C-x b foo RET, M-x python-mode RET,
type in the module name, type M-x find-file-at-point RET, voil?!

kai
-- 
Be indiscrete.  Do it continuously.


From wmiller at mediaone.net  Wed Feb 21 23:48:35 2001
From: wmiller at mediaone.net (Walter Miller)
Date: Wed, 21 Feb 2001 14:48:35 -0800
Subject: How to get device context for com object?
References: <NVxk6.15$uj7.15036@news.pacbell.net> <96ujcc12r6k@news1.newsguy.com> <YQBk6.80$3e2.62458@news.pacbell.net> <96us1u0dj8@news1.newsguy.com> <XuDk6.96953$Wq1.40024851@nnrp5-w.sbc.net> <9701q102nt4@news1.newsguy.com>
Message-ID: <rBXk6.268$FG.70985@news.pacbell.net>

That's great!  Yes, it works but I'm also trying to call
CreateCompatibleDC() and it only works with the device context object not
the (int) handle for the device context.  Is there an api for creating a
device context object based on its own handle?

The ultimate plan is to try an replicate the visual basic example of copying
the screenshot of the client portion of the window and saving it as a bitmap
file.  I have an activex control that does this already but that's too
easy - I'm _actually_ trying to learn all the details of using Python with
win32.

> import win32gui
> hdc = win32gui.GetDC(hwnd)
>
> will set hdc to the (int) handle for the device context
> corresponding to (the client areas of) the window whose
> handle is in (int) hwnd.




From othello at javanet.com  Sun Feb 25 07:43:53 2001
From: othello at javanet.com (Raymond Hettinger)
Date: Sun, 25 Feb 2001 01:43:53 -0500
Subject: data parsing
References: <mailman.982950184.14469.python-list@python.org>
Message-ID: <3A98A9A9.8789B127@javanet.com>

Gnanasekaran Thoppae wrote:

> I have some data in a file 'test', which contains:
>
> Joe|25|30|49|40|
> |28|39|71||
> |30|29|||
> Malcolm|43|60|56||
> |28|37|||
> Amy||70|45||
> |40|30||40
> |40||30||
>
> I want to parse this data and format it in this way:
>
> Joe|25;28;30|30;39;29|49;71|40|
> Malcolm|43;28|60;37|56||
> Amy|40;40|70;30|45;;30|;40;|
>

name = 'unnamed'        # just in-case the file doesn't start with a name
result = {}
for line in inp.split():
    field = line.split('|')
    while len(field) < 6:   # fix misformatted line |40|40||40
        field.append('')
    k = field.pop(0)        # grab name field
    field.pop()             # clear final blank
    if k != '':             # on name change
        name = k
    if not result.has_key(name):
        result[name] = []
    result[name].append(field)

for name in result.keys():
    ans = apply( zip, result[name] )  # tranpose matrix
    ans = map( ';'.join, ans )
    ans.insert(0,name)
    ans.append('')
    print '|'.join(ans)

Raymond



From ycheng at sinica.edu.tw  Fri Feb 23 09:01:22 2001
From: ycheng at sinica.edu.tw (Y. Cheng)
Date: Fri, 23 Feb 2001 08:01:22 +0000 (UTC)
Subject: How python output a unicode string.
Message-ID: <9755ci$1t4m$1@news1.sinica.edu.tw>

Hi,

I try to install a big5.py on /usr/lib/python2.0/encodings
and it seem working fine. Then I try to output an string
which is u"\u9DFD" to stdout by using print. Before that,
I call

	sys.setdefaultencoding("big5")

without any error message.

As I try to output the string, I got error message:

-----------------------
  File "/usr/lib/python2.0/encodings/big5.py", line 18, in encode
    return codecs.charmap_encode(input,errors,encoding_map)
TypeError: character mapping must be in range(256)
-----------------------

is this means that python can't output multi-byte character
with unicode internal encoding ?

Yuan-Chen Cheng


From spamers at must.die  Sun Feb 18 12:21:07 2001
From: spamers at must.die (Fernando Rodríguez)
Date: Sun, 18 Feb 2001 12:21:07 +0100
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net>
Message-ID: <q4bv8t8vc494qq2mqnqkpsu9l162q9rv31@4ax.com>

On 18 Feb 2001 08:00:11 GMT, Jim Eaton
<eatonalive%qwestinternet.net at pop3.qwestinternet.net> wrote:

>I am learning Python as a first language, and I have been wondering what
>is a good language to learn after Python?  I'm thinking of either going
>into C or Java but I'm not sure which one because I've heard many
>arguments either way.  Will going into Java first be any detriment to
>learning C later?  Any suggestions would be greatly appreciated.

A good combination is, IMHO: Python, CommonLisp, C++ and Prolog.  From a
didactic point of view, good second languages are CommonLisp (a more powerful
Python-like language) and Java (an easier to learn C/C++ language).

Remember that there's much more to programming than just learning the syntax
of different languages. If you're new to programming, you should first learn
some basic concepts.  Good beginers books are: "Simply Scheme: Introducing
computer science" and "The Schemer's Guide".  Both use a lisp dialect called
Scheme, that should be easy to undertand (sort of Python with extra
parentheses).





//-----------------------------------------------
//	Fernando Rodriguez Romero
//
//	frr at mindless dot com
//------------------------------------------------


From 1126-810 at online.de  Sun Feb 18 10:18:08 2001
From: 1126-810 at online.de (Stefan Witzgall)
Date: Sun, 18 Feb 2001 10:18:08 +0100
Subject: play_animals and MacPython
Message-ID: <1126-810-F3D9B6.10180818022001@news.online.de>

Hello,

I tried to run the sample program "play_animals" found in "Python for 
the unexperienced". Now I get an error "unslicecable object str"...
Ok, str is None, therefore I can't get some part of it.

Has anyone run this on his machine with python, is this a MacPython 
specific prob, how to get around?

Thank you for any hints.

Bye, Stefan


From lreilly at cs.strath.ac.uk  Tue Feb 20 17:20:03 2001
From: lreilly at cs.strath.ac.uk (Lee Reilly CS1997)
Date: Tue, 20 Feb 2001 16:20:03 +0000
Subject: retrieving file data from an external server - where to start?
Message-ID: <3A929933.AA4F81@cs.strath.ac.uk>

hi there,

i'm using python with zope and am trying to do the following with python
method:

retrieveFileInfo(matric){

  username = getUsername(matric)
  filename = username + ".gif"
  server = "http://blahblahblah/yaddieyaddie/"

  # some code to determine whether or not the file exists
  # where do i start
  # e.g if (retrieve.(server+filename) != none):
  #        do something
}

e.g. given the matric value '9728430' i find that the corresponding user
has username == 'lreilly'. if this person has their image stored on the
external server then it will be called 'lreilly.gif'.

i don't need to actually retrieve the file from python - only determine
whether or not it exists or not. i.e. whether the http request reports a
404 or maybe even if the image size == 0 bytes.

could someone possible be kind enough to give me an idea where to start?
a url? a snippet of code?

thanks very much in advance,

lee


From Norman_Shelley-RRDN60 at email.sps.mot.com  Thu Feb 22 17:47:14 2001
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Thu, 22 Feb 2001 09:47:14 -0700
Subject: nice articles (Re: The price of fame...)
References: <mailman.982834675.5030.python-list@python.org> <87vgq358bu.fsf@joyful.com>
Message-ID: <3A954292.FEED9271@email.sps.mot.com>

Simon Michael wrote:

> I found eg "Text Processing in Python: Tips for Beginners" a very
> useful overview. Thanks!
>
> If I may, an attempt to summarize for my fellow newbies:
>
>   "Charming Python" columns can be found at, eg
>   http://gnosis.cx/publish/tech_index.html .
>   To read, click the title (not the eye logo).
>
> Best regards

Funny, I still get this when I click on the above URL

Not Found
The requested URL /publish/tech_index.html was not found on this server.

Additionally, a 404 Not Found error was encountered while trying to use
an ErrorDocument to handle the request.



From doug at postsmart.net  Sat Feb  3 18:30:14 2001
From: doug at postsmart.net (Doug Ball)
Date: Sat, 3 Feb 2001 11:30:14 -0600
Subject: Python/C++ for graphics vs. other languages?
References: <mailman.980225485.12027.python-list@python.org> <200101231452.IAA15405@starbase.neosoft.com> <mailman.981180244.14710.python-list@python.org> <63bo7t4qrf0oebbj6hnni9tjjdip8hg5u8@4ax.com>
Message-ID: <000501c08e06$f83a6b60$8a00000a@darwin.net>

les schaffer wrote in response to a prior posting:


> >I am working on a real-time 3d simulation of a medical procedure, to run
on
> >an Athlon 1.2 GHz machine under windows 2000.
>
> this sounds interesting. are you at liberty to divulge more details of
> the project?


Thanks for your interest!

We will be working on an interactive simulation of
electromyographically-guided Botulinum toxin injection.  In this procedure,
the practitioner places a needle into a muscle, generally guided by surface
landmarks, then more specifically guided by the electrical signals produced
when the muscle contracts (summated compound action potentials).  Once the
needle is placed, Botulinum toxin is injected, subsequently blocking the
chemical signals than are responsible for controlling muscle contraction (it
blocks the presynaptic vesicle from binding to the presynaptic membrane).
This procedure is typically used for patients with severe muscle spasticity.

The simulation user will have visual and auditory feedback from a 3D model
of the patient and from the EMG-machine monitor.  The user will indicate the
initial location of needle placement, then will request that the patient
model activates the target muscle. This action will result in a simulated
signal on the EMG-monitor (with both visual and auditory components) that
aids the user in further localizing the muscle, while adjusting  the needle
for optimal placement within the muscle.

Our goal is to develop this application for physician education.

Thanks again.

Doug





From annis at biostat.wisc.edu  Thu Feb 15 16:17:16 2001
From: annis at biostat.wisc.edu (William Annis)
Date: 15 Feb 2001 09:17:16 -0600
Subject: Gauging system load
References: <E2Qi6.237705$f36.9512658@news20.bellglobal.com>
Message-ID: <jtsnlgt1qb.fsf@wazor.biostat.wisc.edu>

Bryan Mongeau <bryan at eevolved.com> writes:

> Out of curiosity, does anyone know of a module or a neat trick that will 
> permit me to gauge the system load at a given time?

        What OS?  I have no idea how you'd do this on a PC or a Mac,
but on any unix flavor you should be able to parse the output from
uptime.  And if you're running Solaris, you can grab my PyKstat
package and rip the information right out of the kernel with no
parsing at all.

        http://www.biostat.wisc.edu/~annis/creations/pykstat.html

Under Linux, you can parse the contents of /proc/loadavg.

-- 
William Annis - System Administrator - Biomedical Computing Group
annis at biostat.wisc.edu                       PGP ID:1024/FBF64031
Mi parolas Esperanton - La Internacian Lingvon  www.esperanto.org


From rogersd at bigfoot.com  Wed Feb 14 05:31:58 2001
From: rogersd at bigfoot.com (Dave Rogers)
Date: Tue, 13 Feb 2001 23:31:58 -0500
Subject: Python/ASP   Docs/Questions
Message-ID: <3A8A0A3E.A1EF40E9@bigfoot.com>

I'm a newcomer to this group, but I'm working on a fairly large and exciting web site using Win2K/Python/ASP (win32 135/Python2.0)

I cannot find any docs regarding the COM Scripting object that it uses.

Can anyone (Mark?) refer me to proper docs for it? - I've checked ActiveState and the net in general with no luck.

My two main problems are:

1) I would like for the COM Scripting object to not cache modules for me as I develop my code.  I see the following conceivable solutions:

a) There is some COM method I can call that will turn it off
b) There is some process I can kill or restart (I tried WWW Publishing service) that will knock everything out of memory
c) I can do manual reloads everywhere which are not desirable because I have some open source code that does "from foo import foo" where I cannot reload (foo) and I don't want to mess with the open source code because then I mess up the open source project
d) Reboot the PC

2) I want to set cookies easily using Response.  I would like to go Response.Cookies('name') = 'bob' (for a simple cookie), but instead I am forced to go

import Cookie
c = Cookie.Cookie()
...
Response.AddHeader("Set-Cookie", c[12:])
(or something similar to that)

I know Cookies is a collection - I seem to have this problem setting ASP Object values in general as I have to go Session.SetValue to set session variables.

I really appreciate any help people can give me - these seem like fairly common issues, yet beyond the newbie level.

P.S. One of the neatest things I've seen is Chuck Esterbrook's MiddleKit which was just released under Webware (http://webware.sourceforge.net).  What it does is allow you to map your python objects to a database for design and run time access.  What this means is that you can relate your objects in a 1 to 1 or 1 to many relationship and the MiddleKit will service all of your SQL needs - you do NOT write SQL or use recordsets.  Once you acquire one or more objects from the database, you can follow their object references to other database objects.  In fact I have abandoned ADO in favor of this middle layer which resides in between the database and my web site.  I probably didn't do it justice but you'll just have to check it out.

Thanks,
--
Dave Rogers



From jfontain at winealley.com  Fri Feb 23 12:37:57 2001
From: jfontain at winealley.com (Jean-Luc Fontaine)
Date: Fri, 23 Feb 2001 12:37:57 +0100
Subject: ANNOUNCE: tclpython-1.0
References: <3A958C9A.130C7838@free.fr> <m1NVERAE5il6Ew$D@jessikat.fsnet.co.uk>
Message-ID: <975i2e$pa8$1@s1.read.news.oleane.net>

Robin Becker wrote:

---
How does this compare with the primitive python for tcl package I wrote
long ago?

http://www.jessikat.demon.co.uk/pyserver.zip

The main deficiency of my approach was that it was difficult to get
reverse communication ie callbacks from the python side and that
everything was done via sockets.
---

I did not know that existed: I will check it out, thank you.

The difference is that this is using an embedded python interpreter 
(several I hope in a future version), so communication is done through the 
C APIs.

-- 
Jean-Luc Fontaine mailto:jfontain at winealley.com http://www.winealley.com


From Jerry.Spicklemire at IFLYATA.COM  Wed Feb  7 17:43:21 2001
From: Jerry.Spicklemire at IFLYATA.COM (Spicklemire, Jerry)
Date: Wed, 7 Feb 2001 11:43:21 -0500 
Subject: Red Hat Linux and poplib.py
Message-ID: <977A39E65CFCD3119ABF00D0B741D849E0FEB3@innt-73.ata.com>

Hi Python fans,

I'm hoping someone can point me in the right direction. 
When I try to use a contributed Zope module that calls 
poplib.py, Python v.2.5.2, I get a couple of error messages. 
The first hangs on:

>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 132, in
> ListMessages
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 68, in
> UpdateStatus
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 42, in 
> Connect
>     (Object: POPMailAccountBase)
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 183, in
> pass_
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 146, in
> _shortcmd
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 125, in
> _getresp
> error_proto: (see above)

with an error message of:

> > Error Type: error_proto
> > Error Value: -ERR Not implemented

This makes me think that poplib.py is trying to interact 
with some PopMail feature that isn't implemented in the 
Red Hat v.5.2 distribution.

My guess is that the second error is the result of the first, 
but it really is just a guess. The error is:

>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 132, in
> ListMessages
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 68, in
> UpdateStatus
>     (Object: POPMailAccountBase)
>   File /l01/Zope/lib/python/Products/POPMailBase/POP.py, 
>   line 42, in 
> Connect
>     (Object: POPMailAccountBase)
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 183, in
> pass_
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 146, in
> _shortcmd
>   File /l01/Zope-2.2.1-linux2-x86/lib/python1.5/poplib.py, 
>   line 125, in
> _getresp
> error_proto: (see above)

which is identical to the one above, but with the message:

> > Error Type: error_proto
> > Error Value: -ERR Mailbox in use

It seems like the popmail server isn't updated when the 
client process drops, and keeps the session open. When that 
happens, I can't even connect to the user mailbox from
Eudora, or even directly from the Python prompt.

BTW, I was able to connect to the mailbox via the Python 
prompt once, after the client process was killed manually.
I used some of the examples in the Beazley book, and 
elswhere, to invoke poplib.py, sign on, etc.

I searched the Web for references to Red Hat and PopMail 
problems, and found a short E-Mail thread about this. 
 
 http://plug.skylab.org/200009/msg00187.html

 http://plug.skylab.org/200009/msg00307.html

The recommended solution was to use another PopMail server, 
such as CUCIPOP. Since I'm not the system admin, I don't 
have much access to the host, and such notions aren't often 
warmly received. It was far from painless just to get Python 
and Zope up and running, though we do have one Linux guru on 
staff.

Any ideas would be greatly appreciated!

Thanks,
Jerry S.




From jacobkm at cats.ucsc.edu  Wed Feb 21 19:27:58 2001
From: jacobkm at cats.ucsc.edu (Jacob Kaplan-Moss)
Date: Wed, 21 Feb 2001 10:27:58 -0800
Subject: Rounding Question
References: <jacobkm-07145C.22285820022001@news.ucsc.edu> <96vudj02hnf@news1.newsguy.com>
Message-ID: <jacobkm-4E8467.10275821022001@news.ucsc.edu>

In article <96vudj02hnf at news1.newsguy.com>, "Alex Martelli" 
<aleaxit at yahoo.com> wrote:

[snip a whole bunch of great ideas...]

Thanks so much for the great discussion; although this is really a 
trivial part of what I'm doing, it's nice to be able to do it right...

In the end, I decided that:

> def nextHigherOrEqualMultiple(depth, N):
>     "compute the least multiple of N >= depth"
>     return int((depth+N-1)/N) * N
> 
> with
> 
> tableDepth = nextHigherOrEqualMultiple(depth, 10)

is what I'm going to use; I really only need to make this calculation 
once, so a long function name really isn't a problem, and although I'm 
writing for NAUI tables now, I know that certain other tables use 
increments of 5 feet, so it makes sense to have a generic function 
available.

Thanks once again,

Jacob


From Pekka.Pessi at nokia.com  Mon Feb  5 02:32:34 2001
From: Pekka.Pessi at nokia.com (Pekka Pessi)
Date: 05 Feb 2001 03:32:34 +0200
Subject: . Python 2.1 function attributes
References: <Pine.GSO.4.21.0101271158190.3501-100000@y.glue.umd.edu>, <Pine.GSO.4.21.0101270135190.7677-100000@y.glue.umd.edu> <mailman.980593892.725.python-list@python.org> <$5A+ZwAGjtc6EwE0@jessikat.fsnet.co.uk> <14962.62953.304511.523826@anthem.wooz.org> <mailman.980613639.23439.python-list@python.org> <mailman.980617472.28020.python-list@python.org> <Pine.GSO.4.21.0101271303540.3501-100000@y.glue.umd.edu>
Message-ID: <pv3ddt9aod.fsf@agni.research.nokia.com>

In message <Pine.GSO.4.21.0101271303540.3501-100000 at y.glue.umd.edu> Roy Katz <katz at Glue.umd.edu> writes:
>Let's add bless().  I want to change the class of an object at
>runtime. How 'bout it?  

        How about this:

class sm:
    "State machine accepting ab*"
    def __init__(self):
	self.__class__ = a

    valid = 0

class a(sm):

    def a(self):
	self.__class__ = b

class b(sm):

    def b(self):
	pass

    valid = 1

        BTW, have you never looked in the source code of pickle?
       
                                        Pekka


From fredrik at pythonware.com  Thu Feb 15 08:49:31 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 15 Feb 2001 07:49:31 GMT
Subject: Getting Value From Tkinter Text Object
References: <96faar$30s$1@ins21.netins.net>
Message-ID: <fSLi6.17199$AH6.2346379@newsc.telia.net>

Scott Billings wrote:
> The only option I can find that even seems to be close to what I want, is
> the get() method, which expects a starting and ending point. Starting point
> should be easy enough, but I have no way of knowing what the ending point
> will be.

http://www.pythonware.com/library/tkinter/introduction/text.htm
=> concepts, patterns

    "END (or "end") corresponds to the position just after the last
    character in the buffer.

    "To fetch the text contents of the widget, use the get method:

        contents = text.get(1.0, END)

Cheers /F




From clee at gnwy100.wuh.wustl.edu  Wed Feb 21 20:27:45 2001
From: clee at gnwy100.wuh.wustl.edu (Christopher Lee)
Date: 21 Feb 2001 13:27:45 -0600
Subject: locking files
References: <96v06a$l16$1@solaria.cc.gatech.edu> <970kg2$cn0$1@solaria.cc.gatech.edu>
Message-ID: <7ulmqzu98u.fsf@gnwy100.wuh.wustl.edu>

Joseph Holland King <insanc at cc.gatech.edu> wrote:
: is there anyway to lock a file so that two different programs cannot access
: the file at the same time. ie, my python module is editting file foo, and
: at the same time user x (or program x) tries to read from it. is there 
: anyway that the python module can prevent x from accessing the file for
: a given amount of time?

: to clarify: i am on a unix system, i have tried to use flock however when i
: locked the file i was still able to read, write and copy the file, none of 
: which i want to happen while the file is locked. while using the flock 
: method i was using LOCK_EX. also the third party program is not going to be
: controlled, written, or have anything to do with mine. am i just not using
: flock properly or is there something else that might do the job? thank you.

flock() is an advisory lock, if you don't have cooperation from the other
processes, and if your system supports the SysV interface you can enable
mandatory locks.  "man lockf" or "man fcntl" and search for information on SysV
mandatory locking on your system.

Files are marked as candidates for mandatory file locking by setting the
file mode with SGID but without the "group execute" bits, (chmod g+s; chmod
g-x).  Enabling mandatory locking is system dependent, but you need to use
fnctl/lockf, flock is always advisory.

Under linux, I can do it like this:
[as root]
# enable file locking by remounting /home file system w/ mand flag
mount -o remount -o mand /home   

[as normal user]
> chmod g+s tmp; chmod g-x tmp
> python
>> import fcntl, FCNTL
>> fp = open("tmp", "rw")
>> fd = fp.fileno()
>> fcntl.lockf(fd, FCNTL.F_LOCK)  # secures the lock, prevents
                                  # reading/writing 




From rwklee at home.com  Wed Feb 21 03:23:26 2001
From: rwklee at home.com (Rick Lee)
Date: Wed, 21 Feb 2001 02:23:26 GMT
Subject: Better Threading, starting with Events
References: <jRwk6.270$TW.2706@tor-nn1.netcom.ca>
Message-ID: <3A93269A.8AFFC7BB@home.com>

I like the idea of being able to wait for several events at once.   That's
something I can't do with queues in the Queue module.  So my wish list would
extend to queues as well.  Another thing on my wish list is to be able to join
to several threads at once.  And then finally, to be able to select from any
combination of events, queues, joins, and maybe anything else that blocks.

Maybe you also want to keep on eye on compatibility with Stackless.

- Rick Lee

Warren Postma wrote:

> Here is my little 'event' module __doc__ string:
>
> event module 1.0
>
>   Provides an efficient way for threads to go to sleep while waiting for
>   work to do, freeing up the python interpreter to run the other threads,
>   and yet waking up immediately upon signalling the worker thread
>
>   Uses CreateEvent/SetEvent/WaitForMultipleObjects Win32 functions.
>
>   Usage Sample:
>
>      import event
>      sd = event.new('Shutdown')
>      sd.MyAttribute = 1 # stores something in event dictionary
>      work = event.new('DoSomeWork',1) # 'one-shot'
>      work .work_to_do = func # pass the function to the thread
>      ret = event.select( (sd,work), 5000 ) # wait 5 seconds for signal
>      if ret:  # Got signal?
>          print repr(ret)
>
> ------------------------------------------------------------
>                 Here are the module methods:
> ------------------------------------------------------------
>
> event.new( 'name', manualResetFlag=1, initialStateFlag=0)
>    -> Creates and returns a new event object
>         manualResetFlag = does event reset automatically when received
>         (default=1)
>         initialStateFlag = event signalled or not signalled initially
>         (default=0)
>
>  e = event.new( 'ASampleEvent')
>
>  Using a WinEvent:
>         e.name       name of event <string-attribute>
>         e.wait()     wait indefinitely (until python shutdown) for event
>         e.wait(500)  wait up to 500 milliseconds (0.5 seconds) for event
>         e.state()    query state (non blocking) returns 0 or 1
>         e.set()      set signalled state
>         e.reset()    reset to non-signalled state
>         e.handle()   get internal handle value (as integer)
>
> To wait for several events in parallel, use event.select( (e1,e2,..),
> timeout )
>
> ------------------------------------------------------------
>
> event.select( <event tuple>, timeout=None)
>    -> Waits for a single event or any of a group of events,
>     returns None if the timeout occurs.
>       Note that all waits are interruptable by shutdown,
>       which will cause an exception to be raised.
>
>   Example:
>      ev1 = event.new('DoSomething')
>      ev2 = event.new('DoSomethingElse')
>      ret=event.select( (ev1,ev2), 5000 ) # wait up to 5000 millisec.
>      if ret:
>           print 'got event', ret.name  # which is it? DoSomething, or
> DoSomethingElse
>      else: # none?
>           print 'timeout'
>
> --------------------------- End ----------------------------
>
> Now, for a discussion on my Evil Plan:
>
> Anyone interested in native C Type for Event Signalling? I have written a
> 1.0, or 0.1, that works, it just ain't portable off of Win32. I'll make it
> portable if there is some agreement on the basic idea behind what I'm doing.
>
> I'd like feedback on whether my Evil Plan calls for a PEP.  Before I go
> PEP'ing, I'd like to get some initial comments on other people who have
> implemented worker threads. Is anyone else dissatisfied with threading.py?
>
> My first problem with the existing facilities of threading.py is its
> existence. I don't like that the Primitives are not Primitive, the interface
> is great, the implementation is in Python. Urgh!  Great for Jython/JPython.
> Not so great for CPython guys like me.
>
> Why should Locks, Events, and Signals and Semaphores and so on all live in a
> py file?  Their efficiency is therefore sub-par, and it will negatively
> affect the performance of all apps that use it.  I also feel that a
> consistent API and underlying implementation ought to permeate downwards
> into the core of CPython.  Anyone with me? At that stage, you can see my
> Evil Plan is beginning to get into swing.
>
> But back to reality for a minute. The first thing I needed was extremely
> efficient Event Objects, which could make a thread be "sleepy", waking up
> periodically every 5 seconds to do some housekeeping, then going back to
> sleep, but to be instantly interruptable at any time and immediately start
> doing whatever work is required. They let go of the python interpreter lock,
> and don't make the interpreter do any work, running functions in
> threading.py while the thread ought to be idle, etcetera.
>
> Also, my eventual evil plan involves thread-by-thread signalling in ways not
> currently supported by the current frame objects, and the code in ceval.c.
> It should also involve making various blocking functions "interruptable".
> For example, time.sleep, which calls the C function floatsleep in
> timemodule.c:
>       Py_BEGIN_ALLOW_THREADS
>       Sleep((unsigned long)millisecs);
>       Py_END_ALLOW_THREADS
>
> This should be, IMHO:
>       Py_BEGIN_ALLOW_THREADS
>       ret=WaitForSingleObject( hSystemShutdown, INIFINITE);
>       Py_END_ALLOW_THREADS
>       if (ret != WAIT_TIMEOUT)
>             return NULL; // system is shutting down.
>
> I am basically doing my R&D here anyways to get threading maximally
> efficient on my embedded systems.  But I am more than willing, excited even,
> to try my hand at proposing a PEP.   Not for my little event module, but for
> a whole bunch of threading primitives, in C. Perhaps the CPython version of
> threading.py could be adapted to use them.
>
> The PEP would probably be for an expanded  "thread" module, and a revised
> threading system that improves "pervasive threading" inside CPython.
>
> Anyone? Anyone? Bueller?
>
> Warren



From tjreedy at udel.edu  Tue Feb 20 06:39:08 2001
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 20 Feb 2001 00:39:08 -0500
Subject: Windows/Cygwin/MacOSX import (was RE: python-dev summary, 2001-02-01 - 2001-02-15)
References: <mailman.982384399.26499.python-list@python.org>
Message-ID: <96sve1$69e$1@news.udel.edu>

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.982384399.26499.python-list at python.org...
> [Terry Reedy]
> > One of the hassles with Win9X is that Windows Explorer displays,
> > I believe, some files differently from how they are stored.
>
> This depends on the setting of Explorer's
>
>     View -> Folder Options ... -> View -> Allow all uppercase names
>
> option.  It that's on, file names are shown as stored.  But it's off by
> default.  Then an ALLCAPS name is shown as Allcaps (i.e., first letter
> uppercase, all the rest lowercase), while other names are shown as
stored.

By coincidence, I discovered this 'feature' a couple of days ago, just
after posting this, in the process of turning on the display of hidden
files.

> > So I have sometimes had trouble opening files.
>
> I don't understand that, as open() is case-insensitive on Windows.

I do not believe that this has always been my experience, but the problem I
had was a couple of years ago, so I will wait until I reproduce the problem
before saying more.




From gem at hsv.crc.com  Fri Feb 23 18:55:36 2001
From: gem at hsv.crc.com (Glen Mettler)
Date: Fri, 23 Feb 2001 11:55:36 -0600
Subject: Python with AMS Real Time Project
Message-ID: <9767p9$ear$1@hobbes2.crc.com>

Is anyone using Python with AMS Real Time Project?
I have just acquired both and I am wondering about the ease and practicality
of manipulating RTS from Python.

Glen





From mjackson at wc.eso.mc.xerox.com  Thu Feb 15 21:29:26 2001
From: mjackson at wc.eso.mc.xerox.com (Mark Jackson)
Date: 15 Feb 2001 20:29:26 GMT
Subject: python-dev summary 2001-02-01 - 2001-02-15
References: <mailman.982256267.9885.python-list@python.org>
Message-ID: <96he76$f68$1@news.wrc.xerox.com>

Michael Hudson <mwh21 at cam.ac.uk> writes:

>  This is the first python-dev summary written by Michael Hudson.

Thank you!

>  was proposed.  Discussion went round and round for a while and moved
>  on to more general iteration constructs, prompting Ka-Ping Yee to
>  write a PEP entitled "iterators":
> 
>   <http://python.sourceforge.net/peps/pep-0234.html>
> 
>  Please comment!

With the possible exception of ":spam" and "eggs:" it doesn't look like
it would make code less readable, and the functionality seems good.
Mildly in favor.

>    * Python's release schedule *
> 
>  Skip Montanaro raised some concerns about Python's accelerated
>  release schedule, and it was pointed out that the default Python for
>  both debian unstable and Redhat 7.1 beta was still 1.5.2.  Have
>  *you* upgraded to Python 2.0?  If not, why not?

Ran into difficulties building on Solaris and didn't have the time to
resolve them.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Having your book made into a movie is like having
	your ox made into a bouillon cube.
				- Bill Neely




From binary at eton.powernet.co.uk  Sun Feb 18 18:42:35 2001
From: binary at eton.powernet.co.uk (Richard Heathfield)
Date: Sun, 18 Feb 2001 17:42:35 +0000
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <3A90098B.5EBB3993@eton.powernet.co.uk>

Dave Cross wrote:
> 
> [Please watch the replies on this message as it's heavily
> cross-posted]

Noted.

> 
> I'm doing some comparisons on programming language features and I'd be
> very interested to know how you would handle the following scenarios
> in your programming language of choice.
> 
> 1/ The programmer calls a function that doesn't actually exist within
> the application (or libraries).

In ISO conforming C, he can't. End of story.

> 2/ Can ou filter the input source code before compilation (or
> interpretation) in some way so that language keywords could be changed
> for other strings?

Again in ISO conforming C...

#define ALIAS original

e.g.

#define voiture auto /* :-) */


-- 
Richard Heathfield
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
K&R answers, C books, etc: http://users.powernet.co.uk/eton


From man_at_moon at hotmail.com  Sat Feb 17 08:51:01 2001
From: man_at_moon at hotmail.com (David A.)
Date: Sat, 17 Feb 2001 07:51:01 -0000
Subject: Yet Another Newbie Question (YANQ)
Message-ID: <Xlsj6.95$DC1.647178@newsserver.ip.pt>

Hi to everyone,

    I am playing around with python and I have 1 quetions where I hope
someone can help

I have this problem:

>>> a='a'
>>> prt='print'
>>> t=prt+' '+a
>>> t
'print a'

now when I try to:
>>> eval(t)
Traceback (innermost last):
  File "<pyshell#18>", line 1, in ?
    eval(t)
  File "<string>", line 1
    print a
        ^
SyntaxError: invalid syntax

can someone explain me what I am doing wrong! Why can't I use eval(print)?
Any help is welcome!

David Asfaha
--
"Quidquid latine dictum sit, altum viditur."
Whatever is said in Latin sounds profound.










From knotwell at ix.netcom.com  Thu Feb  1 04:50:07 2001
From: knotwell at ix.netcom.com (Brad Knotwell)
Date: 31 Jan 2001 19:50:07 -0800
Subject: question on threading.py
Message-ID: <86k87brrio.fsf@localhost.my.domain>

Hello all--

Do the _Condition methods notify and wait really need to call 
currentThread?  Unless currentThread has some side-effect I couldn't see
I think these methods don't have any reason to call them.

Lines 182 and 215 of v2.0 threading.py.

--Brad




From max at alcyone.com  Sun Feb 18 11:13:42 2001
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 18 Feb 2001 02:13:42 -0800
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96o0su0mv6@news2.newsguy.com>
Message-ID: <3A8FA056.F7E767B4@alcyone.com>

Alex Martelli wrote:

> There are several possibilities, but I would suggest SQL as the second
> language.  It's incredibly useful AND when used well it urges you to
> think in very different ways from a sequential imperative language
> (such as Python, C, Java, ...) -- thus at the same time broadening
> your horizons AND providing you with LOTS of real-world returns.
> 
> Similarly, for a THIRD language I would suggest some structured markup
> language such as XML (with XPath, XSLT...) or HTML (4.0, with its
> DOM).

I think he meant programming languages, not database query languages or
markup languages.

Java or C++ would probably be up his alley, depending on exactly what he
wanted to do with that further knowledge.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Human love is often but the encounter of two weaknesses.
\__/ Francois Mauriac
    Interstelen / http://www.interstelen.com/
 A multiplayer, strategic, turn-based Web game on an interstellar scale.


From Randy.L.Kemp at motorola.com  Thu Feb  1 18:00:03 2001
From: Randy.L.Kemp at motorola.com (Kemp Randy-W18971)
Date: Thu, 1 Feb 2001 11:00:03 -0600 
Subject: Perl-to-Python converter/translator?
Message-ID: <69BC4C6AE83ED311BC9400805FA7B14AF84DD6@il93exp01.css.mot.com>

Have we converted another to the fray?

-----Original Message-----
From: Gerhard H?ring [mailto:gerhard.nospam at bigfoot.de]
Sent: Wednesday, January 31, 2001 2:48 PM
To: python-list at python.org
Subject: Re: Perl-to-Python converter/translator?


Lance Sloan wrote:
> 
> I'm an experienced Perl programmer who has just been assigned
> a project that must be done in Python.  I'm looking for a
> Perl-to-Python translator to help me out.  I don't expect to
> write the whole project in Perl and translate it, though.  I'm
> just looking for some automation to help me see how my old work
> could be done in Python.
> 
> I didn't find anything useful via Google or Vaults of Parnassus
> searches.  So, I would appreciate any pointers!

Maybe the Perl/Python phrasebook at
http://starship.python.net/~da/jak/cookbook.html is helpful to you.
Intra-document links don't work. Does anybody know a better version of
this document?

Gerhard
-- 
Sorry for the fake email, please use the real one below to reply.
contact: g e r h a r d @ b i g f o o t . d e
web:     http://highqualdev.com





From joconnor at cybermesa.com  Tue Feb  6 22:41:35 2001
From: joconnor at cybermesa.com (Jay O'Connor)
Date: Tue, 06 Feb 2001 21:41:35 GMT
Subject: x = y
References: <JNZf6.2765$O31.16647@newsr01.ono.com>
Message-ID: <3a806eed.5209768@news.cybermesa.com>

On Tue, 06 Feb 2001 21:13:13 GMT, "Jos? Luis Gallego"
<gallegoj at ono.com> wrote:

>Hi all,
>
>I want to create a list x from a list y:
>
>y = [[1,2,3],[4,5,6]]
>x = y
>
>How do I change individual elements of x without changing y?


Make a copy of the list

	x  = list (y)

x will contain the same elements as y, but will be a seperate list, so
you can update either indepently

>>> y = [[1,2,3],[4,5,6]]
>>> x = list (y)
>>> x
[[1, 2, 3], [4, 5, 6]]
>>> y
[[1, 2, 3], [4, 5, 6]]
>>> x [0] = "hi there"
>>> x
['hi there', [4, 5, 6]]
>>> y
[[1, 2, 3], [4, 5, 6]]
>>> 

Take care,


Jay O'Connor
joconnor at cybermesa.com
http://www.cybermesa.com/~joconnor

Python Language Discussion Forum - http://pub1.ezboard.com/fobjectorienteddevelopmentpython

"God himself plays on the bass strings first, when he tunes the soul"


From tim.one at home.com  Wed Feb 28 02:29:07 2001
From: tim.one at home.com (Tim Peters)
Date: Tue, 27 Feb 2001 20:29:07 -0500
Subject: PEP 236:  Back to the __future__
In-Reply-To: <20010227200249.A21243@harmony.cs.rit.edu>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEIPJCAA.tim.one@home.com>

[D-Man]
> ...
> However, I would like to see the use of particular future_statements
> deprecated or even disallowed after the feature becomes mandatory.

It's deliberately not done that way, so that if someone who moves their code
around a lot happens to move to an earlier release, the (earlier) compiler
will complain about a __future__ request it can't satsify.  This is up to
each programmer to deal with in whichever way best fits their usage.  I
expect many users will ignore future_statements entirely, and simply wait for
the next release before fiddling their code.  That's fine too.

Note that there's more than enough info in __future__.py to allow writing a
tool in Python that can reliably search for, identify, and even delete (if
you like) obsolete future_statements wrt to any fixed Python release.  That
was also deliberate.

> As for the future statement being easy to find :
>     grep -r "__future__" my_source_tree/*
> will do the job ;-).

The syntax restrictions listed in the PEP also ensure that a future_statement
begins with "from" starting in the first column (they CANNOT be nested in if
or try blocks, etc).

everything-about-them-is-simple-except-the-implementation<wink>-ly y'rs
    - tim




From jkraska1 at san.rr.com  Sat Feb 24 06:25:20 2001
From: jkraska1 at san.rr.com (Courageous)
Date: Sat, 24 Feb 2001 05:25:20 GMT
Subject: Python training in Finland?
References: <slrn994vil.vjq.ohyvarin@sirppi.helsinki.fi>
Message-ID: <olhe9tk22q9u1b5l7srrg2tcqumje1kb7r@4ax.com>

On 20 Feb 2001 14:25:26 GMT, ohyvarin at cc.helsinki.fi (Otto T Hyvarinen) wrote:

>We starting to use more Python in our department.  So we are looking for
>someone to teach us something :-)

I suggest that you use the inability of a software person to
learn Python without training as a good indicator that you
need to let them go. Anyone who can't pick up Python on
their own is beyond redemption.

`;--)

C//



From jhauser at ifm.uni-kiel.de  Fri Feb  2 14:43:23 2001
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 02 Feb 2001 14:43:23 +0100
Subject: Diamond x Jungle Carpet Python
References: <mailman.981100230.20774.python-list@python.org> <3A7AB6C7.386D14D@engcorp.com>
Message-ID: <877l399p50.fsf@lisboa.ifm.uni-kiel.de>

Peter Hansen <peter at engcorp.com> writes:

> Rod wrote:
> > 
> > I have several Diamond x Jungle Capret Pythons for SALE.
> > 
> > Make me an offer....
> > 
> > Go to: www.qnet20.com
> 
> Which platform?  Version 1.5.2 or earlier?  Doesn't
> ActiveState already make these available as a free
> download provided you don't need to redistribute them?
> 
> Jeez, why don't people ever provide enough detail in 
> their posts?!
> 
> :-)

What? You do not take this serious? I have just ordered two of them
for the sides of my home entrance. And Rod has given me a written
guarantee that these babies will not move and only roll their eyes, as
long as the the outside temperature is below zero degree (hope he
meant celsius). So I think I'm safe for the winter :-)

__Janko

-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany


From phlip_cpp at my-deja.com  Wed Feb 14 06:15:14 2001
From: phlip_cpp at my-deja.com (Phlip)
Date: 14 Feb 2001 05:15:14 GMT
Subject: while loop with f.readline()
References: <BD674CDC0A68C642B60038986FBB56DF015C35@mailsrv.ASPATECH.COM.BR> <mailman.982112534.5154.python-list@python.org>
Message-ID: <96d492$88c@dispatch.concentric.net>

Proclaimed Chris Gonnerman from the mountaintops:

> I have always been a fan of the FileReader class:
> 
> class FileReader:
>     def __init__(self, file):
>         self.file = file
>     def next(self):
>         self.line = self.file.readline()
>         return self.line
> 
> if __name__ == '__main__':
> 
>     f = FileReader(open("c:/autoexec.bat"))
> 
>     while f.next():
>          print f.line,

Came to the thread late can't read all posts don't know if this was covered 
so sue me but...

...what's wrong with this?

        for z in f.readlines():
                print z

That's how I done it since I was a chile'. 

-- 
  Phlip                          phlip_cpp at my-deja.com
============ http://c2.com/cgi/wiki?PhlIp ============
  --  http://www.deja.com/my/pb.xp?member_name=phlip_cpp  --


From rich_somerfield at tertio.com  Mon Feb 26 10:11:59 2001
From: rich_somerfield at tertio.com (Rich Somerfield)
Date: Mon, 26 Feb 2001 09:11:59 -0000
Subject: Help with finding how much memory a variable is using
References: <JAwl6.17037$5n4.338201@news6-win.server.ntlworld.com> <3A986565.BAF4E05B@student.gu.edu.au>
Message-ID: <B5pm6.28048$MN.708724@news2-win.server.ntlworld.com>

<s713221 at student.gu.edu.au> wrote in message
news:3A986565.BAF4E05B at student.gu.edu.au...
> Rich Somerfield wrote:
> >
> > Hi,
> >
> > I am generating a huge, huge list of data (cant think of a decent way to
> > reduce the required storage size and have it in a usable form).
Everytime i
> > try to keep this list for future derivations I get a memory problem from
> > Windows (effectively terminating my python script).
> >
> > I presume this is because of the huge list.  Is it possible to find out
the
> > amount of memory a variable [not a type of the variable, the actual data
> > contained within the variable] is taking up?  This variable is local to
a
> > class and it would appear that it is when i go out of scope of that
class
> > that the memory error is being issued.  Would I be better making this
> > variable global (across all classes)? e.g. would this problem still
happen
> > if the variable was not just local to 1 class?
> >
> > Would a different python datatype solve this problem for me ?
> >
> > If I cant handle this amount of data then I will have to think of a way
of
> > better representing it, this will be unfortunate though as and
compression /
> > reduction of data will [as far as i can think] vastly reduce
performance -
> > which is already a concern.
> >
> > Regards
> > Rich
>
> Do you have to load all the list into memory, or is there a way you can
> load in parts and work on that sequentially. I haven't had a chance to
> use persistance modules yet, but their documentation may be a place to
> start.
>
> Joal Heagney/AncientHart

Unfortunately I need all of the data in memory at the same time.  The only
way I can see to solve this problem is to reduce the memory consumption of
my application.  And the only way I can see how to do that is by vastly
reducing the performance.

Do u know of any information that suggests how large a datatype can be in
Python (are there any upper limits?) ?

Regards
Rich





From shaleh at valinux.com  Thu Feb 22 21:39:52 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Thu, 22 Feb 2001 12:39:52 -0800 (PST)
Subject: executing shell commands
In-Reply-To: <R%dl6.77662$Nw6.251029@NewsReader>
Message-ID: <XFMail.20010222123952.shaleh@valinux.com>

On 22-Feb-2001 chris lamb wrote:
> I am very new to Python and having to learn quickly! I am enjoying it,
> however. It is a flexible and intuitive environment.
> 
> Can someone advise me if there is an equivalent to running a system process
> from inside a script as there is in Perl? I mean in perl if I use backticks
> to enclose a system command, then the command is executed. Is there a
> comparable way of running these commands from Python?
> 

import os
os.system() or os.popen()

system will run the command, but you do not get anything back other than an
exit status.  popen() can launch and app and connect its output to a Python
file handler and let you read from it.



From grante at visi.com  Sat Feb 17 17:21:32 2001
From: grante at visi.com (Grant Edwards)
Date: Sat, 17 Feb 2001 16:21:32 GMT
Subject: ok, so how do I return a sorted list without doing it in place
References: <XFMail.20010216160410.shaleh@valinux.com> <mailman.982369099.4358.python-list@python.org> <96koqk$gv6$1@panix2.panix.com>
Message-ID: <slrn98t98c.g7f.grante@isis.visi.com>

In article <96koqk$gv6$1 at panix2.panix.com>, Aahz Maruch wrote:

>>> I know I can do:
>>> 
>>> list = dict.keys()
>>> list.sort()
>>> .....
>>> 
>>> but it is just plain annoying.
>>
>>I agree. This has bitten me on the butt more than once. It also seems
>>somewhat unPythonic. Usually I expect what happens in Python, but that 
>>sort process always strikes me as unexpected.

On many occasions, I would find it quite handy if there was a sequence
method that returned a sorted shallow copy:

 for k in dict.keys().sorted():
   whatever(k) 

>It's unexpected in some respects, but doing it this way causes the
>fewest surprises overall.  At least you rarely get a downstream bug
>this way.

In my mine the used of the adjective "sorted" as a method name conveys the
fact that it doesn't modify the list in place they way the imperative "sort"
does.  YMMV.

-- 
Grant Edwards                   grante             Yow!  World War III? No
                                  at               thanks!
                               visi.com            


From phd at phd.pp.ru  Fri Feb  9 18:05:21 2001
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 9 Feb 2001 20:05:21 +0300 (MSK)
Subject: I love Python
In-Reply-To: <981736559.1767376098@news.t-online.de>
Message-ID: <Pine.LNX.4.30.0102092004270.27625-100000@fep132.fep.ru>

Pedro Vale Lima <eq3pvl at eq.uc.pt> wrote:
> > I love Python too ;-)
> Yeah, but have you asked Python if he loves you?

   It doesn't matter until *I* am satisfied! :)

Oleg.
----
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From tanzer at swing.co.at  Fri Feb 23 08:04:56 2001
From: tanzer at swing.co.at (Christian Tanzer)
Date: Fri, 23 Feb 2001 08:04:56 +0100
Subject: Nested scopes resolution -- you can breathe again! 
In-Reply-To: Your message of "Thu, 22 Feb 2001 21:59:26 EST."
             <200102230259.VAA19238@cj20424-a.reston1.va.home.com> 
Message-ID: <m14WCHg-000wcEC@swing.co.at>

Guido van Rossum <guido at digicool.com> wrote:

> We (PythonLabs) have received a lot of flak over our plan to introduce
> nested scopes despite the fact that it appears to break a small but
> significant amount of working code.

There were lots of complaints about newly introduced Python features
recently. While I understand concerns about code breakage (myself being
worried about breakage to my code), far more complaints concerned
style issues than breakage.

As I normally don't want to waste time over religious issues -- like
`print >>' being cool or the worst wart ever -- I don't participate in
such discussions on c.l.py.

But IMO Python still is an excellent language and it got even better in the
last year. Thank you and your team for the great work you are doing.

> How do you select nested scopes?  Tim suggested a mechanism that is
> used by the ANSI C committee to enable language features that are
> backwards incompatible: they trigger on the import of a specific
> previously non-existant header file.  (E.g. after #include
> <imaginary.h>, "imaginary" becomes a reserved word.)
> 
> The Python equivalent of this is a magical import that is recognized
> by the compiler; this was also proposed by David Scherer for making
> integer division yield a float.  (See
> http://mail.python.org/pipermail/edu-sig/2000-May/000499.html)  You
> could say that Perl's "use" statement is similar.
> 
> We haven't decided yet which magical import; two proposals are:
> 
>     import __nested_scopes__
>     from __future__ import nested_scopes

Cool.

I hope that `from __future__ import <cool-feature>' will win this
contest -- it defines a common pattern for introducing backwards
incompatible features. 

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92




From andi at opencan.cc  Sun Feb 11 00:09:09 2001
From: andi at opencan.cc (Andreas Poisel)
Date: 10 Feb 2001 23:09:09 GMT
Subject: Vim and python efficiently
References: <mailman.981809351.20580.python-list@python.org> <964ag4$c4d$1@news.tuwien.ac.at>
Message-ID: <964hml$f36$1@news.tuwien.ac.at>

Andreas Poisel <andi at opencan.cc> wrote:
[...]

Sorry for the tabs...
-- 
Andi


From kevin at GamesandTheory.com  Fri Feb 23 16:05:49 2001
From: kevin at GamesandTheory.com (Kevin Douglas)
Date: Fri, 23 Feb 2001 10:05:49 -0500
Subject: Suffering For Your Art
In-Reply-To: <t9cse0s7mra90f@corp.supernews.com>
Message-ID: <DKEEJHGMIIOMNIJFOCAIEEFKDGAA.kevin@GamesandTheory.com>

> > Is it just me or is there a lot of animosity towards Python as
> > a serious programming language ...
>
> When I first started looking at Python a few years back I had to
> get past the mental block that Python was a "scripting" language
> and was therefore inherently less powerful than the more
> traditional "compiled" languages (e.g. C/C++ or Fortran).

This is the first time I have ever posted to this list, in over
a year of being subscribed.

I've built and/or worked on systems that are scalable to the level 
of billions of transactions per month with Python and C and ASM.

The premise is simple - Python is an AUTOMATION and PROTOTYPING
language.  You use it to automate either administrative tasks,
management tasks, monitoring and quality assurance tasks,
data aggregation or analysis tasks, or - more powerfully -
programming tasks (self replicating code, code management, etc).

It's quite easy to build an advertising exchange and collaborative
filtering system such as DoubleClick has (even better than theirs,
actually), or BlueFly or any other such company - in about one
to three months - then work for the next 6 to 12 months on filling
in the details, and native C functionality and inline ASM where
needed.  Such a system usually lasts for 3 to 4 years without
needing much change (from industry experience, including one of
the examples I just mentioned *cough cough*).

I've spoken with executives and venture capitalist that are
responsible for several million dollars worth of technology
expenditures - who re completely confident in the capacity
of Python for Enterprise Level developments of any size - ERP,
HIRS, CRM, Financial Systems, Accounting Systems, Billing
Systems, Trading Systems, data Mining/Warehousing Systems, you
name it.

I was about to build a Python system for Deutsche Bank, the
world's largest bank, with North American, European, and
Asia-Pacific operations, multilingual, HRIS, graphical charts
and analytics, database, security systems, the whole nine
(yes, thin-client Java Swing graphics with distributed load
balancing on the client end, plus ORACLE and encrypted
one-way anonymous data repository for survey data and aggregate
analysis).

Of course, we would never tell the general public that it was
written in Python - that's silly.  Of course, we would never
tell the general public if we were running Java either -
because that is equally as silly.  We run "heterogeneous
systems" - and that's all anyone needs to know - for Risk
Management purposes.

Unfortunately, people generally tend to suck.  So that project
never completed because someone unnamed wasn't willing to wake up
everyday and be their own boss.

And Python coders like myself are starving artists because of
such silly and trivial things.  Mostly because of "grown men"
being considerably more immature than most of my peers.

On an interesting side note, I've noticed that more money seems
to be spent on Welfare Administration than on the actual welfare
itself, given some of the rules and regulations that these
agencies claim are legal and strictly adhere to.  I was turned
down for Food Stamps even though I only have water left right
now - it's interesting how a CTO level programmer gets no respect
because he focused on cost effective realistic solutions instead
of Multi-Level Sales and Marketing tactics like the ones that
destroyed the US Economy.

EMC just announced they'd make $1B less than they thought -
because - yes - all those con artists who were selling "solutions"
based more on merchant partnerships than real solutions - ripped
everyone off and no one is left!

My 2 cents worth.  If I had 2 cents anyway.

Always available for consulting or freelance work or start-up
prototyping, resume available if anyone really cares.


Kevin Douglas
President / Chief Technology Officer
Games and Theory Supercomputing, Inc.






From william_wicker at spectratechnologies.com  Fri Feb 16 19:35:12 2001
From: william_wicker at spectratechnologies.com (William Wicker)
Date: Fri, 16 Feb 2001 18:35:12 -0000
Subject: Win32 COM: Passing NULL as a value
Message-ID: <904A72E79williamwickerspectra@207.126.101.100>

I am trying to work via the COM interface to an object that exposes a 
method something like this:

AnObject.DoSomething(doWhat, toWho)

where toWho is either a dispatch pointer, or NULL, if toWho is not 
significant.

When I try to do this with

obj = win32com.client.Dispatch("AnObject")
obj.DoSomething("kick", None)

or

obj.DoSomething("kick", 0)

I get type mismatch errors back from the COM object.

Is there a way to pass a NULL pointer via win32 COM?

P.S.

I'm trying this with the ActiveState Python 2.0 distribution (python win 
build 202)

Thanks!

    	William.


From fakeaddress at nospammail.com  Wed Feb  7 00:57:08 2001
From: fakeaddress at nospammail.com (Ben Catanzariti)
Date: Wed, 7 Feb 2001 10:57:08 +1100
Subject: Python Users Group Australia
Message-ID: <t43q59.pfd.ln@mail>

Hi,

I am looking for the existence of a Python Users Group in Australia ... does
one exist??

If so how do I get in contact with them:

Elif ... are there other parties interested in forming one?

thanks

Ben




From rickp at telocity.com  Sun Feb 25 23:43:11 2001
From: rickp at telocity.com (Rick Pasotto)
Date: Sun, 25 Feb 2001 22:43:11 GMT
Subject: Tkinter/Pmw callback question
References: <3A997794.66A9D4AB@NoNOSpamtheworld.com>
Message-ID: <slrn99j2ju.qh3.rickp@tc.niof.net>

On Sun, 25 Feb 2001 21:24:11 GMT in comp.lang.python, David Lees wrote:
> I am trying to learn Tkinter and Pmw and am modifying demo Pmw code
> (EntryField.py).  I want to call a Python function that I have defined
> when I push a button and am having problems with the callback.  As
> long as the function has no arguments all is well, but I am unable to
> do a callback on a function that has arguments.  For example this
> works:
> 
>     RunButton = Tkinter.Button(root, text = 'Run',command=foo)
> 
> where foo has no arguments, but when I write:
> 
>     RunButton = Tkinter.Button(root, text = 'Run',command=barf('junk'))
> 
> things do not work correctly.  The 'barf' function seems to execute
> when the program starts up, but not when I click the Run button.
> 
> What goes on here and what is the correct way to do call backs?

I'm just learning myself, but I think what I'm going to say is correct.

'command=' takes the *name* of a function. Adding parens changes the
*name* to an actual function call. The way to get around this is to use
the nameless name 'lambda' for the name of the function and *it* can
call the function you really want. So:

   rb = Button(root, text='Run', command='lambda x="junk":barf(x)')

-- 
"Every great advance in natural knowledge has involved
the absolute rejection of authority."
		-- Thomas Henry Huxley
		   Rick Pasotto email: rickp at telocity.com


From embed at geocities.com  Tue Feb 20 17:21:45 2001
From: embed at geocities.com (Warren Postma)
Date: Tue, 20 Feb 2001 11:21:45 -0500
Subject: Better Threading, starting with Events
Message-ID: <jRwk6.270$TW.2706@tor-nn1.netcom.ca>

Here is my little 'event' module __doc__ string:

event module 1.0

  Provides an efficient way for threads to go to sleep while waiting for
  work to do, freeing up the python interpreter to run the other threads,
  and yet waking up immediately upon signalling the worker thread

  Uses CreateEvent/SetEvent/WaitForMultipleObjects Win32 functions.

  Usage Sample:

     import event
     sd = event.new('Shutdown')
     sd.MyAttribute = 1 # stores something in event dictionary
     work = event.new('DoSomeWork',1) # 'one-shot'
     work .work_to_do = func # pass the function to the thread
     ret = event.select( (sd,work), 5000 ) # wait 5 seconds for signal
     if ret:  # Got signal?
         print repr(ret)

------------------------------------------------------------
                Here are the module methods:
------------------------------------------------------------

event.new( 'name', manualResetFlag=1, initialStateFlag=0)
   -> Creates and returns a new event object
        manualResetFlag = does event reset automatically when received
        (default=1)
        initialStateFlag = event signalled or not signalled initially
        (default=0)

 e = event.new( 'ASampleEvent')

 Using a WinEvent:
        e.name       name of event <string-attribute>
        e.wait()     wait indefinitely (until python shutdown) for event
        e.wait(500)  wait up to 500 milliseconds (0.5 seconds) for event
        e.state()    query state (non blocking) returns 0 or 1
        e.set()      set signalled state
        e.reset()    reset to non-signalled state
        e.handle()   get internal handle value (as integer)

To wait for several events in parallel, use event.select( (e1,e2,..),
timeout )

------------------------------------------------------------

event.select( <event tuple>, timeout=None)
   -> Waits for a single event or any of a group of events,
    returns None if the timeout occurs.
      Note that all waits are interruptable by shutdown,
      which will cause an exception to be raised.

  Example:
     ev1 = event.new('DoSomething')
     ev2 = event.new('DoSomethingElse')
     ret=event.select( (ev1,ev2), 5000 ) # wait up to 5000 millisec.
     if ret:
          print 'got event', ret.name  # which is it? DoSomething, or
DoSomethingElse
     else: # none?
          print 'timeout'


--------------------------- End ----------------------------

Now, for a discussion on my Evil Plan:

Anyone interested in native C Type for Event Signalling? I have written a
1.0, or 0.1, that works, it just ain't portable off of Win32. I'll make it
portable if there is some agreement on the basic idea behind what I'm doing.

I'd like feedback on whether my Evil Plan calls for a PEP.  Before I go
PEP'ing, I'd like to get some initial comments on other people who have
implemented worker threads. Is anyone else dissatisfied with threading.py?

My first problem with the existing facilities of threading.py is its
existence. I don't like that the Primitives are not Primitive, the interface
is great, the implementation is in Python. Urgh!  Great for Jython/JPython.
Not so great for CPython guys like me.

Why should Locks, Events, and Signals and Semaphores and so on all live in a
py file?  Their efficiency is therefore sub-par, and it will negatively
affect the performance of all apps that use it.  I also feel that a
consistent API and underlying implementation ought to permeate downwards
into the core of CPython.  Anyone with me? At that stage, you can see my
Evil Plan is beginning to get into swing.

But back to reality for a minute. The first thing I needed was extremely
efficient Event Objects, which could make a thread be "sleepy", waking up
periodically every 5 seconds to do some housekeeping, then going back to
sleep, but to be instantly interruptable at any time and immediately start
doing whatever work is required. They let go of the python interpreter lock,
and don't make the interpreter do any work, running functions in
threading.py while the thread ought to be idle, etcetera.

Also, my eventual evil plan involves thread-by-thread signalling in ways not
currently supported by the current frame objects, and the code in ceval.c.
It should also involve making various blocking functions "interruptable".
For example, time.sleep, which calls the C function floatsleep in
timemodule.c:
      Py_BEGIN_ALLOW_THREADS
      Sleep((unsigned long)millisecs);
      Py_END_ALLOW_THREADS

This should be, IMHO:
      Py_BEGIN_ALLOW_THREADS
      ret=WaitForSingleObject( hSystemShutdown, INIFINITE);
      Py_END_ALLOW_THREADS
      if (ret != WAIT_TIMEOUT)
            return NULL; // system is shutting down.

I am basically doing my R&D here anyways to get threading maximally
efficient on my embedded systems.  But I am more than willing, excited even,
to try my hand at proposing a PEP.   Not for my little event module, but for
a whole bunch of threading primitives, in C. Perhaps the CPython version of
threading.py could be adapted to use them.

The PEP would probably be for an expanded  "thread" module, and a revised
threading system that improves "pervasive threading" inside CPython.

Anyone? Anyone? Bueller?

Warren




From gzeljko at sezampro.yu  Wed Feb  7 02:31:50 2001
From: gzeljko at sezampro.yu (gzeljko)
Date: Wed, 7 Feb 2001 02:31:50 +0100
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com> <3a806c08.26417386@news.skynet.be>
Message-ID: <95q8l6$htb$1@neptun.beotel.net>

I agree,
but this construct can't go in Python
(what about assigment in this context ?)

Zeljko

Daniel <Daniel.Kinnaer at Advalvas.be> wrote in message
news:3a806c08.26417386 at news.skynet.be...
>
> The part where I find the 'with' command interesting is for
> readability.  Consider :
>
> MySpecialButton := TSpecialButton;
> MySpecialButton.caption := 'Busy';
> MySpecialButton.method2;
> MySpecialButton.caption := "Ok";
>
> or
>
> MySpecialButton := TSpecialButton;
> with MySpecialButton do
> begin
>    caption := 'Busy';
>    method2;
>    caption := 'Done';
> end;
>
> Which code would you prefer?
>





From peter at engcorp.com  Fri Feb  2 14:29:56 2001
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 02 Feb 2001 08:29:56 -0500
Subject: import statement within a function
References: <95dmp6$n4s$1@news.tpi.pl> <1Yte6.11069$AH6.1772843@newsc.telia.net> <95dr93$dcs$1@news.tpi.pl> <rgve6.11079$AH6.1774088@newsc.telia.net> <95e9c2$rmm$1@news.tpi.pl>
Message-ID: <3A7AB654.D6C16746@engcorp.com>

Tomasz Lisowski wrote:
> 
> "Fredrik Lundh" <fredrik at effbot.org>:
> > Tomasz Lisowski wrote:
> > > for the first time is minimal. What are then advantages of importing a
> > > module WITHIN a function.
> >
> > - lazy importing: if you place it within a function, the module
> > is only imported if your program really needs it.
> >
> > - import dependencies (see the recursive import section in the
> > mini-guide).  if you use from-import on the module level, you
> > can end up in situations where the things you import doesn't
> > yet exist.
> >
> > - performance: local lookup is faster than global lookup.
> 
>  Yes, that's really a good point!
> 
> > - readability: it can be easier to grok your code if you import
> > things (i.e. define names) near the place you're using them.
> 
> Thanks, Frederik, that is really a good explanation! BTW, I really
> appreciate your "Introduction to Tkinter". I am using it alot.

The one you highlight, "performance", should probably be
considered the _least_ important of the points Fredrik raised.

The last one, which implies improved maintainability, is
far more important (IMHO), and that alone is a good reason
to put import within a function instead of 'without'
(provided the imported module is not needed throughout
the importing module of course).


From aleaxit at yahoo.com  Sat Feb 24 10:25:27 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Sat, 24 Feb 2001 10:25:27 +0100
Subject: Need to unread or push back bytes to a file
References: <9771ks0ic2@news2.newsguy.com>
Message-ID: <977ukk0uvd@news1.newsguy.com>

"Noah Spurrier" <noah(a)noah.org> wrote in message
news:9771ks0ic2 at news2.newsguy.com...
>
>
> I need to push bytes back into a file object.
> Is there a buffered file wrapper for file objects?
> (Yes, I looked, but I gave up after almost five minutes of searching.)
>
> I'm reading from a pipe, but this could be the same for a file.
> I call an "expect" method which reads until a pattern is found,
> but read() with a pipe will read as much data as there is available
> in the pipe. When I have found a pattern match I want to return
> the stream with the extra character pushed-back.

what about something like:

class NSWrapper:
    def __init__(self, fileob):
        self.fileob = fileob
        self.buffer = ''
    def close(self):
        self.fileob.close()
        self.buffer = ''
    def push_back(self, piece):
        self.buffer = piece + self.buffer
    def readall(self):
        result = self.buffer + self.fileob.read()
        self.buffer = ''
        return result
    def read(self, N=None):
        if N is None: return self.readall()
        avail = len(self.buffer)
        if N>avail:
            result = self.buffer+file.read(N-avail)
            self.buffer = ''
        else:
            result = self.buffer[:N]
            self.buffer = self.buffer[N:]
        return result

needs testing, but this should be roughly what you want...?


Alex





From jurgen.defurne at philips.com  Thu Feb  8 16:23:55 2001
From: jurgen.defurne at philips.com (jurgen.defurne at philips.com)
Date: Thu, 8 Feb 2001 16:23:55 +0100
Subject: Memory leaks or other things ?
Message-ID: <0056900015777418000002L082*@MHS>

Hello, list,

I am having the following problem under NT, with Python 2.0 for NT.

For some time I have written a cron.pl project (yes, Perl) with Cygwin, which
runs fine using fork and exec(). Now, to have easier installation of base programs and
script modules, I would like to everything in Python. I had already (re)written much in
Python, but cron.py was the last.

I implemented cron.py using spawn. The complete application is one script.
The main routine syncs on 00 seconds, and then spawns a subprocess, which
reads and processes crontab. For every process that should be run, the script is
respawned with the appropriate arguments, all using Python 2.0 for NT.

The jobs which I have to run allocate a fair amount of memory (160 Mb). The funny
thing is that the jobs which are still running under cron.pl do not leak memory, but the
same job (different arguments) under cron.py does not give its memory back, although
I can't find any processes anymore in the Task Manager.

I am retesting now, with one modification : using sys.exit(0) to terminate the spawned
scripts properly.

Has anyone an idea ? You have the code here. Review at your own will.

Jurgen

#

# External modules
import re
import os
import sys
import string

from time import *
from BitVector import BitVector

# Global variables
year = month = day = hour = min = sec = None
crontab     = 'c:\\etc\\crontab2'
logfile     = 'c:\\var\\log\\cron2.log';
interpreter = 'd:\\python20\\python'
script      = 'cron.py'
shell       = 'c:\\winnt\\system32\\cmd.exe'
fh          = None
initialised = None

# Class definitions
class JobEntry:
    def __init__(self, line = None):
	if line:
	    M, H, d, m, w, c = string.split(line, sep = None, maxsplit = 5)
	    self.mins = self.BitInit(M, 60)
	    self.hour = self.BitInit(H, 24)
	    self.days = self.BitInit(d, 31)
	    self.mnth = self.BitInit(m, 12)
	    self.dows = self.BitInit(w,  7)
	    self.cmnd = string.strip(c)
	else:
	    self.mins = BitVector(60)
	    self.hour = Bitvector(24)
	    self.days = BitVector(31)
	    self.mnth = BitVector(12)
	    self.dows = BitVector(7)
	    self.cmnd = None
    
    def BitInit(self, value, length):
	rv = BitVector(length)

	if value == '*':
	    for i in range(length):
		rv[i] = 1
	else:
	    args = string.split(value, ',')
	    for arg in args:
		if string.find(arg, '-') != -1:
		    start, end = string.split(arg, '-')
		    for i in range(eval(start), eval(end) + 1):
			rv[i] = 1
		else:
		    rv[eval(arg)] = 1
	
	return rv

# Subroutine definitions

def logmsg(msg):
    fh = open(logfile, 'a')
    systime = ctime(time())
    fh.write('%s -- %s\n' % (systime, msg))
    fh.close()

"""
Running a job using spawn is not as easy as using fork and exec. In the normal
Unix a parent process can wait upon the completion of a subprocess, under
Windows this doesn't work. To provide support for this, the script will be
spawned to execute run_job(), and run_job will provide the necessary services
for controlling the job. The job itself will be run synchronously.
"""

def run_job(argv):
    job          = string.join(argv[2:])

    print job

    # Log activity before job is run
    logmsg('Starting : %s' % job)

    # Run job
    os.system(job)

    # Log activity after job has finished
    logmsg('Job %s finished' % job)

    # Exit system in a good fashion
    sys.exit(0)

def read_crontab():
    job_list = []
    fh       = open(crontab, 'r')

    line = fh.readline()
    while line:
	if re.search('^#',   line):
	    line = fh.readline()
	    continue
	if re.search('^ *$', line):
	    line = fh.readline()
	    continue

	job_list.append(JobEntry(line))
	line = fh.readline()

    fh.close()
    return job_list

def process_jobs(list):
    year, mth, day, hrs, min, sec, dow, None, None = localtime(time())

    year  -= 1 
    mth   -= 1
    day   -= 1
    dow   -= 1

    for entry in list:

	if entry.mins[min] == 0: continue
	if entry.hour[hrs] == 0: continue
	if entry.days[day] == 0: continue
	if entry.mnth[mth] == 0: continue
	if entry.dows[dow] == 0: continue
	
	os.spawnl(os.P_NOWAIT, interpreter, 'python', script, 'run', entry.cmnd)

def check_jobs():
    # Lees inhoud van crontab in tabel
    # Scan tabel voor jobs die moeten worden uitgevoerd
    #	Per job :
    #	 	Spawn Python20, python, script, 'run', job en argumenten
    job_list = read_crontab()
    process_jobs(job_list)

    # Exit this script in an orderly fashion
    sys.exit(0)

def start_check_jobs():
    os.spawnl(os.P_NOWAIT, interpreter, 'python', script, 'check')

def sync():
    global initialised
    
    (None, None, None,
     None, None, sec,
     None, None, None) = localtime(time())

    initial = 60 - sec

    if not initialised:
	if initial == 60: initial = 0
	initialised = 1

    sleep(initial)

def main(argv):
    if len(argv) == 1:
	logmsg('cron.py starting up')
	sync()
	while 1:
	    start_check_jobs()
	    sync()
    elif argv[1] == 'check':
	check_jobs()
    elif argv[1] == 'run':
	run_job(argv)

if __name__ == '__main__':
    main(sys.argv)


From max at alcyone.com  Mon Feb  5 18:09:48 2001
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 05 Feb 2001 09:09:48 -0800
Subject: Please translate this easy snip of C++ to Python
References: <95kcbo$k8s@dispatch.concentric.net> <BTif6.12144$AH6.1880632@newsc.telia.net> <3A7E3C04.9EB02932@engcorp.com> <R2Af6.175419$ge4.60894995@news2.rdc2.tx.home.com> <3A7ECEAB.DB6F8648@cybermesa.com>
Message-ID: <3A7EDE5C.5D3F7050@alcyone.com>

Jay O'Connor wrote:

> Rainer Deyke wrote:
> 
> > Better make that 'except None'.  Unqualified 'except' clauses are
> > generally
> > a bad idea.  In this case, you really don't want to
> > catch...'ComputerExplodingError'.
> 
> I hadn't seen that exception raised yet....is it new?  What do you
> usually do to handle it?

Right past you.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ It is much safer to obey than to rule.
\__/ Thomas a Kempis
    7 sisters productions / http://www.7sisters.com/
 Web design for the future.


From sholden at holdenweb.com  Sun Feb 11 18:20:32 2001
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 11 Feb 2001 12:20:32 -0500
Subject: How to label loops?
References: <981845104.329821352@news.t-online.de>
Message-ID: <jVzh6.3840$n4.28565@e420r-atl2.usenetserver.com>

First, convince me that loops need labelling!

regards
 Steve
--
--
Tools, training and technology to help you meet your information needs



"Zamurai" <zamurai at gmx.net> wrote in message
news:981845104.329821352 at news.t-online.de...




From wware at world.std.com  Wed Feb  7 20:19:41 2001
From: wware at world.std.com (Will Ware)
Date: Wed, 7 Feb 2001 19:19:41 GMT
Subject: embed Tkinter in C
References: <95s3m1$db2$1@hecate.umd.edu>
Message-ID: <G8EJ0t.83J@world.std.com>

Arvind Mani (marvind at glue.umd.edu) wrote:
> I am planning to build a user interface with Tkinter and integrate it with
> an existing C program.
> Can someone tell me where to find links to Tkinter C API and a tutorial

Tkinter is a Python wrapper for the Tk gui toolkit, which is written in
C and runs on Unix and Windows (and the Mac as well, I believe). If your
program is written in C, you might want to connect directly to the Tk
toolkit. Here's some info on Tk's C API:
http://www.bbso.njit.edu/Documentations/TclTkMan/tk4.2b1/tk4.2b1API.html

A possibly better strategy might be to create a Python wrapper for your
C code, and then write Python code that connects your C code to Tkinter.
This will probably be easier and more maintainable. Some resources that
would be useful for this are the Extending-and-Embedding tutorial:
http://www.python.org/doc/current/ext/ext.html
and Swig, an automatic generator for wrapper code:
http://www.swig.org/
http://www.swig.org/Doc1.1/HTML/Python.html

-- 
import string,time,os;print string.join((lambda x:x[:10]+x[8:])(map(
lambda x:string.center("*"*(lambda x:((x<24) ### Seasons Greetings, Will Ware
*(x-3))+3)(x),24),range(1,28, 2))),"\n") ################ wware at world.std.com


From cobrien at Radix.Net  Mon Feb 19 00:30:45 2001
From: cobrien at Radix.Net (Cary O'Brien)
Date: 18 Feb 2001 18:30:45 -0500
Subject: What to do after Python?
References: <3A8F81BC.108488CA@pop3.qwestinternet.net> <96p5mu$3eq$1@gaia.intranet.cdegroot.com> <roy-6DE44A.14135818022001@news.panix.com>
Message-ID: <96plv5$gqg$1@saltmine.radix.net>

In article <roy-6DE44A.14135818022001 at news.panix.com>,
Roy Smith  <roy at panix.com> wrote:
>cg at cdegroot.com wrote:
>> I'd advise C as a second language because it forces you to deal with the 
>> computer behind all that code, which tends to give you a good 
>> perspective for other language.
>
>Actually, the problem with C is that it hides too much of the computer 
>behind all that code.  I spent a few years writing C on alphas, and never 
>even found out how many registers the danged thing had.  If you really want 
>to find out what the hardware is all about, learn assembler, which is 
>perilously close to handing a soldering iron to a software guy.
>

Your problem is that you C compiler was too good.  I learned C and
68000 assembly at the same time with a stupid C compiler.  Stupid
meaning after a while we knew exactly what the C compiler would do.
Often (in the days of vt220 terminals) two of us would work together
debugging, one single-stepping through assembly, and one reading the C
code and keeping in sync.  If you said register you meant register.
Plus if you did you loops just right you could get the 68010 cpu to go
into wicked-fast-loop-mode.

>Another alternative would be to learn lisp.  It may not get you a better 
>job, but will give you the ability to write a better python major mode for 
>emacs :-)
>

Prolog is kinda different.  Plus I like the SQL suggestion.  Go buy
Bruce's PostgreSQL book and go to town.

-- cary

>I don't suppose I could interest you in fortran?










From com-zjensen at msn.com  Sat Feb 10 07:30:31 2001
From: com-zjensen at msn.com (Zac Jensen)
Date: Fri, 9 Feb 2001 23:30:31 -0700
Subject: What is better, JPython or Jython?
References: <mailman.981730586.31400.python-list@python.org>
Message-ID: <962n6a$v9$1@news.xmission.com>

That's pretty cool, since that's not even the question you moron

>"Kemp Randy-W18971" <Randy.L.Kemp at motorola.com> wrote in message
news:mailman.981730586.31400.python-list at python.org...
> Which is better: a boat or a car?  It all depends on whether you are
traveling by land or sea.  Which is better:  python or jpython?
>Are you using it with Java or not?
>





From pen_man at my-deja.com  Wed Feb  7 05:24:46 2001
From: pen_man at my-deja.com (penman)
Date: Wed, 07 Feb 2001 04:24:46 GMT
Subject: Getting one character at a time from sockets
References: <95q044$2f9$1@nnrp1.deja.com> <95qg45$5b0$0@216.39.151.169>
Message-ID: <95qim7$i38$1@nnrp1.deja.com>

In article <95qg45$5b0$0 at 216.39.151.169>,
  Donn Cave <donn at oz.net> wrote:
> Quoth penman <pen_man at my-deja.com>:
> | Using ordinary TCP/IP sockets with the socket module in Python, I
need
> | to get a single character at a time from the client; something
> | like "talk" in unix. How could I do that? (should be platform
> | independent if possible)
> |
> | recv(1) didn't work. It waited till the carriage return came in.
(non-
> | blocking didn't work either)
>
> Don't mess around with non-blocking.  Even recv(8000) will return
> more or less immediately if you send any data at all, so it seems
> kind of likely you're sending no data until the CR.  Substitute a
> test client that has its data built in, and just send('i') to the
> server.  I bet recv() will get it right away, Nagel algorithms
> notwithstanding.
>
> 	Donn Cave, donn at oz.net
>

Thank you, Donn. But the problem is that I have to use ordinary telnet
clients, such as CRT, Netterm, so let the users access to the server
without any specialy programs.

I wonder why nobody ever had the need to recv 1 character "unbuffered"
in Python.


Sent via Deja.com
http://www.deja.com/


From tim.one at home.com  Sat Feb  3 05:28:34 2001
From: tim.one at home.com (Tim Peters)
Date: Fri, 2 Feb 2001 23:28:34 -0500
Subject: "in" operator for strings
In-Reply-To: <95ftl4$o6b$1@tyfon.itea.ntnu.no>
Message-ID: <LNBBLJKPBEHFEDALKOLCGECDINAA.tim.one@home.com>

[Magnus Lie Hetland]
> ...
> What's the built-in string-matching algorithm in Python?
> (Not very important to me, but...)

Brute force, using memcmp.  Note that fancier algorithms that require
precomputing tables of size proportional to the alphabet cardinality aren't
very attractive in a Unicode world.

See mxTextTools for a Boyer-Moore 8-bit string searcher.




From thomas.heller at ion-tof.com  Thu Feb 15 08:43:22 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 15 Feb 2001 08:43:22 +0100
Subject: Python20 and .pth files
References: <3A89FB2B.C1CA8A5E@earthlink.net> <3A8B01F3.1000605@ActiveState.com>
Message-ID: <96g1ar$kuoss$1@ID-59885.news.dfncis.de>

"Mark Hammond" <MarkH at ActiveState.com> wrote in message news:3A8B01F3.1000605 at ActiveState.com...
> greg Landrum wrote:
>
>
> > However... one of the applications I'm working on creates Python COM
> > servers which import Numeric.  These work fine when I call them from Python
> > applications (so none of my unit tests broke when I upgraded to 2.0, so I
> > never noticed the problem).  However, when I create one of these COM
> > objects from within VB, I get an ImportError for Numeric.  After much
> > shrieking and pulling of my hair (I'm giving a demo of these objects on
> > Friday, so I was not pleased to discover that they were mysteriously
> > broken), I worked out a solution: add c:/Python20/Numeric to my PYTHONPATH
> > environment variable.  Now everything is happy.
>
> I believe this is a problem in certain embedding situations.  Specifically, I believe it is because Python does not know its
"home", so doesnt know where to load the .pth files from.

Are you sure site.py is imported at all? There PYTHONPATH is extended by the contents
of the .pth files.

>
> > The Question:
> > Was it an accident that everything just worked before or has something
> > changed?  I know old versions of Numeric used to have an installer that may
> > or may not have set registry keys to help out with this problem.
> > Are .pth files being deprecated so that we (the users) should start
> > badgering developers who still ship modules requiring them?
>
> I dont believe this would ever have worked.  As you suggest, it is quite possible that only new versions of Numeric have .pth
files.
>
New versions have a .pth file (since they are distributed via distutils),
old versions used registry settings.

> This is on my "to-do" list - having Python know its correct home in embedded situations would solve a few other problems too.
>
This is a bug in getpathp.c, see
http://sourceforge.net/bugs/?func=detailbug&bug_id=131064&group_id=5470


> Mark.
>
Mark, please look at the bug. Otherwise you are forcing my to submit a patch ;-)

Thomas




From mrq at for.mat.bham.ac.uk  Thu Feb  8 11:33:45 2001
From: mrq at for.mat.bham.ac.uk (Martyn Quick)
Date: Thu, 8 Feb 2001 10:33:45 +0000
Subject: Some basic questions about Tkinter (probably v easy for experts!)
In-Reply-To: <Igcg6.13212$AH6.1990695@newsc.telia.net>
References: <Pine.GSO.4.21.0102071252070.15785-100000@babbage> <Igcg6.13212$AH6.1990695@newsc.telia.net>
Message-ID: <Pine.GSO.4.21.0102081029430.20775-100000@babbage>

On Wed, 7 Feb 2001, Fredrik Lundh wrote:
> 
> have you read the introduction document?  the first chapter
> seems to answer your questions:
> 
> http://www.pythonware.com/library/tkinter/introduction/hello-tkinter.htm

Thanks very much for your answer Fredrik - I managed to get some menus
working last night, so I think I'm starting to get the hang of
things.  I'll try having a thorough re-read of the Introduction now that
I'm getting an idea of the syntax.

> > armadillo.add_command(lable="Exit", command=exit)
> 
> if you fix the typo, this adds a command item to the armadillo
> menubar.  when that item is selected, Tkinter calls sys.exit.

I've realized one point that was confusing me.
When I use sys.exit to finish a python script I would type

sys.exit()

but in the above you use  exit  rather than  exit().
Is it generally the case that you don't use the brackets in the
comannd=...  part?  (Sorry if this is easily to be found in the
Introduction document!)  Presumably this means that I can't use a command
which relies on parameters?

Thanks once again...

Martyn

--------------------------------------------------------
Dr. Martyn Quick  (Research Fellow in Pure Mathematics)
University of Birmingham, Edgbaston, Birmingham, UK.
http://www.mat.bham.ac.uk/M.R.Quick





From bsass at freenet.edmonton.ab.ca  Tue Feb 27 08:53:45 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 00:53:45 -0700 (MST)
Subject: CPAN functionality for python
In-Reply-To: <20010226180935.B7531@tummy.com>
Message-ID: <Pine.LNX.4.33.0102270016190.28512-100000@bms>

On Mon, 26 Feb 2001, Sean Reifschneider wrote:
> On Mon, Feb 26, 2001 at 11:14:35AM -0700, Bruce Sass wrote:
> >from outside the archive.  The translation from a generic pkg to a
> >distro specific pkg would be a problem for the disto to solve
> >(probably a matter of repacking info about the pkg, then using
>
> The only way having automaticly-generated packages be possible is
> if we really force a conformance from the authors of the packages.
> I've packaged quite a number of RPMs in my day, and you can get pretty
> close to just wrapping a tar file with some meta-information in
> some cases, but these are the minority.

I have assumed that submitted packages would need to meet a
standard of some sort - was that incorrect?

> In most cases you have to provide patches which fix incorrect paths
> in the tar file (for example, the Python tar file uses
> "#!/usr/local/bin/python" all over the place).

That is to be expected... and only the distro is really equipted to
handle these issues.

> If we can force the distributions to be more compatible with being
> packaged, it may be very easy.  Otherwise, we may have to rely on
> people contributing their time to building RPMs if we want to have
> RPMs.

I'm not sure what you mean here.  The distributions are the ones doing
the packaging, the Python version of CPAN should supply all the bits
needed for the distros to perform that function.

> >only way I can see to do that is to not package to anyones spec, but
> >provide enough information so that anyone can package to their
> >fav spec.
>
> I don't agree...  I believe that we should embrace the different packaging
> formats, not act like they don't exist.

Maybe we see "embracing" them a little differently.  I believe
providing everything they need to turn a Python module/package into
their standard format, using their packaging and QC tools, does them
more of a service than providing pre-built packages created by a third
party.  I'm not saying binary style packages shouldn't be supported,
just that the emphasis should be on portable (in that they can easily
be turned into a system's native pkg format) source packages.

> If the distutils package can be
> made to generate native RPMs, that's great.  The *CLIENT* program should
> dictate that.  If it can handle turning a distutil package into an RPM,
> it'll have it's preference be a distutil package.

That sounds like exactly what I was enquiring about - as long as the
distutils pkg provides enough info for every distro to create their
own pkg.

> I don't believe having the archive not able to deal with different package
> formats is acceptable.  If that were the case, then if the package didn't
> provide enough information to build an RPM, the user would have no option
> within the archive of working around it.

Agreed.  The submission should not be accepted if it does not contain
enough information to be packaged.


- Bruce




From dan at eevolved.com  Fri Feb  2 18:15:57 2001
From: dan at eevolved.com (Dan Parisien)
Date: Fri, 02 Feb 2001 17:15:57 GMT
Subject: expanding a list and dict to a *args and **dict
Message-ID: <hXBe6.172472$JT5.6263159@news20.bellglobal.com>

Is it possible to 'expand' a list or dictionary into its components to call 
a function expecting many parameters?

I want to be able to call a function without accessing each individual 
element manually (list[0], list[1], list[2])

I'm sending random parameters over a network (*args and **dict). When I 
unpickle them on the other side, I get a python list and dictionary. I then 
call an arbitrary function (actually a reference to a function that was 
bound at runtime a la 
self.randomfunction = obj.specificfunction
)

That function is not expecting a list and a dictionary, instead it is 
expecting specific parameters(that i don't know at runtime) and I want to 
take the list and turn it into a bunch of parameters like list[0], list[1], 
list[2] (but as stated above, I can't do that)

Ha. I hope someone understood me :)

Dan


From jimholliman at heartsoft.com  Thu Feb  8 22:41:42 2001
From: jimholliman at heartsoft.com (Jim)
Date: Thu, 8 Feb 2001 15:41:42 -0600
Subject: C++ style Stream Operators
Message-ID: <3a831395.0@216.0.152.7>

I'd like to use C++style stream i/o operators, such as "file_object <<
some_stuff" in my python code.
IS this sort of operator setup and what modules would I need to take
advantage of them if so?

Thanks
Jim






From sgriggs at pobox.com  Tue Feb  6 13:16:09 2001
From: sgriggs at pobox.com (sgriggs at pobox.com)
Date: Tue, 06 Feb 2001 12:16:09 GMT
Subject: Pyt;hon 2.1a2 compile problems
Message-ID: <95opu4$t8k$1@nnrp1.deja.com>

Is anyone else having problems getting Python 2.1a2 to compile on
Solaris? I'm using Solaris 2.8 with GCC 2.95.2 and just can't seem to
get it to work. I can compile Python 2.0 on this same box. I'm just
wondering if it's just me at this point. Thanks.

Scott Griggs


Sent via Deja.com
http://www.deja.com/


From bridgman at wyeth.gsfc.nasa.gov  Fri Feb 16 21:09:07 2001
From: bridgman at wyeth.gsfc.nasa.gov (Tom Bridgman)
Date: Fri, 16 Feb 2001 15:09:07 -0500
Subject: Converting an instance to a subclass?
References: <3A8D7636.B8161C45@wyeth.gsfc.nasa.gov> <mailman.982352599.31822.python-list@python.org>
Message-ID: <3A8D88E3.9FF4CA4B@wyeth.gsfc.nasa.gov>

I'm seriously considering your option #2.

These utility scripts are doing things which are really non-standard for
the process I'm developing.  They just exist solely for the purpose of
converting an old data format over to a new one, installing new
defaults, checking that support files are in the right locations, moving
them if they're not, etc.  The methods defined for this should never be
used once we switch over to the new system so I don't want them in my
standard class library to confuse some future maintainer of the code.

Thanks for your input,
Tom

D-Man wrote:
> 
> On Fri, Feb 16, 2001 at 01:49:26PM -0500, Tom Bridgman wrote:
> [snip]
> | and I have a number of methods that return me instances or lists of
> | instances of this class.  These are part of my general library of
> | classes for this project.
> |
> | However, I need to write a utility which will define methods that are
> | only needed by the utility.  I really don't want these methods to be
> | 'permanent' members of the class so I define them in a subclass as part
> | of the utility.
> |
> | Class B(A):
> |    def Cleanup(self,x,y,z):
> |        ...
> |
> 
> This is a good way to add functionality that isn't always needed --
> you can use instances of 'A' when it's not needed, and instances of
> 'B' when it is.
> 
> | Is there a way I can cast the instances of class A into instances of
> | class B so I can use the additional methods?  I can't find anything
> | about it in "Programming Python" but then I'm not quite sure where to
> | look either.
> 
> Since you use the word 'cast', I take it you have some C, C++ or Java
> experience.  Those lanugages are statically typed, so if you declared
> having one type, but really have another, you must explicitly tell the
> compiler with a cast.  Python is dynamically typed, so if you really
> have a "different" type (it can't really be different since you didn't
> declare it in the first place, but) you have it and python is happy.
> 
> What I suspect is happening is that you are creating instances of A
> when you really want B.
> 
> >>> class A : pass
> ...
> >>> class B( A ) : pass
> ...
> >>> obj1 = A()
> >>> obj2 = B()
> >>> obj1.__class__
> <class __main__.A at 007A8934>
> >>> obj2.__class__
> <class __main__.B at 007AA9A4>
> >>>
> 
> In this example, obj2 is an instance of class "B".  Since "B" is a
> subclass of "A", obj2 is also an instance of class "A".  You can use
> obj2 whereever an instance of A or B is expected.  On the other hand,
> obj1 is an instance of class "A", and won't work where an instance of
> class B is expected.  There is no way automatically to turn obj1 into
> an instance of class B.  There are a couple of ways to work around
> though.
> 
> 1)  modify the factory functions to return instances of "B" instead.
>     Since it is a subclass, all existing code will work.
> 
> 2)  Make a conversion that will create a new instance of class "B" and
>     init all data members based on an instance of class "A"
> 
> 3)  Play some obscure tricks using python dynamism to add those
>     functions to obj1 even though they don't belong there
> 
> #1 may be the easiest, but then you might as well just put the
> functions in class "A" and only use them when you need to.
> 
> #2 could be done in B's __init__ function :
> 
>    class B( A ) :
>        def __init__( self , instance_of_A ) :
>            self.value = instance_of_A.value
>            # ...
> 
>    Then when you want an instance of B, you can create it from the
>    given instance of A.
> 
> #3 isn't good since it is obscure and will be difficult to maintain.
> 
> HTH,
> -D

--
Dr. William T."Tom" Bridgman           Scientific Visualization Studio
Global Science & Technology, Inc.      NASA/Goddard Space Flight Center
Email: bridgman at wyeth.gsfc.nasa.gov    Code 935
Phone: 301-286-1346                    Greenbelt, MD 20771
FAX:   TBD                             http://svs.gsfc.nasa.gov/


From janne at oops.nnets.fi  Fri Feb 16 12:59:26 2001
From: janne at oops.nnets.fi (Janne Sinkkonen)
Date: 16 Feb 2001 13:59:26 +0200
Subject: shelve and the best db possible
References: <t05j6.259674$JT5.9093034@news20.bellglobal.com>
Message-ID: <m3ae7m2601.fsf@kinos.nnets.fi>

Dan Parisien <dan at eevolved.com> writes:

> I really love shelve, but dbm doesn't seem to scale very well :( Could 
> anyone point me to a good db with a shelve interface? I've tried to get 
> bsddb3 to go, but the test scripts all die :(

I have used an old bsddb 2.x interface from TCS with a largish data
base (btree, hundreds of MB's, hundreds of thousands of updates)
without any problems. Don't know whether it's still available somewhere.

The module name is bsddb (or TCS.bsddb)

-- 
Janne


From embed at geocities.com  Mon Feb 26 20:28:32 2001
From: embed at geocities.com (Warren Postma)
Date: Mon, 26 Feb 2001 14:28:32 -0500
Subject: Splitting comp.lang.python
References: <mailman.983199848.8916.python-list@python.org> <Hgwm6.2969$8b1.507773@news4.rdc1.on.home.com>
Message-ID: <79ym6.2288$TW.12847@tor-nn1.netcom.ca>

Not again!?  We had this discussion ten times over. Anyone who feels like
weighing in on this one should at least be polite and avoid beating the same
dead horses, over and over again.

I mean, doesn't anyone search Deja anymore? Oh yeah, Deja is gone, and
replaced by the Beta Groups at groups.google.com. Oh well, ironically, now
that we have no Deja.com, I think we are in for a Lot More of that old time
Deja Vu.

Warren Postma





From thomas at xs4all.net  Fri Feb  2 15:03:00 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 2 Feb 2001 15:03:00 +0100
Subject: sys.exit and exceptions (newbie question)
In-Reply-To: <20010202130129.7059DA840@darjeeling.zadka.site.co.il>; from moshez@zadka.site.co.il on Fri, Feb 02, 2001 at 03:01:29PM +0200
References: <20010202002551.S962@xs4all.nl>, <3a7998cf.869169328@localhost> <20010202002551.S962@xs4all.nl> <20010202130129.7059DA840@darjeeling.zadka.site.co.il>
Message-ID: <20010202150300.V962@xs4all.nl>

On Fri, Feb 02, 2001 at 03:01:29PM +0200, Moshe Zadka wrote:
> On Fri, 2 Feb 2001 00:25:51 +0100, Thomas Wouters <thomas at xs4all.net> wrote:

> > You can use sys._exit(). 

> No you can't -- but you can use os._exit().
> Everything else Thomas said is true...

D'oh! I really, really need vacation. I've made *four* errors this week! And
one of them was even a sysadmin error (added an emailalias in the wrong
file, so it didn't work) -- that hasn't happened in years... I think I may
be going senile ;P

Maybe-I-should-just-move-to-management-ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From s713221 at student.gu.edu.au  Sun Feb 18 09:15:20 2001
From: s713221 at student.gu.edu.au (s713221 at student.gu.edu.au)
Date: Sun, 18 Feb 2001 18:15:20 +1000
Subject: how to unimport something
References: <mailman.982471638.25003.python-list@python.org>
Message-ID: <3A8F8498.34B00D23@student.gu.edu.au>

Sean 'Shaleh' Perry wrote:
> 
> Suppose I have a program which has logical segments.  How do I run each segment
> and have modules loaded in that segment get unloaded when I reach the next
> one?

The direct answer is

Import a module into python
>>>import modulename

Delete a module
>>>del modulename

Mind you, you may want to post to the group an example of what you're
doing. Importing modules can take some time to load in, they may be able
to suggest a more efficient and safer method than module swapping to do
what it is you're trying to do.

If you're trying to do what I think you're doing, which is have a series
of modules with similar named functions, and dynamically "plug/unplug"
them from the interpreter, a better way may be to:

>>>import modulea,moduleb,modulec (Import all your plugin modules at the start, drop the cost of module import into startup.)

>>>module = modulea
>>>module.function() (Which is actually modulea.function())
>>>module = moduleb
>>>module.function() (Which is actually moduleb.function())
>>>module = modulec
>>>module.function() (Which is actually modulec.function())
Unless you had the same functions and classes in each module, and they
could accept each others argument patterns, there's the risk of your
script failing because a piece of calling code called module.function()
expecting modulea, but getting moduleb. Messy

If you were planning to add different functionalities to the interpreter
at the same time and remove that functionality later, you would
seriously risk writing code that calls modulea functionality, only to
crash your script because another piece of code wiped out modulea
earlier.

Finally, if you're thinking to only have the modules you need loaded
into the interpreter at any time, this may save some memory, but will
cost you in module loading time. Usually not worth it unless you're
writing scripts which eg. run for days on a resource-scarce server and
it only swaps modules a couple of times a day, or are running the
scripts in a severely memory scarce environment.

Joal Heagney/AncientHart


From clgonsal at keeshah.penguinpowered.com  Thu Feb  1 08:27:02 2001
From: clgonsal at keeshah.penguinpowered.com (C.Laurence Gonsalves)
Date: Thu, 01 Feb 2001 07:27:02 GMT
Subject: [Q] ftplib: How to compare date?
References: <9I6e6.7505$p8.1228772@typhoon.southeast.rr.com>
Message-ID: <slrn97i4kh.abo.clgonsal@keeshah.penguinpowered.com>

On Thu, 01 Feb 2001 05:43:33 GMT, Daehyok Shin <DSHIN at nc.rr.com> wrote:
>I like to compare the creation dates of two files using ftplib.
>But, I cannot find any function to get the date of remote files.
>How can I do it?

I don't think there's any way to do this reliably. You can try and get a
list of the directory, but different FTP servers will return the listing
in different formats. If you only want it to work on known FTP servers,
you can parse the format yourself though (string.split and time.strptime
would probably come in handy at this point).

Also, if you want to compare the time stamps of files on the local
system with the time stamps of files on a remote FTP server, be aware
that the clocks probably aren't in sync, and might not even be in the
same time zone.

-- 
  C. Laurence Gonsalves                "Any sufficiently advanced
  clgonsal at kami.com                     technology is indistinguishable
  http://cryogen.com/clgonsal/          from magic." -- Arthur C. Clarke


From ssthapa at classes.cs.uchicago.edu  Tue Feb 20 18:53:50 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Tue, 20 Feb 2001 17:53:50 GMT
Subject: Python 2.x breaks cmp() (was Re: A suspected bug)
References: <mailman.982467140.8836.python-list@python.org> <Z6Oj6.18070$AH6.2468100@newsc.telia.net> <slrn98vrqv.i0.amk@mira.erols.com> <96os85$1qe$1@panix3.panix.com>
Message-ID: <slrn993luj.24c.ssthapa@ntcs-ip45.uchicago.edu>

Aahz Maruch <aahz at panix.com> wrote:
>>>>> L=[1j, 2j]
>>>>> L.sort()
>>Traceback (most recent call last):
>>  File "<stdin>", line 1, in ?
>>TypeError: cannot compare complex numbers using <, <=, >, >=
>>>>>
>
>Is there any chance this can be treated as a bug and fixed for the
>release of 2.1?  Alternatively, given that we're already breaking code
>with the change in the way complex numbers are handled, should cmp() now
>raise an exception *every* time the type/class differs?

  If you are referring to the comparision of complex numbers, I'm not
sure how the ordering would be done.   For example consider 1 and 1j, 
they have the same magnitude so mathematically speaking there is no
way to consistently order them like the real numbers where magnitude
is a more useful measure. IIRC, for any given magnitude n there are an
infinite number of complex numbers with that magnitude, namely all n*e^(ix)
for real x.


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From junaftnoon at nospamplzyahoo.com  Wed Feb  7 23:01:05 2001
From: junaftnoon at nospamplzyahoo.com (June Kim)
Date: Thu, 8 Feb 2001 07:01:05 +0900
Subject: 2nd Python Seminar in Seoul Was a Success
References: <95n463$anj$1@news.nuri.net> <95q39h$hji$5@newshost.accu.uu.nl> <95q8tj$1c8$1@news.nuri.net> <95rgng$8ul$1@nnrp1.deja.com> <3dn1bygrmt.fsf@ute.cnri.reston.va.us>
Message-ID: <95sgmd$j8v$1@news.nuri.net>

"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3dn1bygrmt.fsf at ute.cnri.reston.va.us...
> John Grayson <johngrayson at home.com> writes:
> > I received a package from my publisher at Manning last night:
> > Inside was the Korean translation of Python and Tkinter Programming!
> > It is published by Infobook ISBN 89-8054-424-3.
>
> If people can point me at information about the Korean translations,
> I'll happily add them to the list of non-English books in the Python
> bookstore.
>
> --amk
>

There are a few Korean translations and original works as well. Below are
the book titles and original publishers, followed by the translators, Korean
publisher, and the URL of the web page for the book. (This list isn't
comprehensive but it is all I know of, AFAIK)

Python Pocket Reference, O'Reilly, translated by Kim, Tae-Sun. Hanbit Media
http://www.hanbitbook.co.kr/look.htm?book_code=001016-00001

Learning Python, O'Reilly, translated by Lee, Kang-Sung. Hanbit Media
http://www.hanbitbook.co.kr/look.htm?book_code=010201-00001
(The translator is the current president of Korea Python Users Group.)

Python and Tkinter Programming, Manning, translated by Ryu Jung-Wook.
Infobook.
http://www.infobook.co.kr/cgi-bin/Vmk_BookInfo.cgi?bookinfo_but=????&in_book
id=ISBN_89_8054_424_3
(This URL includes Korean characters, which means you couldn't access the
page. In this case, follow this alternative URL of an online book store in
Korea :
http://www.wowbook.com/generic/book/info/book_detail.asp?isbn=ISBN1-884777-8
1-3 )

Internet Programming with Python, (by Aaron & Guido), translated by Kwak,
Joon-Ki. Samkakhyoung. (translated in 1997 but out of print now)
http://www.wowbook.com/generic/book/info/book_detail.asp?isbn=ISBN89-7467-43
5-1

About XML, (originally in Korean), Park Jae-Ho. Youngjin.
https://www.youngjin.com/shop/ycom-s-global002-pre.asp?url=ycom-s-002-01-pre
.asp&pcode=9788931415575
http://www.wowbook.com/generic/book/info/book_detail.asp?isbn=ISBN89-314-155
7-5
(This book is about XML but uses Python substantially)

In addition to these, there are a number of on-going works to write about
Python; moreover, I'm leading an open content project(in Korean), with
collective authorship, for writing An Introduction to Computer Programming
with Python -- it is geared towards complete newbies in computer programming
and it's in the spirit of CP4E. I hope open content projects would evolve
along with Python.

Best regards,

June




From tjg at exceptionalminds.com  Tue Feb 27 20:08:30 2001
From: tjg at exceptionalminds.com (Timothy Grant)
Date: Tue, 27 Feb 2001 11:08:30 -0800
Subject: IP Math anyone?
In-Reply-To: <003f01c0a0c2$ac255040$a100000a@local>; from chris.gonnerman@usa.net on Tue, Feb 27, 2001 at 07:38:56AM -0600
References: <20010226170223.B29494@trufflehunter.avalongroup.net> <73931622.983270231@[10.0.0.101]> <003f01c0a0c2$ac255040$a100000a@local>
Message-ID: <20010227110830.I29494@trufflehunter.avalongroup.net>

Thanks to all of you who made this thread much more
enlightening than expected! As I expected, it is a somewhat
complex problem, for which I will probably throw together a
class with a nice interface but internals of which are a crude
hack, and then refine the internals as time allows.


On Tue, Feb 27, 2001 at 07:38:56AM -0600, Chris Gonnerman wrote:
> You neglected to tell us where to download your class from, or to attach
> the source.  I at least would like to see it.
> 
> ----- Original Message -----
> From: "Tino Wildenhain" <tino at wildenhain.de>
> Subject: Re: IP Math anyone?
> > Hi Timothy,
> >
> > for teaching I made a class for this.
> >
> > you can instantiate it with a=ip('192.168.0.254/24') or
> > a=ip('192.168.0.254','255.255.255.0') just using a=ip('192.168.0.254')
> > would assign a appropriate subnet mask for this class (class B results in
> > /16 or 255.255.0.0)
> >
> > Since IP-adresses are only 32bit non signed integers, they are internally
> > represented
> > like that.
> > If you make a class like that you can simply add the support to add
> > integers to
> > it according to the subnet-mask you use.
> >
> > Regards
> > Tino

-- 
Stand Fast,
    tjg.

Timothy Grant                         tjg at exceptionalminds.com
Red Hat Certified Engineer            www.exceptionalminds.com
Avalon Technology Group, Inc.         <><       (503) 246-3630
>>>>>>>>>>>>>Linux, because rebooting is *NOT* normal<<<<<<<<<
>>>>This machine was last rebooted:  41 days 23:20 hours ago<<



From chris.gonnerman at usa.net  Sun Feb  4 02:11:30 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sat, 3 Feb 2001 19:11:30 -0600
Subject: Gratuitous Change (Was: Re: "in" operator for strings)
References: <hK8e6.141262$y9.26215366@typhoon.we.rr.com> <2jae6.119$o3.4587@news.world-online.no> <95beco$e0u$1@tyfon.itea.ntnu.no> <95birk0r0d@news1.newsguy.com> <QPee6.1195$o91.132115@typhoon.nyroc.rr.com> <95eb4r0csk@news1.newsguy.com> <ztGe6.8408$o5.48355@e420r-atl1.usenetserver.com> <006701c08dfa$08ed8c60$a100000a@local> <20010203094706.B30977@glacier.fnational.com>
Message-ID: <007701c08e47$696012a0$a100000a@local>

----- Original Message -----
From: "Neil Schemenauer" <nas at arctrix.com>
Subject: Re: Gratuitous Change (Was: Re: "in" operator for strings)


> Can you give some examples of changes you have to "put up with"?
> The only controversial changes in 2.1a2 that I know of are nested
> scopes, function attributes and weak references.  The nested
> scope changes are the only ones that affect existing code.  You
> have to forgive me if I think your spreading FUD.

Well, that hurts, but I guess I have it coming.  FUD is my enemy, let me
spread no FUD.

My primary complaint is against changes to the language semantics.  I am
against change in a language in general when the language is already IMHO
the finest I've ever worked with.

You are probably right, in that these changes in 2.1 are not so bad.  On the
other hand, I probably won't use them.  Heck, I still import string.  I'm
not
against the changes in 2.0, although I don't use most of them; I still have
servers
running 1.5.2 that would be a pain to upgrade.

One thing I see a lot in 2.0 code that is legal but UGLY is:

    nstr = ''.join((s,t,u,v))

(for instance) but I guess you can write ugly code in any language.

My only real complaint is the need to rebuild ALL MY EXTENSIONS
every time the version is upgraded.  The build process is painless under
Linux but a pain under Windows and I must live in both worlds.  Even
under Linux, I wind up rebuilding a LOT of software because I use a LOT
of Python.

Yeah, I'm rambling.  The point is (yes, I have one) that I think all changes
are potentially dangerous.  I do respect and trust GvR though, so perhaps
I'll try 2.1 when it's not alpha anymore.

BTW I've lost track... is 2.1 stackless or was that just a rumor?







From ullrich at math.okstate.edu  Fri Feb  2 15:58:08 2001
From: ullrich at math.okstate.edu (David C. Ullrich)
Date: Fri, 02 Feb 2001 14:58:08 GMT
Subject: pyXML support for XSL tranformations??
References: <rtga6.18098$bb.708443@news1.rdc1.tx.home.com> <94d1tj$6sa$1@nnrp1.deja.com> <3a6aef10.3738521@nntp.sprynet.com> <3A7232DB.C576B547@ogbuji.net> <94vg3k$ed9$1@nnrp1.deja.com> <3A7442FF.E1CFDF6D@ogbuji.net>
Message-ID: <3a7ac7f1.3651537@nntp.sprynet.com>

On Sun, 28 Jan 2001 16:04:18 GMT, Uche Ogbuji <uche at ogbuji.net> wrote:

>"David C. Ullrich" wrote:
>
>> Once I got it working it went right onto my I-don't-see-
>> how-you-guys-can-give-this-stuff-away list, btw - hope
>> you don't have the idea I've been wanting my money
>> back or anything.
>
>No, I just wanted to have things clear on the thread because there *was*
>a time when installing 4Suite was a huge pain, and I'd like all to know
>we've worked very hard to make it a breeze (with some help from Greg
>Ward and distutils).

The install (probably a year or so ago) _was_ very easy. I may
as well say what was so stupid about me: There was a spot
in the instructions about unzipping to a "suitable" directory,
with no definition of "suitable". Didn't take me long to figure
out what was unsuitable about the directory I used. That
led to copying some things over some things. I didn't replace
a certain subdirectory that already existed cuz I wasn't
sure whether anything was using that stuff already. The
XSLT didn't work. Then I replaced python\xml\dom with
the one that appeared when I unzipped your package
and everything worked fine.

The "uncleanliness" I was alluding to was having to
modify part of an existing installation. But if I were
not an idiot I would have chosen a suitable directory
to begin with and I wouldn't have noticed the change.
Or I would have noticed that the stuff in the directory
I was reluctant to replace was _also_ an xml\dom
from 4Thought - I thought for no good reason that
it was a module from an entirely independent
source so that replacing it would likely break something
else.

>> > There is no
>> > reason why you can't download 4Suite-0.10.1.tar.gz, untar it, run
>> > "python setup.py install", and be completely ready to go.
>> 
>> You only think that because you're not stupid enough to find
>> ways to screw it up...
>
>Well, we'd actually like 4Suite to be easy to install even for
>non-Pythoneers. 

I think I just slipped through the cracks there - knew just
enough to do it wrong. Not your fault.

> That's because some of the features are completely
>unique in XML space, and we think we can actually draw some XML users to
>Python if they don't have to work too hard to get started.  That's why
>we put together the HOWTOs which even walk you though installing Python.
>
>> > But just in case, there are a 4Suite Installation HOWTOs for UNIX and
>> > Windows.
>> >
>> > See
>> >
>> > http://services.4Suite.org/topics/rdf/Top/Documentation/HOWTOs
>
>
>-- 
>Uche Ogbuji
>Personal:   uche at ogbuji.net		http://uche.ogbuji.net
>Work:       uche.ogbuji at fourthought.com	http://Fourthought.com



From SBrunning at trisystems.co.uk  Fri Feb  9 15:45:05 2001
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 9 Feb 2001 14:45:05 -0000 
Subject: What is better, JPython or Jython?
Message-ID: <31575A892FF6D1118F5800600846864D5B17DE@intrepid>

> From:	Zamurai [SMTP:zamurai at gmx.net]
> It seems that JPython and Jython are quite similar, but what are the
> differences between both? And which one is better?
 
Jython supersedes JPython, so use the latter.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.



From samschul at pacbell.net  Mon Feb  5 09:07:41 2001
From: samschul at pacbell.net (Sam Schulenburg)
Date: Mon, 05 Feb 2001 08:07:41 GMT
Subject: Install PythonWin on top of Python 2.0
References: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <95ln0b$7vb$1@nnrp1.deja.com>

In article <95l4en$rbs$1 at nnrp1.deja.com>,
  Deja User <binary3418 at my-deja.com> wrote:

You need to get the 2.0 build I think it is build 135

> Hi, there:
>
>   This is a question from a new user. I recently installed the BeOpen
> Python 2.0 on my Win98 PC. Now I'd like to install the Win32 extension
> so I download "win32all-125.exe". Unfortunately, I cannot install it
> because this build insists on the presence of Python 1.5.x. Now, do I
> have to install 1.5 so as to install PythonWin? Would the author of
> PythonWin please do us a favor to relax the version check in
PythonWin?
>
>   Thanks,
>
> Bin
>
> Sent via Deja.com
> http://www.deja.com/
>


Sent via Deja.com
http://www.deja.com/


From idfx at my-deja.com  Fri Feb  9 20:29:14 2001
From: idfx at my-deja.com (idfx)
Date: Fri, 09 Feb 2001 19:29:14 GMT
Subject: Python/C++ interface
References: <3A79D28B.1BEB46A0@cybermesa.com>
Message-ID: <961gdt$fm4$1@nnrp1.deja.com>

In article <3A79D28B.1BEB46A0 at cybermesa.com>,
  Jay O'Connor <joconnor at cybermesa.com> wrote:
> All,
>
> I was doing some testing with the Python->C++ interface and ran into
> something odd.
>
> The example shows calling a C function from python with
>
> 	import spam
> 	spam.system ("ls -l")
>
> and on the C++ side... the signature reads
> 	PyObject* spam_system(PyObject *self, PyObject * args) {
> 		...
> 	}
>
> Now.  On the C++ side, args is considered, and parsed, as a Tuple.
> However, what I found was that a Tuple was not being passed, just a
> String.  The only time a Tuple was passed was if there was more than
one
> argument.  The whole argument list would be converted to a Tuple,
then.
>
> Is this an error in the documentation?  Or has the passing method
> changed between 1.5.x and 2.0 (I'm using 2.0) and the documentation is
> just out of date?
>
> Also,  is there a good way of printing the name of the type of a
> PyObject?

     I may be wrong, or over-simplifying, but I think you've gotten
confused by the terms, understandably.  Yes, args is read into a tuple,
always, but if there's only one argument, then it becomes a tuple of
length 1.

ie:    tpl = ("yourstring",)

that way, you can refer to as many arguments as there are,

( numargs=len(tpl) )

instead of worrying about special cases for 1 or 0 arguments.
Your args is still a string, it's just that it's a string that happens
to be an element of a tuple.


   OTOH, I may be totally mis-reading your problem.  The other
possibility is that, for a system call, (e.g. 'ls -l'), C++ makes a
string of the argument(s) to preserve the command you are passing to
the system, which should be a string anyway, or so I'm told.


    As for the printing of type names, this is something I had trouble
with, too.  Eventually, I came up with this, although I'm sure there
are more elegant ways to address the problem:

def printype( obj ):
	naym = `type(obj)`       #    string it,
	i = naym.find("'")
	naym = naym[(i+1):]      #    trim it,
	i = naym.find("'")
	naym = naym[:i]
	return naym              #    return it!   (as a string)

    Hey, what can I say? It works. Usually.    ; )

Hope that helps.

--
yrs,
in_sanity,

idfx


Sent via Deja.com
http://www.deja.com/


From adaminthedomaindevtty.net  Sat Feb 10 00:34:56 2001
From: adaminthedomaindevtty.net (Adam Logghe)
Date: Fri, 9 Feb 2001 15:34:56 -0800
Subject: PEP status and python-dev summaries
References: <mailman.981742518.20824.python-list@python.org> <3dg0hn8zlr.fsf@ute.cnri.reston.va.us>
Message-ID: <3a847ca3$1_1@news.nwlink.com>

Thanks for the summaries Andrew. I will be very sorry to see them go.

While I didn't see any need to comment, I was very glad to have an idea of
what was in the pipe. I would very much like to see the PEPs themselves as
they come through so I can plan ahead for changes to my own code as Python
moves along.

I would suggest that the full text of the PEP should be posted for every
change.

While this would be painful on a mailing list, I would not find it to be an
issue whatsoever on a newsgroup.

As a guide to what has changed, ideally a diff from the old version would be
attached on the end of the new PEP version. This would allow a quick visual
look at the changes. It may not be easy to make perfect sense of the change,
it would quickly point out whether the PEP had undergone minor or major
change.

I would suggest that publishing often provides at least two advantages.

A. Eliminating the decision of whether or not to publish will eliminate the
tendency for the  PEP undergo significant "creep" as minor changes add up to
significant changes in actual function. I think anyone that programs can
agree that minor syntax changes can and do have profound effects.

B. Having the PEP show up in front of people at multiple points during it's
life will expose it to more eyeballs.

If full posting is not pursued perhaps you might do something like-

A weekly summary of the mail traffic on the list simply by listing of the
active PEPs and then the number of times that PEP number appears in email
subject or body over the week. I would hope that people try to include the
PEP number in the subject line as a matter of courtesy anyway.

While this may not be the most accurate it would give us a simple
representation of the activity around a PEP and would perhaps spur further
investigations.

Thanks again. Sorry to see the Summary go but you shouldn't let it get in
the way of coding. ; )

I will look forward to however Jeremy and Barry decide to proceed.

Adam Logghe
adam in the domain devtty.net


"Andrew Kuchling" <akuchlin at mems-exchange.org> wrote in message
news:3dg0hn8zlr.fsf at ute.cnri.reston.va.us...
> Jeremy Hylton <jeremy at alum.mit.edu> writes:
> > - Regular announcements of PEP creation and PEP status changes should
> >   be posted to comp.lang.python and c.l.p.a.
>
> Should the full text of PEPs be posted at some point?  Perhaps the
> first draft, and significant revisions, should be posted so that
> people can follow-up directly to the PEP.  "Significant" will be left
> up to the PEP author; you wouldn't repost after every typo or minor
> clarification, but if your first approach was completely bogus, you'd
> post the massively revised version.
>
> --amk




From richard at iopen.co.nz  Wed Feb 14 22:40:14 2001
From: richard at iopen.co.nz (Richard Waid)
Date: Thu, 15 Feb 2001 10:40:14 +1300
Subject: CPAN functionality for python
References: <mailman.982142776.25775.python-list@python.org>
Message-ID: <96eu1l$1m3r$1@raewyn.paradise.net.nz>

In article <mailman.982142776.25775.python-list at python.org>, "Oleg
Broytmann" <phd at phd.pp.ru> wrote:

> On Tue, 13 Feb 2001, Doug Hellmann wrote:
>> >    Ok, so we've found one of these 2 or 3 mirrors.
>> >
>> > >     - Each mirror runs a ZEO client program and its own Zope
>> > >     server.
>> >
>> >    What will be these other 1 or 2 mirror sites that will run Zope?
>>
>> Why do you assume no one will run Zope?
> 
>    I didn't assume. I know for sure there are not much Zope
>    installations, and most FTP mirrors will not agree to run Zope for us.

Something that seems to be overlooked here - Zope might not be run by
large numbers of sites, but the sites that do run it are usually by
people fairly keen on Python. I'd suggest that while we might not get
huge numbers of mirrors initially, we'd get _enough_ (at least 1-3
per continent to begin with).

The advantage of Zope is that you could prototype the system quickly and
easily, as well as being an (almost) pure python solution, which just
seems 'right' somehow :)

Has anyone mentioned using XMLRPC to transfer the metadata and requests
around? It would be trivial to leverage that feature with Zope. The
actual transfers could be done out-of-band, using ordinary FTP or HTTP.

I'd suggest that there were two kinds of mirrors - meta data, and actual
data. The actual data could be stored on a regular FTP server, the
metadata could be accessed via XMLRPC from a Zope server. A little like
www.rpmfind.org in a way.

Richard Waid
Network/Software Engineer
http://iopen.co.nz


From quinn at lira.ugcs.caltech.edu  Mon Feb 12 00:53:26 2001
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 11 Feb 2001 23:53:26 GMT
Subject: Closing a file before or after return?
References: <960t2j$9n9$1@cubacola.tninet.se>
Message-ID: <slrn98e9fm.qne.quinn@lira.ugcs.caltech.edu>

On Fri, 9 Feb 2001 14:58:39 +0100, Gustaf Liljegren <gustafl at algonet.se> wrote:
>I'm using Python 2.0.
>
>Have a look at this function. It checks if a particular user (mail adress)
>exist in a .htpasswd file. The user name is what comes before ':' on each
>line.
>
># Check if user exists
>def user_exist(user):
>  f = open('.htpasswd', 'r')
>  for line in f.readlines():
>    i = string.find(line, ':')
>    if line[:i] == user:
>      return 1
>
>I'd like to close the file too, but I don't know where to put that
>statement. Of course, the file should be closed in both cases -- not just
>when the user is found.

# Check if user exists
def user_exist(user):
  f = open('.htpasswd', 'r')
  try:
      for line in f.readlines():
        i = string.find(line, ':')
        if line[:i] == user:
          return 1
  finally:
    f.close()

... will work.  But it's not necessary, since python will close the file for
you when the last reference disappears.  Explicitly deallocating anything more
complicated than memory or fds is usually good practice, though.


From MarkH at ActiveState.com  Sat Feb 24 02:44:27 2001
From: MarkH at ActiveState.com (Mark Hammond)
Date: Sat, 24 Feb 2001 01:44:27 GMT
Subject: win32com and makepy
References: <88id9tgj54l5q8rgvtss4tc9u9k91kd0kn@4ax.com>
Message-ID: <3A97111F.9050705@ActiveState.com>

Sean Laurent wrote:

> Howdy  All,
> I'm having a strange problem with win32com.client.Dispatch().  I used
> win32com\client\makepy.py to successfully generate Python sources.
> Unfortunately, after I call win32com.client.Dispatch(), the object I
> get back is of type "COMObject" instead of one of the generated
> classes.

This comes up fairly regularly, and unfortunately it depends on the object in question.

The simplest work around is to use the classes in that generated module directly.  Eg:


mod = gencache.EnsureModule('{000C1092-0000-0000-C000-000000000046}',1033, 1, 0)

i = mod.WhateverTheClassNameIs()

And to make life more painful:

>>>> db = i.OpenDatabase("d:\\temp\\test.msi", 0)

Will need to done like:

db = i.OpenDatabase("d:\\temp\\test.msi", 0)
db = mod.Database(db) # Assuming "Database" is the class name

Mark.



From shaleh at valinux.com  Sat Feb 17 08:07:52 2001
From: shaleh at valinux.com (Sean 'Shaleh' Perry)
Date: Fri, 16 Feb 2001 23:07:52 -0800
Subject: a few linux/unix related questions
In-Reply-To: <3a8e191f$0$25513$7f31c96c@news01.syd.optusnet.com.au>; from c941520@alinga.newcastle.edu.au on Sun, Feb 18, 2001 at 05:19:38PM +1100
References: <3a8e191f$0$25513$7f31c96c@news01.syd.optusnet.com.au>
Message-ID: <20010216230752.A14988@valinux.com>

On Sun, Feb 18, 2001 at 05:19:38PM +1100, Ben de Luca wrote:
> I was wondering how i catch various signals in nix python so i can exit my
> programs gracefully?
>

import signal
 
> also how might i start  another application then have two way communication
> between it's shell and the python program that started it?
>

os.popen() # just like in C

DESCRIPTION
       The  popen()  function opens a process by creating a pipe,
       forking, and invoking the shell.  Since a pipe is by defi-
       nition  unidirectional, the type argument may specify only
       reading or writing, not both; the resulting stream is cor-
       respondingly read-only or write-only.
 



From aahz at panix.com  Wed Feb  7 14:35:03 2001
From: aahz at panix.com (Aahz Maruch)
Date: 7 Feb 2001 05:35:03 -0800
Subject: Variable depth of nesting
References: <g0k872soog.fsf@scumbag.ecs.soton.ac.uk>
Message-ID: <95riu7$6f4$1@panix3.panix.com>

In article <g0k872soog.fsf at scumbag.ecs.soton.ac.uk>,
Jacek Generowicz  <jmg at ecs.soton.ac.uk> wrote:
>
>Is it somehow possible to write loops with variable nesting depths?

Try recursion.
-- 
                      --- Aahz (Copyright 2001 by aahz at pobox.com)

Androgynous poly kinky vanilla queer het    <*>     http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6

Life as a contractor: when they tell you it's going to rain, sometimes
it drizzles and sometimes it pours.


From Noah at noah.org  Thu Feb 22 09:27:02 2001
From: Noah at noah.org (Noah)
Date: Thu, 22 Feb 2001 00:27:02 -0800
Subject: Non-blocking pipe read under Windows NT
Message-ID: <q34l6.297$0Q5.207467@news.pacbell.net>

Hmmm... Windows Python2.0 does not have fcntl.
How do I do a non-blocking select on a Pipe file descriptor?

Under UNIX I can do something like this:

# hard way to read a file

import os, fcntl, FCNTL

# Open a pipe to read a file.
(fin, fout, ferr) = os.popen3 ('cat testfile')

# Turn off blocking on file. read() will then return -1 if no data
flags = fcntl.fcntl (fout.fileno(), FCNTL.F_GETFL, 0)
flags = flags | FCNTL.O_NONBLOCK
fcntl.fcntl (fout.fileno(), FCNTL.F_SETFL, flags)

# Print out the file without blocking.
done = 0
while !done:
    (r,w,e) = select.select ([fout], [], [], None)
    if len(r) > 0:
        data = r[0].read()
        if data == -1:
            done = 1
        else:
            print data




From fellowsd at cs.man.ac.uk  Thu Feb 22 10:54:16 2001
From: fellowsd at cs.man.ac.uk (Donal K. Fellows)
Date: Thu, 22 Feb 2001 09:54:16 +0000
Subject: Weird Language Features
In-Reply-To: <008F0A63472BD311AF9800104BCD102561CCA6@minirex.vmlabs.com>
	(message from John Schmitt on Wed, 21 Feb 2001 20:19:05 -0800)
References: <008F0A63472BD311AF9800104BCD102561CCA6@minirex.vmlabs.com>
Message-ID: <E14VsS0-0003n9-00@athene>

John Schmitt <jschmitt at vmlabs.com>
> Doesn't Applescript do something like this?  I recall browsing the docs
> where they mentioned that you can program Applescript in more than one
> dialect.  I guessed from the docs that you can write Applescript program in
> English, French, and Japanese.  If you write an Applescript program in one
> language (ie English), change your dialect (ie to French) and load the
> program in your editor again, the source code will show up in the new
> language  (ie French).  Do I understand that correctly?

It's a Mac, so I've no idea!  I have heard rumours to this effect
though.

However, to me this says that the raw program text is a binary form
that is never actually displayed to anyone in the normal course of
things.  Furthermore, this auto-translation is deeply unlikely to
affect the bits of the program that usually need attention; strings
shown to users (and possibly the names of identifiers in the program
if you need it to be maintained by someone with a different native
language.)  Somehow I'm fairly sure that AppleScript can't change the
language of those automatically (since no-one else in the world can
either, as many manuals are ample demonstration of...  :^)

A downside of this is that it makes it difficult for people in
different countries to discuss applescript properly, since a correct
example fragment for one person might be complete gibberish for
another.  You'd have to pass structured documents about instead of
just plain text messages...  :^(

Donal.
-- 
Donal K. Fellows, Department of Computer Science, University of Manchester, UK.
(work) fellowsd at cs.man.ac.uk     Tel: +44-161-275-6137  (preferred email addr.)
(home) donal at ugglan.demon.co.uk  Tel: +44-1274-401017   Mobile: +44-7957-298955
http://www.cs.man.ac.uk/~fellowsd/  (Don't quote my .sig; I've seen it before!)



From tim_one at email.msn.com  Thu Feb  1 07:36:19 2001
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 1 Feb 2001 01:36:19 -0500
Subject: "in" for dicts (was: Python 2.1 function attributes)
In-Reply-To: <20010130095928.A962@xs4all.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEIEIMAA.tim_one@email.msn.com>

[Tim]
> we also haven't pulled the platform-dependent tricks Perl does
> to optimize char-at-a-time reading the way vendors *should*
> optimize fgets() but almost never do (Tru64 Unix appears to be
> the sole probable exception to date).

[Thomas Wouters]
> FreeBSD, actually, had insane performance in that area as well.
> (getc() being just as fast as getc_unlocked(), in processes with
> only a single thread active.)

This one doesn't have anything to do with locking or not, it has to do with
whether the inner loop is optimized to use register shadows of _iobuf
members, and to use the platform's moral equivalent of the _cnt member to
avoid needing to check for EOF in the inner loop.  It's not "legal" to do
that in user-level code, because it requires breaking into the FILE*
abstraction.  Vendors can do it under the covers, although apparently almost
never do.  The unique thing about Tru64 Unix was the report that using the
fgets() method was substantially faster than using the getc_unlocked()
method; that's darned hard to account for unless Tru64 optimizes the inner
loop of fgets() (as Perl does).

IOW, just using getc_unlocked() still leaves a pile of *potential*
improvement on the floor, even on a single-thread run, but improvement you
can't get at without cheating (unless the vendor libc authors were savvy
enough to do it for you).

If Perl while(<>) is still substantially faster on FreeBSD than Python 2.1
using the fgets() method in single-thread runs, platform failure to optimize
fgets() in this way is a top contender for "why".




From syver at NOSPAMcyberwatcher.com  Fri Feb 16 02:00:14 2001
From: syver at NOSPAMcyberwatcher.com (Syver Enstad)
Date: Fri, 16 Feb 2001 02:00:14 +0100
Subject: Python emacs mode (newbie)
Message-ID: <96hu7d$4pq$1@troll.powertech.no>

Is there anybody who has any idea on how to implement a module loading
function in GNU Emacs like the one Pythonwin has?

The basic idea is that instead of writing a file name you write a module
name, like ex:
httplib or win32com.client.dynamic and the editor will load the specified
module.

I guess the implementation would work around searching the python path, but
I am total newbie in elisp (and lisp itself for that matter) and could use
some pointers as to how to search the filesystem from elisp, get the
registry keys or environment variables for the pythonpath, and how to load a
file into the current window.




From h4rv3st.m00n at gmx.net  Thu Feb  1 20:00:50 2001
From: h4rv3st.m00n at gmx.net (Harvest T. Moon)
Date: Thu, 1 Feb 2001 20:00:50 +0100
Subject: Q: elements in a list [newbie]
Message-ID: <95cbfs$gu0iq$1@ID-22517.news.dfncis.de>

i hope this is not some kind of 'very stupid' question, how can i find out
how many elements are in a list?
(generated by file.readlines())

Harvest T. Moon




From peter at engcorp.com  Thu Feb  8 06:42:44 2001
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 08 Feb 2001 00:42:44 -0500
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au>
Message-ID: <3A8231D4.1E5951A6@engcorp.com>

Austin Wilson wrote:
> 
> I have just downloaded py2exe and I am trying to freeze a program.  However,
> I am getting the following output.  Can anyone help.
> 
> BTW:  I am using Python20 and wxWindows 2.2.2 on Win98.
> 
> changing back to 'C:\Program Files\Python20'
[snip]
> ValueError: list.remove(x): x not in list

I got the same thing ("list.remove(x) raising ValueError") 
but happily py2exe _still_ managed to generate the 
executable successfully and I was able to run the resulting
program without difficulty!

Not that this helps much, but I have to say I was
pretty pleased that something that came without
guarantees, which I couldn't take the time to
fix when it failed, still managed to do the job
I needed in a pinch.

A big Thank You to the py2exe crew!  :-)


From bill at libc.org  Mon Feb 12 21:00:55 2001
From: bill at libc.org (Bill Anderson)
Date: Tue, 13 Feb 2001 03:00:55 +0700
Subject: Embedding UNIX Commands
References: <13EE655665F4D311AB4D0008C789498A01EE2C3C@zei02exm02.cork.cig.mot.com> <mailman.982008497.29571.python-list@python.org> <kuwvavxvyu.fsf@lasipalatsi.fi>
Message-ID: <rB7i6.1898$AX.389928@news03.micron.net>

In article <kuwvavxvyu.fsf at lasipalatsi.fi>, "Erno Kuusela"
<erno-news at erno.iki.fi> wrote:

> In article <mailman.982008497.29571.python-list at python.org>, D-Man
> <dsh8290 at rit.edu> writes:
> 
> | I don't think you can use setenv -- that is a csh thing, and system()
> | won't even start up a shell (AFAIK).  Even if it did, it wouldn't do
> | you any good since that shell would terminate and its environment
> | would be meaningless.  (Also, on Linux systems bash is the default
> | shell)
> 
> actually, system() (and popen()) pass the command line to /bin/sh under
> unix.

Which _is_ bash (on most Linux Dists):
ucntcme at locutus in /home/ucntcme 
$ ll /bin/sh lrwxrwxrwx    1 root     root            4 Jan 13 19:58 /bin/sh -> bash*


From thomas.heller at ion-tof.com  Thu Feb  8 10:48:05 2001
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 8 Feb 2001 10:48:05 +0100
Subject: Problem with py2exe
References: <95s54j$c72$1@gossamer.itmel.bhp.com.au>
Message-ID: <95tq0o$ie1b4$1@ID-59885.news.dfncis.de>

"Austin Wilson" <wilson.austin.aj at bhp.com.au> wrote in message news:95s54j$c72$1 at gossamer.itmel.bhp.com.au...
> Hi
>
> I have just downloaded py2exe and I am trying to freeze a program.  However,
> I am getting the following output.  Can anyone help.
>
> BTW:  I am using Python20 and wxWindows 2.2.2 on Win98.
>
> Thanks
> Austin
>
>
> changing back to 'C:\Program Files\Python20'
> creating dist\oww\oww.exe
> Using stub 'py2exe\run.exe'
> Resolving binary dependencies
> Traceback (most recent call last):
>   File "setup.py", line 6, in ?
>     scripts=["oww.py"],
>   File "c:\program files\python20\lib\distutils\core.py", line 138, in setup
>     dist.run_commands()
>   File "c:\program files\python20\lib\distutils\dist.py", line 829, in
> run_commands
>     self.run_command(cmd)
>   File "c:\program files\python20\lib\distutils\dist.py", line 849, in
> run_command
>     cmd_obj.run()
>   File "py2exe\py2exe.py", line 365, in run
>     self.copy_dependend_dlls(final_dir, use_runw, dlls)
>   File "py2exe\py2exe.py", line 417, in copy_dependend_dlls
>     alldlls.remove(self.get_exe_stub(use_runw))
> ValueError: list.remove(x): x not in list
>
>
>
Which py2exe version did you use?
Can you send me your script by private email so that I can take a look?

Thanks for trying it,

Thomas




From db3l at fitlinxx.com  Wed Feb 14 02:18:45 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 13 Feb 2001 20:18:45 -0500
Subject: breasking out of a while loop (in an if statement)
References: <3A899AD8.B4182CA4@cs.strath.ac.uk>
Message-ID: <u66ieoydm.fsf@ctwd0143.fitlinxx.com>

Lee Reilly CS1997 <lreilly at cs.strath.ac.uk> writes:

> Hi, I wonder if anyone could be kind enough to point out what I am doing
> wrong here (please see code below).
(...)
> -=---=-=--=-==-=-=-=-==
> while (j!=0):
>     if (radiobutton=='comma'):
>         tempString = string.split(text[i-1], ",")
>     else:
>         tempString = string.split(text[i-1], "\t")
> 
>     matric = string.rstrip(string.lstrip(tempString[0]))
>     SQL = "select * from STUDENTS where matric= '" + matric + "'"
>     result=db_conn.query(SQL)
>     result = len(result)
>     if (result>=3):
>        return "Student with matric value " + matric + " already exists
> in the database"
>        # break
>     else:
>         j=j-1
> -=---=-=--=-==-=-=-=-==
> 
> => Error Type: IndexError
> => Error Value: list index out of range
> 
> I'm using Python with Zope BTW.
> 
> Can anyone see my problem? Any help would be greatly appreciated ;-)

It would help to have the full traceback since that would highlight
the line where the actual error is occuring.  Also, it looks like this
code depends on surrounding code (e.g., where is "i" or "text" set?),
so having more detail about that surrounding code might help.

But given the error, and assuming that the error is actually within
this specific code, then it would seem to me to either be arising from
your reference to "text[i-1]" or "tempString[0]".

In the former case, you should verify that "text" has the contents you
expect and that it has enough entries in any pass through the loop as
indexed by "i-1".

I think the latter case is safe, since from what I can see, using
string.split with a dividing character will always at least produce a
single entry in the list, although it may be an empty string.  But you
might want to verify it in this actual code as well, since at least in
Python 1.5.2, string.split does return an empty list (e.g., if I split
an empty string on the default separator) in some cases, which could
then yield the IndexError when you try to access element 0.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From ssthapa at classes.cs.uchicago.edu  Wed Feb 21 02:12:28 2001
From: ssthapa at classes.cs.uchicago.edu (ssthapa at classes.cs.uchicago.edu)
Date: Wed, 21 Feb 2001 01:12:28 GMT
Subject: What of Siphon?
References: <mailman.982689203.3370.python-list@python.org> <Npyk6.96825$Wq1.39953628@nnrp5-w.sbc.net> <3A92C86D.282751E3@andrewcooke.free-online.co.uk>
Message-ID: <slrn995obe.4nc.ssthapa@ntcs-ip45.uchicago.edu>

Andrew Cooke <andrew at andrewcooke.free-online.co.uk> wrote:
>
>My ISP seems to have decided the rest of the world doesn't exist (apart
>from usenet, which makes it a very odd world that remains), but isn't
>this connected with the CPAN-for-Python project?  The name rings a bell.
>
>May be completely wrong...
>Andrew

  Yes, it is.  


-- 
----------------------------------------------------------------------------
	   		              |
Suchandra Thapa                       | "There are only two kinds of math . 
s-thapa-11 at NOSPAMalumni.uchicago.edu  | books. Those you cannot read beyond 
			              | the first sentence, and those you 
			              | can not read beyond the first page."
			              |                     -C.N. Yang
----------------------------------------------------------------------------


From marc.vollmer at mahr.de  Tue Feb 27 08:52:22 2001
From: marc.vollmer at mahr.de (Marc Vollmer)
Date: Tue, 27 Feb 2001 08:52:22 +0100
Subject: Newbie: Problems with PythonWin and wxPython
Message-ID: <97fmla$p8o62$1@ID-59219.news.dfncis.de>

Hello,

My system is:
Windows 2000
PythonWin 2.0 (win32all build 135)
wxPython

I start the example >>wxpython\lib\wxPlotCanvas.py<< with F5. At the first
and second time the program runs normally, but by the third start pythonwin
cancel with memory (read) problems.

How I can elimate this problem?

Thanks

Marc Vollmer




From stephen_purcell at yahoo.com  Wed Feb 14 11:31:29 2001
From: stephen_purcell at yahoo.com (Steve Purcell)
Date: Wed, 14 Feb 2001 11:31:29 +0100
Subject: Getting an instance's class name?
References: <96dkeb$fcj$1@news.rz.uni-karlsruhe.de>
Message-ID: <20010214113129.A9420@freedom.puma-ag.com>

Bjoern Giesler wrote:
> can anyone tell me how to obtain the name(s) of the class(es) that
> constitute an instance's type?
> 

With the '__class__' attribute of the instance, and the '__name__' attribute
of the class:

>>> p = Exception()
>>> p.__class__.__name__
'Exception'
>>> 

Note that you'll get an attribute error for objects that are not class
instances. For non-class instances, a handy way to get the type name is:

>>> p = 1
>>> type(p).__name__
'int'
>>>


-Steve

-- 
Steve Purcell, Pythangelist
http://pyunit.sourceforge.net/
http://pyserv.sourceforge.net/
Available for consulting and training.
"Even snakes are afraid of snakes." -- Steven Wright



From db3l at fitlinxx.com  Wed Feb 28 02:26:13 2001
From: db3l at fitlinxx.com (David Bolen)
Date: 27 Feb 2001 20:26:13 -0500
Subject: Nested scopes resolution -- you can breathe again!
References: <LNBBLJKPBEHFEDALKOLCMEEPJBAA.tim.one@home.com> <mailman.982897324.9109.python-list@python.org> <V0yJ0qA4Ejl6EwIt@jessikat.fsnet.co.uk> <ulmqt158l.fsf@ctwd0143.fitlinxx.com> <B8hhIIArT4m6EwC7@jessikat.demon.co.uk>
Message-ID: <uu25fmwcq.fsf@ctwd0143.fitlinxx.com>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:

> yes, but it would be just as easy to get the failure to do that. The
> message may come from windows, but the extension (in our case zlib)

The reason I suggested placing a python15.dll on the system is to get
rid of the Windows message, leaving it just to the Python ImportError.

> still wasn't loaded and that should have been enough, but we had a
> try/except around the particular import. Finding a matching pyd/dll
> which then fails to load is surely not 'just an import error' even if it
> eventually gets raised as one.

Well, but it doesn't have to be too much work.  I'm thinking of the
specific ImportError of:

 "Module use of python15.dll conflicts with this version of Python."

in which case you know that you have an old extension (which has to be
on Python's sys.path somewhere), and that one of three things are
true:

1. The imported module directly depends on python15.dll (probably most
   likely for C extension modules being directly imported, like zlib).
2. The module you imported itself imports some other module with the
   dependency - in that case you have the traceback showing which import
   of which module failed, which reverts to the prior 1.
3. The module has a Windows DLL dependency which eventually pulls in
   python15.dll.  In this case, you know the top level module, so use
   a standard windows tool like depends to check the dependencies.

It's really only case (3) where you need something more than Python
itself provides, but even in (3) you know the module to check, and
it's a relatively straight forward path to fully identifying the root
cause.  It should also be reasonably rare since most extension modules
directly use python15.dll rather than linking to another DLL which itself
is the extension.

Of course, none of this is to excuse the mess that is Windows DLL-land :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From guido at digicool.com  Mon Feb 26 19:22:45 2001
From: guido at digicool.com (Guido van Rossum)
Date: Mon, 26 Feb 2001 13:22:45 -0500
Subject: [Distutils] Re: CPAN functionality for python
In-Reply-To: Your message of "Mon, 26 Feb 2001 11:14:35 MST."
             <Pine.LNX.4.33.0102261043060.28411-100000@bms> 
References: <Pine.LNX.4.33.0102261043060.28411-100000@bms> 
Message-ID: <200102261822.NAA26272@cj20424-a.reston1.va.home.com>

> There are enough different packaging schemes around that it seems
> unnecessarily heavy-handed to pick one (or two or three) over the
> others - Python should not appear to favor any paticular system.  The
> only way I can see to do that is to not package to anyones spec, but
> provide enough information so that anyone can package to their
> fav spec.

But practicality beats purity: if a large fraction of users,
especially newbie users, are using a specific platform, it makes a lot
of sense to provide pre-packaged distributions for them!  E.g. Windows
installers, Red Hat RPMs.  (So far my experience with Windows
installers is much more positive than with RPMs though -- RPMs forever
seem to depend on some version of some other RPM that you don't have.)

--Guido van Rossum (home page: http://www.python.org/~guido/)



From owen at astrono.spamwashington.emu  Fri Feb 23 17:41:26 2001
From: owen at astrono.spamwashington.emu (Russell E. Owen)
Date: Fri, 23 Feb 2001 08:41:26 -0800
Subject: package question: getting names one level deep?
Message-ID: <owen-2302010841550001@cs205-86.fsmodem.washington.edu>

I like to break my code up into lots of files -- typically one file contains a class and turn a related set of files/classes into a package. When a file contains a class, naturally I give the file the same name as the class.

However, I've not found a nice way to refer to the resulting classes. I hope I'm missing some easy, standard trick.

Say I have a directory structure such as:
apackage/
  __init__.py  which contains __all__ = [list of all files]
  classa.py
  classb.py
  ...

To use these classes I write:

import apackage
x = package.classa.classa(etc)

It's that duplicated file-name.identical-class-name that I'd like to eliminate, so I could say:

import apackage (or something)
x = apackage.classa(etc)

Basically I'd like to do an import * into the "apackage" namespace, and ideally I'd like "import apackage" to somehow magically do it. In other words, ideally I'd like to set up my packages so they act as if all their code (at least all intended-to-be-visible-to-the-outside-world code) was in a single module.

Any hints?

-- Russell


From adolfo.benin at riskmap.it  Thu Feb 15 19:55:33 2001
From: adolfo.benin at riskmap.it (Adolfo Benin)
Date: Thu, 15 Feb 2001 19:55:33 +0100
Subject: COM server info
In-Reply-To: <E14TOS4-00065t-00@mail.python.org>
Message-ID: <5.0.2.1.0.20010215193232.00a63eb0@mail1.cs.interbusiness.it>

Hi,
	I've developed a COM object in Python, then I regitered it as a DCOM on my 
machine.
Every time a client makes a request to the server DCOM, the Python 
interpreter is launched and the COM is created. But when the client ends, 
the COM server also ends.
So every call to the server creates a new python interpreter instance in 
memory!

Is there a way to keep the COM server alive using always the same COM 
object instance?
I have already performed this task succesfully using CORBA (single servant 
CORBA object), but I need to expose the server as a COM object too.

Have you got any suggestion?

Thanks,
Adolfo.




From dbrueck at edgix.com  Tue Feb 13 16:50:12 2001
From: dbrueck at edgix.com (Dave Brueck)
Date: Tue, 13 Feb 2001 08:50:12 -0700
Subject: spawnv question
In-Reply-To: <bcki8tc9bmoophugldu9k4mfju3emm4jc2@4ax.com>
Message-ID: <BLEOJGEMBNBNEAKMIFDHCEDJPJAA.dbrueck@edgix.com>

argv[0] is reserved for the program name.

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Daniel Klein
> Sent: Tuesday, February 13, 2001 8:32 AM
> To: python-list at python.org
> Subject: spawnv question
> 
> 
> Here is a snippet of code I'm using with the 'spawnv' command:
> 
> import os
> args = ('http://www.python.org', 'http://www.python.org') # This works
> os.spawnv(os.P_NOWAIT, 'c:\progra~1\intern~1\iexplore', args)
> print "done"
> 
> The Python Reference for the 'spawnv' command says:
> 
>  "Execute the program path in a new process, passing the 
> arguments specified in
> args as command-line parameters. args may be a list or a tuple."
> 
> My question is, why do I have to specify the url argument twice 
> in the 'args'
> variable. It doesn't seem to work properly if I specify it as a 1-element
> tuple, ie...
> 
> args = ('http://www.python.org',) # This doesn't work
> 
> Thanks,
> Daniel Klein
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list



From dsavitsk at e-coli.net  Fri Feb  2 23:38:50 2001
From: dsavitsk at e-coli.net (dsavitsk)
Date: Fri, 2 Feb 2001 16:38:50 -0600
Subject: ADO/ODBC call via Python/IIS
References: <95f558$jgl$1@nnrp1.deja.com> <E%Fe6.7998$o5.47333@e420r-atl1.usenetserver.com>
Message-ID: <%FGe6.31314$Ch.7369562@newsrump.sjc.telocity.net>

> >         adoConn.Open('pyasp')

> Is it possobile pyasp is a user DSN, when ASP expects systems DSNs?

if this is the issue, asp and python work well (faster in fact on a web
server) with dsn-less connection strings.

ds




From adam at deprince.net  Wed Feb 21 21:56:57 2001
From: adam at deprince.net (Adam DePrince)
Date: Wed, 21 Feb 2001 15:56:57 -0500
Subject: newbie - concatanating 2 lists
References: <mailman.982776626.16974.python-list@python.org> <slrn999ood.219.scarblac@pino.selwerd.nl>
Message-ID: <3A942B99.490C286C@deprince.net>

Remco Gerlich wrote:
> Sean 'Shaleh' Perry <shaleh at valinux.com> wrote in comp.lang.python:
> > On 21-Feb-2001 Gnanasekaran Thoppae wrote:
> > > i am just beginning to use python.
> > > i have:
> > > li1 = ['a', 'b', 'c']
> > > li2 = ['x', 'y', 'z']
> > > i want:
> > > li3 = ['ax', 'by', 'cz']
> > > how do i do it?
> > python 2 has a lovely function called zip(), maybe it could help.  Otherwise, a
> > quick map would probably do this.
> For Python 2, with zip and list comprehensions, it would look like
> li3 = [a+b for a,b in zip(li1,li2)]
> Hmm, or even
> li3 = map(''.join, zip(l1,l2))
> I like
> li3 = map(operator.add, li1, li2)
> best (why do people write their own lambda for operator.add? :)).
> But all of these are trying to get it to fit in the least number of
> characters (and they're pretty fast, although I have doubts about the zip
> things).
> A newbie should also understand the basic way:
> 
> A = ['a','b','c']
> B = ['d','e','f']
> C = []
> for i in range(len(A)):

And if you want to handle mismatched lists gracefully ... 

  for i in range(min((len(A),len(B)))):


>    C.append(A[i]+B[i])
> 
> This will give an exception if l1 is longer than l2, but all the above
> methods have problems with different length lists as well.
> 
> zip(), list comprehensions, string methods etc are cool, but we end up
> with a lot of ways to do it...
> --
> Remco Gerlich


From greg at perceval.be  Thu Feb  8 12:54:00 2001
From: greg at perceval.be (Gregoire Welraeds)
Date: Thu, 8 Feb 2001 12:54:00 +0100 (CET)
Subject: None assigment
Message-ID: <Pine.BSF.4.21.0102081248050.4685-100000@localhost>

While playing a bit with python 2.0, I found that I can assign some value
to None

EG: None= 2

which I found to be very pernicious because then the following, for
example, won't work any more:

>>> b= filter(None, [1,2,'',3])

Bug or feature ?

If it is a feature (I suppose it is not a bug :), what is the interest of
this ?

Gregoire Welraeds
Perceval Development team
-------------------------------------------------------------------------------
Perceval Technologies sa/nv	Tel: +32-2-6409194		
Rue Tenbosch, 9			Fax: +32-2-6403154		
B-1000 Brussels			general information:   info at perceval.net
BELGIUM				technical information: helpdesk at perceval.net
URL: http://www.perceval.be/
-------------------------------------------------------------------------------




From uche at ogbuji.net  Wed Feb 14 15:12:30 2001
From: uche at ogbuji.net (Uche Ogbuji)
Date: Wed, 14 Feb 2001 14:12:30 GMT
Subject: DOM creation
References: <m3bss8k70l.fsf@boss.dreamsoft.com>
Message-ID: <3A8A9254.C4F9FDC@ogbuji.net>

Venkatesh Prasad Ranganath wrote:

> I have a question on how DOM for a XML document conforming to DOM 2 should be
> constructed?
> 
> Now if there are no namespaces specified in the document then should attributes
> be added to DOM using
> setAttributeNS('', Name, Value)
> or
> setAttribute(Name, Value)?
> 
> The problem I am facing is when reading in a XML document with no explicit
> namespace specified in it through PyXML the attributes are added to the DOM
> using setAttributeNS with an empty NameSpace.  So, I wanted to clarify if this
> is a problem with PyXML or is this how other DOM Constructors work.

This is correct behavior.  Of course, if you use the
xml.dom.ext.reader.Sax reader, you get a tree with no namespace
specifiers at all, which is also correct.

If you plan to migrate to namespaces in future, or to mix namespace with
non-namespace behavior, I'd suggest sticking to the PyExpat and Sax2
readers and using the DOm Level 2 methods (with appended "NS").

> waiting for reply,

This reminds me.  I'm not sure I sent a reply to your last enquiry.

I had a few questions, such as which Reader you were trying to use (It
looked as if you didn't paste all of your example code in).

However, I'd suggest trying the latest 4Suite 0.10.2 beta that was
announced (since I noticed you're using XPath), and see if you still
have those problems.  If so, pleace copy follow-ups to
xml-sig at python.org, which I check more regularly than this newsgroup.

Thanks.

-- 
Uche Ogbuji
Personal:   uche at ogbuji.net		http://uche.ogbuji.net
Work:       uche.ogbuji at fourthought.com	http://Fourthought.com


From aleaxit at yahoo.com  Thu Feb  8 23:35:34 2001
From: aleaxit at yahoo.com (Alex Martelli)
Date: Thu, 8 Feb 2001 23:35:34 +0100
Subject: C++ style Stream Operators
References: <3a831395.0@216.0.152.7>
Message-ID: <95v75u021ll@news2.newsguy.com>

"Jim" <jimholliman at heartsoft.com> wrote in message
news:3a831395.0 at 216.0.152.7...
> I'd like to use C++style stream i/o operators, such as "file_object <<
> some_stuff" in my python code.
> IS this sort of operator setup and what modules would I need to take
> advantage of them if so?

You can get a somewhat similar look with the new-ish
    print >> file_object, some_stuff
syntax.  Or, go whole hog (the expression somehow
seems particularly appropriate) and write a class with
an overloaded __lshift__ operator:

class Stream:
    def __init__(self, file):
        self.file = file
    def __lshift__(self, stuff):
        self.file.write(str(stuff))
        return self
    def __repr__(self):
        return ''

import sys
cout = Stream(sys.stdout)

so you can code beauties like cout<<"Hi there!".  The
'return self' at the end of __lshift__ lets you chain your
'<<'s, and the __repr__ even lets you try it out at the
interactive interpreter command line.

I think you'd be very badly advised to pursue this route,
spending energy to enable doubtful syntax sugar which
many Pythonistas will find quite unpleasant, but, in any
case, Python offers you "enough rope to shoot yourself
in the foot" -- and you could argue that "print>>flob"
is a precedent that legitimates what you're doing:-).


Alex





From digitig at cix.co.uk  Sun Feb 25 22:17:00 2001
From: digitig at cix.co.uk (Tim Rowe)
Date: Sun, 25 Feb 2001 21:17 +0000 (GMT Standard Time)
Subject: equivalent of NULL or OM in Python
References: <L1bj6.16746$GA5.260574@e420r-atl1.usenetserver.com>
Message-ID: <memo.20010225211730.34111A@digitig.compulink.co.uk>

In article <L1bj6.16746$GA5.260574 at e420r-atl1.usenetserver.com>, 
sholden at holdenweb.com (Steve Holden) wrote:

<snip>

> Choice of data representation can be critical in algorithm design, and 
> this
> is quite a good example. However, I have not fully considered the 
> symbolic
> manipulations you want to perform, so I may be putting my foot in my 
> mouth
> here. That's how I know I have a size 10 mouth: the foot fits perfectly.
> 
> It would appear to be more regular to represent absent terms as a
> coefficient of zero. Furthermore, if you reverse the order of the terms,
> putting x^0 at the left, you can then use the index into the list as the
> power. So your examples could be represented as follows:
> 
> 3X^2+4X+4 : [4, 4, 3]
> 3x^2+4: [4, 0, 3]

It depends what he wants to do. If he's doing things like CRC calculations 
the lists are likely to have a lot of zeroes. 

I think it would be worth considering numpy, as polynomial conversions are 
easily represented as array operations, and ISTR numpy is clued up on 
sparse arrays.


From neo at thezion.net  Wed Feb 21 20:10:04 2001
From: neo at thezion.net (Neo)
Date: Wed, 21 Feb 2001 20:10:04 +0100
Subject: Unicode error with ado in win 32
Message-ID: <9713oe$mse8b$1@ID-35431.news.dfncis.de>

I've created a simple script which read and print on the interactive window
the data from a database field (sql server 7).
It work properly but when into the rs.fields('azienda') theres the string
"abb?" the script returns the following message.

Traceback (most recent call last):
  File "g:\programmi\Python20\Pythonwin\pywin\framework\scriptutils.py",
line 298, in RunScript
    debugger.run(codeObject, __main__.__dict__, start_stepping=0)
  File "g:\programmi\Python20\Pythonwin\pywin\debugger\__init__.py", line
60, in run
    _GetCurrentDebugger().run(cmd, globals,locals, start_stepping)
  File "g:\programmi\Python20\Pythonwin\pywin\debugger\debugger.py", line
582, in run
    _doexec(cmd, globals, locals)
  File "g:\programmi\Python20\Pythonwin\pywin\debugger\debugger.py", line
921, in _doexec
    exec cmd in globals, locals
  File "G:\Programmi\Python20\fede.py", line 12, in ?
    print unicode(a)
  File "g:\programmi\Python20\win32com\client\dynamic.py", line 165, in
__str__
    return str(self.__call__())
UnicodeError: ASCII encoding error: ordinal not in range(128)

I've try to translate the string into my regional unicode character set, but
it not work.

Help me please.

The strange kind is that using an ODBC connection this error not appear.
Thanks.



--
Neo
****************************
Follow the White Rabbit...
Knock Knock Neo...
www.thezion.net
****************************

Below there's the block code that i've used.

Bye


dataconn.Open (strConn)
sql="select * from GpAzienda ;"
rs.Open(sql, dataconn)
while not rs.eof:
    a = gettext.gettext(rs.Fields('azienda'))
    print a
    rs.MoveNext()

del rs
del dataconn





From thomas at xs4all.net  Fri Feb  2 19:46:09 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Fri, 2 Feb 2001 19:46:09 +0100
Subject: nested scopes
In-Reply-To: <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk>; from robin@jessikat.fsnet.co.uk on Fri, Feb 02, 2001 at 04:05:58PM +0000
References: <msXosJAT0oe6Ewmn@jessikat.fsnet.co.uk> <Pine.SOL.4.30.0102021550090.28216-100000@mimosa.csv.warwick.ac.uk> <55q$5KAmrte6EwVo@jessikat.fsnet.co.uk>
Message-ID: <20010202194609.W962@xs4all.nl>

On Fri, Feb 02, 2001 at 04:05:58PM +0000, Robin Becker wrote:
> In article <Pine.SOL.4.30.0102021550090.28216-100000 at mimosa.csv.warwick.
> ac.uk>, John J. Lee <phrxy at csv.warwick.ac.uk> writes
> >On Fri, 2 Feb 2001, Robin Becker wrote:
> >
> >> Is it really true that J Hylton's nested scopes will cause code like
> >>
> >> def bingo():
> >>   from string import *
> >>   ....
> >>
> >> to be declared bad? I'm already against nested scopes, but requiring
> >> this to be wrong seems awful. We seem to be replacing a simple 2-level
> >> system by a more complex one & restrictions.
> >[...]
> >
> >Are nested scopes actually going to end up in Python proper, then?  As
> >opposed to remaining as an optional patch?

> don't know about that, apparently 2.1a has it

It won't be optional, it'll be part of Python. I guess that if there is
a *really* *really* *really* lot of code out there that will break, it could
be withheld until Python 2.2, and the practice of 'from foo import *' in
function/class scopes be made to issue a warning in Python 2.1, but frankly
I doubt there'll be that much code breakage. At least it's not a *silent*
breakage ! It's a compiler error, so the first run of a script under the new
Python will detect it. It's not like it's a new kind of exception being
raised.

> >And why should the above become bad rather than just import into the
> >appropriate namespace?
> >
> ...
> I have a feeling that in order to make nested scopes workable/acceptable
> the byte compiler is being hacked and this has had some unfortunate side
> effects.

Now, now, 'hacked' is a bit strong. Jeremy did a fairly extensive rewrite of
the compiler, complete and decent enough not to be called a 'hack' for sure.
Besides, developers never 'hack', they 'extend' :-) The problem with 'from
foo import *' in a function or class scope, though, is that it makes it
impossible for the compiler to see what names will be defined in that scope,
and hence what names a nested scope uses will resolve to that. I don't like
it at all, either. In fact, I was the first to whine about it, incessantly,
on python-dev. I think I have to buy Jeremy a beer (or whatever's his
poison) on the Python conference to make up ;)

The situation does lend itself to a workaround (a *true* hack, mind
you:) but I haven't looked into implementing it, yet. (Jeremy might be doing
that, but I'm not sure): The compiler could try and live with 'from foo
import *' as long as the function doesn't define a scope nested in that
function. Unfortunately, this isn't going to fix every instance of the
problem: a single lambda statement inside the function will make the import
illegal again. Disabling nested-scoping in the presense of 'from foo import
*' might be technically possible, but would create enormous amounts of
confusion, I'm sure.

And lastly, I'd like to point out (as Jeremy does :) that the reference
manual already states that 'from foo import *' in anything but a module
scope is illegal. Even if the nested scopes patch was withdrawn from Python
2.1, 'from foo import *' in local scopes *will* generate a warning (warnings
are also new in 2.1) so your best bet, either way, is sanitizing your code. 

(Don't get me wrong: I seriously think 'from foo import *' is overused. The
language reference makes it pretty clear what it was intended for, and I
think even the common 'from Tkinter import *' is wrong use -- but I don't
like code breakage, regardless.)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!



From hamish_lawson at yahoo.co.uk  Thu Feb 22 10:22:46 2001
From: hamish_lawson at yahoo.co.uk (hamish_lawson at yahoo.co.uk)
Date: Thu, 22 Feb 2001 09:22:46 -0000
Subject: Problems using Grail.
In-Reply-To: <3A943B33.CD618A9B@olen.to>
Message-ID: <972lp6+iiqq@eGroups.com>

Joonas Paalasmaa <joonas at o...> wrote:

> When I tried to start Grail by typing "python grail-0.6\grail.py"
> Python gave the following error message.

>   File "C:\Python16\grail-0.6\Stylesheet.py", line 45, in load
>     massaged.append((g, c), v % fparms_dict)
> TypeError: append requires exactly 1 argument; 2 given

The append() call is officially meant to take one argument, but in 
versions of Python before 1.6, multiple arguments would be understood 
as comprising a tuple. However this behaviour was never officially 
documented or approved, and was dropped in Python 1.6 to bring the 
actual behaviour in line with the official documentation. This line 
of code will therefore need to be modified to use an explicit tuple 
in the append argument:

    massaged.append(((g, c), v % fparms_dict))

Hamish Lawson





From bryan at eevolved.com  Sun Feb 18 15:32:40 2001
From: bryan at eevolved.com (Bryan Mongeau)
Date: Sun, 18 Feb 2001 14:32:40 GMT
Subject: 3D animation
References: <aJPj6.171612$KP3.43764577@news3.rdc1.on.home.com>
Message-ID: <c2Rj6.148997$Pm2.2652810@news20.bellglobal.com>

shaka wrote:

> 
> 
> Ya, I am planning to write a program to present a project at school.  I
> need
> a language that will allow me to do 3D animations and other animation. 
> And I am struggling between writting the program in java or using python.
> Can someone help me?  If python is the ultimate choice then can you give
> me some instructions on what libraries I should use, because I am quite a
> beginner here>
> 
> Thanks.
> 
> Shaka. Z
> 
> 

Honestly couldn't tell you if Java has anything better for this purpose but 
I can tell you python has what you need.

First you need a real GUI toolkit. Try:

http://sourceforge.net/projects/foxgui/
Make sure to compile with OpenGL support.

with the python bindings here:

http://sourceforge.net/projects/fxpy/

Grab the openGL bindings:

http://sourceforge.net/projects/pyopengl/

Now you should be set to look at the funky GL tests that come in the FXPy 
bindings. This should give you a great example of what you can do easily in 
python.

Good luck.
-- 
<==================================>
Bryan Mongeau
Lead Developer, Director
eEvolved Real-Time Technologies Inc.
http://www.eevolved.com
<==================================>

"This is one of the hardest lessons for humans to learn. We cannot admit 
that things might be neither good nor evil, neither cruel nor kind, but 
simply callous - indifferent to all suffering, lacking all purpose." -- 
Richard Dawkins



From johngrayson at home.com  Sun Feb 11 00:54:35 2001
From: johngrayson at home.com (John Grayson)
Date: Sat, 10 Feb 2001 23:54:35 GMT
Subject: New to Tkinter
References: <961qd3$50l$1@hecate.umd.edu>
Message-ID: <964kbq$su9$1@nnrp1.deja.com>

In article <961qd3$50l$1 at hecate.umd.edu>,
  "Arvind Mani" <marvind at glue.umd.edu> wrote:
> Hi,
>
> I  need a menu based GUI. Clicking on a menu item should open a new dialog
> box. I tried to do this - only the application main window disappears when
> the dialog box corresponding to a menu item pops up. Also it has the same
> title as the application main window.
>
> Thanks,
> Arvind
>

If you want help, you'll need to post some (or all) of your code
so that we can see what you're doing. In case you're not aware,
menus work pretty well...

    John


Sent via Deja.com
http://www.deja.com/


From kirschh at lionbioscience.com  Fri Feb  2 14:08:20 2001
From: kirschh at lionbioscience.com (Harald Kirsch)
Date: 02 Feb 2001 14:08:20 +0100
Subject: execfile('bla.py'), can bla.py know its full path
Message-ID: <yv2g0hxb5bv.fsf@lionsp093.lion-ag.de>

When a script is called by execfile, can it find out the full absolute
path name by which it was called?


  Harald Kirsch

-- 
----------------+------------------------------------------------------
Harald Kirsch   | kirschh at lionbioscience.com | "How old is the epsilon?"
LION Bioscience | +49 6221 4038 172          |        -- Paul Erd?s


From elflord at panix.com  Sun Feb 18 23:28:52 2001
From: elflord at panix.com (Donovan Rebbechi)
Date: 18 Feb 2001 22:28:52 GMT
Subject: Weird Language Features
References: <nnhv8t4obq30ljfgseplp7pi4khhsf9g9n@4ax.com>
Message-ID: <slrn990j54.pcq.elflord@panix6.panix.com>

On Sun, 18 Feb 2001 13:16:49 +0000, Dave Cross wrote:

>1/ The programmer calls a function that doesn't actually exist within
>the application (or libraries). Is the a feature whereby the
>programmer can create a "catch-all" function which is called in cases
>like these? Can this function examine the list of existing functions
>and call the most appropriate one? Or create a new function on the fly
>and install it into the application?

Posting this from comp.lang.c++. In C++, the object model itself does 
not support this.

In both C and C++, one uses dynamic loading to get this kind of behaviour.
Indeed, interpreters written in C and C++ tend to rely heavily on dynamic
loading.

To get the kind of functionality you're talking about, one would probably
store a function table in a map/tree and one could put a "default"
method in there that acts as a catch all.

However, typically one uses exceptions to deal with failed requests, the
idea being that if a request fails, the caller is in a better position 
than the library author to work out what to do.

>2/ Can ou filter the input source code before compilation (or
>interpretation) in some way so that language keywords could be changed
>for other strings? Imagine you wanted to allow someone to program your
>language of choice in, say, French. How would you go about translating
>French keywords into ones that the compiler (or interpreter) could
>understand. What if the translation wasn't one-to-one or fixed? Could
>you put enough intelligence into the translator so that it could
>handle certain strings differently depending on where they appeared in
>the source code?

One could do this with macros but it's a fairly primitive solution ...

>  Perl does
>have these features (see the AUTOLOAD function and source filters) and
>I'm interested in seeing how widespread they are in other languages.

... of course one could always write a preprocessor in perl (-;

>Of course, if you'd like to tell me just why you consider it's a good
>thing that your language of choice doesn't have these features, then
>I'd be only too happy to hear that too.

C and C++ are standardised (as opposed to implementation defined) languages,
the former is meant to be small, the latter is big enough as it is. 
Introducing support for this kind of thing in the language (either
C or C++) would probably be a mistake IMO.

-- 
Donovan Rebbechi * http://pegasus.rutgers.edu/~elflord/ * 
elflord at panix dot com


From footech at get2net.dk  Fri Feb 23 13:09:08 2001
From: footech at get2net.dk (Mikkel Rasmussen)
Date: Fri, 23 Feb 2001 13:09:08 +0100
Subject: Newbie: Large dictionaries
References: <WUpl6.43$Gv2.2355@news.get2net.dk> <mailman.982924567.27809.python-list@python.org>
Message-ID: <vpsl6.134$Gv2.4218@news.get2net.dk>

Hi!

I can also put more than 4 million entries *into* the dictionary. But I can
only get less than 65.000 back out!

What do you get with: len(dict.keys()) ?

And with a loop like

count = 0
for elem in dict:
    count = count + 1
print count

The keys are ordinary words like an ordinary dictionary :-) with an average
length of about 6 characters.


Mikkel Rasmussen


Steve Purcell <stephen_purcell at yahoo.com> wrote in message
news:mailman.982924567.27809.python-list at python.org...
> Mikkel Rasmussen wrote:
> > Why can't there be more than about 65.000 entries in a dictionary, or
more
> > importantly: how can I make a large dictionary with room for at least
> > 500.000
> > elements?
>
> It's probably a memory issue: how big are the keys and/or values?
>
> On my machine, I can easily get a million entries into a dictionary:-
>
> >>> dict = {}
> >>> try:
> ...   for i in xrange(1000000):
> ...     dict[i] = None
> ... except:
> ...   print "stopped at", i
> ...   raise
> ...
> >>>
>
> -Steve
>
> --
> Steve Purcell, Pythangelist
> Get testing at http://pyunit.sourceforge.net/
> Get servlets at http://pyserv.sourceforge.net/
> "Even snakes are afraid of snakes." -- Steven Wright
>




From bsass at freenet.edmonton.ab.ca  Tue Feb 27 23:37:21 2001
From: bsass at freenet.edmonton.ab.ca (Bruce Sass)
Date: Tue, 27 Feb 2001 15:37:21 -0700 (MST)
Subject: CPAN functionality for python - requirements
In-Reply-To: <20010227125351.H1781@tummy.com>
Message-ID: <Pine.LNX.4.33.0102271450220.30946-100000@bms>

I'm a little concerned about the privacy aspect of this...

> That decision is up to the client.  If the client has the smarts to turn
> a distutils package into an RPM, then the client would list it's preferred
> format as a distutils package and it would handle the rest.  If it can't,
> you can either select an RPM, or fall back to a distutils package.

Why should the client need to "list" anything.

client: what do you have?
server: this(deb,rh-5.rpm,rh-6.2.rpm,rh-7.rpm) that() other(hqx)
client: send me this(deb), that(mandrake.rpm), other()
server: sorry don't have that(mandrake.rpm)
client: send me that(deb)
server: sorry don't have that(deb)
client: send me that()

This is how I would handle a curious server...
client: what do you have?
server: what format do you want?
<click>

In other words... the server is just that - a server - it does not ask
questions, it does not keep track of any preferences, it serves up
files.

This could work...
client: only show me (deb) and ()
client: what do you have?
...

The server should NOT be usable as a tool to track Python users and
their habits, and making it do so should require a conscious effort on
the operators part (so there is no opportunity for the operator to
say, "I don't track you, that's just how the software works").


- Bruce




From adam at deprince.net  Fri Feb 23 07:34:04 2001
From: adam at deprince.net (Adam DePrince)
Date: Fri, 23 Feb 2001 01:34:04 -0500
Subject: may i get sample source ?
References: <mailman.982905784.23883.python-list@python.org>
Message-ID: <3A96045C.FC182C5D@deprince.net>

jayhawks at netsgo.com wrote:
> 
> HI?
> 
> may i get sample source about python + informix ?
> or
> where get i sample ?
> 
> i don't know wel about using python+informix .
> 
> i am new to python .
> 
> thank

You can look up python addons here: http://www.vex.net/parnassus/

Two interfaces to informix are here: 
http://www.vex.net/parnassus/apyllo.py?find=informix

Informixdb claims to support DB1.0; documentation on DB1.0 can be found
here: http://www.python.org/topics/database/DatabaseAPI-1.0.html

Adam DePrince
Starmedia Network, Inc.
Email:
zlib.decompress('x\332KLI\314\325KI-(\312\314KNu(.I,\312MM\311L\324K\316\317\005\000\221\331\012s')


From m.hadfield at niwa.cri.nz  Wed Feb 28 00:06:19 2001
From: m.hadfield at niwa.cri.nz (Mark Hadfield)
Date: Wed, 28 Feb 2001 12:06:19 +1300
Subject: output of shell command
References: <3A99FC5D.158A407B@al.com.au> <ci0k9tkke44ule7hlg2q7j6b8v5vsddlnk@4ax.com> <97diqu$nq7ks$1@ID-11957.news.dfncis.de> <UYAm6.21417$P87.223264@e420r-atl2.usenetserver.com> <983237981.694338@clam-ext> <u3dczkcpk.fsf@roundpoint.com> <983312067.173004@clam-ext> <pgWm6.169$jv4.27104@e420r-atl2.usenetserver.com>
Message-ID: <983315235.473229@clam-ext>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:pgWm6.169$jv4.27104 at e420r-atl2.usenetserver.com...
> Dunno. Still won't work for me on Windows 98. Neither, come to that, does
> the parenthesized version
>
>     a=(os.popen("dir")).readlines()
>
> I get the same error.

OK so the parentheses were a red herring. It's a difference in the behaviour
of piping in windowed applications vs console applications under Windows 9x.

---
Mark Hadfield
m.hadfield at niwa.cri.nz  http://katipo.niwa.cri.nz/~hadfield
National Institute for Water and Atmospheric Research




From tim.one at home.com  Mon Feb  5 04:52:03 2001
From: tim.one at home.com (Tim Peters)
Date: Sun, 4 Feb 2001 22:52:03 -0500
Subject: Install PythonWin on top of Python 2.0
In-Reply-To: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEIHINAA.tim.one@home.com>

[Deja User]
>   This is a question from a new user. I recently installed the
> BeOpen Python 2.0 on my Win98 PC. Now I'd like to install the Win32
> xtension so I download "win32all-125.exe". Unfortunately, I cannot
> install it because this build insists on the presence of Python 1.5.x.

That's because 125 only knew about 1.5.x -- unfortunately, it wasn't able to
see into the future so that it could work correctly with a release of Python
that didn't yet exist <wink>.

> Now, do I have to install 1.5 so as to install PythonWin? Would the
> author of PythonWin please do us a favor to relax the version check
> in PythonWin?

Heh.  Some favor!  That would probably send your machine crashing in flames.

You simply need to get a newer version of win32all.  Look here:

http://www.activestate.com/Products/ActivePython/win32all.html

The version for use with 2.0 is clearly marked.

Or you can uninstall BeOpen 2.0 and get ActiveState's 2.0 package (which
includes the Win32 extensions) instead:

http://www.activestate.com/Products/ActivePython/Download.html

consumer-choice-is-a-wonderful-thing-although-it's-just-confusing-
    at-first-ly y'rs  - tim




From daniel.dittmar at sap.com  Tue Feb  6 17:46:15 2001
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Tue, 6 Feb 2001 17:46:15 +0100
Subject: While everyone is saying what they want in Python :)
References: <3A7D64D8.74E71826@cybermesa.com> <3A7D9C1D.781F8E5@alcyone.com> <3A7DB154.87E04245@cybermesa.com> <axIf6.35318$p8.7719400@typhoon.southeast.rr.com> <3a8012f5.7625600@nntp.sprynet.com>
Message-ID: <95p9om$h23$1@news1.wdf.sap-ag.de>

In VB, although WITH can be nested, only the innermost is valid. Together
with the .dot notation makes your intentions pretty explicit. In Pascal, you
have
- nested with
- multiple expressions allowed in the with clause
- no dots
which leads to very messy code.

Although cascading in Smalltalk and with in VB are very different, in most
cases, they lead to similar code (Pythonified):

sys.stdout
    .write (...);
    .write (...);
    .write (...)

with sys.stdout:
    .write (...);
    .write (...);
    .write (...)

The main exception is object creation, where in Smalltalk, you can create an
object and send it a bunch of cascaded initialization messages.

One problem with cascading is that I don't know yet of a syntax using
indentation. And using ';' as a delimiter? There are already far too many
people writing 'if <cond>: thenstmt', all in one line.

Daniel

--
Daniel Dittmar
daniel.dittmar at sap.com
SAP DB, SAP Labs Berlin
http://www.sapdb.org/





From ben.hutchings at roundpoint.com  Wed Feb 21 01:07:45 2001
From: ben.hutchings at roundpoint.com (Ben Hutchings)
Date: 20 Feb 2001 16:07:45 -0800
Subject: Optimisation problem - in C?
References: <mailman.982698204.12938.python-list@python.org>
Message-ID: <uhf1o3nla.fsf@roundpoint.com>

Tim Churches <tchur at optushome.com.au> writes:
<snip>
> def makevector(sourcelist,elementlist):
> 	resultvector = []
> 	for element in elementlist:
> 		resultvector.append(sourcelist[element])
> 	return resultvector
<snip>
> My test data has about 300,000 elements in sourcelist and between about
> 10,000 and 70,000 elements in elementlist, but larger lists are
> envisaged, hence the earnest desire for maximum speed.
> 
> Questions:
> Can this function (which essentially does the fairly generic operation
> of "fetch the elements of sequence a using the element indexes contained
> in sequence b") be optimised without resorting to a C module?
<snip>

I think it can.  Since a list's contents are contiguous in memory (in
order to support random access), appending to it will cause it to
reallocate and move all its contents periodically.  If I remember
correctly, Python's lists increase their memory allocations by
constant amounts rather than by a constant factor, so this loop will
run in quadratic time, i.e. proportional to the square of the length
of elementlist!

Each of the following definitions should be equivalent to the above,
and I think at least one will run faster:

    def makevector(sourcelist,elementlist):
        resultvector = [None] * len(elementlist)
        for i in range(len(elementlist)):
            resultvector[i] = sourcelist[elementlist[i]]
        return resultvector

    def makevector(sourcelist,elementlist):
        return [sourcelist[element] for element in elementlist]

    def makevector(sourcelist,elementlist):
        return map(lambda i,l=sourcelist: l[i], elementlist)

-- 
Any opinions expressed are my own and not necessarily those of Roundpoint.


From fredrik at pythonware.com  Wed Feb 28 15:21:18 2001
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 28 Feb 2001 14:21:18 GMT
Subject: CPAN functionality for python - requirements
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022623140707.15046@branagan> <20010226214909.A10411@tummy.com> <01022706564001.20069@branagan> <mailman.983302211.26581.python-list@python.org>
Message-ID: <yP7n6.15965$Qb7.2607352@newsb.telia.net>

Sean Reifschneider wrote:
> Don't get me wrong, I'm all for documentation.  The lack of "javadoc" for
> Python has really hurt though...

current CVS version:

>>> import pydoc
>>> print pydoc.__doc__
Generate Python documentation in HTML or as text for interactive use.

At the shell command line outside of Python, run "pydoc <name>" to show
documentation on something.  <name> may be the name of a Python function,
module, package, or a dotted reference to a class or function within a
module or module in a package.  Alternatively, the argument can be the
path to a Python source file.

Or, at the shell prompt, run "pydoc -k <keyword>" to search for a keyword
in the one-line descriptions of modules.

Or, at the shell prompt, run "pydoc -p <port>" to start an HTTP server
on a given port on the local machine to generate documentation web pages.

Or, at the shell prompt, run "pydoc -w <name>" to write out the HTML
documentation for a module to a file named "<name>.html".

In the Python interpreter, do "from pydoc import help" to provide online
help.  Calling help(thing) on a Python object documents the object.

Cheers /F




From cerutti at together.net  Mon Feb 12 20:02:55 2001
From: cerutti at together.net (Neil Cerutti)
Date: 12 Feb 2001 19:02:55 GMT
Subject: Tkinter event
References: <963ekd$7si$1@buty.wanadoo.nl>
Message-ID: <slrn98gd9b.7m.cerutti@fiad06.norwich.edu>

knock knock posted:
>got a canvas with items (lines). the mouseover event should give a specifc
>desciption for every item in the canvas. so i did a mouseover bind to a
>unique tag for every item! the idea is to make function that listens to the
>event and look ups the description. how can this method find out which item
>(tag) sent the event?? this must be possible, right?

Use the widget method of your event object.

-- 
Neil Cerutti <cerutti at together.net>


From chris.gonnerman at usa.net  Mon Feb  5 04:39:38 2001
From: chris.gonnerman at usa.net (Chris Gonnerman)
Date: Sun, 4 Feb 2001 21:39:38 -0600
Subject: Install PythonWin on top of Python 2.0
References: <95l4en$rbs$1@nnrp1.deja.com>
Message-ID: <008601c08f25$48e987c0$a100000a@local>

----- Original Message -----
From: "Deja User" <binary3418 at my-deja.com>
Subject: Install PythonWin on top of Python 2.0


> Hi, there:
>
>   This is a question from a new user. I recently installed the BeOpen
> Python 2.0 on my Win98 PC. Now I'd like to install the Win32 extension
> so I download "win32all-125.exe". Unfortunately, I cannot install it
> because this build insists on the presence of Python 1.5.x. Now, do I
> have to install 1.5 so as to install PythonWin? Would the author of
> PythonWin please do us a favor to relax the version check in PythonWin?

You have the wrong version.  Sadly the python.org site is a bit confusing;
ActiveState
distributes this software, and you need the 2.0 compatible version.  Visit:

    http://www.activestate.com/Products/ActivePython/win32all.html

for downloads and info.






From embed at geocities.com  Fri Feb  9 15:17:51 2001
From: embed at geocities.com (Warren Postma)
Date: Fri, 9 Feb 2001 09:17:51 -0500
Subject: types.FunctionType vs types.LambdaType
References: <mailman.981710703.17559.python-list@python.org>
Message-ID: <y%Sg6.1908$D3.8413@tor-nn1.netcom.ca>

"Xavier Defrang" deftly observed:
> I was exploring the wonderful world of the Python
> documentation when I saw that the standard library
> types module had two distinct function types :
> FunctionType and LambdaType.  Both of these are
> references to the type 'function' (see code snippet).


Maybe it's historical baggage. It's been the same actual object, with two
'name bindings' in the 'types' module at least since 1.5.2.

Warren




From jafo at tummy.com  Tue Feb 27 05:49:09 2001
From: jafo at tummy.com (Sean Reifschneider)
Date: Mon, 26 Feb 2001 21:49:09 -0700
Subject: CPAN functionality for python - requirements
In-Reply-To: <01022623140707.15046@branagan>; from doughellmann@home.com on Mon, Feb 26, 2001 at 10:45:38PM -0500
References: <Pine.LNX.4.33.0102260150060.28098-100000@bms> <01022606501904.15046@branagan> <20010226183606.F7531@tummy.com> <01022623140707.15046@branagan>
Message-ID: <20010226214909.A10411@tummy.com>

On Mon, Feb 26, 2001 at 10:45:38PM -0500, Doug Hellmann wrote:
>The command "rpm -qa" only produces meaningful results on a system where RPM is
>the default and primary package installation system.  There are more OSes where
>that is *not* the case (and where RPM isn't installed at all) than where it is.

Apaprently I wasn't canonical enough...  On an rpm-based system, it's
bogus if "rpm -qa" doesn't list packages installed by CPAN.  Just as it's
bogus for the similar to happen on Debian, HP, Sun, Windows, etc...

>It still isn't clear to me why we would want to use more than one distribution
>package format.  It appears that one argument is so that sysadmins can use
>those tools to track what is installed on the system.  But isn't that part of

That's one argument, and I'd say it's a good one at that...  Other reasons
are:

	The admin doesn't have to learn new packaging formats to deal with the
	Python packages.

	If we don't leverage off the platform's package management tool, we have
	to write our own.  Who's volunteering?

	If the packages aren't listed in the native packaging system, they are
	much more likely to get overlooked if admins search for things to
	update and the like.

>1. Locate Python packages you want on the net.

By "packages" do you mean modules only?  I envision it also handling
distribution of Python applications and even Python itself.

>2. Resolve dependencies between what you want, what you have, and what you
>   need.
>3. Download the appropriate packages to give you what you want.
>4. Install those packages.
>
>Do we agree these are the primary features?

Yes.

>I sense a consensus that the "install" part should be handled by distutils.  Is
>that right?  

That seems to be where we diverge...  If somone has made an RPM of it, I'd
rather have that than some files winding up hanging around my file-system.

>3. Need to include documentation along with source for packages.

Not gonna happen...  Until there's tools and standards for such documentation,
it's not really possible to deal with them...  In the cases where docstrings
are used, installing the module produces the documentation, but that seems
to be the exception...

>complexity of the new tool.  (How do we handle RPMs on MacOS?  How do we handle

My schema lists the package format, architecture and platform.  If you
search for a package for MacOS, and are returned an RPM, then apparently
somone has built an RPM for MacOS.  Your client should list the package
formats that it desires and prefers, which means that if you don't have
RPM on your Mac, you will pull down something else -- without penalizing
those who *DO* have RPM running on their Mac...

>HQX files on Win32?  What about for Pippy, where files as such might not even
>be appropriate?  Does Jython have any special requirements?)  Do we deal with

My setup handles that -- set the platform to "pilot" and you get access to
only the things that will work there (in theory).

Again, there are two distinct things here.  The archive network and the
tools which make use of it...

>this complexity by deciding on a per-platform basis what format to use, or do
>we force the user to specify the format they want downloaded?

You say "we force the user to specify".  I say "we allow the user the choose".

>If we're modeling what we're doing based on CPAN, maybe we should look at
>that design.  What format does CPAN use when downloading Perl packages?  What
>features does CPAN have that we want?  What does it not have that we want?

CPAN is a Unix-like directory structure, files are downloaded as .tar bundles
which are extracted and a "perl Makefile.pl; make; make test; make install"
is run.  Does Pippy have "make"?

Sean
-- 
 Gone Postal Sort: Iterate over elements, any element that is out of order you
 blow away.  -- Evelyn, Kevin, and Sean, watching Monty Python and reading DDJ
Sean Reifschneider, Inimitably Superfluous <jafo at tummy.com>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python



From johngrayson at home.com  Fri Feb  9 12:56:40 2001
From: johngrayson at home.com (John Grayson)
Date: Fri, 09 Feb 2001 11:56:40 GMT
Subject: shift tab Tkinter
References: <95voh4$oj$1@nnrp1.deja.com>
Message-ID: <960ltm$mrn$1@nnrp1.deja.com>

In article <95voh4$oj$1 at nnrp1.deja.com>,
  bernard604 at my-deja.com wrote:
> I can not get Tkinter to respond to any double key events.

This little snippet works...

from Tkinter import *

root = Tk()

def gotit(event):
    print 'Got Shift-Tab'

frame = Frame(root)
entry = Entry(frame)
entry.pack(anchor=CENTER)
frame.pack()

root.bind_class('Entry', '<Shift-Tab>', gotit)

root.mainloop()


   John Grayson


Sent via Deja.com
http://www.deja.com/


From root at rainerdeyke.com  Tue Feb  6 22:45:39 2001
From: root at rainerdeyke.com (Rainer Deyke)
Date: Tue, 06 Feb 2001 21:45:39 GMT
Subject: nested scopes
References: <mailman.981176465.13517.python-list@python.org> <95lu1i$ebs@gap.cco.caltech.edu> <mailman.981395193.31631.python-list@python.org> <3A7EF349.407C71E1@san.rr.com> <mailman.981470381.12491.python-list@python.org> <3A804026.AD3D1540@san.rr.com> <m3r91b3b5b.fsf@atrus.jesus.cam.ac.uk> <vSYf6.182681$ge4.63154971@news2.rdc2.tx.home.com> <m3lmrj35z6.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <7g_f6.182817$ge4.63254223@news2.rdc2.tx.home.com>

"Michael Hudson" <mwh21 at cam.ac.uk> wrote in message
news:m3lmrj35z6.fsf at atrus.jesus.cam.ac.uk...
> "Rainer Deyke" <root at rainerdeyke.com> writes:
> > Some Python functions (including 'globals' and 'locals') have behavior
> > similar to that of dynamic scoping.
>
> How so?  Note I'm using "dynamic scope" in a pretty technical sense
> here - "indefinite scope and dynamic extent".  Python variables have
> definite scope and dynamic extent (although the objects they point to
> have indefinite extent).

I am referring to the fact that these functions access the namespace of the
calling function.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor




From jon_warner at my-deja.com  Thu Feb  8 18:39:38 2001
From: jon_warner at my-deja.com (jon_warner at my-deja.com)
Date: Thu, 08 Feb 2001 17:39:38 GMT
Subject: anydbm can't find type of dbm file
References: <95s05c$njd$1@nnrp1.deja.com> <mailman.981614822.29236.python-list@python.org>
Message-ID: <95ulko$2eg$1@nnrp1.deja.com>

In article <mailman.981614822.29236.python-list at python.org>,
  Sean Reifschneider <jafo at tummy.com> wrote:
> >  File "/usr/lib/python1.5/anydbm.py", line 83, in open
> >    raise error, "db type could not be determined"
> >anydbm.error: db type could not be determined
>
> Please see http://www.tummy.com/radiuscontext/errata.html for a
discussion
> of this and possible ways to fix it.

Thanks Sean,

I'll install a newer Python.

Jonathan


Sent via Deja.com
http://www.deja.com/


From Bill.Scherer at VerizonWireless.com  Tue Feb 13 17:01:07 2001
From: Bill.Scherer at VerizonWireless.com (Bill Scherer)
Date: Tue, 13 Feb 2001 11:01:07 -0500
Subject: Python 2.1 release schedule
References: <mailman.981764955.17485.python-list@python.org> <3A895617.C7EDB5C8@javanet.com>
Message-ID: <3A895A43.6C2DA72@VerizonWireless.com>

Raymond Hettinger wrote:
> 
> Jeremy Hylton wrote:
> 
> > I updated the Python 2.1 release schedule (PEP 226):
> > <snip>
> > second beta in mid- to late-March, and aim for a final release
> > sometime in April.
> 
> Call me superstitious, but going final on Friday, April 13th
>  in the first year of a new millineum, at tax time seems a
> trifle risky.
> 
> RH

I disagree.  Could you imagine a better confluence of external events to
coincide with the release of a wonderful new Python?  There has to be a
balance between good and evil, right?  ;-)



-- 
William K. Scherer
Sr. Member of Applications Staff - Verizon Wireless
Bill.Scherer_at_VerizonWireless.com



From amk at mira.erols.com  Tue Feb 20 01:54:43 2001
From: amk at mira.erols.com (A.M. Kuchling)
Date: 20 Feb 2001 00:54:43 GMT
Subject: What's up with enscript for syntax highlighting?
References: <96p7f2$bqi@dispatch.concentric.net> <96p7rk$sub@dispatch.concentric.net> <mailman.982530861.24974.python-list@python.org> <96pf0m$sur@dispatch.concentric.net>
Message-ID: <slrn993gtk.hb.amk@207-172-111-136.s136.tnt1.ann.va.dialup.rcn.com>

On 18 Feb 2001 21:32:06 GMT, Phlip <phlip_cpp at my-deja.com> wrote:
>Proclaimed Thomas Wouters from the mountaintops:
>> one everyone commits at some point in life. Furthermore, RTFM is much more
>> effective if you do it gently and make them feel nicely embarassed, rather
>> than having them just say 'well, fuck you too' when reading the first
>> insult, and not learn a thing.
>
>Thanks. I'l try to keep that in mind the next time I flame myself.

Isn't comp.lang.python a great newsgroup? :)

--amk



From thomas at xs4all.net  Sun Feb 18 22:13:36 2001
From: thomas at xs4all.net (Thomas Wouters)
Date: Sun, 18 Feb 2001 22:13:36 +0100
Subject: What's up with enscript for syntax highlighting?
In-Reply-To: <96p7rk$sub@dispatch.concentric.net>; from phlip_cpp@my-deja.com on Sun, Feb 18, 2001 at 07:29:56PM +0000
References: <96p7f2$bqi@dispatch.concentric.net> <96p7rk$sub@dispatch.concentric.net>
Message-ID: <20010218221336.C14292@xs4all.nl>

On Sun, Feb 18, 2001 at 07:29:56PM +0000, Phlip wrote:
> Proclaimed Phlip from the mountaintops:

> > Is there a class
>